Diff [7030dab75c493bbd89de53cb88e496019c794ff4:71d6bd8f286296bcd18128eea5d0f04ce5968247] for / – Cforall

Jenkinsfile_disabled

-              r7030dab
+              r71d6bd8
+                        }
                         sh "${SrcDir}/configure CXX=${Settings.Compiler.CXX} CC=${Settings.Compiler.CC} ${Settings.Architecture.flags} AR=gcc-ar RANLIB=gcc-ranlib ${targets} --quiet --prefix=${BuildDir}"
+                        sh "${SrcDir}/configure CXX=${Settings.Compiler.CXX} CC=${Settings.Compiler.CC} ${Settings.Architecture.flags} ${targets} --quiet"
                         // Configure libcfa
 …
                 dir (BuildDir) {
                         sh "make -j 8 --no-print-directory -C libcfa/${Settings.Architecture.name}-nodebug"
+                }
+        }
-        build_stage('Build : install', true) {
-                // Build outside of the src tree to ease cleaning
-                dir (BuildDir) {
-                        sh "make -j 8 --no-print-directory install"
+                }
+        }
 …
                 echo "Archiving core dumps"
                 dir (BuildDir) {
                         archiveArtifacts artifacts: "tests/crashes/**/*,lib/**/lib*.so*", fingerprint: true
+                        archiveArtifacts artifacts: "tests/crashes/**/*", fingerprint: true
+                }
                 throw err
 …
                 //Then publish the results
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'compile'        , groupCompile    , false, 'Compilation')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'compile.diff'   , groupCompile    , true , 'Compilation (relative)')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'ctxswitch'      , groupConcurrency, false, 'Context Switching')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'ctxswitch.diff' , groupConcurrency, true , 'Context Switching (relative)')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'mutex'          , groupConcurrency, false, 'Mutual Exclusion')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'mutex.diff'     , groupConcurrency, true , 'Mutual Exclusion (relative)')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'scheduling'     , groupConcurrency, false, 'Internal and External Scheduling')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'scheduling.diff', groupConcurrency, true , 'Internal and External Scheduling (relative)')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'compile'       , groupCompile    , false, 'Compilation')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'compile.diff'  , groupCompile    , true , 'Compilation (relative)')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'ctxswitch'     , groupConcurrency, false, 'Context Switching')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'ctxswitch.diff', groupConcurrency, true , 'Context Switching (relative)')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'mutex'         , groupConcurrency, false, 'Mutual Exclusion')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'mutex.diff'    , groupConcurrency, true , 'Mutual Exclusion (relative)')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'signal'        , groupConcurrency, false, 'Internal and External Scheduling')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'signal.diff'   , groupConcurrency, true , 'Internal and External Scheduling (relative)')
+        }
+}
 …
         public String CXX
         public String CC
+        public String lto
+        CC_Desc(String name, String CXX, String CC, String lto) {
+        CC_Desc(String name, String CXX, String CC) {
                 this.name = name
                 this.CXX = CXX
+                this.CC  = CC
+                this.lto = lto
+                this.CC = CC
+        }
+}
 …
                 switch( param.Compiler ) {
                         case 'gcc-9':
                                 this.Compiler = new CC_Desc('gcc-9', 'g++-9', 'gcc-9', '-flto=auto')
+                                this.Compiler = new CC_Desc('gcc-9', 'g++-9', 'gcc-9')
                         break
                         case 'gcc-8':
                                 this.Compiler = new CC_Desc('gcc-8', 'g++-8', 'gcc-8', '-flto=auto')
+                                this.Compiler = new CC_Desc('gcc-8', 'g++-8', 'gcc-8')
                         break
                         case 'gcc-7':
                                 this.Compiler = new CC_Desc('gcc-7', 'g++-7', 'gcc-7', '-flto=auto')
+                                this.Compiler = new CC_Desc('gcc-7', 'g++-7', 'gcc-7')
                         break
                         case 'gcc-6':
                                 this.Compiler = new CC_Desc('gcc-6', 'g++-6', 'gcc-6', '-flto=auto')
+                                this.Compiler = new CC_Desc('gcc-6', 'g++-6', 'gcc-6')
                         break
                         case 'gcc-5':
                                 this.Compiler = new CC_Desc('gcc-5', 'g++-5', 'gcc-5', '-flto=auto')
+                                this.Compiler = new CC_Desc('gcc-5', 'g++-5', 'gcc-5')
                         break
                         case 'gcc-4.9':
                                 this.Compiler = new CC_Desc('gcc-4.9', 'g++-4.9', 'gcc-4.9', '-flto=auto')
+                                this.Compiler = new CC_Desc('gcc-4.9', 'g++-4.9', 'gcc-4.9')
                         break
                         case 'clang':
                                 this.Compiler = new CC_Desc('clang', 'clang++-6.0', 'gcc-6', '-flto=thin -flto-jobs=0')
+                                this.Compiler = new CC_Desc('clang', 'clang++-6.0', 'gcc-6')
                         break
                         default :
 …
         // prepare the properties
         properties ([                                                                                                   \
-                buildDiscarder(logRotator(                                                                              \
-                        artifactDaysToKeepStr: '',                                                                      \
-                        artifactNumToKeepStr: '',                                                                       \
-                        daysToKeepStr: '730',                                                                           \
-                        numToKeepStr: '1000'                                                                            \
-                )),                                                                                                             \
                 [$class: 'ParametersDefinitionProperty',                                                                \
                         parameterDefinitions: [                                                                         \

benchmark/Makefile.am

-              r7030dab
+              r71d6bd8
 ## Created On       : Sun May 31 09:08:15 2015
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Tue Mar 10 11:41:18 2020
 ## Update Count     : 258
+## Last Modified On : Mon Jul 29 18:02:19 2019
+## Update Count     : 54
 ###############################################################################
 …
 BENCH_V_CFA = $(__bench_v_CFA_$(__quiet))
 BENCH_V_CXX = $(__bench_v_CXX_$(__quiet))
+BENCH_V_GOC = $(__bench_v_GOC_$(__quiet))
+BENCH_V_JAVAC = $(__bench_v_JAVAC_$(__quiet))
 BENCH_V_UPP = $(__bench_v_UPP_$(__quiet))
-BENCH_V_GOC = $(__bench_v_GOC_$(__quiet))
-BENCH_V_PY = $(__bench_v_PY_$(__quiet))
-BENCH_V_RUSTC = $(__bench_v_RUSTC_$(__quiet))
-BENCH_V_NODEJS = $(__bench_v_NODEJS_$(__quiet))
-BENCH_V_JAVAC = $(__bench_v_JAVAC_$(__quiet))
 __quiet = verbose
 …
 __bench_v_CFA_quiet = @
 __bench_v_CXX_quiet = @
+__bench_v_GOC_quiet = @
+__bench_v_JAVAC_quiet = @
 __bench_v_UPP_quiet = @
-__bench_v_GOC_quiet = @
-__bench_v_RUSTC_quiet = @
-__bench_v_JAVAC_quiet = @
 __bench_v_CC_verbose = $(AM_V_CC)
 __bench_v_CFA_verbose = $(AM_V_CFA)
 __bench_v_CXX_verbose = $(AM_V_CXX)
+__bench_v_GOC_verbose = $(AM_V_GOC)
+__bench_v_JAVAC_verbose = $(AM_V_JAVAC)
 __bench_v_UPP_verbose = $(AM_V_UPP)
-__bench_v_GOC_verbose = $(AM_V_GOC)
-__bench_v_PY_verbose = $(AM_V_PY)
-__bench_v_RUSTC_verbose = $(AM_V_RUST)
-__bench_v_NODEJS_verbose = $(AM_V_NODEJS)
-__bench_v_JAVAC_verbose = $(AM_V_JAVAC)
 …
 STATS    = ${abs_top_srcdir}/tools/stat.py
 # NEED AT LEAST 4 DATA VALUES FOR BENCHMARKS BECAUSE THE MAX AND MIN VALUES ARE REMOVED
 repeats  = 13 # 31 for benchmarks
+repeats  = 5 # 31 for benchmarks
 arch     = x64
 skipcompile = no
 …
 dummyC.c:
         echo "int main() { return 0; }" > ${@}
+        @echo "int main() { return 0; }" > ${@}
 dummyCXX.cpp:
+        echo "int main() { return 0; }" > ${@}
+.SILENT:                # do not print recipe
+        @echo "int main() { return 0; }" > ${@}
 .NOTPARALLEL:
+.PHONY: jenkins cleancsv
+## =========================================================================================================
+all : basic$(EXEEXT) ctxswitch$(EXEEXT) mutex$(EXEEXT) schedint$(EXEEXT) schedext$(EXEEXT) creation$(EXEEXT)
+basic_loop_DURATION = 15000000000
+basic_function_DURATION = 10000000000
+basic_tls_fetch_add_DURATION = 10000000000
+basic_DURATION = 250000000
+ctxswitch_pthread_DURATION = 25000000
+ctxswitch_rust_thread_DURATION = $(ctxswitch_pthread_DURATION)
+ctxswitch_cfa_generator_DURATION = 5000000000
+ctxswitch_nodejs_await_DURATION = 5000000
+ctxswitch_DURATION = 100000000
+#mutex_java_DURATION = 10000000
+mutex_DURATION = 50000000
+schedint_pthread_DURATION = 1000000
+schedint_java_DURATION = $(schedint_pthread_DURATION)
+schedint_rust_DURATION = $(schedint_pthread_DURATION)
+schedint_DURATION = 10000000
+schedext_DURATION = 10000000
+creation_pthread_DURATION = 250000
+creation_rust_thread_DURATION = ${creation_pthread_DURATION}
+creation_java_thread_DURATION = ${creation_pthread_DURATION}
+creation_cfa_coroutine_DURATION = 100000000
+creation_cfa_coroutine_eager_DURATION = 10000000
+creation_cfa_generator_DURATION = 1000000000
+creation_upp_coroutine_DURATION = ${creation_cfa_coroutine_eager_DURATION}
+creation_cfa_thread_DURATION = 10000000
+creation_upp_thread_DURATION = ${creation_cfa_thread_DURATION}
+creation_DURATION = 10000000
+.PHONY: compile.csv ctxswitch.csv mutex.csv signal.csv
+## =========================================================================================================
+all : ctxswitch$(EXEEXT) mutex$(EXEEXT) signal$(EXEEXT) waitfor$(EXEEXT) creation$(EXEEXT)
 %.run : %$(EXEEXT) ${REPEAT}
+        rm -f .result.log
+        echo "------------------------------------------------------"
+        echo $<
+        ${REPEAT} ${repeats} -- ./a.out\
+                $(if ${$(subst -,_,$(basename $@))_DURATION},\
+                        ${$(subst -,_,$(basename $@))_DURATION},\
+                        ${$(firstword $(subst -, ,$(basename $@)))_DURATION}) | tee -a .result.log
+        ${STATS} .result.log
+        echo "------------------------------------------------------"
+        rm -f a.out .result.log *.class
+#       ${REPEAT} ${repeats} -- /usr/bin/time -f "%Uu %Ss %Er %Mkb" ./a.out
+        @rm -f .result.log
+        @echo "------------------------------------------------------"
+        @echo $<
+        @${REPEAT} ${repeats} ./a.out | tee -a .result.log
+        @${STATS} .result.log
+        @echo "------------------------------------------------------"
+        @rm -f a.out .result.log *.class
 %.runquiet :
         +make $(basename $@) CFLAGS="-w" __quiet=quiet
         taskset -c 1 ./a.out
         rm -f a.out
+        @+make $(basename $@) CFLAGS="-w" __quiet=quiet
+        @taskset -c 1 ./a.out
+        @rm -f a.out
 %.make :
         printf "${PRINT_FORMAT}" $(basename $(subst compile-,,$@))
         +/usr/bin/time -f ${TIME_FORMAT} make $(basename $@) 2>&1
+        @printf "${PRINT_FORMAT}" $(basename $(subst compile-,,$@))
+        @+/usr/bin/time -f ${TIME_FORMAT} make $(basename $@) 2>&1
 ${REPEAT} :
         +make -C ${abs_top_builddir}/tools repeat
+        @+make -C ${abs_top_builddir}/tools repeat
 ## =========================================================================================================
 …
 FIX_NEW_LINES = cat $@ | tr "\n" "\t" | sed -r 's/\t,/,/' | tr "\t" "\n" > $@
+cleancsv:
+        rm -f compile.csv basic.csv ctxswitch.csv mutex.csv scheduling.csv
+jenkins$(EXEEXT): cleancsv
+jenkins$(EXEEXT):
 @DOifskipcompile@
         +make compile.csv
         -+make compile.diff.csv
+        @+make compile.csv
+        @-+make compile.diff.csv
 @DOendif@
         +make ctxswitch.csv
         -+make ctxswitch.diff.csv
         +make mutex.csv
         -+make mutex.diff.csv
         +make scheduling.csv
         -+make scheduling.diff.csv
+        @+make ctxswitch.csv
+        @-+make ctxswitch.diff.csv
+        @+make mutex.csv
+        @-+make mutex.diff.csv
+        @+make signal.csv
+        @-+make signal.diff.csv
 @DOifskipcompile@
         cat compile.csv
 …
         cat mutex.csv
         -cat mutex.diff.csv
         cat scheduling.csv
         -cat scheduling.diff.csv
+        cat signal.csv
+        -cat signal.diff.csv
 compile.csv:
+        echo "building $@"
+        echo "array,attributes,empty,expression,io,monitor,operators,typeof" > $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-array.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-attributes.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-empty.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-expression.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-io.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-monitor.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-operators.make >> $@
+        +make TIME_FORMAT='%e' PRINT_FORMAT='' compile-typeof.make >> $@
+        $(srcdir)/fixcsv.sh $@
+        @echo "array,attributes,empty,expression,io,monitor,operators,typeof" > $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-array.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-attributes.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-empty.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-expression.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-io.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-monitor.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-operators.make >> $@
+        @+make TIME_FORMAT='%e' PRINT_FORMAT='' compile-typeof.make >> $@
+        @$(srcdir)/fixcsv.sh $@
 ctxswitch.csv:
+        echo "building $@"
+        echo "generator,coroutine,thread" > $@
+        +make ctxswitch-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
+        +make ctxswitch-cfa_coroutine.runquiet >> $@ && echo -n ',' >> $@
+        +make ctxswitch-cfa_thread.runquiet >> $@
+        $(srcdir)/fixcsv.sh $@
+        @echo "generator,coroutine,thread" > $@
+        @+make ctxswitch-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
+        @+make ctxswitch-cfa_coroutine.runquiet >> $@ && echo -n ',' >> $@
+        @+make ctxswitch-cfa_thread.runquiet >> $@
+        @$(srcdir)/fixcsv.sh $@
 mutex.csv:
+        echo "building $@"
+        echo "1-monitor,2-monitor" > $@
+        +make mutex-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        +make mutex-cfa2.runquiet >> $@
+        $(srcdir)/fixcsv.sh $@
+scheduling.csv:
+        echo "building $@"
+        echo "schedint-1,schedint-2,schedext-1,schedext-2" > $@
+        +make schedint-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        +make schedint-cfa2.runquiet >> $@ && echo -n ',' >> $@
+        +make schedext-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        +make schedext-cfa2.runquiet >> $@
+        $(srcdir)/fixcsv.sh $@
+        @echo "1-monitor,2-monitor" > $@
+        @+make mutex-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        @+make mutex-cfa2.runquiet >> $@
+        @$(srcdir)/fixcsv.sh $@
+signal.csv:
+        @echo "signal-1,signal-2,waitfor-1,waitfor-2" > $@
+        @+make signal-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        @+make signal-cfa2.runquiet >> $@ && echo -n ',' >> $@
+        @+make waitfor-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        @+make waitfor-cfa2.runquiet >> $@
+        @$(srcdir)/fixcsv.sh $@
 %.diff.csv: %.csv
+        test -e $(srcdir)/baselines/$(arch)/$< || (echo "Error : Missing baseline for ${<}" && false)
+        $(srcdir)/baselines/calc.py $(srcdir)/baselines/$(arch)/$(<) $(<) > $@
+## =========================================================================================================
+BASIC_DEPEND  =                                 \
+        basic-loop.run                          \
+        basic-function.run                      \
+        basic-fetch_add.run                     \
+        basic-ttst_lock.run                     \
+        basic-tls-fetch_add.run
+basic-loop$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/loop.c
+basic-function$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/function.c
+basic-fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/fetch_add.c
+basic-ttst_lock$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/ttst_lock.c
+basic-tls-fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/tls_fetch_add.c
+basic$(EXEEXT): $(BASIC_DEPEND)
+## =========================================================================================================
+CTXSWITCH_DEPEND  =                     \
+        @test -e $(srcdir)/baselines/$(arch)/$< || (echo "Error : Missing baseline for ${<}" && false)
+        @$(srcdir)/baselines/calc.py $(srcdir)/baselines/$(arch)/$(<) $(<) > $@
+## =========================================================================================================
+loop$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=5000000000 $(srcdir)/loop.c
+function$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=5000000000 $(srcdir)/function.c
+fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=500000000  $(srcdir)/fetch_add.c
+ttst_lock$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=500000000  $(srcdir)/ttst_lock.c
+tls-fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=500000000  $(srcdir)/tls-fetch_add.c
+## =========================================================================================================
+CTXSWITCH_DEPEND  =                 \
+        loop.run                                \
+        function.run                    \
+        fetch_add.run                   \
+        ttst_lock.run                   \
+        tls-fetch_add.run                       \
+        ctxswitch-pthread.run           \
         ctxswitch-cfa_generator.run     \
         ctxswitch-cfa_coroutine.run     \
 …
         ctxswitch-upp_coroutine.run     \
         ctxswitch-upp_thread.run        \
+        ctxswitch-python_coroutine.run  \
+        ctxswitch-nodejs_coroutine.run  \
+        ctxswitch-nodejs_await.run      \
+        ctxswitch-goroutine_thread.run  \
+        ctxswitch-rust_thread.run       \
+        ctxswitch-nodejs_coroutine.run  \
+        ctxswitch-java_thread.run       \
+        ctxswitch-pthread.run
+        ctxswitch-goroutine.run         \
+        ctxswitch-java_thread.run
 if WITH_LIBFIBRE
 CTXSWITCH_DEPEND  +=                    \
         ctxswitch-kos_fibre.run         \
+CTXSWITCH_DEPEND  +=           \
+        ctxswitch-kos_fibre.run  \
         ctxswitch-kos_fibre2.run
 ctxswitch-kos_fibre$(EXEEXT):
 …
 ctxswitch$(EXEEXT): $(CTXSWITCH_DEPEND)
+ctxswitch-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=50000000 $(srcdir)/ctxswitch/pthreads.c
 ctxswitch-cfa_generator$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_gen.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_gen.cfa
 ctxswitch-cfa_coroutine$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_cor.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_cor.cfa
 ctxswitch-cfa_thread$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_thrd.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_thrd.cfa
 ctxswitch-cfa_thread2$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_thrd2.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_thrd2.cfa
 ctxswitch-upp_coroutine$(EXEEXT):
         $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/ctxswitch/upp_cor.cc
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/upp_cor.cc
 ctxswitch-upp_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/ctxswitch/upp_thrd.cc
+ctxswitch-python_coroutine$(EXEEXT):
+        $(BENCH_V_PY)echo "#!/bin/sh" > a.out
+        echo "python3.7 $(srcdir)/ctxswitch/python_cor.py" >> a.out
+        chmod a+x a.out
+ctxswitch-nodejs_coroutine$(EXEEXT):
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
+        echo "nodejs $(srcdir)/ctxswitch/node_cor.js" >> a.out
+        chmod a+x a.out
+ctxswitch-nodejs_await$(EXEEXT):
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
+        echo "nodejs $(srcdir)/ctxswitch/node_await.js" >> a.out
+        chmod a+x a.out
+ctxswitch-goroutine_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/upp_thrd.cc
+ctxswitch-goroutine$(EXEEXT):
         $(BENCH_V_GOC)go build -o a.out $(srcdir)/ctxswitch/goroutine.go
-ctxswitch-rust_thread$(EXEEXT):
-        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/ctxswitch/rust_thrd.rs
 ctxswitch-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/ctxswitch/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+ctxswitch-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/ctxswitch/pthreads.c
+## =========================================================================================================
+mutex$(EXEEXT) :                \
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+## =========================================================================================================
+mutex$(EXEEXT) :\
+        loop.run                        \
+        function.run            \
+        fetch_add.run           \
+        mutex-pthread_lock.run  \
+        mutex-upp.run           \
         mutex-cfa1.run          \
         mutex-cfa2.run          \
         mutex-cfa4.run          \
+        mutex-upp.run           \
+        mutex-go.run            \
+        mutex-rust.run          \
+        mutex-java.run          \
+        mutex-pthread.run
+mutex-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/mutex/pthreads.c
+        mutex-java_thread.run
+mutex-pthread_lock$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=50000000 $(srcdir)/mutex/pthreads.c
+mutex-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/mutex/upp.cc
 mutex-cfa1$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/mutex/cfa1.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=5000000  $(srcdir)/mutex/cfa1.cfa
 mutex-cfa2$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/mutex/cfa2.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=5000000  $(srcdir)/mutex/cfa2.cfa
 mutex-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/mutex/cfa4.cfa
+mutex-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/mutex/upp.cc
+mutex-go$(EXEEXT):
+        $(BENCH_V_GOC)go build -o a.out $(srcdir)/mutex/goroutine.go
+mutex-rust$(EXEEXT):
+        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/mutex/rust.rs
+mutex-java$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=5000000  $(srcdir)/mutex/cfa4.cfa
+mutex-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/mutex/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+## =========================================================================================================
+schedint$(EXEEXT) :             \
+        schedint-cfa1.run       \
+        schedint-cfa2.run       \
+        schedint-cfa4.run       \
+        schedint-upp.run        \
+        schedint-rust.run       \
+        schedint-java.run       \
+        schedint-pthread.run
+schedint-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedint/cfa1.cfa
+schedint-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedint/cfa2.cfa
+schedint-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedint/cfa4.cfa
+schedint-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/schedint/upp.cc
+schedint-rust$(EXEEXT):
+        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/schedint/rust.rs
+schedint-java$(EXEEXT):
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+## =========================================================================================================
+signal$(EXEEXT) :\
+        signal-pthread_cond.run \
+        signal-upp.run          \
+        signal-cfa1.run         \
+        signal-cfa2.run         \
+        signal-cfa4.run         \
+        signal-java_thread.run
+signal-pthread_cond$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=500000  $(srcdir)/schedint/pthreads.c
+signal-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=5000000 $(srcdir)/schedint/upp.cc
+signal-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedint/cfa1.cfa
+signal-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedint/cfa2.cfa
+signal-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedint/cfa4.cfa
+signal-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/schedint/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+schedint-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/schedint/pthreads.c
+## =========================================================================================================
+schedext$(EXEEXT) :             \
+        schedext-cfa1.run       \
+        schedext-cfa2.run       \
+        schedext-cfa4.run       \
+        schedext-upp.run        \
+        schedext-goroutine.run
+schedext-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedext/cfa1.cfa
+schedext-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedext/cfa2.cfa
+schedext-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedext/cfa4.cfa
+schedext-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/schedext/upp.cc
+schedext-goroutine$(EXEEXT):
+        $(BENCH_V_GOC)go build -o a.out $(srcdir)/schedext/goroutine.go
+## =========================================================================================================
+creation$(EXEEXT) :                             \
+        creation-cfa_generator.run              \
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+## =========================================================================================================
+waitfor$(EXEEXT) :\
+        waitfor-upp.run         \
+        waitfor-cfa1.run                \
+        waitfor-cfa2.run                \
+        waitfor-cfa4.run
+waitfor-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=5000000 $(srcdir)/schedext/upp.cc
+waitfor-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedext/cfa1.cfa
+waitfor-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedext/cfa2.cfa
+waitfor-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedext/cfa4.cfa
+## =========================================================================================================
+creation$(EXEEXT) :\
+        creation-pthread.run                    \
         creation-cfa_coroutine.run              \
         creation-cfa_coroutine_eager.run        \
 …
         creation-upp_coroutine.run              \
         creation-upp_thread.run                 \
+        creation-python_coroutine.run           \
+        creation-nodejs_coroutine.run           \
+        creation-goroutine_thread.run           \
+        creation-rust_thread.run                \
+        creation-java_thread.run                \
+        creation-pthread.run
+creation-cfa_generator$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_gen.cfa
+        creation-goroutine.run                  \
+        creation-java_thread.run
 creation-cfa_coroutine$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_cor.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=10000000 $(srcdir)/creation/cfa_cor.cfa
 creation-cfa_coroutine_eager$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_cor.cfa  -DEAGER
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=10000000 $(srcdir)/creation/cfa_cor.cfa  -DEAGER
 creation-cfa_thread$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_thrd.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=10000000 $(srcdir)/creation/cfa_thrd.cfa
 creation-upp_coroutine$(EXEEXT):
         $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/creation/upp_cor.cc
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/creation/upp_cor.cc
 creation-upp_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/creation/upp_thrd.cc
+creation-python_coroutine$(EXEEXT):
+        $(BENCH_V_PY)echo "#!/bin/sh" > a.out
+        echo "python3.7 $(srcdir)/creation/python_cor.py" >> a.out
+        chmod a+x a.out
+creation-nodejs_coroutine$(EXEEXT):
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
+        echo "nodejs $(srcdir)/creation/node_cor.js" >> a.out
+        chmod a+x a.out
+creation-goroutine_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/creation/upp_thrd.cc
+creation-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=250000   $(srcdir)/creation/pthreads.c
+creation-goroutine$(EXEEXT):
         $(BENCH_V_GOC)go build -o a.out $(srcdir)/creation/goroutine.go
-creation-rust_thread$(EXEEXT):
-        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/creation/rust_thrd.rs
 creation-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/creation/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+creation-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/creation/pthreads.c
+## =========================================================================================================
+compile$(EXEEXT) :              \
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+## =========================================================================================================
+compile$(EXEEXT) :\
         compile-array.make      \
         compile-attributes.make \
 …
         compile-typeof.make
 testdir = $(top_srcdir)/tests
 compile-array$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/array.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/array.cfa
 compile-attributes$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/attributes.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/attributes.cfa
 compile-empty$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(srcdir)/compile/empty.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(srcdir)/compile/empty.cfa
 compile-expression$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/expression.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/expression.cfa
 compile-io$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/io1.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/io1.cfa
 compile-monitor$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/monitor.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/monitor.cfa
 compile-operators$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/operators.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/operators.cfa
 compile-thread$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/thread.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/thread.cfa
 compile-typeof$(EXEEXT):
+        $(CFACOMPILE) -fsyntax-only -w $(testdir)/typeof.cfa
+## =========================================================================================================
+size$(EXEEXT) : size-cfa.runquiet
+size-cfa$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/size/size.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/typeof.cfa

benchmark/Makefile.in

-              r7030dab
+              r71d6bd8
 EXTRA_PROGRAMS = dummy$(EXEEXT)
 @WITH_LIBFIBRE_TRUE@am__append_1 = \
 @WITH_LIBFIBRE_TRUE@    ctxswitch-kos_fibre.run         \
+@WITH_LIBFIBRE_TRUE@    ctxswitch-kos_fibre.run  \
 @WITH_LIBFIBRE_TRUE@    ctxswitch-kos_fibre2.run
 …
 LTCFACOMPILE = $(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) \
         $(LIBTOOLFLAGS) --mode=compile $(CFACC) $(DEFS) \
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(AM_CFLAGS) $(CFAFLAGS) $(CFLAGS)
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) \
+        $(AM_CFLAGS) $(CFLAGS)
 AM_V_CFA = $(am__v_CFA_@AM_V@)
 …
 am__v_CFA_0 = @echo "  CFA     " $@;
 am__v_CFA_1 =
+AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
+am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
+am__v_JAVAC_0 = @echo "  JAVAC   " $@;
+am__v_JAVAC_1 =
+AM_V_GOC = $(am__v_GOC_@AM_V@)
+am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
+am__v_GOC_0 = @echo "  GOC     " $@;
+am__v_GOC_1 =
 UPPCC = u++
 UPPCOMPILE = $(UPPCC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_UPPFLAGS) $(UPPFLAGS) $(AM_CXXFLAGS) $(CXXFLAGS) $(AM_CFLAGS) $(CFLAGS)
 …
 am__v_UPP_0 = @echo "  UPP     " $@;
 am__v_UPP_1 =
-AM_V_GOC = $(am__v_GOC_@AM_V@)
-am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
-am__v_GOC_0 = @echo "  GOC     " $@;
-am__v_GOC_1 =
-AM_V_PY = $(am__v_PY_@AM_V@)
-am__v_PY_ = $(am__v_PY_@AM_DEFAULT_V@)
-am__v_PY_0 = @echo "  PYTHON  " $@;
-am__v_PY_1 =
-AM_V_RUST = $(am__v_RUST_@AM_V@)
-am__v_RUST_ = $(am__v_RUST_@AM_DEFAULT_V@)
-am__v_RUST_0 = @echo "  RUST    " $@;
-am__v_RUST_1 =
-AM_V_NODEJS = $(am__v_NODEJS_@AM_V@)
-am__v_NODEJS_ = $(am__v_NODEJS_@AM_DEFAULT_V@)
-am__v_NODEJS_0 = @echo "  NODEJS  " $@;
-am__v_NODEJS_1 =
-AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
-am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
-am__v_JAVAC_0 = @echo "  JAVAC   " $@;
-am__v_JAVAC_1 =
 # applies to both programs
 …
 BENCH_V_CFA = $(__bench_v_CFA_$(__quiet))
 BENCH_V_CXX = $(__bench_v_CXX_$(__quiet))
+BENCH_V_GOC = $(__bench_v_GOC_$(__quiet))
+BENCH_V_JAVAC = $(__bench_v_JAVAC_$(__quiet))
 BENCH_V_UPP = $(__bench_v_UPP_$(__quiet))
-BENCH_V_GOC = $(__bench_v_GOC_$(__quiet))
-BENCH_V_PY = $(__bench_v_PY_$(__quiet))
-BENCH_V_RUSTC = $(__bench_v_RUSTC_$(__quiet))
-BENCH_V_NODEJS = $(__bench_v_NODEJS_$(__quiet))
-BENCH_V_JAVAC = $(__bench_v_JAVAC_$(__quiet))
 __quiet = verbose
 __bench_v_CC_quiet = @
 __bench_v_CFA_quiet = @
 __bench_v_CXX_quiet = @
+__bench_v_GOC_quiet = @
+__bench_v_JAVAC_quiet = @
 __bench_v_UPP_quiet = @
-__bench_v_GOC_quiet = @
-__bench_v_RUSTC_quiet = @
-__bench_v_JAVAC_quiet = @
 __bench_v_CC_verbose = $(AM_V_CC)
 __bench_v_CFA_verbose = $(AM_V_CFA)
 __bench_v_CXX_verbose = $(AM_V_CXX)
+__bench_v_GOC_verbose = $(AM_V_GOC)
+__bench_v_JAVAC_verbose = $(AM_V_JAVAC)
 __bench_v_UPP_verbose = $(AM_V_UPP)
-__bench_v_GOC_verbose = $(AM_V_GOC)
-__bench_v_PY_verbose = $(AM_V_PY)
-__bench_v_RUSTC_verbose = $(AM_V_RUST)
-__bench_v_NODEJS_verbose = $(AM_V_NODEJS)
-__bench_v_JAVAC_verbose = $(AM_V_JAVAC)
 TOOLSDIR = ${abs_top_builddir}/tools/
 REPEAT = ${abs_top_builddir}/tools/repeat
 STATS = ${abs_top_srcdir}/tools/stat.py
 # NEED AT LEAST 4 DATA VALUES FOR BENCHMARKS BECAUSE THE MAX AND MIN VALUES ARE REMOVED
 repeats = 13 # 31 for benchmarks
+repeats = 5 # 31 for benchmarks
 arch = x64
 skipcompile = no
 …
 PRINT_FORMAT = %20s: #Comments needed for spacing
 dummy_SOURCES = dummyC.c dummyCXX.cpp
-basic_loop_DURATION = 15000000000
-basic_function_DURATION = 10000000000
-basic_tls_fetch_add_DURATION = 10000000000
-basic_DURATION = 250000000
-ctxswitch_pthread_DURATION = 25000000
-ctxswitch_rust_thread_DURATION = $(ctxswitch_pthread_DURATION)
-ctxswitch_cfa_generator_DURATION = 5000000000
-ctxswitch_nodejs_await_DURATION = 5000000
-ctxswitch_DURATION = 100000000
-#mutex_java_DURATION = 10000000
-mutex_DURATION = 50000000
-schedint_pthread_DURATION = 1000000
-schedint_java_DURATION = $(schedint_pthread_DURATION)
-schedint_rust_DURATION = $(schedint_pthread_DURATION)
-schedint_DURATION = 10000000
-schedext_DURATION = 10000000
-creation_pthread_DURATION = 250000
-creation_rust_thread_DURATION = ${creation_pthread_DURATION}
-creation_java_thread_DURATION = ${creation_pthread_DURATION}
-creation_cfa_coroutine_DURATION = 100000000
-creation_cfa_coroutine_eager_DURATION = 10000000
-creation_cfa_generator_DURATION = 1000000000
-creation_upp_coroutine_DURATION = ${creation_cfa_coroutine_eager_DURATION}
-creation_cfa_thread_DURATION = 10000000
-creation_upp_thread_DURATION = ${creation_cfa_thread_DURATION}
-creation_DURATION = 10000000
 FIX_NEW_LINES = cat $@ | tr "\n" "\t" | sed -r 's/\t,/,/' | tr "\t" "\n" > $@
+BASIC_DEPEND = \
+        basic-loop.run                          \
+        basic-function.run                      \
+        basic-fetch_add.run                     \
+        basic-ttst_lock.run                     \
+        basic-tls-fetch_add.run
+CTXSWITCH_DEPEND = ctxswitch-cfa_generator.run \
+        ctxswitch-cfa_coroutine.run ctxswitch-cfa_thread.run \
+        ctxswitch-cfa_thread2.run ctxswitch-upp_coroutine.run \
+        ctxswitch-upp_thread.run ctxswitch-python_coroutine.run \
+        ctxswitch-nodejs_coroutine.run ctxswitch-nodejs_await.run \
+        ctxswitch-goroutine_thread.run ctxswitch-rust_thread.run \
+        ctxswitch-nodejs_coroutine.run ctxswitch-java_thread.run \
+        ctxswitch-pthread.run $(am__append_1)
+CTXSWITCH_DEPEND = loop.run function.run fetch_add.run ttst_lock.run \
+        tls-fetch_add.run ctxswitch-pthread.run \
+        ctxswitch-cfa_generator.run ctxswitch-cfa_coroutine.run \
+        ctxswitch-cfa_thread.run ctxswitch-cfa_thread2.run \
+        ctxswitch-upp_coroutine.run ctxswitch-upp_thread.run \
+        ctxswitch-goroutine.run ctxswitch-java_thread.run \
+        $(am__append_1)
 testdir = $(top_srcdir)/tests
 all: all-am
 …
 dummyC.c:
         echo "int main() { return 0; }" > ${@}
+        @echo "int main() { return 0; }" > ${@}
 dummyCXX.cpp:
+        echo "int main() { return 0; }" > ${@}
+.SILENT:                # do not print recipe
+        @echo "int main() { return 0; }" > ${@}
 .NOTPARALLEL:
 .PHONY: jenkins cleancsv
 all : basic$(EXEEXT) ctxswitch$(EXEEXT) mutex$(EXEEXT) schedint$(EXEEXT) schedext$(EXEEXT) creation$(EXEEXT)
+.PHONY: compile.csv ctxswitch.csv mutex.csv signal.csv
+all : ctxswitch$(EXEEXT) mutex$(EXEEXT) signal$(EXEEXT) waitfor$(EXEEXT) creation$(EXEEXT)
 %.run : %$(EXEEXT) ${REPEAT}
+        rm -f .result.log
+        echo "------------------------------------------------------"
+        echo $<
+        ${REPEAT} ${repeats} -- ./a.out\
+                $(if ${$(subst -,_,$(basename $@))_DURATION},\
+                        ${$(subst -,_,$(basename $@))_DURATION},\
+                        ${$(firstword $(subst -, ,$(basename $@)))_DURATION}) | tee -a .result.log
+        ${STATS} .result.log
+        echo "------------------------------------------------------"
+        rm -f a.out .result.log *.class
+#       ${REPEAT} ${repeats} -- /usr/bin/time -f "%Uu %Ss %Er %Mkb" ./a.out
+        @rm -f .result.log
+        @echo "------------------------------------------------------"
+        @echo $<
+        @${REPEAT} ${repeats} ./a.out | tee -a .result.log
+        @${STATS} .result.log
+        @echo "------------------------------------------------------"
+        @rm -f a.out .result.log *.class
 %.runquiet :
         +make $(basename $@) CFLAGS="-w" __quiet=quiet
         taskset -c 1 ./a.out
         rm -f a.out
+        @+make $(basename $@) CFLAGS="-w" __quiet=quiet
+        @taskset -c 1 ./a.out
+        @rm -f a.out
 %.make :
         printf "${PRINT_FORMAT}" $(basename $(subst compile-,,$@))
         +/usr/bin/time -f ${TIME_FORMAT} make $(basename $@) 2>&1
+        @printf "${PRINT_FORMAT}" $(basename $(subst compile-,,$@))
+        @+/usr/bin/time -f ${TIME_FORMAT} make $(basename $@) 2>&1
 ${REPEAT} :
+        +make -C ${abs_top_builddir}/tools repeat
+cleancsv:
+        rm -f compile.csv basic.csv ctxswitch.csv mutex.csv scheduling.csv
+jenkins$(EXEEXT): cleancsv
+        @+make -C ${abs_top_builddir}/tools repeat
+jenkins$(EXEEXT):
 @DOifskipcompile@
         +make compile.csv
         -+make compile.diff.csv
+        @+make compile.csv
+        @-+make compile.diff.csv
 @DOendif@
         +make ctxswitch.csv
         -+make ctxswitch.diff.csv
         +make mutex.csv
         -+make mutex.diff.csv
         +make scheduling.csv
         -+make scheduling.diff.csv
+        @+make ctxswitch.csv
+        @-+make ctxswitch.diff.csv
+        @+make mutex.csv
+        @-+make mutex.diff.csv
+        @+make signal.csv
+        @-+make signal.diff.csv
 @DOifskipcompile@
         cat compile.csv
 …
         cat mutex.csv
         -cat mutex.diff.csv
         cat scheduling.csv
         -cat scheduling.diff.csv
+        cat signal.csv
+        -cat signal.diff.csv
 compile.csv:
+        echo "building $@"
+        echo "array,attributes,empty,expression,io,monitor,operators,typeof" > $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-array.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-attributes.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-empty.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-expression.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-io.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-monitor.make >> $@
+        +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-operators.make >> $@
+        +make TIME_FORMAT='%e' PRINT_FORMAT='' compile-typeof.make >> $@
+        $(srcdir)/fixcsv.sh $@
+        @echo "array,attributes,empty,expression,io,monitor,operators,typeof" > $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-array.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-attributes.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-empty.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-expression.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-io.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-monitor.make >> $@
+        @+make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-operators.make >> $@
+        @+make TIME_FORMAT='%e' PRINT_FORMAT='' compile-typeof.make >> $@
+        @$(srcdir)/fixcsv.sh $@
 ctxswitch.csv:
+        echo "building $@"
+        echo "generator,coroutine,thread" > $@
+        +make ctxswitch-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
+        +make ctxswitch-cfa_coroutine.runquiet >> $@ && echo -n ',' >> $@
+        +make ctxswitch-cfa_thread.runquiet >> $@
+        $(srcdir)/fixcsv.sh $@
+        @echo "generator,coroutine,thread" > $@
+        @+make ctxswitch-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
+        @+make ctxswitch-cfa_coroutine.runquiet >> $@ && echo -n ',' >> $@
+        @+make ctxswitch-cfa_thread.runquiet >> $@
+        @$(srcdir)/fixcsv.sh $@
 mutex.csv:
+        echo "building $@"
+        echo "1-monitor,2-monitor" > $@
+        +make mutex-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        +make mutex-cfa2.runquiet >> $@
+        $(srcdir)/fixcsv.sh $@
+scheduling.csv:
+        echo "building $@"
+        echo "schedint-1,schedint-2,schedext-1,schedext-2" > $@
+        +make schedint-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        +make schedint-cfa2.runquiet >> $@ && echo -n ',' >> $@
+        +make schedext-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        +make schedext-cfa2.runquiet >> $@
+        $(srcdir)/fixcsv.sh $@
+        @echo "1-monitor,2-monitor" > $@
+        @+make mutex-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        @+make mutex-cfa2.runquiet >> $@
+        @$(srcdir)/fixcsv.sh $@
+signal.csv:
+        @echo "signal-1,signal-2,waitfor-1,waitfor-2" > $@
+        @+make signal-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        @+make signal-cfa2.runquiet >> $@ && echo -n ',' >> $@
+        @+make waitfor-cfa1.runquiet >> $@ && echo -n ',' >> $@
+        @+make waitfor-cfa2.runquiet >> $@
+        @$(srcdir)/fixcsv.sh $@
 %.diff.csv: %.csv
+        test -e $(srcdir)/baselines/$(arch)/$< || (echo "Error : Missing baseline for ${<}" && false)
+        $(srcdir)/baselines/calc.py $(srcdir)/baselines/$(arch)/$(<) $(<) > $@
+basic-loop$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/loop.c
+basic-function$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/function.c
+basic-fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/fetch_add.c
+basic-ttst_lock$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/ttst_lock.c
+basic-tls-fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/basic/tls_fetch_add.c
+basic$(EXEEXT): $(BASIC_DEPEND)
+        @test -e $(srcdir)/baselines/$(arch)/$< || (echo "Error : Missing baseline for ${<}" && false)
+        @$(srcdir)/baselines/calc.py $(srcdir)/baselines/$(arch)/$(<) $(<) > $@
+loop$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=5000000000 $(srcdir)/loop.c
+function$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=5000000000 $(srcdir)/function.c
+fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=500000000  $(srcdir)/fetch_add.c
+ttst_lock$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=500000000  $(srcdir)/ttst_lock.c
+tls-fetch_add$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) -DBENCH_N=500000000  $(srcdir)/tls-fetch_add.c
 @WITH_LIBFIBRE_TRUE@ctxswitch-kos_fibre$(EXEEXT):
 …
 ctxswitch$(EXEEXT): $(CTXSWITCH_DEPEND)
+ctxswitch-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=50000000 $(srcdir)/ctxswitch/pthreads.c
 ctxswitch-cfa_generator$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_gen.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_gen.cfa
 ctxswitch-cfa_coroutine$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_cor.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_cor.cfa
 ctxswitch-cfa_thread$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_thrd.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_thrd.cfa
 ctxswitch-cfa_thread2$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/ctxswitch/cfa_thrd2.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/cfa_thrd2.cfa
 ctxswitch-upp_coroutine$(EXEEXT):
         $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/ctxswitch/upp_cor.cc
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/upp_cor.cc
 ctxswitch-upp_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/ctxswitch/upp_thrd.cc
+ctxswitch-python_coroutine$(EXEEXT):
+        $(BENCH_V_PY)echo "#!/bin/sh" > a.out
+        echo "python3.7 $(srcdir)/ctxswitch/python_cor.py" >> a.out
+        chmod a+x a.out
+ctxswitch-nodejs_coroutine$(EXEEXT):
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
+        echo "nodejs $(srcdir)/ctxswitch/node_cor.js" >> a.out
+        chmod a+x a.out
+ctxswitch-nodejs_await$(EXEEXT):
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
+        echo "nodejs $(srcdir)/ctxswitch/node_await.js" >> a.out
+        chmod a+x a.out
+ctxswitch-goroutine_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/ctxswitch/upp_thrd.cc
+ctxswitch-goroutine$(EXEEXT):
         $(BENCH_V_GOC)go build -o a.out $(srcdir)/ctxswitch/goroutine.go
-ctxswitch-rust_thread$(EXEEXT):
-        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/ctxswitch/rust_thrd.rs
 ctxswitch-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/ctxswitch/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+ctxswitch-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/ctxswitch/pthreads.c
+mutex$(EXEEXT) :                \
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+mutex$(EXEEXT) :\
+        loop.run                        \
+        function.run            \
+        fetch_add.run           \
+        mutex-pthread_lock.run  \
+        mutex-upp.run           \
         mutex-cfa1.run          \
         mutex-cfa2.run          \
         mutex-cfa4.run          \
+        mutex-upp.run           \
+        mutex-go.run            \
+        mutex-rust.run          \
+        mutex-java.run          \
+        mutex-pthread.run
+mutex-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/mutex/pthreads.c
+        mutex-java_thread.run
+mutex-pthread_lock$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=50000000 $(srcdir)/mutex/pthreads.c
+mutex-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/mutex/upp.cc
 mutex-cfa1$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/mutex/cfa1.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=5000000  $(srcdir)/mutex/cfa1.cfa
 mutex-cfa2$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/mutex/cfa2.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=5000000  $(srcdir)/mutex/cfa2.cfa
 mutex-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/mutex/cfa4.cfa
+mutex-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/mutex/upp.cc
+mutex-go$(EXEEXT):
+        $(BENCH_V_GOC)go build -o a.out $(srcdir)/mutex/goroutine.go
+mutex-rust$(EXEEXT):
+        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/mutex/rust.rs
+mutex-java$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=5000000  $(srcdir)/mutex/cfa4.cfa
+mutex-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/mutex/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+schedint$(EXEEXT) :             \
+        schedint-cfa1.run       \
+        schedint-cfa2.run       \
+        schedint-cfa4.run       \
+        schedint-upp.run        \
+        schedint-rust.run       \
+        schedint-java.run       \
+        schedint-pthread.run
+schedint-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedint/cfa1.cfa
+schedint-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedint/cfa2.cfa
+schedint-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedint/cfa4.cfa
+schedint-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/schedint/upp.cc
+schedint-rust$(EXEEXT):
+        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/schedint/rust.rs
+schedint-java$(EXEEXT):
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+signal$(EXEEXT) :\
+        signal-pthread_cond.run \
+        signal-upp.run          \
+        signal-cfa1.run         \
+        signal-cfa2.run         \
+        signal-cfa4.run         \
+        signal-java_thread.run
+signal-pthread_cond$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=500000  $(srcdir)/schedint/pthreads.c
+signal-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=5000000 $(srcdir)/schedint/upp.cc
+signal-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedint/cfa1.cfa
+signal-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedint/cfa2.cfa
+signal-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedint/cfa4.cfa
+signal-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/schedint/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+schedint-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/schedint/pthreads.c
+schedext$(EXEEXT) :             \
+        schedext-cfa1.run       \
+        schedext-cfa2.run       \
+        schedext-cfa4.run       \
+        schedext-upp.run        \
+        schedext-goroutine.run
+schedext-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedext/cfa1.cfa
+schedext-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedext/cfa2.cfa
+schedext-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/schedext/cfa4.cfa
+schedext-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/schedext/upp.cc
+schedext-goroutine$(EXEEXT):
+        $(BENCH_V_GOC)go build -o a.out $(srcdir)/schedext/goroutine.go
+creation$(EXEEXT) :                             \
+        creation-cfa_generator.run              \
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+waitfor$(EXEEXT) :\
+        waitfor-upp.run         \
+        waitfor-cfa1.run                \
+        waitfor-cfa2.run                \
+        waitfor-cfa4.run
+waitfor-upp$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=5000000 $(srcdir)/schedext/upp.cc
+waitfor-cfa1$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedext/cfa1.cfa
+waitfor-cfa2$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedext/cfa2.cfa
+waitfor-cfa4$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=500000  $(srcdir)/schedext/cfa4.cfa
+creation$(EXEEXT) :\
+        creation-pthread.run                    \
         creation-cfa_coroutine.run              \
         creation-cfa_coroutine_eager.run        \
 …
         creation-upp_coroutine.run              \
         creation-upp_thread.run                 \
+        creation-python_coroutine.run           \
+        creation-nodejs_coroutine.run           \
+        creation-goroutine_thread.run           \
+        creation-rust_thread.run                \
+        creation-java_thread.run                \
+        creation-pthread.run
+creation-cfa_generator$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_gen.cfa
+        creation-goroutine.run                  \
+        creation-java_thread.run
 creation-cfa_coroutine$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_cor.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=10000000 $(srcdir)/creation/cfa_cor.cfa
 creation-cfa_coroutine_eager$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_cor.cfa  -DEAGER
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=10000000 $(srcdir)/creation/cfa_cor.cfa  -DEAGER
 creation-cfa_thread$(EXEEXT):
         $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/creation/cfa_thrd.cfa
+        $(BENCH_V_CFA)$(CFACOMPILE) -DBENCH_N=10000000 $(srcdir)/creation/cfa_thrd.cfa
 creation-upp_coroutine$(EXEEXT):
         $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/creation/upp_cor.cc
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/creation/upp_cor.cc
 creation-upp_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) $(srcdir)/creation/upp_thrd.cc
+creation-python_coroutine$(EXEEXT):
+        $(BENCH_V_PY)echo "#!/bin/sh" > a.out
+        echo "python3.7 $(srcdir)/creation/python_cor.py" >> a.out
+        chmod a+x a.out
+creation-nodejs_coroutine$(EXEEXT):
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
+        echo "nodejs $(srcdir)/creation/node_cor.js" >> a.out
+        chmod a+x a.out
+creation-goroutine_thread$(EXEEXT):
+        $(BENCH_V_UPP)$(UPPCOMPILE) -DBENCH_N=50000000 $(srcdir)/creation/upp_thrd.cc
+creation-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE)    -DBENCH_N=250000   $(srcdir)/creation/pthreads.c
+creation-goroutine$(EXEEXT):
         $(BENCH_V_GOC)go build -o a.out $(srcdir)/creation/goroutine.go
-creation-rust_thread$(EXEEXT):
-        $(BENCH_V_RUSTC)rustc -C opt-level=3 -o a.out $(srcdir)/creation/rust_thrd.rs
 creation-java_thread$(EXEEXT):
         $(BENCH_V_JAVAC)javac -d $(builddir) $(srcdir)/creation/JavaThread.java
+        echo "#!/bin/sh" > a.out
+        echo "java JavaThread" >> a.out
+        chmod a+x a.out
+creation-pthread$(EXEEXT):
+        $(BENCH_V_CC)$(COMPILE) $(srcdir)/creation/pthreads.c
+compile$(EXEEXT) :              \
+        @echo "#!/bin/sh" > a.out
+        @echo "java JavaThread" >> a.out
+        @chmod a+x a.out
+compile$(EXEEXT) :\
         compile-array.make      \
         compile-attributes.make \
 …
 compile-array$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/array.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/array.cfa
 compile-attributes$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/attributes.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/attributes.cfa
 compile-empty$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(srcdir)/compile/empty.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(srcdir)/compile/empty.cfa
 compile-expression$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/expression.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/expression.cfa
 compile-io$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/io1.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/io1.cfa
 compile-monitor$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/monitor.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/monitor.cfa
 compile-operators$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/operators.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/operators.cfa
 compile-thread$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/thread.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/concurrent/thread.cfa
 compile-typeof$(EXEEXT):
+        $(CFACOMPILE) -fsyntax-only -w $(testdir)/typeof.cfa
+size$(EXEEXT) : size-cfa.runquiet
+size-cfa$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/size/size.cfa
+        @$(CFACOMPILE) -fsyntax-only -w $(testdir)/typeof.cfa
 # Tell versions [3.59,3.63) of GNU make to not export all variables.

benchmark/bench.h

-              r7030dab
+              r71d6bd8
 #endif
         #include <stdlib.h>
+        #include <stdint.h>                             // uint64_t
+        #include <unistd.h>                             // sysconf
+        #include <unistd.h>                                     // sysconf
 #if ! defined(__cforall)
         #include <time.h>
 …
+static inline uint64_t bench_time() {
+        struct timespec ts;
+        clock_gettime( CLOCK_THREAD_CPUTIME_ID, &ts );
+        return 1000000000LL * ts.tv_sec + ts.tv_nsec;
+} // bench_time
+static inline unsigned long long int bench_time() {
+    struct timespec ts;
+    clock_gettime(
+#if defined( __linux__ )
+         CLOCK_THREAD_CPUTIME_ID,
+#elif defined( __freebsd__ )
+         CLOCK_PROF,
+#elif defined( __solaris__ )
+         CLOCK_HIGHRES,
+#else
+    #error uC++ : internal error, unsupported architecture
+#endif
+         &ts );
+    return 1000000000LL * ts.tv_sec + ts.tv_nsec;
+} // Time
 #ifndef BENCH_N
 #define BENCH_N 10000000
+#define BENCH_N 500 //10000000
 #endif
-size_t times = BENCH_N;
-#define BENCH_START()                           \
-        if ( argc > 2 ) exit( EXIT_FAILURE );   \
-        if ( argc == 2 ) {                      \
-                times = atoi( argv[1] );        \
+        }
 #define BENCH(statement, output)                \
+        uint64_t StartTime, EndTime;            \
+        size_t n = BENCH_N;                     \
+        if( argc > 2 ) return 1;                \
+        if( argc == 2 ) {                               \
+                n = atoi(argv[1]);              \
+        }                                               \
+        long long int StartTime, EndTime;       \
         StartTime = bench_time();               \
         statement;                              \
+        statement;                                      \
         EndTime = bench_time();                 \
         double output = (double)( EndTime - StartTime ) / times;
+        double output =         \
+            (double)( EndTime - StartTime ) / n;
 #if defined(__cforall)
 …
+}
 #endif
-#if defined(__U_CPLUSPLUS__)
-unsigned int uDefaultPreemption() {
-        return 0;
+}
-#endif

benchmark/creation/JavaThread.java

-              r7030dab
+              r71d6bd8
         static int x = 2;
         static private int times = Integer.parseInt("10000") ;
+        static private final int NoOfTimes = Integer.parseInt("10000") ;
         public static class MyThread extends Thread {
 …
+        }
         public static void helper() throws InterruptedException {
                 for(int i = 1; i <= times; i += 1) {
+                for(int i = 1; i <= NoOfTimes; i += 1) {
                         MyThread m = new MyThread();
                         x = nextRandom( x );
 …
                 helper();
                 long end = System.nanoTime();
                 System.out.println( (end - start) / times );
+                System.out.println( (end - start) / NoOfTimes );
+        }
         public static void main(String[] args) throws InterruptedException {
+                if ( args.length > 2 ) System.exit( 1 );
+                if ( args.length == 2 ) { times = Integer.parseInt(args[1]); }
+                for (int i = Integer.parseInt("5"); --i >= 0 ; ) {
+                for (int n = Integer.parseInt("5"); --n >= 0 ; ) {
                         InnerMain();
                         Thread.sleep(2000);             // 2 seconds
+                        Thread.sleep(2000);     // 2 seconds
                         x = nextRandom(x);
+                }
 …
+        }
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/creation/cfa_cor.cfa

-              r7030dab
+              r71d6bd8
 coroutine MyCoroutine {};
 void ?{}( MyCoroutine & this ) {
+void ?{} (MyCoroutine & this) {
 #ifdef EAGER
         resume( this );
+        resume(this);
 #endif
+}
 void main( MyCoroutine & ) {}
+void main(MyCoroutine &) {}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for ( times ) {
                         MyCoroutine c;
+                for ( i; n ) {
+                        MyCoroutine m;
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/creation/cfa_thrd.cfa

-              r7030dab
+              r71d6bd8
 void main(MyThread &) {}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for ( times ) {
+                for ( i; n ) {
                         MyThread m;
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/creation/goroutine.go

-              r7030dab
+              r71d6bd8
 import (
+        "fmt"
+        "time"
+        "os"
+        "strconv"
+    "fmt"
+    "time"
+)
 …
 func main() {
+        var times int = 10000000
+        if len( os.Args ) > 2 { os.Exit( 1 ) }
+        if len( os.Args ) == 2 { times, _ = strconv.Atoi(os.Args[1]) }
+        const NoOfTimes = 500000
         start := time.Now()
         for i := 1; i <= times; i += 1 {
+        for i := 1; i <= NoOfTimes; i += 1 {
                 go noop()               // creation
-                <- shake                // wait for completion
+        }
         end := time.Now()
+        fmt.Printf( "%d\n", end.Sub(start) / time.Duration(times) )
+        fmt.Printf("%d\n", end.Sub(start) / time.Duration(NoOfTimes))
+        <- shake
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/creation/pthreads.c

-              r7030dab
+              r71d6bd8
 #include "bench.h"
 static void * foo(void *arg) {
+static void *foo(void *arg) {
     return arg;
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for (size_t i = 0; i < times; i++) {
+                for (size_t i = 0; i < n; i++) {
                         pthread_t thread;
                         if (pthread_create(&thread, NULL, foo, NULL) < 0) {
 …
                                 return 1;
+                        }
                         if (pthread_join( thread, NULL) < 0) {
                                 perror( "failure" );
 …
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/creation/upp_cor.cc

-              r7030dab
+              r71d6bd8
 _Coroutine MyCor {
         void main() {}
-  public:
-        MyCor() { resume(); }
 };
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for (size_t i = 0; i < times; i++) {
+                for (size_t i = 0; i < n; i++) {
                         MyCor m;
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/creation/upp_thrd.cc

-              r7030dab
+              r71d6bd8
 };
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for (size_t i = 0; i < times; i++) {
+                for (size_t i = 0; i < n; i++) {
                         MyThread m;
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/JavaThread.java

-              r7030dab
+              r71d6bd8
         static int x = 2;
         static private int times = Integer.parseInt("100000");
+        static private final int NoOfTimes = Integer.parseInt("1000000") ;
         public static void helper() {
                 for(int i = 1; i <= times; i += 1) {
+                for(int i = 1; i <= NoOfTimes; i += 1) {
                         Thread.yield();
+                }
 …
                 helper();
                 long end = System.nanoTime();
                 System.out.println( (end - start) / times );
+                System.out.println( (end - start) / NoOfTimes );
+        }
         public static void main(String[] args) throws InterruptedException {
+                if ( args.length > 2 ) System.exit( 1 );
+                if ( args.length == 2 ) { times = Integer.parseInt(args[1]); }
+                for (int i = Integer.parseInt("5"); --i >= 0 ; ) {
+                for (int n = Integer.parseInt("5"); --n >= 0 ; ) {
                         InnerMain();
                         Thread.sleep(2000);     // 2 seconds
+                        Thread.sleep(2000);     // 2 seconds
                         x = nextRandom(x);
+                }
 …
+        }
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/cfa_cor.cfa

-              r7030dab
+              r71d6bd8
 #include <thread.hfa>
 #include "../bench.h"
+#include "bench.h"
+coroutine C {};
+void main( __attribute__((unused)) C & ) {
+        for () {
+                suspend;
+coroutine GreatSuspender {};
+void ?{}( GreatSuspender & this ) {
+        prime(this);
+}
+void main( __attribute__((unused)) GreatSuspender & this ) {
+        while( true ) {
+                suspend();
+        }
+}
+int main( int argc, char * argv[] ) {
+        C c;
+        BENCH_START()
+int main(int argc, char* argv[]) {
+        GreatSuspender s;
         BENCH(
                 for ( times ) {
                         resume( c );
+                for ( i; n ) {
+                        resume( s );
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/cfa_cor_then.cfa

-              r7030dab
+              r71d6bd8
 void noOp(void) {}
 coroutine C {} c;
+coroutine GreatSuspender {};
 void ?{}( C & this ) {
+void ?{}( GreatSuspender & this ) {
         prime(this);
+}
 void main( __attribute__((unused)) C & this ) {
         while () {
+void main( __attribute__((unused)) GreatSuspender & this ) {
+        while( true ) {
                 suspend_then(noOp);
+        }
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
+        GreatSuspender s;
         BENCH(
                 for ( times ) {
                         resume( c );
+                for ( i; n ) {
+                        resume( s );
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/cfa_gen.cfa

-              r7030dab
+              r71d6bd8
 #include "../bench.h"
+generator G {};
+void main( G & ) {
+typedef struct {
+        void * next;
+} GreatSuspender;
+void comain( GreatSuspender * this ) {
+    if ( __builtin_expect(this->next != 0, 1) ) goto *(this->next);
+    this->next = &&s1;
         for () {
+                suspend;
+            return;
+          s1: ;
+        }
+}
 int main( int argc, char * argv[] ) {
         G g;
+        BENCH_START()
+int main(int argc, char* argv[]) {
+    GreatSuspender s = { 0 };
         BENCH(
                 for ( times ) {
                         resume( g );
+                for ( i; n ) {
+                        comain( &s );
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/cfa_thrd.cfa

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for ( times ) {
+                for ( i; n ) {
                         yield();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/cfa_thrd2.cfa

-              r7030dab
+              r71d6bd8
 void main(__attribute__((unused)) Fibre & this) {
         while ( ! done ) {
+        while(!done) {
                 yield();
+        }
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         Fibre f1;
         BENCH(
                 for ( times ) {
+                for ( i; n ) {
                         yield();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
         done = true;
+        return 0;
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/goroutine.go

-              r7030dab
+              r71d6bd8
 import (
+        "fmt"
+        "time"
+        "os"
+        "strconv"
+        "runtime"
+    "fmt"
+    "runtime"
+    "time"
+)
 …
 func main() {
+        var times int = 10000000
+        if len( os.Args ) > 2 { os.Exit( 1 ) }
+        if len( os.Args ) == 2 { times, _ = strconv.Atoi(os.Args[1]) }
+        go ContextSwitch( times )               // context switch
+        const NoOfTimes = 10000000
+        go ContextSwitch( NoOfTimes )           // context switch
         <- shake
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/kos_fibre.cpp

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for (size_t i = 0; i < times; i++) {
+                for (size_t i = 0; i < n; i++) {
                         Fibre::yield();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+        return 0;
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/kos_fibre2.cpp

-              r7030dab
+              r71d6bd8
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         Fibre* f1 = (new Fibre)->run(f1main);
         BENCH(
                 for (size_t i = 0; i < times; i++) {
+                for (size_t i = 0; i < n; i++) {
                         Fibre::yield();
                 },
                 result
+        )
         printf( "%g\n", result );
+        printf("%g\n", result);
         done = true;
         Fibre::yield();
         f1->join();
+        return 0;
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/pthreads.c

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for (size_t i = 0; i < times; i++) {
+                for (size_t i = 0; i < n; i++) {
                         sched_yield();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}

benchmark/ctxswitch/upp_cor.cc

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+_Coroutine C {
+_Coroutine GreatSuspender {
+public:
+        GreatSuspender() {
+                resume();
+        }
+        void do_resume() {
+                resume();
+        }
+private:
         void main() {
                 while( true ) {
 …
+                }
+        }
+  public:
+        void do_resume() {
+                resume();
+        }
+} c;
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+};
+int main(int argc, char* argv[]) {
+        GreatSuspender s;
         BENCH(
                 for (size_t i = 0; i < times; i++) {
                         c.do_resume();
+                for (size_t i = 0; i < n; i++) {
+                        s.do_resume();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/ctxswitch/upp_thrd.cc

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for (size_t i = 0; i < times; i++) {
+                for (size_t i = 0; i < n; i++) {
                         uThisTask().yield();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/mutex/JavaThread.java

-              r7030dab
+              r71d6bd8
         static int x = 2;
         static private int times = Integer.parseInt("100000000");
+        static private final int NoOfTimes = Integer.parseInt("100000000") ;
         public synchronized void noop() {
 …
                 // Inhibit biased locking ...
                 x = (j.hashCode() ^ System.identityHashCode(j)) | 1 ;
                 for(int i = 1; i <= times; i += 1) {
+                for(int i = 1; i <= NoOfTimes; i += 1) {
                         x = nextRandom(x);
                         j.noop();
 …
                 helper();
                 long end = System.nanoTime();
                 System.out.println( (end - start) / times );
+                System.out.println( (end - start) / NoOfTimes );
+        }
         public static void main(String[] args) throws InterruptedException {
-                if ( args.length > 2 ) System.exit( 1 );
-                if ( args.length == 2 ) { times = Integer.parseInt(args[1]); }
                 for (int n = Integer.parseInt("5"); --n >= 0 ; ) {
                         InnerMain();
 …
+        }
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/mutex/cfa1.cfa

-              r7030dab
+              r71d6bd8
 #include "bench.h"
 monitor M {} m1;
 void __attribute__((noinline)) call( M & mutex p1 ) {}
+monitor M {};
+void __attribute__((noinline)) call( M & mutex m ) {}
 int main( int argc, char * argv[] ) {
         BENCH_START()
+int main(int argc, char* argv[]) {
+        M m;
         BENCH(
                 for ( times ) {
                         call( m1 );
+                for ( i; n ) {
+                        call(m);
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/mutex/cfa2.cfa

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+monitor M {} m1, m2;
+monitor M {};
+void __attribute__((noinline)) call( M & mutex m1, M & mutex m2 ) {}
+void __attribute__((noinline)) call( M & mutex p1, M & mutex p2 ) {}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
+        M m1, m2;
         BENCH(
                 for ( times ) {
                         call( m1, m2 );
+                for ( i; n ) {
+                        call(m1, m2);
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/mutex/cfa4.cfa

-              r7030dab
+              r71d6bd8
 monitor M {} m1, m2, m3, m4;
 void __attribute__((noinline)) call( M & mutex p1, M & mutex p2, M & mutex p3, M & mutex p4 ) {}
+monitor M {};
+void __attribute__((noinline)) call( M & mutex m1, M & mutex m2, M & mutex m3, M & mutex m4 ) {}
 int main( int argc, char * argv[] ) {
         BENCH_START()
+int main(int argc, char* argv[]) {
+        M m1, m2, m3, m4;
         BENCH(
                 for ( times ) {
                         call( m1, m2, m3, m4 );
+                for ( i; n ) {
+                        call(m1, m2, m3, m4);
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/mutex/pthreads.c

-              r7030dab
+              r71d6bd8
 void __attribute__((noinline)) call() {
          pthread_mutex_lock( &mutex );
          pthread_mutex_unlock( &mutex );
+         pthread_mutex_lock  (&mutex);
+         pthread_mutex_unlock(&mutex);
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         BENCH(
                 for ( size_t i = 0; i < times; i++ ) {
+                for (size_t i = 0; i < n; i++) {
                         call();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/mutex/upp.cc

-              r7030dab
+              r71d6bd8
 };
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int argc, char* argv[]) {
         MyMonitor m;
         BENCH(
                 for ( size_t i = 0; i < times; i++ ) {
+                for (size_t i = 0; i < n; i++) {
                         m.call();
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedext/cfa1.cfa

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
 #include "../bench.h"
+#include "bench.h"
+monitor M {} m1;
+int argc;
+char** argv;
+volatile int go = 0;
+void __attribute__((noinline)) call( M & mutex p1 ) {}
+void __attribute__((noinline)) wait( M & mutex p1 ) {
+        for ( times ) {
+                waitfor( call : p1 );
+        }
+monitor M {};
+M m1;
+void __attribute__((noinline)) call( M & mutex a1 ) {}
+int  __attribute__((noinline)) wait( M & mutex a1 ) {
+        go = 1;
+        BENCH(
+                for ( i; n ) {
+                        waitfor(call, a1);
+                },
+                result
+        )
+        printf("%g\n", result);
+        go = 0;
+        return 0;
+}
 thread T {};
+void ^?{}( T & mutex this ) {}
 void main( T & ) {
+        BENCH(
+                for ( times ) { call( m1 ); },
+                result
+        )
+        printf( "%g\n", result );
+        while(go == 0) { yield(); }
+        while(go == 1) { call(m1); }
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
         T t;
         wait( m1 );
+        return wait(m1);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedext/cfa2.cfa

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
 #include "../bench.h"
+#include "bench.h"
+monitor M {} m1, m2;
+int argc;
+char** argv;
+volatile int go = 0;
+void __attribute__((noinline)) call( M & mutex p1, M & mutex p2 ) {}
+void __attribute__((noinline)) wait( M & mutex p1, M & mutex p2 ) {
+        for ( times ) {
+                waitfor( call : p1, p2 );
+        }
+}
+thread T {};
+void main( T & ) {
+monitor M {};
+M m1, m2;
+void __attribute__((noinline)) call( M & mutex a1, M & mutex a2 ) {}
+int  __attribute__((noinline)) wait( M & mutex a1, M & mutex a2 ) {
+        go = 1;
         BENCH(
                 for ( times ) {
                         call( m1, m2 );
+                for ( i; n ) {
+                        waitfor(call, a1, a2);
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+        go = 0;
+        return 0;
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+        T t;
+        wait( m1, m2 );
+thread T {};
+void ^?{}( T & mutex this ) {}
+void main( T & ) {
+        while(go == 0) { yield(); }
+        while(go == 1) { call(m1, m2); }
+}
+// Local Variables: //
+// tab-width: 4 //
+// End: //
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
+        T t;
+        return wait(m1, m2);
+}

benchmark/schedext/cfa4.cfa

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
 #include "../bench.h"
+#include "bench.h"
+monitor M {} m1, m2, m3, m4;
+int argc;
+char** argv;
+volatile int go = 0;
+void __attribute__((noinline)) call( M & mutex p1, M & mutex p2, M & mutex p3, M & mutex p4 ) {}
+void __attribute__((noinline)) wait( M & mutex p1, M & mutex p2, M & mutex p3, M & mutex p4 ) {
+        for ( times ) {
+                waitfor( call : p1, p2, p3, p4 );
+        }
+}
+thread T {};
+void main( T & ) {
+monitor M {};
+M m1, m2, m3, m4;
+void __attribute__((noinline)) call( M & mutex a1, M & mutex a2, M & mutex a3, M & mutex a4 ) {}
+int  __attribute__((noinline)) wait( M & mutex a1, M & mutex a2, M & mutex a3, M & mutex a4 ) {
+        go = 1;
         BENCH(
                 for ( times ) {
                         call( m1, m2, m3, m4 );
+                for ( i; n ) {
+                        waitfor(call, a1, a2, a3, a4);
                 },
                 result
+        )
+        printf( "%g\n", result );
+        printf("%g\n", result);
+        go = 0;
+        return 0;
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+        T t;
+        wait( m1, m2, m3, m4 );
+thread T {};
+void ^?{}( T & mutex this ) {}
+void main( T & ) {
+        while(go == 0) { yield(); }
+        while(go == 1) { call(m1, m2, m3, m4); }
+}
+// Local Variables: //
+// tab-width: 4 //
+// End: //
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
+        T t;
+        return wait(m1, m2, m3, m4);
+}

benchmark/schedext/upp.cc

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+int argc;
+char** argv;
+volatile int go = 0;
 _Monitor M {
 public:
         void __attribute__((noinline)) call() {}
+        void __attribute__((noinline)) wait() {
+                for ( size_t i = 0; i < times; i++ ) {
+                        _Accept(call);
+                }
+        int __attribute__((noinline)) wait() {
+                go = 1;
+                BENCH(
+                        for (size_t i = 0; i < n; i++) {
+                                _Accept(call);
+                        },
+                        result
+                )
+                printf("%g\n", result);
+                go = 0;
+                return 0;
+        }
+} m;
+};
+M m;
 _Task T {
         void main() {
+                BENCH(
+                        for ( size_t i = 0; i < times; i++ ) {
+                                m.call();
+                        },
+                        result
+                )
+                printf( "%g\n", result );
+                while(go == 0) { yield(); }
+                while(go == 1) { m.call(); }
+        }
 };
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(int margc, char* margv[]) {
+        argc = margc;
+        argv = margv;
         T t;
         m.wait();
+        return m.wait();
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedint/JavaThread.java

-              r7030dab
+              r71d6bd8
         static int x = 2;
         static private int times = Integer.parseInt("1000000");
+        static private final int NoOfTimes = Integer.parseInt("1000000") ;
         public static void helper( Monitor m ) throws InterruptedException {
                 for(int i = 1; i <= times; i += 1) {
+                for(int i = 1; i <= NoOfTimes; i += 1) {
                         m.wait();               // relase monitor lock
                         m.next = true;
 …
                 synchronized(m) {
                         s.start();
                         while( ! Monitor.go ) { // waiter must start first
+                        while( !Monitor.go ) {
                                 Thread.yield();
+                        }
 …
                 Monitor.go = false;
                 s.join();
                 System.out.println( (end - start) / times);
+                System.out.println( (end - start) / NoOfTimes);
+        }
         public static void main(String[] args) throws InterruptedException {
-                if ( args.length > 2 ) System.exit( 1 );
-                if ( args.length == 2 ) { times = Integer.parseInt(args[1]); }
                 for (int n = Integer.parseInt("5"); --n >= 0 ; ) {
                         InnerMain();
 …
+        }
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedint/cfa1.cfa

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
 #include "../bench.h"
+#include "bench.h"
+int argc;
+char** argv;
 volatile int go = 0;
 condition c;
+monitor M {} m1;
+monitor M {};
+M m1;
 void __attribute__((noinline)) call( M & mutex p1 ) {
         signal( c );
+void __attribute__((noinline)) call( M & mutex a1 ) {
+        signal(c);
+}
+void __attribute__((noinline)) wait( M & mutex p1 ) {
+int  __attribute__((noinline)) wait( M & mutex a1 ) {
         go = 1;
+        for ( times ) {
+                wait( c );
+        }
+        BENCH(
+                for ( i; n ) {
+                        wait(c);
+                },
+                result
+        )
+        printf("%g\n", result);
+        go = 0;
+        return 0;
+}
 thread T {};
+void ^?{}( T & mutex ) {}
 void main( T & ) {
+        while ( go == 0 ) { yield(); } // waiter must start first
+        BENCH(
+                for ( times ) { call( m1 ); },
+                result
+        )
+        printf( "%g\n", result );
+        while(go == 0) { yield(); }
+        while(go == 1) { call(m1); }
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
         T t;
         wait( m1 );
+        return wait(m1);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedint/cfa2.cfa

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
 #include "../bench.h"
+#include "bench.h"
+int argc;
+char** argv;
 volatile int go = 0;
 condition c;
+monitor M {} m1, m2;
+monitor M {};
+M m1, m2;
 void __attribute__((noinline)) call( M & mutex p1, M & mutex p2 ) {
         signal( c );
+void __attribute__((noinline)) call( M & mutex a1, M & mutex a2 ) {
+        signal(c);
+}
+void __attribute__((noinline)) wait( M & mutex p1, M & mutex p2 ) {
+int  __attribute__((noinline)) wait( M & mutex a1, M & mutex a2 ) {
         go = 1;
+        for ( times ) {
+                wait( c );
+        }
+        BENCH(
+                for ( i; n ) {
+                        wait(c);
+                },
+                result
+        )
+        printf("%g\n", result);
+        go = 0;
+        return 0;
+}
 thread T {};
+void ^?{}( T & mutex this ) {}
 void main( T & ) {
+        while ( go == 0 ) { yield(); } // waiter must start first
+        BENCH(
+                for ( times ) { call( m1, m2 ); },
+                result
+        )
+        printf( "%g\n", result );
+        while(go == 0) { yield(); }
+        while(go == 1) { call(m1, m2); }
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
         T t;
         wait( m1, m2 );
+        return wait(m1, m2);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedint/cfa4.cfa

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
 #include "../bench.h"
+#include "bench.h"
+int argc;
+char** argv;
 volatile int go = 0;
 condition c;
+monitor M {} m1, m2, m3, m4;
+monitor M {};
+M m1, m2, m3, m4;
 void __attribute__((noinline)) call( M & mutex p1, M & mutex p2, M & mutex p3, M & mutex p4 ) {
         signal( c );
+void __attribute__((noinline)) call( M & mutex a1, M & mutex a2, M & mutex a3, M & mutex a4 ) {
+        signal(c);
+}
+void __attribute__((noinline)) wait( M & mutex p1, M & mutex p2, M & mutex p3, M & mutex p4 ) {
+int  __attribute__((noinline)) wait( M & mutex a1, M & mutex a2, M & mutex a3, M & mutex a4 ) {
         go = 1;
+        for ( times ) {
+                wait( c );
+        }
+        BENCH(
+                for ( i; n ) {
+                        wait(c);
+                },
+                result
+        )
+        printf("%g\n", result);
+        go = 0;
+        return 0;
+}
 thread T {};
+void ^?{}( T & mutex this ) {}
 void main( T & ) {
+        while ( go == 0 ) { yield(); } // waiter must start first
+        BENCH(
+                for ( times ) { call( m1, m2, m3, m4 ); },
+                result
+        )
+        printf( "%g\n", result );
+        while(go == 0) { yield(); }
+        while(go == 1) { call(m1, m2, m3, m4); }
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
         T t;
         wait( m1, m2, m3, m4 );
+        return wait(m1, m2, m3, m4);
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedint/pthreads.c

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+int argc;
+char** argv;
 volatile int go = 0;
+pthread_cond_t c;
 pthread_mutex_t m;
-pthread_cond_t c;
 void __attribute__((noinline)) call() {
         pthread_mutex_lock( &m );
         pthread_cond_signal( &c );
         pthread_mutex_unlock( &m );
+        pthread_mutex_lock(&m);
+        pthread_cond_signal(&c);
+        pthread_mutex_unlock(&m);
+}
 void __attribute__((noinline)) wait() {
+int  __attribute__((noinline)) wait() {
         pthread_mutex_lock(&m);
         go = 1;
+        for ( size_t i = 0; i < times; i++ ) {
+                pthread_cond_wait( &c, &m );
+        }
+        BENCH(
+                for (size_t i = 0; i < n; i++) {
+                        pthread_cond_wait(&c, &m);
+                },
+                result
+        )
+        printf("%g\n", result);
         go = 0;
+        pthread_mutex_unlock( &m );
+        pthread_mutex_unlock(&m);
+        return 0;
+}
+void * thread_main( __attribute__((unused)) void * arg ) {
+        while ( go == 0 ) { sched_yield(); } // waiter must start first
+        // barging for lock acquire => may not execute N times
+        BENCH(
+                while ( go == 1 ) { call(); },
+                result
+        )
+        printf( "%g\n", result );
+void* thread_main(__attribute__((unused)) void * arg ) {
+        while(go == 0) { sched_yield(); }
+        while(go == 1) { call(); }
         return NULL;
+}
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
         pthread_t thread;
         if ( pthread_create( &thread, NULL, thread_main, NULL ) < 0 ) {
+        if (pthread_create(&thread, NULL, thread_main, NULL) < 0) {
                 perror( "failure" );
                 return 1;
+        }
         wait();
         if ( pthread_join( thread, NULL ) < 0 ) {
+        if (pthread_join( thread, NULL) < 0) {
                 perror( "failure" );
                 return 1;
+        }
+        return 0;
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

benchmark/schedint/upp.cc

-              r7030dab
+              r71d6bd8
 #include "bench.h"
+int argc;
+char** argv;
 volatile int go = 0;
 …
                 cond.signal();
+        }
+        void __attribute__((noinline)) wait() {
+        int __attribute__((noinline)) wait() {
                 go = 1;
+                for ( size_t i = 0; i < times; i++ ) {
+                        cond.wait();
+                }
+                BENCH(
+                        for (size_t i = 0; i < n; i++) {
+                                cond.wait();
+                        },
+                        result
+                )
+                printf("%g\n", result);
+                go = 0;
+                return 0;
+        }
+} m;
+};
+M m;
 _Task T {
         void main() {
+                while ( go == 0 ) { yield(); } // waiter must start first
+                BENCH(
+                        for ( size_t i = 0; i < times; i++ ) {
+                                m.call();
+                        },
+                        result
+                )
+                printf( "%g\n", result );
+                while(go == 0) { yield(); }
+                while(go == 1) { m.call(); }
+        }
 };
+int main( int argc, char * argv[] ) {
+        BENCH_START()
+int main(__attribute__((unused)) int argc, __attribute__((unused)) char* argv[]) {
         T t;
         m.wait();
+        return m.wait();
+}
-// Local Variables: //
-// tab-width: 4 //
-// End: //

configure

r7030dab	r71d6bd8
2557	2557	# don't use the default CFLAGS as they unconditonnaly add -O2
2558	2558	: ${CFLAGS=""}
2559		~~: ${CXXFLAGS=""}~~
2560	2559
2561	2560	am__api_version='1.15'

configure.ac

r7030dab	r71d6bd8
14	14	# don't use the default CFLAGS as they unconditonnaly add -O2
15	15	: ${CFLAGS=""}
16		~~: ${CXXFLAGS=""}~~
17	16
18	17	AM_INIT_AUTOMAKE([subdir-objects])

doc/bibliography/pl.bib

-              r7030dab
+              r71d6bd8
 %    Predefined journal names:
 %  acmcs: Computing Surveys             acta: Acta Infomatica
+@string{acta="Acta Infomatica"}
 %  cacm: Communications of the ACM
 %  ibmjrd: IBM J. Research & Development ibmsj: IBM Systems Journal
 …
 %  tcs: Theoretical Computer Science
-@string{acta="Acta Infomatica"}
 string{ieeepds="IEEE Transactions on Parallel and Distributed Systems"}
 @string{ieeepds="IEEE Trans. Parallel Distrib. Syst."}
 …
     series      = {ACM Distinguished Dissertations},
     year        = 1983,
+}
-@article{Zhang19,
-    keywords    = {Algebraic effects, dynamic scoping, exceptions, parametricity, type systems},
-    author      = {Zhang, Yizhou and Myers, Andrew C.},
-    title       = {Abstraction-safe Effect Handlers via Tunneling},
-    journal     = {Proc. ACM Program. Lang.},
-    issue_date  = {January 2019},
-    volume      = {3},
-    number      = {POPL},
-    month       = jan,
-    year        = {2019},
-    issn        = {2475-1421},
-    pages       = {5:1--5:29},
-    articleno   = {5},
-    publisher   = {ACM},
-    address     = {New York, NY, USA},
+}
-@inproceedings{Zhang16,
-    keywords    = {Exception tunneling, Genus, exception handling},
-    author      = {Zhang, Yizhou and Salvaneschi, Guido and Beightol, Quinn and Liskov, Barbara and Myers, Andrew C.},
-    title       = {Accepting Blame for Safe Tunneled Exceptions},
-    booktitle   = {Proceedings of the 37th ACM SIGPLAN Conference on Programming Language Design and Implementation},
-    series      = {PLDI'16},
-    year        = {2016},
-    location    = {Santa Barbara, CA, USA},
-    pages       = {281--295},
-    publisher   = {ACM},
-    address     = {New York, NY, USA},
+}
 …
     journal     = sigplan,
     year        = 1981,
+    month       = feb,
+    volume      = 16,
+    number      = 2,
+    pages       = {48-52},
+    month       = feb, volume = 16, number = 2, pages = {48-52},
     comment     = {
         A one-pass, top-down algorithm for overload resolution.  Input is a
 …
     title       = {An Alternative to Subclassing},
     journal     = sigplan,
+    volume      = {21},
+    number      = {11},
+    volume      = {21},    number = {11},
     pages       = {424-428},
+    month       = nov,
+    year        = 1986,
+    month       = nov, year = 1986,
     comment     = {
         The Smalltalk class hierarchy has three uses: factoring out code;
 …
     isbn        = {3-540-66538-2},
     location    = {Toulouse, France},
+    doi         = {http://doi.acm.org/10.1145/318773.319251},
     publisher   = {Springer},
     address     = {London, UK},
 …
     year        = 2010,
     pages       = {39--50},
+    numpages    = {12},
     publisher   = {IEEE Computer Society},
     address     = {Washington, DC, USA},
 …
+}
-@manual{C99,
-    keywords    = {ISO/IEC C 9899},
-    contributer = {pabuhr@plg},
-    key         = {C99},
-    title       = {C Programming Language {ISO/IEC} 9899:1999(E)},
-    edition     = {2nd},
-    publisher   = {International Standard Organization},
-    address     = {\href{https://webstore.ansi.org/Standards/INCITS/INCITSISOIEC98991999R2005}{https://webstore.ansi.org/\-Standards/\-INCITS/\-INCITSISOIEC98991999R2005}},
-    year        = 1999,
+}
 @manual{C11,
     keywords    = {ISO/IEC C 11},
 …
     location    = {London, United Kingdom},
     pages       = {41--53},
+    numpages    = {13},
+    url         = {http://doi.acm.org/10.1145/360204.360207},
+    doi         = {10.1145/360204.360207},
+    acmid       = {360207},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     year        = 1993,
     pages       = {201--208},
+    url         = {http://doi.acm.org/10.1145/155360.155580},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     location    = {Boulder, Colorado, USA},
     pages       = {91--97},
+    numpages    = {7},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     issn        = {0004-5411},
     pages       = {215--225},
+    numpages    = {11},
+    url         = {http://doi.acm.org/10.1145/321879.321884},
+    doi         = {10.1145/321879.321884},
+    acmid       = {321884},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
+}
-@misc{Drepper13,
-    keywords    = {thread-local storage},
-    contributer = {pabuhr@plg},
-    author      = {Ulrich Drepper},
-    title       = {{ELF} Handling For Thread-Local Storage},
-    year        = 2013,
-    month       = aug,
-    note        = {WikipediA},
-    howpublished= {\href{http://www.akkadia.org/drepper/tls.pdf}
-                  {http://\-www.akkadia.org/\-drepper/\-tls.pdf}},
+}
 @misc{Turley99,
     keywords    = {embedded system, micrprocessor},
 …
     howpublished= {\href{https://www.eetimes.com/author.asp?sectionid=36&doc_id=1287712}
                   {https://\-www.eetimes.com/\-author.asp?sectionid=\-36&doc_id=1287712}},
+}
-@article{Xiao19,
-    keywords    = {bug classification, fault trigger, Linux operating system, regression bug},
-    contributer = {pabuhr@plg},
-    author      = {Guanping Xiao and Zheng Zheng and Beibei Yin and Kishor S. Trivedi and Xiaoting Du and Kai-Yuan Cai},
-    title       = {An Empirical Study of Fault Triggers in the Linux Operating System: An Evolutionary Perspective},
-    journal     = {IEEE Transactions on Reliability},
-    month       = dec,
-    year        = 2019,
-    volume      = 68,
-    number      = 4,
-    pages       = {1356-1383},
+}
 …
+}
-@inproceedings{Palix11,
-    keywords    = {Linux, fault-finding tools},
-    contributer = {pabuhr@plg},
-    author      = {Nicolas Palix and Ga\"el Thomas and Suman Saha and Christophe Calv\`es and Julia Lawall and Gilles Muller},
-    title       = {Faults in Linux: Ten Years Later},
-    booktitle   = {Proc. of the 16 International Conf. on Arch. Support for Prog. Lang. and Oper. Sys.},
-    series      = {ASPLOS'11},
-    month       = mar,
-    year        = 2011,
-    location    = {Newport Beach, California, USA},
-    pages       = {305-318},
-    publisher   = {ACM},
-    address     = {New York, NY, USA},
+}
 @article{Lamport87,
     keywords    = {software solutions, mutual exclusion, fast},
 …
     issn        = {0001-0782},
     pages       = {107--115},
+    numpages    = {9},
+    url         = {http://doi.acm.org/10.1145/1538788.1538814},
+    doi         = {10.1145/1538788.1538814},
+    acmid       = {1538814},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
+}
-@mastersthesis{Radhakrishnan19,
-    author      = {Srihari Radhakrishnan},
-    title       = {High Performance Web Servers: A Study In Concurrent Programming Models},
-    school      = {School of Computer Sc., University of Waterloo},
-    year        = 2019,
-    optaddress  = {Waterloo, Ontario, Canada, N2L 3G1},
-    note        = {\href{https://uwspace.uwaterloo.ca/handle/10012/14706}{https://\-uwspace.uwaterloo.ca/\-handle/\-10012/\-14706}},
+}
 @article{katzenelson83b,
     contributer = {gjditchfield@plg},
 …
     pages       = {115-138},
     year        = 1971,
+}
-@inproceedings{Hagersten03,
-    keywords    = {cache storage, parallel architectures, performance evaluation, shared memory systems},
-    author      = {Zoran Radovi\'{c} and Erik Hagersten},
-    title       = {Hierarchical backoff locks for nonuniform communication architectures},
-    booktitle   = {Proceedings of the Ninth International Symposium on High-Performance Computer Architecture},
-    year        = {2003},
-    location    = {Anaheim, CA, USA},
-    pages       = {241-252},
-    publisher   = {IEEE},
+}
 …
+}
-@misc{gccValueLabels,
-    keywords    = {gcc extension, value labels},
-    contributer = {pabuhr@plg},
-    key         = {Labels as Values},
-    author      = {{gcc Extension}},
-    title       = {Labels as Values},
-    year        = {since gcc-3},
-    howpublished= {\href{https://gcc.gnu.org/onlinedocs/gcc/Labels-as-Values.html}
-                  {https:\-//gcc.gnu.org/\-onlinedocs/\-gcc/\-Labels-as-Values.html}},
+}
 @mastersthesis{Clarke90,
     keywords    = {concurrency, postponing requests},
 …
 @article{Pierce00,
     keywords    = {Scala, polymorphism, subtyping, type inference},
+    keywords    = {Scala},
     contributer = {a3moss@uwaterloo.ca},
     author      = {Pierce, Benjamin C. and Turner, David N.},
 …
     issn        = {0164-0925},
     pages       = {1--44},
+    numpages    = {44},
+    url         = {http://doi.acm.org/10.1145/345099.345100},
+    doi         = {10.1145/345099.345100},
+    acmid       = {345100},
     publisher   = {ACM},
     address     = {New York, NY, USA},
+    keywords    = {polymorphism, subtyping, type inference},
+}
-@article{Dice15,
-    keywords    = {Concurrency, NUMA, hierarchical locks, locks, multicore, mutex, mutual exclusion, spin locks},
-    author      = {Dice, David and Marathe, Virendra J. and Shavit, Nir},
-    title       = {Lock Cohorting: A General Technique for Designing NUMA Locks},
-    journal     = {ACM Trans. Parallel Comput.},
-    issue_date  = {January 2015},
-    volume      = 1,
-    number      = 2,
-    month       = feb,
-    year        = 2015,
-    pages       = {13:1--13:42},
-    publisher   = {ACM},
-    address     = {New York, NY, USA},
+}
 @article{Sundell08,
 …
     journal     = sigplan,
     year        = 1989,
+    month       = jun,
+    volume      = 24,
+    number      = 6,
+    pages       = {37-48},
+    month       = jun, volume = 24, number = 6, pages = {37-48},
     abstract    = {
         This paper describes a scheme we have used to manage a large
 …
     year        = 1986,
     pages       = {313--326},
+    numpages    = {14},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     year        = 1986,
     pages       = {327--348},
+    numpages    = {22},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     year        = 2005,
     pages       = {146-196},
+    numpages    = {51},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     year        = 2000,
     pages       = {29-46},
     note        = {OOPSLA'00, Oct. 15--19, 2000, Minneapolis, Minn., U.S.A.},
+    note        = {OOPSLA'00, Oct. 15--19, 2000, Minneapolis, Minnesota, U.S.A.},
+}
 …
     location    = {San Diego, California, USA},
     pages       = {101--112},
+    numpages    = {12},
+    url         = {http://doi.acm.org/10.1145/2535838.2535878},
+    doi         = {10.1145/2535838.2535878},
+    acmid       = {2535878},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     issn        = {0362-1340},
     pages       = {30--42},
+    numpages    = {13},
+    url         = {http://doi.acm.org/10.1145/947586.947589},
+    doi         = {10.1145/947586.947589},
     publisher   = {ACM},
     address     = {New York, NY, USA}
 …
     month       = 9,
     year        = 2005,
+}
-@article{Bauer15,
-    keywords    = {resumption exceptions, theory},
-    contributer = {pabuhr@plg},
-    author      = {Andrej Bauer and Matija Pretnar},
-    title       = {Programming with Algebraic Effects and Handlers},
-    journal     = {Journal of Logical and Algebraic Methods in Programming},
-    publisher   = {Elsevier BV},
-    volume      = 84,
-    number      = 1,
-    month       = jan,
-    year        = 2015,
-    pages       = {108-123},
+}
 …
     issn        = {0164-0925},
     pages       = {429-475},
+    url         = {http://doi.acm.org/10.1145/1133651.1133653},
+    doi         = {10.1145/1133651.1133653},
+    acmid       = {1133653},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     issn        = {0001-0782},
     pages       = {565--569},
+    numpages    = {5},
+    url         = {http://doi.acm.org/10.1145/359545.359566},
+    doi         = {10.1145/359545.359566},
+    acmid       = {359566},
     publisher   = {ACM},
     address     = {New York, NY, USA}
 …
     issn        = {0362-1340},
     pages       = {145--147},
+    numpages    = {3},
+    url         = {http://doi.acm.org/10.1145/122598.122614},
+    doi         = {10.1145/122598.122614},
+    acmid       = {122614},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     issn        = {0362-1340},
     pages       = {82--87},
+    numpages    = {6},
+    url         = {http://doi.acm.org/10.1145/947680.947688},
+    doi         = {10.1145/947680.947688},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
+}
-@article{Cascaval08,
-    author      = {Cascaval, Calin and Blundell, Colin and Michael, Maged and Cain, Harold W. and Wu, Peng and Chiras, Stefanie and Chatterjee, Siddhartha},
-    title       = {Software Transactional Memory: Why Is It Only a Research Toy?},
-    journal     = {Queue},
-    volume      = {6},
-    number      = {5},
-    month       = sep,
-    year        = {2008},
-    pages       = {40:46--40:58},
-    publisher   = {ACM},
-    address     = {New York, NY, USA},
+}
 @article{Dijkstra65a,
     keywords    = {N-thread software-solution mutual exclusion},
 …
     year        = 1974,
     pages       = {261-301},
+    issn        = {0360-0300},
+    doi         = {http://doi.acm.org/10.1145/356635.356640},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     publisher   = {ACM Press},
     address     = {New York, NY, USA},
+    doi         = {http://doi.acm.org/10.1145/356586.356588},
+}
 …
     howpublished= {\href{https://projects.eclipse.org/proposals/trace-compass}{https://\-projects.eclipse.org/\-proposals/\-trace-compass}},
+}
+@inproceedings{Boehm09,
+    author      = {Boehm, Hans-J.},
+    title       = {Transactional Memory Should Be an Implementation Technique, Not a Programming Interface},
+    booktitle   = {Proceedings of the First USENIX Conference on Hot Topics in Parallelism},
+    series      = {HotPar'09},
+    year        = {2009},
+    location    = {Berkeley, California},
+    publisher   = {USENIX Association},
+    address     = {Berkeley, CA, USA},
+}
 @article{Leroy00,
     keywords    = {type-systems, exceptions},
 …
     number      = {2},
     pages       = {204-214},
+    month       = apr,
+    year        = 1988,
+    month       = apr, year = 1988,
     comment     = {
         Extended record types add fields to their base record.  Assignment
 …
     issn        = {0004-5411},
     pages       = {245--281},
+    numpages    = {37},
+    url         = {http://doi.acm.org/10.1145/62.2160},
+    doi         = {10.1145/62.2160},
+    acmid       = {2160},
     publisher   = {ACM},
     address     = {New York, NY, USA},
 …
     contributer = {pabuhr@plg},
     author      = {Boehm, Hans-J. and Adve, Sarita V.},
     title       = {You Don't Know Jack About Shared Variables or Memory Models},
+    title       = {You Don'T Know Jack About Shared Variables or Memory Models},
     journal     = cacm,
     volume      = 55,

doc/papers/concurrency/Paper.tex

-              r7030dab
+              r71d6bd8
 \newcommand{\CCseventeen}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}17\xspace} % C++17 symbolic name
 \newcommand{\CCtwenty}{\textrm{C}\kern-.1em\hbox{+\kern-.25em+}20\xspace} % C++20 symbolic name
 \newcommand{\Csharp}{C\raisebox{-0.7ex}{\large$^\sharp$}\xspace} % C# symbolic name
+\newcommand{\Csharp}{C\raisebox{-0.7ex}{\Large$^\sharp$}\xspace} % C# symbolic name
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 …
 \newcommand*{\etc}{%
         \@ifnextchar{.}{\ETC}%
                 {\ETC.\xspace}%
+        {\ETC.\xspace}%
 }}{}%
 \@ifundefined{etal}{
 \newcommand{\ETAL}{\abbrevFont{et}~\abbrevFont{al}}
 \newcommand*{\etal}{%
         \@ifnextchar{.}{\ETAL}%
                 {\ETAL.\xspace}%
+        \@ifnextchar{.}{\protect\ETAL}%
+                {\protect\ETAL.\xspace}%
 }}{}%
 \@ifundefined{viz}{
 …
                 __float80, float80, __float128, float128, forall, ftype, generator, _Generic, _Imaginary, __imag, __imag__,
                 inline, __inline, __inline__, __int128, int128, __label__, monitor, mutex, _Noreturn, one_t, or,
                 otype, restrict, resume, __restrict, __restrict__, __signed, __signed__, _Static_assert, suspend, thread,
+                otype, restrict, __restrict, __restrict__, __signed, __signed__, _Static_assert, thread,
                 _Thread_local, throw, throwResume, timeout, trait, try, ttype, typeof, __typeof, __typeof__,
                 virtual, __volatile, __volatile__, waitfor, when, with, zero_t},
         moredirectives={defined,include_next},
         % replace/adjust listing characters that look bad in sanserif
         literate={-}{\makebox[1ex][c]{\raisebox{0.5ex}{\rule{0.8ex}{0.1ex}}}}1 {^}{\raisebox{0.6ex}{$\scriptstyle\land\,$}}1
+        literate={-}{\makebox[1ex][c]{\raisebox{0.4ex}{\rule{0.8ex}{0.1ex}}}}1 {^}{\raisebox{0.6ex}{$\scriptstyle\land\,$}}1
                 {~}{\raisebox{0.3ex}{$\scriptstyle\sim\,$}}1 % {`}{\ttfamily\upshape\hspace*{-0.1ex}`}1
                 {<}{\textrm{\textless}}1 {>}{\textrm{\textgreater}}1
 …
                 _Else, _Enable, _Event, _Finally, _Monitor, _Mutex, _Nomutex, _PeriodicTask, _RealTimeTask,
                 _Resume, _Select, _SporadicTask, _Task, _Timeout, _When, _With, _Throw},
+}
+\lstdefinelanguage{Golang}{
+        morekeywords=[1]{package,import,func,type,struct,return,defer,panic,recover,select,var,const,iota,},
+        morekeywords=[2]{string,uint,uint8,uint16,uint32,uint64,int,int8,int16,int32,int64,
+                bool,float32,float64,complex64,complex128,byte,rune,uintptr, error,interface},
+        morekeywords=[3]{map,slice,make,new,nil,len,cap,copy,close,true,false,delete,append,real,imag,complex,chan,},
+        morekeywords=[4]{for,break,continue,range,goto,switch,case,fallthrough,if,else,default,},
+        morekeywords=[5]{Println,Printf,Error,},
+        sensitive=true,
+        morecomment=[l]{//},
+        morecomment=[s]{/*}{*/},
+        morestring=[b]',
+        morestring=[b]",
+        morestring=[s]{`}{`},
+}
 …
 {}
 \lstnewenvironment{uC++}[1][]
 {\lstset{language=uC++,moredelim=**[is][\protect\color{red}]{`}{`},#1}\lstset{#1}}
+{\lstset{#1}}
 {}
 \lstnewenvironment{Go}[1][]
 …
+}
 \newsavebox{\myboxA}
 \newsavebox{\myboxB}
 \newsavebox{\myboxC}
 \newsavebox{\myboxD}
+\newbox\myboxA
+\newbox\myboxB
+\newbox\myboxC
+\newbox\myboxD
 \title{\texorpdfstring{Advanced Control-flow and Concurrency in \protect\CFA}{Advanced Control-flow in Cforall}}
 …
 \CFA is a polymorphic, non-object-oriented, concurrent, backwards-compatible extension of the C programming language.
 This paper discusses the design philosophy and implementation of its advanced control-flow and concurrent/parallel features, along with the supporting runtime written in \CFA.
 These features are created from scratch as ISO C has only low-level and/or unimplemented concurrency, so C programmers continue to rely on library approaches like pthreads.
+These features are created from scratch as ISO C has only low-level and/or unimplemented concurrency, so C programmers continue to rely on library features like pthreads.
 \CFA introduces modern language-level control-flow mechanisms, like generators, coroutines, user-level threading, and monitors for mutual exclusion and synchronization.
 % Library extension for executors, futures, and actors are built on these basic mechanisms.
 …
 \begin{document}
 \linenumbers                            % comment out to turn off line numbering
+\linenumbers                                            % comment out to turn off line numbering
 \maketitle
 …
 \section{Introduction}
+\CFA~\cite{Moss18,Cforall} is a modern, polymorphic, non-object-oriented\footnote{
+\CFA has object-oriented features, such as constructors, destructors, virtuals and simple trait/interface inheritance.
+% Go interfaces, Rust traits, Swift Protocols, Haskell Type Classes and Java Interfaces.
+% "Trait inheritance" works for me. "Interface inheritance" might also be a good choice, and distinguish clearly from implementation inheritance.
+% You'll want to be a little bit careful with terms like "structural" and "nominal" inheritance as well. CFA has structural inheritance (I think Go as well) -- it's inferred based on the structure of the code. Java, Rust, and Haskell (not sure about Swift) have nominal inheritance, where there needs to be a specific statement that "this type inherits from this type".
+This paper discusses the design philosophy and implementation of advanced language-level control-flow and concurrent/parallel features in \CFA~\cite{Moss18,Cforall} and its runtime, which is written entirely in \CFA.
+\CFA is a modern, polymorphic, non-object-oriented\footnote{
+\CFA has features often associated with object-oriented programming languages, such as constructors, destructors, virtuals and simple inheritance.
 However, functions \emph{cannot} be nested in structures, so there is no lexical binding between a structure and set of functions (member/method) implemented by an implicit \lstinline@this@ (receiver) parameter.},
 backwards-compatible extension of the C programming language.
+In many ways, \CFA is to C as Scala~\cite{Scala} is to Java, providing a \emph{research vehicle} for new typing and control-flow capabilities on top of a highly popular programming language\footnote{
+The TIOBE index~\cite{TIOBE} for December 2019 ranks the top five \emph{popular} programming languages as Java 17\%, C 16\%, Python 10\%, and \CC 6\%, \Csharp 5\% = 54\%, and over the past 30 years, C has always ranked either first or second in popularity.}
+allowing immediate dissemination.
+This paper discusses the design philosophy and implementation of advanced language-level control-flow and concurrent/parallel features in \CFA and its runtime, which is written entirely in \CFA.
+The \CFA control-flow framework extends ISO \Celeven~\cite{C11} with new call/return and concurrent/parallel control-flow.
+% The call/return extensions retain state between callee and caller versus losing the callee's state on return;
+% the concurrency extensions allow high-level management of threads.
+Call/return control-flow with argument/parameter passing appeared in the first programming languages.
+Over the past 50 years, call/return has been augmented with features like static/dynamic call, exceptions (multi-level return) and generators/coroutines (retain state between calls).
+While \CFA has mechanisms for dynamic call (algebraic effects) and exceptions\footnote{
+\CFA exception handling will be presented in a separate paper.
+The key feature that dovetails with this paper is nonlocal exceptions allowing exceptions to be raised across stacks, with synchronous exceptions raised among coroutines and asynchronous exceptions raised among threads, similar to that in \uC~\cite[\S~5]{uC++}}, this work only discusses retaining state between calls via generators/coroutines.
+\newterm{Coroutining} was introduced by Conway~\cite{Conway63} (1963), discussed by Knuth~\cite[\S~1.4.2]{Knuth73V1}, implemented in Simula67~\cite{Simula67}, formalized by Marlin~\cite{Marlin80}, and is now popular and appears in old and new programming languages: CLU~\cite{CLU}, \Csharp~\cite{Csharp}, Ruby~\cite{Ruby}, Python~\cite{Python}, JavaScript~\cite{JavaScript}, Lua~\cite{Lua}, \CCtwenty~\cite{C++20Coroutine19}.
+Coroutining is sequential execution requiring direct handoff among coroutines, \ie only the programmer is controlling execution order.
+If coroutines transfer to an internal event-engine for scheduling the next coroutines, the program transitions into the realm of concurrency~\cite[\S~3]{Buhr05a}.
+Coroutines are only a stepping stone towards concurrency where the commonality is that coroutines and threads retain state between calls.
+\Celeven/\CCeleven define concurrency~\cite[\S~7.26]{C11}, but it is largely wrappers for a subset of the pthreads library~\cite{Pthreads}.\footnote{Pthreads concurrency is based on simple thread fork/join in a function and mutex/condition locks, which is low-level and error-prone}
+Interestingly, almost a decade after the \Celeven standard, neither gcc-9, clang-9 nor msvc-19 (most recent versions) support the \Celeven include @threads.h@, indicating no interest in the C11 concurrency approach (possibly because of the recent effort to add concurrency to \CC).
+While the \Celeven standard does not state a threading model, the historical association with pthreads suggests implementations would adopt kernel-level threading (1:1)~\cite{ThreadModel}, as for \CC.
+In many ways, \CFA is to C as Scala~\cite{Scala} is to Java, providing a \emph{research vehicle} for new typing and control-flow capabilities on top of a highly popular programming language allowing immediate dissemination.
+Within the \CFA framework, new control-flow features are created from scratch because ISO \Celeven defines only a subset of the \CFA extensions, where the overlapping features are concurrency~\cite[\S~7.26]{C11}.
+However, \Celeven concurrency is largely wrappers for a subset of the pthreads library~\cite{Butenhof97,Pthreads}, and \Celeven and pthreads concurrency is simple, based on thread fork/join in a function and mutex/condition locks, which is low-level and error-prone;
+no high-level language concurrency features are defined.
+Interestingly, almost a decade after publication of the \Celeven standard, neither gcc-8, clang-9 nor msvc-19 (most recent versions) support the \Celeven include @threads.h@, indicating little interest in the C11 concurrency approach (possibly because the effort to add concurrency to \CC).
+Finally, while the \Celeven standard does not state a threading model, the historical association with pthreads suggests implementations would adopt kernel-level threading (1:1)~\cite{ThreadModel}.
 In contrast, there has been a renewed interest during the past decade in user-level (M:N, green) threading in old and new programming languages.
 As multi-core hardware became available in the 1980/90s, both user and kernel threading were examined.
 Kernel threading was chosen, largely because of its simplicity and fit with the simpler operating systems and hardware architectures at the time, which gave it a performance advantage~\cite{Drepper03}.
 Libraries like pthreads were developed for C, and the Solaris operating-system switched from user (JDK 1.1~\cite{JDK1.1}) to kernel threads.
 As a result, many current languages implementations adopt the 1:1 kernel-threading model, like Java (Scala), Objective-C~\cite{obj-c-book}, \CCeleven~\cite{C11}, C\#~\cite{Csharp} and Rust~\cite{Rust}, with a variety of presentation mechanisms.
 From 2000 onwards, several language implementations have championed the M:N user-threading model, like Go~\cite{Go}, Erlang~\cite{Erlang}, Haskell~\cite{Haskell}, D~\cite{D}, and \uC~\cite{uC++,uC++book}, including putting green threads back into Java~\cite{Quasar}, and many user-threading libraries have appeared~\cite{Qthreads,MPC,Marcel}.
 The main argument for user-level threading is that it is lighter weight than kernel threading (locking and context switching do not cross the kernel boundary), so there is less restriction on programming styles that encourages large numbers of threads performing medium-sized work to facilitate load balancing by the runtime~\cite{Verch12}.
+As a result, languages like Java, Scala, Objective-C~\cite{obj-c-book}, \CCeleven~\cite{C11}, and C\#~\cite{Csharp} adopt the 1:1 kernel-threading model, with a variety of presentation mechanisms.
+From 2000 onwards, languages like Go~\cite{Go}, Erlang~\cite{Erlang}, Haskell~\cite{Haskell}, D~\cite{D}, and \uC~\cite{uC++,uC++book} have championed the M:N user-threading model, and many user-threading libraries have appeared~\cite{Qthreads,MPC,Marcel}, including putting green threads back into Java~\cite{Quasar}.
+The main argument for user-level threading is that it is lighter weight than kernel threading (locking and context switching do not cross the kernel boundary), so there is less restriction on programming styles that encourage large numbers of threads performing medium work units to facilitate load balancing by the runtime~\cite{Verch12}.
 As well, user-threading facilitates a simpler concurrency approach using thread objects that leverage sequential patterns versus events with call-backs~\cite{Adya02,vonBehren03}.
 Finally, performant user-threading implementations (both time and space) meet or exceed direct kernel-threading implementations, while achieving the programming advantages of high concurrency levels and safety.
 A further effort over the past two decades is the development of language memory models to deal with the conflict between language features and compiler/hardware optimizations, \eg some language features are unsafe in the presence of aggressive sequential optimizations~\cite{Buhr95a,Boehm05}.
+A further effort over the past two decades is the development of language memory models to deal with the conflict between language features and compiler/hardware optimizations, \ie some language features are unsafe in the presence of aggressive sequential optimizations~\cite{Buhr95a,Boehm05}.
 The consequence is that a language must provide sufficient tools to program around safety issues, as inline and library code is all sequential to the compiler.
 One solution is low-level qualifiers and functions (\eg @volatile@ and atomics) allowing \emph{programmers} to explicitly write safe (race-free~\cite{Boehm12}) programs.
+A safer solution is high-level language constructs so the \emph{compiler} knows the concurrency boundaries (where mutual exclusion and synchronization are acquired/released) and provide implicit safety at and across these boundaries.
+While the optimization problem is best known with respect to concurrency, it applies to other complex control-flow, like exceptions and coroutines.
+As well, language solutions allow matching the language paradigm with the approach, \eg matching the functional paradigm with data-flow programming or the imperative paradigm with thread programming.
+Finally, it is important for a language to provide safety over performance \emph{as the default}, allowing careful reduction of safety (unsafe code) for performance when necessary.
+Two concurrency violations of this philosophy are \emph{spurious wakeup} (random wakeup~\cite[\S~9]{Buhr05a}) and \emph{barging}\footnote{
+Barging is competitive succession instead of direct handoff, \ie after a lock is released both arriving and preexisting waiter threads compete to acquire the lock.
+Hence, an arriving thread can temporally \emph{barge} ahead of threads already waiting for an event, which can repeat indefinitely leading to starvation of waiter threads.
+A safer solution is high-level language constructs so the \emph{compiler} knows the optimization boundaries, and hence, provides implicit safety.
+This problem is best known with respect to concurrency, but applies to other complex control-flow, like exceptions\footnote{
+\CFA exception handling will be presented in a separate paper.
+The key feature that dovetails with this paper is nonlocal exceptions allowing exceptions to be raised across stacks, with synchronous exceptions raised among coroutines and asynchronous exceptions raised among threads, similar to that in \uC~\cite[\S~5]{uC++}
+} and coroutines.
+Finally, language solutions allow matching constructs with language paradigm, \ie imperative and functional languages often have different presentations of the same concept to fit their programming model.
+Finally, it is important for a language to provide safety over performance \emph{as the default}, allowing careful reduction of safety for performance when necessary.
+Two concurrency violations of this philosophy are \emph{spurious wakeup} (random wakeup~\cite[\S~8]{Buhr05a}) and \emph{barging}\footnote{
+The notion of competitive succession instead of direct handoff, \ie a lock owner releases the lock and an arriving thread acquires it ahead of preexisting waiter threads.
 } (signals-as-hints~\cite[\S~8]{Buhr05a}), where one is a consequence of the other, \ie once there is spurious wakeup, signals-as-hints follow.
+(Author experience teaching concurrency is that students are confused by these semantics.)
+However, spurious wakeup is \emph{not} a foundational concurrency property~\cite[\S~9]{Buhr05a};
+it is a performance design choice.
+We argue removing spurious wakeup and signals-as-hints make concurrent programming simpler and safer as there is less local non-determinism to manage.
+If barging acquisition is allowed, its specialized performance advantage should be available as an option not the default.
+\CFA embraces language extensions for advanced control-flow, user-level threading, and safety as the default.
+We present comparative examples to support our argument that the \CFA control-flow extensions are as expressive and safe as those in other concurrent imperative programming languages, and perform experiments to show the \CFA runtime is competitive with other similar mechanisms.
+However, spurious wakeup is \emph{not} a foundational concurrency property~\cite[\S~8]{Buhr05a}, it is a performance design choice.
+Similarly, signals-as-hints are often a performance decision.
+We argue removing spurious wakeup and signals-as-hints make concurrent programming significantly safer because it removes local non-determinism and matches with programmer expectation.
+(Author experience teaching concurrency is that students are highly confused by these semantics.)
+Clawing back performance, when local non-determinism is unimportant, should be an option not the default.
+\begin{comment}
+Most augmented traditional (Fortran 18~\cite{Fortran18}, Cobol 14~\cite{Cobol14}, Ada 12~\cite{Ada12}, Java 11~\cite{Java11}) and new languages (Go~\cite{Go}, Rust~\cite{Rust}, and D~\cite{D}), except \CC, diverge from C with different syntax and semantics, only interoperate indirectly with C, and are not systems languages, for those with managed memory.
+As a result, there is a significant learning curve to move to these languages, and C legacy-code must be rewritten.
+While \CC, like \CFA, takes an evolutionary approach to extend C, \CC's constantly growing complex and interdependent features-set (\eg objects, inheritance, templates, etc.) mean idiomatic \CC code is difficult to use from C, and C programmers must expend significant effort learning \CC.
+Hence, rewriting and retraining costs for these languages, even \CC, are prohibitive for companies with a large C software-base.
+\CFA with its orthogonal feature-set, its high-performance runtime, and direct access to all existing C libraries circumvents these problems.
+\end{comment}
+\CFA embraces user-level threading, language extensions for advanced control-flow, and safety as the default.
+We present comparative examples so the reader can judge if the \CFA control-flow extensions are better and safer than those in other concurrent, imperative programming languages, and perform experiments to show the \CFA runtime is competitive with other similar mechanisms.
 The main contributions of this work are:
 \begin{itemize}[topsep=3pt,itemsep=0pt]
+\begin{itemize}[topsep=3pt,itemsep=1pt]
 \item
+a set of fundamental execution properties that dictate which language-level control-flow features need to be supported,
+language-level generators, coroutines and user-level threading, which respect the expectations of C programmers.
 \item
+integration of these language-level control-flow features, while respecting the style and expectations of C programmers,
+monitor synchronization without barging, and the ability to safely acquiring multiple monitors \emph{simultaneously} (deadlock free), while seamlessly integrating these capabilities with all monitor synchronization mechanisms.
 \item
+monitor synchronization without barging, and the ability to safely acquiring multiple monitors \emph{simultaneously} (deadlock free), while seamlessly integrating these capabilities with all monitor synchronization mechanisms,
+\item
+providing statically type-safe interfaces that integrate with the \CFA polymorphic type-system and other language features,
+providing statically type-safe interfaces that integrate with the \CFA polymorphic type-system and other language features.
 % \item
 % library extensions for executors, futures, and actors built on the basic mechanisms.
 \item
+a runtime system without spurious wake-up and no performance loss,
+a runtime system with no spurious wakeup.
 \item
+a dynamic partitioning mechanism to segregate groups of executing user and kernel threads performing specialized work (\eg web-server or compute engine) or requiring different scheduling (\eg NUMA or real-time).
+a dynamic partitioning mechanism to segregate the execution environment for specialized requirements.
 % \item
 % a non-blocking I/O library
 \item
 experimental results showing comparable performance of the \CFA features with similar mechanisms in other languages.
+experimental results showing comparable performance of the new features with similar mechanisms in other programming languages.
 \end{itemize}
+Section~\ref{s:FundamentalExecutionProperties} presents the compositional hierarchy of execution properties directing the design of control-flow features in \CFA.
+Section~\ref{s:StatefulFunction} begins advanced control by introducing sequential functions that retain data and execution state between calls producing constructs @generator@ and @coroutine@.
+Section~\ref{s:Concurrency} begins concurrency, or how to create (fork) and destroy (join) a thread producing the @thread@ construct.
+Section~\ref{s:StatefulFunction} begins advanced control by introducing sequential functions that retain data and execution state between calls, which produces constructs @generator@ and @coroutine@.
+Section~\ref{s:Concurrency} begins concurrency, or how to create (fork) and destroy (join) a thread, which produces the @thread@ construct.
 Section~\ref{s:MutualExclusionSynchronization} discusses the two mechanisms to restricted nondeterminism when controlling shared access to resources (mutual exclusion) and timing relationships among threads (synchronization).
 Section~\ref{s:Monitor} shows how both mutual exclusion and synchronization are safely embedded in the @monitor@ and @thread@ constructs.
 Section~\ref{s:CFARuntimeStructure} describes the large-scale mechanism to structure (cluster) threads and virtual processors (kernel threads).
+Section~\ref{s:Performance} uses a series of microbenchmarks to compare \CFA threading with pthreads, Java 11.0.6, Go 1.12.6, Rust 1.37.0, Python 3.7.6, Node.js 12.14.1, and \uC 7.0.0.
+\section{Fundamental Execution Properties}
+\label{s:FundamentalExecutionProperties}
+The features in a programming language should be composed from a set of fundamental properties rather than an ad hoc collection chosen by the designers.
+To this end, the control-flow features created for \CFA are based on the fundamental properties of any language with function-stack control-flow (see also \uC~\cite[pp.~140-142]{uC++}).
+The fundamental properties are execution state, thread, and mutual-exclusion/synchronization (MES).
+These independent properties can be used alone, in pairs, or in triplets to compose different language features, forming a compositional hierarchy where the most advanced feature has all the properties (state/thread/MES).
+While it is possible for a language to only support the most advanced feature~\cite{Hermes90}, this unnecessarily complicates and makes inefficient solutions to certain classes of problems.
+As is shown, each of the (non-rejected) composed features solves a particular set of problems, and hence, has a defensible position in a programming language.
+If a compositional feature is missing, a programmer has too few/many fundamental properties resulting in a complex and/or is inefficient solution.
+In detail, the fundamental properties are:
+\begin{description}[leftmargin=\parindent,topsep=3pt,parsep=0pt]
+\item[\newterm{execution state}:]
+is the state information needed by a control-flow feature to initialize, manage compute data and execution location(s), and de-initialize.
+State is retained in fixed-sized aggregate structures and dynamic-sized stack(s), often allocated in the heap(s) managed by the runtime system.
+The lifetime of the state varies with the control-flow feature, where longer life-time and dynamic size provide greater power but also increase usage complexity and cost.
+Control-flow transfers among execution states occurs in multiple ways, such as function call, context switch, asynchronous await, etc.
+Because the programming language determines what constitutes an execution state, implicitly manages this state, and defines movement mechanisms among states, execution state is an elementary property of the semantics of a programming language.
+% An execution-state is related to the notion of a process continuation \cite{Hieb90}.
+\item[\newterm{threading}:]
+is execution of code that occurs independently of other execution, \ie the execution resulting from a thread is sequential.
+Multiple threads provide \emph{concurrent execution};
+concurrent execution becomes parallel when run on multiple processing units (hyper-threading, cores, sockets).
+There must be language mechanisms to create, block/unblock, and join with a thread.
+\item[\newterm{MES}:]
+is the concurrency mechanisms to perform an action without interruption and establish timing relationships among multiple threads.
+These two properties are independent, \ie mutual exclusion cannot provide synchronization and vice versa without introducing additional threads~\cite[\S~4]{Buhr05a}.
+Limiting MES, \eg no access to shared data, results in contrived solutions and inefficiency on multi-core von Neumann computers where shared memory is a foundational aspect of its design.
+\end{description}
+These properties are fundamental because they cannot be built from existing language features, \eg a basic programming language like C99~\cite{C99} cannot create new control-flow features, concurrency, or provide MES using atomic hardware mechanisms.
+\subsection{Execution Properties}
+Table~\ref{t:ExecutionPropertyComposition} shows how the three fundamental execution properties: state, thread, and mutual exclusion compose a hierarchy of control-flow features needed in a programming language.
+(When doing case analysis, not all combinations are meaningful.)
+Note, basic von Neumann execution requires at least one thread and an execution state providing some form of call stack.
+For table entries missing these minimal components, the property is borrowed from the invoker (caller).
+Case 1 is a function that borrows storage for its state (stack frame/activation) and a thread from its invoker and retains this state across \emph{callees}, \ie function local-variables are retained on the stack across calls.
+Case 2 is case 1 with access to shared state so callers are restricted during update (mutual exclusion) and scheduling for other threads (synchronization).
+Case 3 is a stateful function supporting resume/suspend along with call/return to retain state across \emph{callers}, but has some restrictions because the function's state is stackless.
+Note, stackless functions still borrow the caller's stack and thread, where the stack is used to preserve state across its callees.
+Case 4 is cases 2 and 3 with protection to shared state for stackless functions.
+Cases 5 and 6 are the same as 3 and 4 but only the thread is borrowed as the function state is stackful, so resume/suspend is a context switch from the caller's to the function's stack.
+Cases 7 and 8 are rejected because a function that is given a new thread must have its own stack where the thread begins and stack frames are stored for calls, \ie there is no stack to borrow.
+Cases 9 and 10 are rejected because a thread with a fixed state (no stack) cannot accept calls, make calls, block, or be preempted, all of which require an unknown amount of additional dynamic state.
+Hence, once started, this kind of thread must execute to completion, \ie computation only, which severely restricts runtime management.
+Cases 11 and 12 have a stackful thread with and without safe access to shared state.
+Execution properties increase the cost of creation and execution along with complexity of usage.
+\begin{table}
+\caption{Execution property composition}
+\centering
+\label{t:ExecutionPropertyComposition}
+\renewcommand{\arraystretch}{1.25}
+%\setlength{\tabcolsep}{5pt}
+\begin{tabular}{c|c||l|l}
+\multicolumn{2}{c||}{execution properties} & \multicolumn{2}{c}{mutual exclusion / synchronization} \\
+\hline
+stateful                        & thread        & \multicolumn{1}{c|}{No} & \multicolumn{1}{c}{Yes} \\
+\hline
+\hline
+No                                      & No            & \textbf{1}\ \ \ function                              & \textbf{2}\ \ \ @monitor@ function    \\
+\hline
+Yes (stackless)         & No            & \textbf{3}\ \ \ @generator@                   & \textbf{4}\ \ \ @monitor@ @generator@ \\
+\hline
+Yes (stackful)          & No            & \textbf{5}\ \ \ @coroutine@                   & \textbf{6}\ \ \ @monitor@ @coroutine@ \\
+\hline
+No                                      & Yes           & \textbf{7}\ \ \ {\color{red}rejected} & \textbf{8}\ \ \ {\color{red}rejected} \\
+\hline
+Yes (stackless)         & Yes           & \textbf{9}\ \ \ {\color{red}rejected} & \textbf{10}\ \ \ {\color{red}rejected} \\
+\hline
+Yes (stackful)          & Yes           & \textbf{11}\ \ \ @thread@                             & \textbf{12}\ \ @monitor@ @thread@             \\
+\end{tabular}
+\end{table}
+Given the execution-properties taxonomy, programmers can now answer three basic questions: is state necessary across calls and how much, is a separate thread necessary, is access to shared state necessary.
+The answers define the optimal language feature need for implementing a programming problem.
+The next sections discusses how \CFA fills in the table with language features, while other programming languages may only provide a subset of the table.
+\subsection{Design Requirements}
+The following design requirements largely stem from building \CFA on top of C.
+\begin{itemize}[topsep=3pt,parsep=0pt]
+\item
+All communication must be statically type checkable for early detection of errors and efficient code generation.
+This requirement is consistent with the fact that C is a statically-typed programming-language.
+\item
+Direct interaction among language features must be possible allowing any feature to be selected without restricting comm\-unication.
+For example, many concurrent languages do not provide direct communication (calls) among threads, \ie threads only communicate indirectly through monitors, channels, messages, and/or futures.
+Indirect communication increases the number of objects, consuming more resources, and require additional synchronization and possibly data transfer.
+\item
+All communication is performed using function calls, \ie data is transmitted from argument to parameter and results are returned from function calls.
+Alternative forms of communication, such as call-backs, message passing, channels, or communication ports, step outside of C's normal form of communication.
+\item
+All stateful features must follow the same declaration scopes and lifetimes as other language data.
+For C that means at program startup, during block and function activation, and on demand using dynamic allocation.
+\item
+MES must be available implicitly in language constructs as well as explicitly for specialized requirements, because requiring programmers to build MES using low-level locks often leads to incorrect programs.
+Furthermore, reducing synchronization scope by encapsulating it within language constructs further reduces errors in concurrent programs.
+\item
+Both synchronous and asynchronous communication are needed.
+However, we believe the best way to provide asynchrony, such as call-buffering/chaining and/or returning futures~\cite{multilisp}, is building it from expressive synchronous features.
+\item
+Synchronization must be able to control the service order of requests including prioritizing selection from different kinds of outstanding requests, and postponing a request for an unspecified time while continuing to accept new requests.
+Otherwise, certain concurrency problems are difficult, e.g.\ web server, disk scheduling, and the amount of concurrency is inhibited~\cite{Gentleman81}.
+\end{itemize}
+We have satisfied these requirements in \CFA while maintaining backwards compatibility with the huge body of legacy C programs.
+% In contrast, other new programming languages must still access C programs (\eg operating-system service routines), but do so through fragile C interfaces.
+\subsection{Asynchronous Await / Call}
+Asynchronous await/call is a caller mechanism for structuring programs and/or increasing concurrency, where the caller (client) postpones an action into the future, which is subsequently executed by a callee (server).
+The caller detects the action's completion through a \newterm{future}/\newterm{promise}.
+The benefit is asynchronous caller execution with respect to the callee until future resolution.
+For single-threaded languages like JavaScript, an asynchronous call passes a callee action, which is queued in the event-engine, and continues execution with a promise.
+When the caller needs the promise to be fulfilled, it executes @await@.
+A promise-completion call-back can be part of the callee action or the caller is rescheduled;
+in either case, the call back is executed after the promise is fulfilled.
+While asynchronous calls generate new callee (server) events, we content this mechanism is insufficient for advanced control-flow mechanisms like generators or coroutines (which are discussed next).
+Specifically, control between caller and callee occurs indirectly through the event-engine precluding direct handoff and cycling among events, and requires complex resolution of a control promise and data.
+Note, @async-await@ is just syntactic-sugar over the event engine so it does not solve these deficiencies.
+For multi-threaded languages like Java, the asynchronous call queues a callee action with an executor (server), which subsequently executes the work by a thread in the executor thread-pool.
+The problem is when concurrent work-units need to interact and/or block as this effects the executor, \eg stops threads.
+While it is possible to extend this approach to support the necessary mechanisms, \eg message passing in Actors, we show monitors and threads provide an equally competitive approach that does not deviate from normal call communication and can be used to build asynchronous call, as is done in Java.
+Section~\ref{s:Performance} uses a series of microbenchmarks to compare \CFA threading with pthreads, Java OpenJDK-9, Go 1.12.6 and \uC 7.0.0.
 …
 \label{s:StatefulFunction}
+A \emph{stateful function} has the ability to remember state between calls, where state can be either data or execution, \eg plugin, device driver, finite-state machine (FSM).
+A simple technique to retain data state between calls is @static@ declarations within a function, which is often implemented by hoisting the declarations to the global scope but hiding the names within the function using name mangling.
+However, each call starts the function at the top making it difficult to determine the last point of execution in an algorithm, and requiring multiple flag variables and testing to reestablish the continuation point.
+Hence, the next step of generalizing function state is implicitly remembering the return point between calls and reentering the function at this point rather than the top, called \emph{generators}\,/\,\emph{iterators} or \emph{stackless coroutines}.
+For example, a Fibonacci generator retains data and execution state allowing it to remember prior values needed to generate the next value and the location in the algorithm to compute that value.
+The next step of generalization is instantiating the function to allow multiple named instances, \eg multiple Fibonacci generators, where each instance has its own state, and hence, can generate an independent sequence of values.
+Note, a subset of generator state is a function \emph{closure}, \ie the technique of capturing lexical references when returning a nested function.
+A further generalization is adding a stack to a generator's state, called a \emph{coroutine}, so it can suspend outside of itself, \eg call helper functions to arbitrary depth before suspending back to its resumer without unwinding these calls.
+For example, a coroutine iterator for a binary tree can stop the traversal at the visit point (pre, infix, post traversal), return the node value to the caller, and then continue the recursive traversal from the current node on the next call.
+There are two styles of activating a stateful function, \emph{asymmetric} or \emph{symmetric}, identified by resume/suspend (no cycles) and resume/resume (cycles).
+These styles \emph{do not} cause incremental stack growth, \eg a million resume/suspend or resume/resume cycles do not remember each cycle just the last resumer for each cycle.
+Selecting between stackless/stackful semantics and asymmetric/symmetric style is a tradeoff between programming requirements, performance, and design, where stackless is faster and smaller (modified call/return between closures), stackful is more general but slower and larger (context switching between distinct stacks), and asymmetric is simpler control-flow than symmetric.
+Additionally, storage management for the closure/stack (especially in unmanaged languages, \ie no garbage collection) must be factored into design and performance.
+Note, creation cost (closure/stack) is amortized across usage, so activation cost (resume/suspend) is usually the dominant factor.
+% The stateful function is an old idea~\cite{Conway63,Marlin80} that is new again~\cite{C++20Coroutine19}, where execution is temporarily suspended and later resumed, \eg plugin, device driver, finite-state machine.
+% Hence, a stateful function may not end when it returns to its caller, allowing it to be restarted with the data and execution location present at the point of suspension.
+% If the closure is fixed size, we call it a \emph{generator} (or \emph{stackless}), and its control flow is restricted, \eg suspending outside the generator is prohibited.
+% If the closure is variable size, we call it a \emph{coroutine} (or \emph{stackful}), and as the names implies, often implemented with a separate stack with no programming restrictions.
+% Hence, refactoring a stackless coroutine may require changing it to stackful.
+% A foundational property of all \emph{stateful functions} is that resume/suspend \emph{do not} cause incremental stack growth, \ie resume/suspend operations are remembered through the closure not the stack.
+% As well, activating a stateful function is \emph{asymmetric} or \emph{symmetric}, identified by resume/suspend (no cycles) and resume/resume (cycles).
+% A fixed closure activated by modified call/return is faster than a variable closure activated by context switching.
+% Additionally, any storage management for the closure (especially in unmanaged languages, \ie no garbage collection) must also be factored into design and performance.
+% Therefore, selecting between stackless and stackful semantics is a tradeoff between programming requirements and performance, where stackless is faster and stackful is more general.
+% nppNote, creation cost is amortized across usage, so activation cost is usually the dominant factor.
+For example, Python presents asymmetric generators as a function object, \uC presents symmetric coroutines as a \lstinline[language=C++]|class|-like object, and many languages present threading using function pointers, @pthreads@~\cite{Butenhof97}, \Csharp~\cite{Csharp}, Go~\cite{Go}, and Scala~\cite{Scala}.
+\begin{center}
+\begin{tabular}{@{}l|l|l@{}}
+\multicolumn{1}{@{}c|}{Python asymmetric generator} & \multicolumn{1}{c|}{\uC symmetric coroutine} & \multicolumn{1}{c@{}}{Pthreads thread} \\
+\hline
+\begin{python}
+`def Gen():` $\LstCommentStyle{\color{red}// function}$
+        ... yield val ...
+gen = Gen()
+for i in range( 10 ):
+        print( next( gen ) )
+\end{python}
+&
+\begin{uC++}
+`_Coroutine Cycle {` $\LstCommentStyle{\color{red}// class}$
+        Cycle * p;
+        void main() { p->cycle(); }
+        void cycle() { resume(); }  `};`
+Cycle c1, c2; c1.p=&c2; c2.p=&c1; c1.cycle();
+\end{uC++}
+&
+\begin{cfa}
+void * rtn( void * arg ) { ... }
+int i = 3, rc;
+pthread_t t; $\C{// thread id}$
+$\LstCommentStyle{\color{red}// function pointer}$
+rc=pthread_create(&t, `rtn`, (void *)i);
+\end{cfa}
+\end{tabular}
+\end{center}
+\CFA's preferred presentation model for generators/coroutines/threads is a hybrid of functions and classes, giving an object-oriented flavour.
+Essentially, the generator/coroutine/thread function is semantically coupled with a generator/coroutine/thread custom type via the type's name.
+The custom type solves several issues, while accessing the underlying mechanisms used by the custom types is still allowed for flexibility reasons.
+Each custom type is discussed in detail in the following sections.
+\subsection{Generator}
+Stackless generators (Table~\ref{t:ExecutionPropertyComposition} case 3) have the potential to be very small and fast, \ie as small and fast as function call/return for both creation and execution.
+The \CFA goal is to achieve this performance target, possibly at the cost of some semantic complexity.
+A series of different kinds of generators and their implementation demonstrate how this goal is accomplished.\footnote{
+The \CFA operator syntax uses \lstinline|?| to denote operands, which allows precise definitions for pre, post, and infix operators, \eg \lstinline|?++|, \lstinline|++?|, and \lstinline|?+?|, in addition \lstinline|?\{\}| denotes a constructor, as in \lstinline|foo `f` = `\{`...`\}`|, \lstinline|^?\{\}| denotes a destructor, and \lstinline|?()| is \CC function call \lstinline|operator()|.
+Operator \lstinline+|+ is overloaded for printing, like bit-shift \lstinline|<<| in \CC.
+The \CFA \lstinline|with| clause opens an aggregate scope making its fields directly accessible, like Pascal \lstinline|with|, but using parallel semantics;
+multiple aggregates may be opened.
+\CFA has rebindable references \lstinline|int i, & ip = i, j; `&ip = &j;`| and non-rebindable references \lstinline|int i, & `const` ip = i, j; `&ip = &j;` // disallowed|.
+}%
+The stateful function is an old idea~\cite{Conway63,Marlin80} that is new again~\cite{C++20Coroutine19}, where execution is temporarily suspended and later resumed, \eg plugin, device driver, finite-state machine.
+Hence, a stateful function may not end when it returns to its caller, allowing it to be restarted with the data and execution location present at the point of suspension.
+This capability is accomplished by retaining a data/execution \emph{closure} between invocations.
+If the closure is fixed size, we call it a \emph{generator} (or \emph{stackless}), and its control flow is restricted, \eg suspending outside the generator is prohibited.
+If the closure is variable size, we call it a \emph{coroutine} (or \emph{stackful}), and as the names implies, often implemented with a separate stack with no programming restrictions.
+Hence, refactoring a stackless coroutine may require changing it to stackful.
+A foundational property of all \emph{stateful functions} is that resume/suspend \emph{do not} cause incremental stack growth, \ie resume/suspend operations are remembered through the closure not the stack.
+As well, activating a stateful function is \emph{asymmetric} or \emph{symmetric}, identified by resume/suspend (no cycles) and resume/resume (cycles).
+A fixed closure activated by modified call/return is faster than a variable closure activated by context switching.
+Additionally, any storage management for the closure (especially in unmanaged languages, \ie no garbage collection) must also be factored into design and performance.
+Therefore, selecting between stackless and stackful semantics is a tradeoff between programming requirements and performance, where stackless is faster and stackful is more general.
+Note, creation cost is amortized across usage, so activation cost is usually the dominant factor.
 \begin{figure}
 …
         int fn = f->fn; f->fn = f->fn1;
                 f->fn1 = f->fn + fn;
         return fn;
+}
 int main() {
 …
 void `main(Fib & fib)` with(fib) {
         [fn1, fn] = [1, 0];
         for () {
 …
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
 typedef struct {
         int `restart`, fn1, fn;
+        int fn1, fn;  void * `next`;
 } Fib;
 #define FibCtor { `0`, 1, 0 }
+#define FibCtor { 1, 0, NULL }
 Fib * comain( Fib * f ) {
+        `static void * states[] = {&&s0, &&s1};`
+        `goto *states[f->restart];`
+  s0: f->`restart` = 1;
+        if ( f->next ) goto *f->next;
+        f->next = &&s1;
         for ( ;; ) {
                 return f;
           s1:; int fn = f->fn + f->fn1;
                 f->fn1 = f->fn; f->fn = fn;
+                        f->fn1 = f->fn; f->fn = fn;
+        }
+}
 …
 \end{lrbox}
 \subfloat[C]{\label{f:CFibonacci}\usebox\myboxA}
+\subfloat[C asymmetric generator]{\label{f:CFibonacci}\usebox\myboxA}
 \hspace{3pt}
 \vrule
 \hspace{3pt}
 \subfloat[\CFA]{\label{f:CFAFibonacciGen}\usebox\myboxB}
+\subfloat[\CFA asymmetric generator]{\label{f:CFAFibonacciGen}\usebox\myboxB}
 \hspace{3pt}
 \vrule
 \hspace{3pt}
 \subfloat[C generated code for \CFA version]{\label{f:CFibonacciSim}\usebox\myboxC}
+\subfloat[C generator implementation]{\label{f:CFibonacciSim}\usebox\myboxC}
 \caption{Fibonacci (output) asymmetric generator}
 \label{f:FibonacciAsymmetricGenerator}
 …
 };
 void ?{}( Fmt & fmt ) { `resume(fmt);` } // constructor
 void ^?{}( Fmt & f ) with(f) { $\C[2.25in]{// destructor}$
+void ^?{}( Fmt & f ) with(f) { $\C[1.75in]{// destructor}$
         if ( g != 0 || b != 0 ) sout | nl; }
 void `main( Fmt & f )` with(f) {
 …
                 for ( ; g < 5; g += 1 ) { $\C{// groups}$
                         for ( ; b < 4; b += 1 ) { $\C{// blocks}$
                                 do { `suspend;` $\C{// wait for character}$
                                 while ( ch == '\n' ); // ignore newline
                                 sout | ch;                      $\C{// print character}$
                         } sout | " ";  $\C{// block separator}$
                 } sout | nl; $\C{// group separator}$
+                                `suspend;` $\C{// wait for character}$
+                                while ( ch == '\n' ) `suspend;` // ignore
+                                sout | ch;                                              // newline
+                        } sout | " ";  // block spacer
+                } sout | nl; // group newline
+        }
+}
 …
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
 typedef struct {
         int `restart`, g, b;
+        void * next;
         char ch;
+        int g, b;
 } Fmt;
 void comain( Fmt * f ) {
+        `static void * states[] = {&&s0, &&s1};`
+        `goto *states[f->restart];`
+  s0: f->`restart` = 1;
+        if ( f->next ) goto *f->next;
+        f->next = &&s1;
         for ( ;; ) {
                 for ( f->g = 0; f->g < 5; f->g += 1 ) {
                         for ( f->b = 0; f->b < 4; f->b += 1 ) {
                                 do { return;  s1: ;
                                 } while ( f->ch == '\n' );
+                                return;
+                          s1:;  while ( f->ch == '\n' ) return;
                                 printf( "%c", f->ch );
                         } printf( " " );
 …
+}
 int main() {
         Fmt fmt = { `0` };  comain( &fmt ); // prime
+        Fmt fmt = { NULL };  comain( &fmt ); // prime
         for ( ;; ) {
                 scanf( "%c", &fmt.ch );
 …
 \end{lrbox}
 \subfloat[\CFA]{\label{f:CFAFormatGen}\usebox\myboxA}
 \hspace{35pt}
+\subfloat[\CFA asymmetric generator]{\label{f:CFAFormatGen}\usebox\myboxA}
+\hspace{3pt}
 \vrule
 \hspace{3pt}
 \subfloat[C generated code for \CFA version]{\label{f:CFormatGenImpl}\usebox\myboxB}
+\subfloat[C generator simulation]{\label{f:CFormatSim}\usebox\myboxB}
 \hspace{3pt}
 \caption{Formatter (input) asymmetric generator}
 …
 \end{figure}
+Figure~\ref{f:FibonacciAsymmetricGenerator} shows an unbounded asymmetric generator for an infinite sequence of Fibonacci numbers written (left to right) in C, \CFA, and showing the underlying C implementation for the \CFA version.
+Stateful functions appear as generators, coroutines, and threads, where presentations are based on function objects or pointers~\cite{Butenhof97, C++14, MS:VisualC++, BoostCoroutines15}.
+For example, Python presents generators as a function object:
+\begin{python}
+def Gen():
+        ... `yield val` ...
+gen = Gen()
+for i in range( 10 ):
+        print( next( gen ) )
+\end{python}
+Boost presents coroutines in terms of four functor object-types:
+\begin{cfa}
+asymmetric_coroutine<>::pull_type
+asymmetric_coroutine<>::push_type
+symmetric_coroutine<>::call_type
+symmetric_coroutine<>::yield_type
+\end{cfa}
+and many languages present threading using function pointers, @pthreads@~\cite{Butenhof97}, \Csharp~\cite{Csharp}, Go~\cite{Go}, and Scala~\cite{Scala}, \eg pthreads:
+\begin{cfa}
+void * rtn( void * arg ) { ... }
+int i = 3, rc;
+pthread_t t; $\C{// thread id}$
+`rc = pthread_create( &t, rtn, (void *)i );` $\C{// create and initialized task, type-unsafe input parameter}$
+\end{cfa}
+% void mycor( pthread_t cid, void * arg ) {
+%       int * value = (int *)arg;                               $\C{// type unsafe, pointer-size only}$
+%       // thread body
+% }
+% int main() {
+%       int input = 0, output;
+%       coroutine_t cid = coroutine_create( &mycor, (void *)&input ); $\C{// type unsafe, pointer-size only}$
+%       coroutine_resume( cid, (void *)input, (void **)&output ); $\C{// type unsafe, pointer-size only}$
+% }
+\CFA's preferred presentation model for generators/coroutines/threads is a hybrid of objects and functions, with an object-oriented flavour.
+Essentially, the generator/coroutine/thread function is semantically coupled with a generator/coroutine/thread custom type.
+The custom type solves several issues, while accessing the underlying mechanisms used by the custom types is still allowed.
+\subsection{Generator}
+Stackless generators have the potential to be very small and fast, \ie as small and fast as function call/return for both creation and execution.
+The \CFA goal is to achieve this performance target, possibly at the cost of some semantic complexity.
+A series of different kinds of generators and their implementation demonstrate how this goal is accomplished.
+Figure~\ref{f:FibonacciAsymmetricGenerator} shows an unbounded asymmetric generator for an infinite sequence of Fibonacci numbers written in C and \CFA, with a simple C implementation for the \CFA version.
 This generator is an \emph{output generator}, producing a new result on each resumption.
 To compute Fibonacci, the previous two values in the sequence are retained to generate the next value, \ie @fn1@ and @fn@, plus the execution location where control restarts when the generator is resumed, \ie top or middle.
 …
 The C version only has the middle execution state because the top execution state is declaration initialization.
 Figure~\ref{f:CFAFibonacciGen} shows the \CFA approach, which also has a manual closure, but replaces the structure with a custom \CFA @generator@ type.
+Each generator type must have a function named \lstinline|main|,
+% \footnote{
+% The name \lstinline|main| has special meaning in C, specifically the function where a program starts execution.
+% Leveraging starting semantics to this name for generator/coroutine/thread is a logical extension.}
+called a \emph{generator main} (leveraging the starting semantics for program @main@ in C), which is connected to the generator type via its single reference parameter.
+This generator type is then connected to a function that \emph{must be named \lstinline|main|},\footnote{
+The name \lstinline|main| has special meaning in C, specifically the function where a program starts execution.
+Hence, overloading this name for other starting points (generator/coroutine/thread) is a logical extension.}
+called a \emph{generator main},which takes as its only parameter a reference to the generator type.
 The generator main contains @suspend@ statements that suspend execution without ending the generator versus @return@.
+For the Fibonacci generator-main,
+For the Fibonacci generator-main,\footnote{
+The \CFA \lstinline|with| opens an aggregate scope making its fields directly accessible, like Pascal \lstinline|with|, but using parallel semantics.
+Multiple aggregates may be opened.}
 the top initialization state appears at the start and the middle execution state is denoted by statement @suspend@.
 Any local variables in @main@ \emph{are not retained} between calls;
 …
 Resuming an ended (returned) generator is undefined.
 Function @resume@ returns its argument generator so it can be cascaded in an expression, in this case to print the next Fibonacci value @fn@ computed in the generator instance.
+Figure~\ref{f:CFibonacciSim} shows the C implementation of the \CFA asymmetric generator.
+Only one execution-state field, @restart@, is needed to subscript the suspension points in the generator.
+At the start of the generator main, the @static@ declaration, @states@, is initialized to the N suspend points in the generator (where operator @&&@ dereferences/references a label~\cite{gccValueLabels}).
+Next, the computed @goto@ selects the last suspend point and branches to it.
+The  cost of setting @restart@ and branching via the computed @goto@ adds very little cost to the suspend/resume calls.
+An advantage of the \CFA explicit generator type is the ability to allow multiple type-safe interface functions taking and returning arbitrary types.
+Figure~\ref{f:CFibonacciSim} shows the C implementation of the \CFA generator only needs one additional field, @next@, to handle retention of execution state.
+The computed @goto@ at the start of the generator main, which branches after the previous suspend, adds very little cost to the resume call.
+Finally, an explicit generator type provides both design and performance benefits, such as multiple type-safe interface functions taking and returning arbitrary types.\footnote{
+The \CFA operator syntax uses \lstinline|?| to denote operands, which allows precise definitions for pre, post, and infix operators, \eg \lstinline|++?|, \lstinline|?++|, and \lstinline|?+?|, in addition \lstinline|?\{\}| denotes a constructor, as in \lstinline|foo `f` = `\{`...`\}`|, \lstinline|^?\{\}| denotes a destructor, and \lstinline|?()| is \CC function call \lstinline|operator()|.
+}%
 \begin{cfa}
 int ?()( Fib & fib ) { return `resume( fib )`.fn; } $\C[3.9in]{// function-call interface}$
+int ?()( Fib & fib, int N ) { for ( N - 1 ) `fib()`; return `fib()`; } $\C{// add parameter to skip N values}$
+double ?()( Fib & fib ) { return (int)`fib()` / 3.14159; } $\C{// different return type, cast prevents recursive call}$
+Fib f;  int i;  double d;
+i = f();  i = f( 2 );  d = f();                                         $\C{// alternative interfaces}\CRT$
+int ?()( Fib & fib, int N ) { for ( N - 1 ) `fib()`; return `fib()`; } $\C{// use function-call interface to skip N values}$
+double ?()( Fib & fib ) { return (int)`fib()` / 3.14159; } $\C{// different return type, cast prevents recursive call}\CRT$
+sout | (int)f1() | (double)f1() | f2( 2 ); // alternative interface, cast selects call based on return type, step 2 values
 \end{cfa}
 Now, the generator can be a separately compiled opaque-type only accessed through its interface functions.
 For contrast, Figure~\ref{f:PythonFibonacci} shows the equivalent Python Fibonacci generator, which does not use a generator type, and hence only has a single interface, but an implicit closure.
+\begin{figure}
+%\centering
+\newbox\myboxA
+\begin{lrbox}{\myboxA}
+\begin{python}[aboveskip=0pt,belowskip=0pt]
+def Fib():
+        fn1, fn = 0, 1
+        while True:
+                `yield fn1`
+                fn1, fn = fn, fn1 + fn
+f1 = Fib()
+f2 = Fib()
+for i in range( 10 ):
+        print( next( f1 ), next( f2 ) )
+\end{python}
+\end{lrbox}
+\newbox\myboxB
+\begin{lrbox}{\myboxB}
+\begin{python}[aboveskip=0pt,belowskip=0pt]
+def Fmt():
+        try:
+                while True:                                             $\C[2.5in]{\# until destructor call}$
+                        for g in range( 5 ):            $\C{\# groups}$
+                                for b in range( 4 ):    $\C{\# blocks}$
+                                        while True:
+                                                ch = (yield)    $\C{\# receive from send}$
+                                                if '\n' not in ch: $\C{\# ignore newline}$
+                                                        break
+                                        print( ch, end='' )     $\C{\# print character}$
+                                print( '  ', end='' )   $\C{\# block separator}$
+                        print()                                         $\C{\# group separator}$
+        except GeneratorExit:                           $\C{\# destructor}$
+                if g != 0 | b != 0:                             $\C{\# special case}$
+                        print()
+fmt = Fmt()
+`next( fmt )`                                                   $\C{\# prime, next prewritten}$
+for i in range( 41 ):
+        `fmt.send( 'a' );`                                      $\C{\# send to yield}$
+\end{python}
+\end{lrbox}
+\hspace{30pt}
+\subfloat[Fibonacci]{\label{f:PythonFibonacci}\usebox\myboxA}
+\hspace{3pt}
+\vrule
+\hspace{3pt}
+\subfloat[Formatter]{\label{f:PythonFormatter}\usebox\myboxB}
+\caption{Python generator}
+\label{f:PythonGenerator}
+\end{figure}
+Having to manually create the generator closure by moving local-state variables into the generator type is an additional programmer burden (removed by the coroutine in Section~\ref{s:Coroutine}).
+This manual requirement follows from the generality of allowing variable-size local-state, \eg local state with a variable-length array requires dynamic allocation as the array size is unknown at compile time.
+Having to manually create the generator closure by moving local-state variables into the generator type is an additional programmer burden.
+(This restriction is removed by the coroutine in Section~\ref{s:Coroutine}.)
+This requirement follows from the generality of variable-size local-state, \eg local state with a variable-length array requires dynamic allocation because the array size is unknown at compile time.
 However, dynamic allocation significantly increases the cost of generator creation/destruction and is a showstopper for embedded real-time programming.
 But more importantly, the size of the generator type is tied to the local state in the generator main, which precludes separate compilation of the generator main, \ie a generator must be inlined or local state must be dynamically allocated.
 With respect to safety, we believe static analysis can discriminate persistent generator state from temporary generator-main state and raise a compile-time error for temporary usage spanning suspend points.
 Our experience using generators is that the problems have simple data state, including local state, but complex execution state, so the burden of creating the generator type is small.
+With respect to safety, we believe static analysis can discriminate local state from temporary variables in a generator, \ie variable usage spanning @suspend@, and generate a compile-time error.
+Finally, our current experience is that most generator problems have simple data state, including local state, but complex execution state, so the burden of creating the generator type is small.
 As well, C programmers are not afraid of this kind of semantic programming requirement, if it results in very small, fast generators.
 …
 The example takes advantage of resuming a generator in the constructor to prime the loops so the first character sent for formatting appears inside the nested loops.
 The destructor provides a newline, if formatted text ends with a full line.
+Figure~\ref{f:CFormatGenImpl} shows the C implementation of the \CFA input generator with one additional field and the computed @goto@.
+For contrast, Figure~\ref{f:PythonFormatter} shows the equivalent Python format generator with the same properties as the format generator.
+% https://dl-acm-org.proxy.lib.uwaterloo.ca/
+Figure~\ref{f:DeviceDriverGen} shows an important application for an asymmetric generator, a device-driver, because device drivers are a significant source of operating-system errors: 85\% in Windows XP~\cite[p.~78]{Swift05} and 51.6\% in Linux~\cite[p.~1358,]{Xiao19}. %\cite{Palix11}
+Swift \etal~\cite[p.~86]{Swift05} restructure device drivers using the Extension Procedure Call (XPC) within the kernel via functions @nooks_driver_call@ and @nooks_kernel_call@, which have coroutine properties context switching to separate stacks with explicit hand-off calls;
+however, the calls do not retain execution state, and hence always start from the top.
+The alternative approach for implementing device drivers is using stack-ripping.
+However, Adya \etal~\cite{Adya02} argue against stack ripping in Section 3.2 and suggest a hybrid approach in Section 4 using cooperatively scheduled \emph{fibers}, which is coroutining.
+As an example, the following protocol:
+Figure~\ref{f:CFormatSim} shows the C implementation of the \CFA input generator with one additional field and the computed @goto@.
+For contrast, Figure~\ref{f:PythonFormatter} shows the equivalent Python format generator with the same properties as the Fibonacci generator.
+Figure~\ref{f:DeviceDriverGen} shows a \emph{killer} asymmetric generator, a device-driver, because device drivers caused 70\%-85\% of failures in Windows/Linux~\cite{Swift05}.
+Device drives follow the pattern of simple data state but complex execution state, \ie finite state-machine (FSM) parsing a protocol.
+For example, the following protocol:
 \begin{center}
 \ldots\, STX \ldots\, message \ldots\, ESC ETX \ldots\, message \ldots\, ETX 2-byte crc \ldots
 \end{center}
 is for a simple network message beginning with the control character STX, ending with an ETX, and followed by a 2-byte cyclic-redundancy check.
+is a network message beginning with the control character STX, ending with an ETX, and followed by a 2-byte cyclic-redundancy check.
 Control characters may appear in a message if preceded by an ESC.
 When a message byte arrives, it triggers an interrupt, and the operating system services the interrupt by calling the device driver with the byte read from a hardware register.
+The device driver returns a status code of its current state, and when a complete message is obtained, the operating system read the message accumulated in the supplied buffer.
+Hence, the device driver is an input/output generator, where the cost of resuming the device-driver generator is the same as call/return, so performance in an operating-system kernel is excellent.
+The key benefits of using a generator are correctness, safety, and maintenance because the execution states are transcribed directly into the programming language rather than table lookup or stack ripping.
+The conclusion is that FSMs are complex and occur in important domains, so direct generator support is important in a system programming language.
+The device driver returns a status code of its current state, and when a complete message is obtained, the operating system knows the message is in the message buffer.
+Hence, the device driver is an input/output generator.
+Note, the cost of creating and resuming the device-driver generator, @Driver@, is virtually identical to call/return, so performance in an operating-system kernel is excellent.
+As well, the data state is small, where variables @byte@ and @msg@ are communication variables for passing in message bytes and returning the message, and variables @lnth@, @crc@, and @sum@ are local variable that must be retained between calls and are manually hoisted into the generator type.
+% Manually, detecting and hoisting local-state variables is easy when the number is small.
+In contrast, the execution state is large, with one @resume@ and seven @suspend@s.
+Hence, the key benefits of the generator are correctness, safety, and maintenance because the execution states are transcribed directly into the programming language rather than using a table-driven approach.
+Because FSMs can be complex and frequently occur in important domains, direct generator support is important in a system programming language.
 \begin{figure}
 \centering
+\newbox\myboxA
+\begin{lrbox}{\myboxA}
+\begin{python}[aboveskip=0pt,belowskip=0pt]
+def Fib():
+        fn1, fn = 0, 1
+        while True:
+                `yield fn1`
+                fn1, fn = fn, fn1 + fn
+f1 = Fib()
+f2 = Fib()
+for i in range( 10 ):
+        print( next( f1 ), next( f2 ) )
+\end{python}
+\end{lrbox}
+\newbox\myboxB
+\begin{lrbox}{\myboxB}
+\begin{python}[aboveskip=0pt,belowskip=0pt]
+def Fmt():
+        try:
+                while True:
+                        for g in range( 5 ):
+                                for b in range( 4 ):
+                                        print( `(yield)`, end='' )
+                                print( '  ', end='' )
+                        print()
+        except GeneratorExit:
+                if g != 0 | b != 0:
+                        print()
+fmt = Fmt()
+`next( fmt )`                    # prime, next prewritten
+for i in range( 41 ):
+        `fmt.send( 'a' );`      # send to yield
+\end{python}
+\end{lrbox}
+\subfloat[Fibonacci]{\label{f:PythonFibonacci}\usebox\myboxA}
+\hspace{3pt}
+\vrule
+\hspace{3pt}
+\subfloat[Formatter]{\label{f:PythonFormatter}\usebox\myboxB}
+\caption{Python generator}
+\label{f:PythonGenerator}
+\bigskip
 \begin{tabular}{@{}l|l@{}}
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
 …
 `generator` Driver {
         Status status;
         char byte, * msg; // communication
         int lnth, sum;      // local state
         short int crc;
+        unsigned char byte, * msg; // communication
+        unsigned int lnth, sum;      // local state
+        unsigned short int crc;
 };
 void ?{}( Driver & d, char * m ) { d.msg = m; }
 …
 (The trivial cycle is a generator resuming itself.)
 This control flow is similar to recursion for functions but without stack growth.
 Figure~\ref{f:PingPongFullCoroutineSteps} shows the steps for symmetric control-flow are creating, executing, and terminating the cycle.
+The steps for symmetric control-flow are creating, executing, and terminating the cycle.
 Constructing the cycle must deal with definition-before-use to close the cycle, \ie, the first generator must know about the last generator, which is not within scope.
 (This issue occurs for any cyclic data structure.)
 The example creates the generators, @ping@/@pong@, and then assigns the partners that form the cycle.
 % (Alternatively, the constructor can assign the partners as they are declared, except the first, and the first-generator partner is set after the last generator declaration to close the cycle.)
 Once the cycle is formed, the program main resumes one of the generators, @ping@, and the generators can then traverse an arbitrary cycle using @resume@ to activate partner generator(s).
+% The example creates all the generators and then assigns the partners that form the cycle.
+% Alternatively, the constructor can assign the partners as they are declared, except the first, and the first-generator partner is set after the last generator declaration to close the cycle.
+Once the cycle is formed, the program main resumes one of the generators, and the generators can then traverse an arbitrary cycle using @resume@ to activate partner generator(s).
 Terminating the cycle is accomplished by @suspend@ or @return@, both of which go back to the stack frame that started the cycle (program main in the example).
-Note, the creator and starter may be different, \eg if the creator calls another function that starts the cycle.
 The starting stack-frame is below the last active generator because the resume/resume cycle does not grow the stack.
+Also, since local variables are not retained in the generator function, there are no objects with destructors to be called, so the cost is the same as a function return.
+Destructor cost occurs when the generator instance is deallocated by the creator.
+Also, since local variables are not retained in the generator function, it does not contain any objects with destructors that must be called, so the  cost is the same as a function return.
+Destructor cost occurs when the generator instance is deallocated, which is easily controlled by the programmer.
+Figure~\ref{f:CPingPongSim} shows the implementation of the symmetric generator, where the complexity is the @resume@, which needs an extension to the calling convention to perform a forward rather than backward jump.
+This jump-starts at the top of the next generator main to re-execute the normal calling convention to make space on the stack for its local variables.
+However, before the jump, the caller must reset its stack (and any registers) equivalent to a @return@, but subsequently jump forward.
+This semantics is basically a tail-call optimization, which compilers already perform.
+The example shows the assembly code to undo the generator's entry code before the direct jump.
+This assembly code depends on what entry code is generated, specifically if there are local variables and the level of optimization.
+To provide this new calling convention requires a mechanism built into the compiler, which is beyond the scope of \CFA at this time.
+Nevertheless, it is possible to hand generate any symmetric generators for proof of concept and performance testing.
+A compiler could also eliminate other artifacts in the generator simulation to further increase performance, \eg LLVM has various coroutine support~\cite{CoroutineTS}, and \CFA can leverage this support should it fork @clang@.
 \begin{figure}
 …
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
 `generator PingPong` {
-        int N, i;                               // local state
         const char * name;
+        int N;
+        int i;                          // local state
         PingPong & partner; // rebindable reference
 };
 void `main( PingPong & pp )` with(pp) {
         for ( ; i < N; i += 1 ) {
                 sout | name | i;
 …
 \begin{cfa}[escapechar={},aboveskip=0pt,belowskip=0pt]
 typedef struct PingPong {
-        int restart, N, i;
         const char * name;
+        int N, i;
         struct PingPong * partner;
+        void * next;
 } PingPong;
 #define PPCtor(name, N) {0, N, 0, name, NULL}
+#define PPCtor(name, N) {name,N,0,NULL,NULL}
 void comain( PingPong * pp ) {
+        static void * states[] = {&&s0, &&s1};
+        goto *states[pp->restart];
+  s0: pp->restart = 1;
+        if ( pp->next ) goto *pp->next;
+        pp->next = &&cycle;
         for ( ; pp->i < pp->N; pp->i += 1 ) {
                 printf( "%s %d\n", pp->name, pp->i );
                 asm( "mov  %0,%%rdi" : "=m" (pp->partner) );
                 asm( "mov  %rdi,%rax" );
+                asm( "add  $16, %rsp" );
+                asm( "popq %rbp" );
+                asm( "popq %rbx" );
                 asm( "jmp  comain" );
           s1: ;
+          cycle: ;
+        }
+}
 …
 \end{figure}
+\begin{figure}
+\centering
+\input{FullCoroutinePhases.pstex_t}
+\vspace*{-10pt}
+\caption{Symmetric coroutine steps: Ping / Pong}
+\label{f:PingPongFullCoroutineSteps}
+\end{figure}
+Figure~\ref{f:CPingPongSim} shows the C implementation of the \CFA symmetric generator, where there is still only one additional field, @restart@, but @resume@ is more complex because it does a forward rather than backward jump.
+Before the jump, the parameter for the next call @partner@ is placed into the register used for the first parameter, @rdi@, and the remaining registers are reset for a return.
+The @jmp comain@ restarts the function but with a different parameter, so the new call's behaviour depends on the state of the coroutine type, i.e., branch to restart location with different data state.
+While the semantics of call forward is a tail-call optimization, which compilers perform, the generator state is different on each call rather a common state for a tail-recursive function (i.e., the parameter to the function never changes during the forward calls.
+However, this assembler code depends on what entry code is generated, specifically if there are local variables and the level of optimization.
+Hence, internal compiler support is necessary for any forward call (or backwards return), \eg LLVM has various coroutine support~\cite{CoroutineTS}, and \CFA can leverage this support should it eventually fork @clang@.
+For this reason, \CFA does not support general symmetric generators at this time, but, it is possible to hand generate any symmetric generators (as in Figure~\ref{f:CPingPongSim}) for proof of concept and performance testing.
+Finally, part of this generator work was inspired by the recent \CCtwenty coroutine proposal~\cite{C++20Coroutine19}, which uses the general term coroutine to mean generator.
+Finally, part of this generator work was inspired by the recent \CCtwenty generator proposal~\cite{C++20Coroutine19} (which they call coroutines).
 Our work provides the same high-performance asymmetric generators as \CCtwenty, and extends their work with symmetric generators.
 An additional \CCtwenty generator feature allows @suspend@ and @resume@ to be followed by a restricted compound statement that is executed after the current generator has reset its stack but before calling the next generator, specified with \CFA syntax:
 …
 \label{s:Coroutine}
 Stackful coroutines (Table~\ref{t:ExecutionPropertyComposition} case 5) extend generator semantics, \ie there is an implicit closure and @suspend@ may appear in a helper function called from the coroutine main.
+Stackful coroutines extend generator semantics, \ie there is an implicit closure and @suspend@ may appear in a helper function called from the coroutine main.
 A coroutine is specified by replacing @generator@ with @coroutine@ for the type.
 Coroutine generality results in higher cost for creation, due to dynamic stack allocation, for execution, due to context switching among stacks, and for terminating, due to possible stack unwinding and dynamic stack deallocation.
+Coroutine generality results in higher cost for creation, due to dynamic stack allocation, execution, due to context switching among stacks, and terminating, due to possible stack unwinding and dynamic stack deallocation.
 A series of different kinds of coroutines and their implementations demonstrate how coroutines extend generators.
 First, the previous generator examples are converted to their coroutine counterparts, allowing local-state variables to be moved from the generator type into the coroutine main.
+\begin{center}
+\begin{tabular}{@{}l|l|l|l@{}}
+\multicolumn{1}{c|}{Fibonacci} & \multicolumn{1}{c|}{Formatter} & \multicolumn{1}{c|}{Device Driver} & \multicolumn{1}{c}{PingPong} \\
+\hline
+\begin{description}
+\item[Fibonacci]
+Move the declaration of @fn1@ to the start of coroutine main.
 \begin{cfa}[xleftmargin=0pt]
 void main( Fib & fib ) ...
+void main( Fib & fib ) with(fib) {
         `int fn1;`
+\end{cfa}
+&
+\end{cfa}
+\item[Formatter]
+Move the declaration of @g@ and @b@ to the for loops in the coroutine main.
 \begin{cfa}[xleftmargin=0pt]
 for ( `g`; 5 ) {
         for ( `b`; 4 ) {
+\end{cfa}
+&
+\end{cfa}
+\item[Device Driver]
+Move the declaration of @lnth@ and @sum@ to their points of initialization.
 \begin{cfa}[xleftmargin=0pt]
+status = CONT;
+`int lnth = 0, sum = 0;`
+...
+`short int crc = byte << 8;`
+\end{cfa}
+&
+        status = CONT;
+        `unsigned int lnth = 0, sum = 0;`
+        ...
+        `unsigned short int crc = byte << 8;`
+\end{cfa}
+\item[PingPong]
+Move the declaration of @i@ to the for loop in the coroutine main.
 \begin{cfa}[xleftmargin=0pt]
 void main( PingPong & pp ) ...
+void main( PingPong & pp ) with(pp) {
         for ( `i`; N ) {
+\end{cfa}
+\end{tabular}
+\end{center}
+\end{cfa}
+\end{description}
 It is also possible to refactor code containing local-state and @suspend@ statements into a helper function, like the computation of the CRC for the device driver.
 \begin{cfa}
 int Crc() {
+unsigned int Crc() {
         `suspend;`
         short int crc = byte << 8;
+        unsigned short int crc = byte << 8;
         `suspend;`
         status = (crc | byte) == sum ? MSG : ECRC;
 …
 \begin{comment}
 Figure~\ref{f:Coroutine3States} creates a @coroutine@ type, @`coroutine` Fib { int fn; }@, which provides communication, @fn@, for the \newterm{coroutine main}, @main@, which runs on the coroutine stack, and possibly multiple interface functions, \eg @restart@.
+Figure~\ref{f:Coroutine3States} creates a @coroutine@ type, @`coroutine` Fib { int fn; }@, which provides communication, @fn@, for the \newterm{coroutine main}, @main@, which runs on the coroutine stack, and possibly multiple interface functions, \eg @next@.
 Like the structure in Figure~\ref{f:ExternalState}, the coroutine type allows multiple instances, where instances of this type are passed to the (overloaded) coroutine main.
 The coroutine main's stack holds the state for the next generation, @f1@ and @f2@, and the code represents the three states in the Fibonacci formula via the three suspend points, to context switch back to the caller's @resume@.
 The interface function @restart@, takes a Fibonacci instance and context switches to it using @resume@;
+The interface function @next@, takes a Fibonacci instance and context switches to it using @resume@;
 on restart, the Fibonacci field, @fn@, contains the next value in the sequence, which is returned.
 The first @resume@ is special because it allocates the coroutine stack and cocalls its coroutine main on that stack;
 …
 \begin{figure}
 \centering
+\lstset{language=CFA,escapechar={},moredelim=**[is][\protect\color{red}]{`}{`}}% allow $
 \begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
 \begin{cfa}
 `coroutine` Prod {
         Cons & c;                       $\C[1.5in]{// communication}$
+        Cons & c;                       // communication
         int N, money, receipt;
 };
 void main( Prod & prod ) with( prod ) {
+        for ( i; N ) {          $\C{// 1st resume}\CRT$
+        // 1st resume starts here
+        for ( i; N ) {
                 int p1 = random( 100 ), p2 = random( 100 );
+                sout | p1 | " " | p2;
                 int status = delivery( c, p1, p2 );
+                sout | " $" | money | nl | status;
                 receipt += 1;
+        }
         stop( c );
+        sout | "prod stops";
+}
 int payment( Prod & prod, int money ) {
 …
 \begin{cfa}
 `coroutine` Cons {
         Prod & p;                       $\C[1.5in]{// communication}$
+        Prod & p;                       // communication
         int p1, p2, status;
         bool done;
 };
 void ?{}( Cons & cons, Prod & p ) {
         &cons.p = &p;           $\C{// reassignable reference}$
+        &cons.p = &p; // reassignable reference
         cons.[status, done ] = [0, false];
+}
 void main( Cons & cons ) with( cons ) {
+        int money = 1, receipt; $\C{// 1st resume}\CRT$
+        // 1st resume starts here
+        int money = 1, receipt;
         for ( ; ! done; ) {
+                sout | p1 | " " | p2 | nl | " $" | money;
                 status += 1;
                 receipt = payment( p, money );
+                sout | " #" | receipt;
                 money += 1;
+        }
+        sout | "cons stops";
+}
 int delivery( Cons & cons, int p1, int p2 ) {
 …
 This example is illustrative because both producer/consumer have two interface functions with @resume@s that suspend execution in these interface (helper) functions.
 The program main creates the producer coroutine, passes it to the consumer coroutine in its initialization, and closes the cycle at the call to @start@ along with the number of items to be produced.
+The call to @start@ is the first @resume@ of @prod@, which remembers the program main as the starter and creates @prod@'s stack with a frame for @prod@'s coroutine main at the top, and context switches to it.
+@prod@'s coroutine main starts, creates local-state variables that are retained between coroutine activations, and executes $N$ iterations, each generating two random values, calling the consumer's @deliver@ function to transfer the values, and printing the status returned from the consumer.
+The first @resume@ of @prod@ creates @prod@'s stack with a frame for @prod@'s coroutine main at the top, and context switches to it.
+@prod@'s coroutine main starts, creates local-state variables that are retained between coroutine activations, and executes $N$ iterations, each generating two random values, calling the consumer to deliver the values, and printing the status returned from the consumer.
 The producer call to @delivery@ transfers values into the consumer's communication variables, resumes the consumer, and returns the consumer status.
+Similarly on the first resume, @cons@'s stack is created and initialized, holding local-state variables retained between subsequent activations of the coroutine.
+The symmetric coroutine cycle forms when the consumer calls the producer's @payment@ function, which resumes the producer in the consumer's delivery function.
+When the producer calls @delivery@ again, it resumes the consumer in the @payment@ function.
+Both interface function than return to the their corresponding coroutine-main functions for the next cycle.
+On the first resume, @cons@'s stack is created and initialized, holding local-state variables retained between subsequent activations of the coroutine.
+The consumer iterates until the @done@ flag is set, prints the values delivered by the producer, increments status, and calls back to the producer via @payment@, and on return from @payment@, prints the receipt from the producer and increments @money@ (inflation).
+The call from the consumer to @payment@ introduces the cycle between producer and consumer.
+When @payment@ is called, the consumer copies values into the producer's communication variable and a resume is executed.
+The context switch restarts the producer at the point where it last context switched, so it continues in @delivery@ after the resume.
+@delivery@ returns the status value in @prod@'s coroutine main, where the status is printed.
+The loop then repeats calling @delivery@, where each call resumes the consumer coroutine.
+The context switch to the consumer continues in @payment@.
+The consumer increments and returns the receipt to the call in @cons@'s coroutine main.
+The loop then repeats calling @payment@, where each call resumes the producer coroutine.
 Figure~\ref{f:ProdConsRuntimeStacks} shows the runtime stacks of the program main, and the coroutine mains for @prod@ and @cons@ during the cycling.
-As a consequence of a coroutine retaining its last resumer for suspending back, these reverse pointers allow @suspend@ to cycle \emph{backwards} around a symmetric coroutine cycle.
 \begin{figure}
 …
 \caption{Producer / consumer runtime stacks}
 \label{f:ProdConsRuntimeStacks}
+\medskip
+\begin{center}
+\input{FullCoroutinePhases.pstex_t}
+\end{center}
+\vspace*{-10pt}
+\caption{Ping / Pong coroutine steps}
+\label{f:PingPongFullCoroutineSteps}
 \end{figure}
 Terminating a coroutine cycle is more complex than a generator cycle, because it requires context switching to the program main's \emph{stack} to shutdown the program, whereas generators started by the program main run on its stack.
+Furthermore, each deallocated coroutine must execute all destructors for object allocated in the coroutine type \emph{and} allocated on the coroutine's stack at the point of suspension, which can be arbitrarily deep.
+In the example, termination begins with the producer's loop stopping after N iterations and calling the consumer's @stop@ function, which sets the @done@ flag, resumes the consumer in function @payment@, terminating the call, and the consumer's loop in its coroutine main.
+% (Not shown is having @prod@ raise a nonlocal @stop@ exception at @cons@ after it finishes generating values and suspend back to @cons@, which catches the @stop@ exception to terminate its loop.)
+When the consumer's main ends, its stack is already unwound so any stack allocated objects with destructors are finalized.
+The question now is where does control continue?
+Furthermore, each deallocated coroutine must guarantee all destructors are run for object allocated in the coroutine type \emph{and} allocated on the coroutine's stack at the point of suspension, which can be arbitrarily deep.
+When a coroutine's main ends, its stack is already unwound so any stack allocated objects with destructors have been finalized.
 The na\"{i}ve semantics for coroutine-cycle termination is to context switch to the last resumer, like executing a @suspend@/@return@ in a generator.
 However, for coroutines, the last resumer is \emph{not} implicitly below the current stack frame, as for generators, because each coroutine's stack is independent.
 Unfortunately, it is impossible to determine statically if a coroutine is in a cycle and unrealistic to check dynamically (graph-cycle problem).
 Hence, a compromise solution is necessary that works for asymmetric (acyclic) and symmetric (cyclic) coroutines.
+Our solution is to retain a coroutine's starter (first resumer), and context switch back to the starter when the coroutine ends.
 Hence, the consumer restarts its first resumer, @prod@, in @stop@, and when the producer ends, it restarts its first resumer, program main, in @start@ (see dashed lines from the end of the coroutine mains in Figure~\ref{f:ProdConsRuntimeStacks}).
+Our solution is to context switch back to the first resumer (starter) once the coroutine ends.
 This semantics works well for the most common asymmetric and symmetric coroutine usage patterns.
+For asymmetric coroutines, it is common for the first resumer (starter) coroutine to be the only resumer;
+for symmetric coroutines, it is common for the cycle creator to persist for the lifetime of the cycle.
+For asymmetric coroutines, it is common for the first resumer (starter) coroutine to be the only resumer.
+All previous generators converted to coroutines have this property.
+For symmetric coroutines, it is common for the cycle creator to persist for the lifetime of the cycle.
+Hence, the starter coroutine is remembered on the first resume and ending the coroutine resumes the starter.
+Figure~\ref{f:ProdConsRuntimeStacks} shows this semantic by the dashed lines from the end of the coroutine mains: @prod@ starts @cons@ so @cons@ resumes @prod@ at the end, and the program main starts @prod@ so @prod@ resumes the program main at the end.
 For other scenarios, it is always possible to devise a solution with additional programming effort, such as forcing the cycle forward (backward) to a safe point before starting termination.
+Note, the producer/consumer example does not illustrate the full power of the starter semantics because @cons@ always ends first.
+Assume generator @PingPong@ in Figure~\ref{f:PingPongSymmetricGenerator} is converted to a coroutine.
+Unlike generators, coroutines have a starter structure with multiple levels, where the program main starts @ping@ and @ping@ starts @pong@.
+By adjusting $N$ for either @ping@/@pong@, it is possible to have either finish first.
+If @pong@ ends first, it resumes its starter @ping@ in its coroutine main, then @ping@ ends and resumes its starter the program main on return;
+if @ping@ ends first, it resumes its starter the program main on return.
+Regardless of the cycle complexity, the starter structure always leads back to the program main, but the path can be entered at an arbitrary point.
+Once back at the program main (creator), coroutines @ping@ and @pong@ are deallocated, runnning any destructors for objects within the coroutine and possibly deallocating any coroutine stacks for non-terminated coroutines, where stack deallocation implies stack unwinding to find destructors for allocated objects on the stack.
+Hence, the \CFA termination semantics for the generator and coroutine ensure correct deallocation semnatics, regardless of the coroutine's state (terminated or active), like any other aggregate object.
+The producer/consumer example does not illustrate the full power of the starter semantics because @cons@ always ends first.
+Assume generator @PingPong@ is converted to a coroutine.
+Figure~\ref{f:PingPongFullCoroutineSteps} shows the creation, starter, and cyclic execution steps of the coroutine version.
+The program main creates (declares) coroutine instances @ping@ and @pong@.
+Next, program main resumes @ping@, making it @ping@'s starter, and @ping@'s main resumes @pong@'s main, making it @pong@'s starter.
+Execution forms a cycle when @pong@ resumes @ping@, and cycles $N$ times.
+By adjusting $N$ for either @ping@/@pong@, it is possible to have either one finish first, instead of @pong@ always ending first.
+If @pong@ ends first, it resumes its starter @ping@ in its coroutine main, then @ping@ ends and resumes its starter the program main in function @start@.
+If @ping@ ends first, it resumes its starter the program main in function @start@.
+Regardless of the cycle complexity, the starter stack always leads back to the program main, but the stack can be entered at an arbitrary point.
+Once back at the program main, coroutines @ping@ and @pong@ are deallocated.
+For generators, deallocation runs the destructors for all objects in the generator type.
+For coroutines, deallocation deals with objects in the coroutine type and must also run the destructors for any objects pending on the coroutine's stack for any unterminated coroutine.
+Hence, if a coroutine's destructor detects the coroutine is not ended, it implicitly raises a cancellation exception (uncatchable exception) at the coroutine and resumes it so the cancellation exception can propagate to the root of the coroutine's stack destroying all local variable on the stack.
+So the \CFA semantics for the generator and coroutine, ensure both can be safely deallocated at any time, regardless of their current state, like any other aggregate object.
+Explicitly raising normal exceptions at another coroutine can replace flag variables, like @stop@, \eg @prod@ raises a @stop@ exception at @cons@ after it finishes generating values and resumes @cons@, which catches the @stop@ exception to terminate its loop.
+Finally, there is an interesting effect for @suspend@ with symmetric coroutines.
+A coroutine must retain its last resumer to suspend back because the resumer is on a different stack.
+These reverse pointers allow @suspend@ to cycle \emph{backwards}, which may be useful in certain cases.
+However, there is an anomaly if a coroutine resumes itself, because it overwrites its last resumer with itself, losing the ability to resume the last external resumer.
+To prevent losing this information, a self-resume does not overwrite the last resumer.
 …
 Users wanting to extend custom types or build their own can only do so in ways offered by the language.
 Furthermore, implementing custom types without language support may display the power of a programming language.
 \CFA blends the two approaches, providing custom type for idiomatic \CFA code, while extending and building new custom types is still possible, similar to Java concurrency with builtin and library (@java.util.concurrent@) monitors.
+\CFA blends the two approaches, providing custom type for idiomatic \CFA code, while extending and building new custom types is still possible, similar to Java concurrency with builtin and library.
 Part of the mechanism to generalize custom types is the \CFA trait~\cite[\S~2.3]{Moss18}, \eg the definition for custom-type @coroutine@ is anything satisfying the trait @is_coroutine@, and this trait both enforces and restricts the coroutine-interface functions.
 …
 forall( `dtype` T | is_coroutine(T) ) void $suspend$( T & ), resume( T & );
 \end{cfa}
+Note, copying generators/coroutines/threads is undefined because muliple objects cannot execute on a shared stack and stack copying does not work in unmanaged languages (no garbage collection), like C, because the stack may contain pointers to objects within it that require updating for the copy.
+Note, copying generators/coroutines/threads is not meaningful.
+For example, both the resumer and suspender descriptors can have bidirectional pointers;
+copying these coroutines does not update the internal pointers so behaviour of both copies would be difficult to understand.
+Furthermore, two coroutines cannot logically execute on the same stack.
+A deep coroutine copy, which copies the stack, is also meaningless in an unmanaged language (no garbage collection), like C, because the stack may contain pointers to object within it that require updating for the copy.
 The \CFA @dtype@ property provides no \emph{implicit} copying operations and the @is_coroutine@ trait provides no \emph{explicit} copying operations, so all coroutines must be passed by reference (pointer).
 The function definitions ensure there is a statically typed @main@ function that is the starting point (first stack frame) of a coroutine, and a mechanism to get (read) the coroutine descriptor from its handle.
 …
 The combination of custom types and fundamental @trait@ description of these types allows a concise specification for programmers and tools, while more advanced programmers can have tighter control over memory layout and initialization.
 Figure~\ref{f:CoroutineMemoryLayout} shows different memory-layout options for a coroutine (where a thread is similar).
+Figure~\ref{f:CoroutineMemoryLayout} shows different memory-layout options for a coroutine (where a task is similar).
 The coroutine handle is the @coroutine@ instance containing programmer specified type global/communication variables across interface functions.
 The coroutine descriptor contains all implicit declarations needed by the runtime, \eg @suspend@/@resume@, and can be part of the coroutine handle or separate.
 The coroutine stack can appear in a number of locations and be fixed or variable sized.
 Hence, the coroutine's stack could be a variable-length structure (VLS)\footnote{
 We are examining VLSs, where fields can be variable-sized structures or arrays.
+Hence, the coroutine's stack could be a VLS\footnote{
+We are examining variable-sized structures (VLS), where fields can be variable-sized structures or arrays.
 Once allocated, a VLS is fixed sized.}
 on the allocating stack, provided the allocating stack is large enough.
 For a VLS stack allocation/deallocation is an inexpensive adjustment of the stack pointer, modulo any stack constructor costs (\eg initial frame setup).
 For stack allocation in the heap, allocation/deallocation is an expensive allocation, where the heap can be a shared resource, modulo any stack constructor costs.
 It is also possible to use a split (segmented) stack calling convention, available with gcc and clang, allowing a variable-sized stack via a set of connected blocks in the heap.
+For heap stack allocation, allocation/deallocation is an expensive heap allocation (where the heap can be a shared resource), modulo any stack constructor costs.
+With heap stack allocation, it is also possible to use a split (segmented) stack calling convention, available with gcc and clang, so the stack is variable sized.
 Currently, \CFA supports stack/heap allocated descriptors but only fixed-sized heap allocated stacks.
 In \CFA debug-mode, the fixed-sized stack is terminated with a write-only page, which catches most stack overflows.
 Experience teaching concurrency with \uC~\cite{CS343} shows fixed-sized stacks are rarely an issue for students.
 Split-stack allocation is under development but requires recompilation of legacy code, which is not always possible.
+Split-stack allocation is under development but requires recompilation of legacy code, which may be impossible.
 \begin{figure}
 …
 Concurrency is nondeterministic scheduling of independent sequential execution paths (threads), where each thread has its own stack.
 A single thread with multiple stacks, \ie coroutining, does \emph{not} imply concurrency~\cite[\S~3]{Buhr05a}.
 Coroutining self-schedule the thread across stacks so execution is deterministic.
+A single thread with multiple call stacks, \newterm{coroutining}~\cite{Conway63,Marlin80}, does \emph{not} imply concurrency~\cite[\S~2]{Buhr05a}.
+In coroutining, coroutines self-schedule the thread across stacks so execution is deterministic.
 (It is \emph{impossible} to generate a concurrency error when coroutining.)
+The transition to concurrency, even for a single thread with multiple stacks, occurs when coroutines context switch to a \newterm{scheduling coroutine}, introducing non-determinism from the coroutine perspective~\cite[\S~3]{Buhr05a}.
+However, coroutines are a stepping stone towards concurrency.
+The transition to concurrency, even for a single thread with multiple stacks, occurs when coroutines context switch to a \newterm{scheduling coroutine}, introducing non-determinism from the coroutine perspective~\cite[\S~3,]{Buhr05a}.
 Therefore, a minimal concurrency system requires coroutines \emph{in conjunction with a nondeterministic scheduler}.
+The resulting execution system now follows a cooperative threading-model~\cite{Adya02,libdill} because context-switching points to the scheduler (blocking) are known, but the next unblocking point is unknown due to the scheduler.
+Adding \newterm{preemption} introduces \newterm{non-cooperative} or \newterm{preemptive} scheduling, where context switching points to the scheduler are unknown as they can occur randomly between any two instructions often based on a timer interrupt.
+The resulting execution system now follows a cooperative threading model~\cite{Adya02,libdill}, called \newterm{non-preemptive scheduling}.
+Adding \newterm{preemption} introduces non-cooperative scheduling, where context switching occurs randomly between any two instructions often based on a timer interrupt, called \newterm{preemptive scheduling}.
+While a scheduler introduces uncertain execution among explicit context switches, preemption introduces uncertainty by introducing implicit context switches.
 Uncertainty gives the illusion of parallelism on a single processor and provides a mechanism to access and increase performance on multiple processors.
 The reason is that the scheduler/runtime have complete knowledge about resources and how to best utilized them.
 However, the introduction of unrestricted nondeterminism results in the need for \newterm{mutual exclusion} and \newterm{synchronization}~\cite[\S~4]{Buhr05a}, which restrict nondeterminism for correctness;
+However, the introduction of unrestricted nondeterminism results in the need for \newterm{mutual exclusion} and \newterm{synchronization}, which restrict nondeterminism for correctness;
 otherwise, it is impossible to write meaningful concurrent programs.
 Optimal concurrent performance is often obtained by having as much nondeterminism as mutual exclusion and synchronization correctness allow.
 A scheduler can also be stackless or stackful.
+A scheduler can either be a stackless or stackful.
 For stackless, the scheduler performs scheduling on the stack of the current coroutine and switches directly to the next coroutine, so there is one context switch.
 For stackful, the current coroutine switches to the scheduler, which performs scheduling, and it then switches to the next coroutine, so there are two context switches.
 …
 \label{s:threads}
 Threading (Table~\ref{t:ExecutionPropertyComposition} case 11) needs the ability to start a thread and wait for its completion.
+Threading needs the ability to start a thread and wait for its completion.
 A common API for this ability is @fork@ and @join@.
+\vspace{4pt}
+\par\noindent
+\begin{tabular}{@{}l|l|l@{}}
+\multicolumn{1}{c|}{\textbf{Java}} & \multicolumn{1}{c|}{\textbf{\Celeven}} & \multicolumn{1}{c}{\textbf{pthreads}} \\
+\hline
+\begin{cfa}
+class MyThread extends Thread {...}
+mythread t = new MyThread(...);
+\begin{cquote}
+\begin{tabular}{@{}lll@{}}
+\multicolumn{1}{c}{\textbf{Java}} & \multicolumn{1}{c}{\textbf{\Celeven}} & \multicolumn{1}{c}{\textbf{pthreads}} \\
+\begin{cfa}
+class MyTask extends Thread {...}
+mytask t = new MyTask(...);
 `t.start();` // start
 // concurrency
 …
+&
 \begin{cfa}
 class MyThread { ... } // functor
 MyThread mythread;
 `thread t( mythread, ... );` // start
+class MyTask { ... } // functor
+MyTask mytask;
+`thread t( mytask, ... );` // start
 // concurrency
 `t.join();` // wait
 …
 \end{cfa}
 \end{tabular}
+\vspace{1pt}
+\par\noindent
+\end{cquote}
 \CFA has a simpler approach using a custom @thread@ type and leveraging declaration semantics (allocation/deallocation), where threads implicitly @fork@ after construction and @join@ before destruction.
 \begin{cfa}
 thread MyThread {};
 void main( MyThread & this ) { ... }
+thread MyTask {};
+void main( MyTask & this ) { ... }
 int main() {
         MyThread team`[10]`; $\C[2.5in]{// allocate stack-based threads, implicit start after construction}$
+        MyTask team`[10]`; $\C[2.5in]{// allocate stack-based threads, implicit start after construction}$
         // concurrency
 } $\C{// deallocate stack-based threads, implicit joins before destruction}$
 …
 Arbitrary topologies are possible using dynamic allocation, allowing threads to outlive their declaration scope, identical to normal dynamic allocation.
 \begin{cfa}
 MyThread * factory( int N ) { ... return `anew( N )`; } $\C{// allocate heap-based threads, implicit start after construction}$
+MyTask * factory( int N ) { ... return `anew( N )`; } $\C{// allocate heap-based threads, implicit start after construction}$
 int main() {
         MyThread * team = factory( 10 );
+        MyTask * team = factory( 10 );
         // concurrency
         `delete( team );` $\C{// deallocate heap-based threads, implicit joins before destruction}\CRT$
 …
 Threads in \CFA are user level run by runtime kernel threads (see Section~\ref{s:CFARuntimeStructure}), where user threads provide concurrency and kernel threads provide parallelism.
 Like coroutines, and for the same design reasons, \CFA provides a custom @thread@ type and a @trait@ to enforce and restrict the thread-interface functions.
+Like coroutines, and for the same design reasons, \CFA provides a custom @thread@ type and a @trait@ to enforce and restrict the task-interface functions.
 \begin{cquote}
 \begin{tabular}{@{}c@{\hspace{3\parindentlnth}}c@{}}
 …
 \label{s:MutualExclusionSynchronization}
 Unrestricted nondeterminism is meaningless as there is no way to know when a result is completed and safe to access.
+Unrestricted nondeterminism is meaningless as there is no way to know when the result is completed without synchronization.
 To produce meaningful execution requires clawing back some determinism using mutual exclusion and synchronization, where mutual exclusion provides access control for threads using shared data, and synchronization is a timing relationship among threads~\cite[\S~4]{Buhr05a}.
+The shared data protected by mutual exlusion is called a \newterm{critical section}~\cite{Dijkstra65}, and the protection can be simple (only 1 thread) or complex (only N kinds of threads, \eg group~\cite{Joung00} or readers/writer~\cite{Courtois71}).
+Without synchronization control in a critical section, an arriving thread can barge ahead of preexisting waiter threads resulting in short/long-term starvation, staleness/freshness problems, and/or incorrect transfer of data.
+Preventing or detecting barging is a challenge with low-level locks, but made easier through higher-level constructs.
+This challenge is often split into two different approaches: barging \emph{avoidance} and \emph{prevention}.
+Approaches that unconditionally releasing a lock for competing threads to acquire must use barging avoidance with flag/counter variable(s) to force barging threads to wait;
+approaches that conditionally hold locks during synchronization, \eg baton-passing~\cite{Andrews89}, prevent barging completely.
+At the lowest level, concurrent control is provided by atomic operations, upon which different kinds of locking mechanisms are constructed, \eg spin locks, semaphores~\cite{Dijkstra68b}, barriers, and path expressions~\cite{Campbell74}.
+However, for productivity it is always desirable to use the highest-level construct that provides the necessary efficiency~\cite{Hochstein05}.
+A significant challenge with locks is composability because it takes careful organization for multiple locks to be used while preventing deadlock.
+Easing composability is another feature higher-level mutual-exclusion mechanisms can offer.
+Some concurrent systems eliminate mutable shared-state by switching to non-shared communication like message passing~\cite{Thoth,Harmony,V-Kernel,MPI} (Erlang, MPI), channels~\cite{CSP} (CSP,Go), actors~\cite{Akka} (Akka, Scala), or functional techniques (Haskell).
+Some concurrent systems eliminate mutable shared-state by switching to stateless communication like message passing~\cite{Thoth,Harmony,V-Kernel,MPI} (Erlang, MPI), channels~\cite{CSP} (CSP,Go), actors~\cite{Akka} (Akka, Scala), or functional techniques (Haskell).
 However, these approaches introduce a new communication mechanism for concurrency different from the standard communication using function call/return.
 Hence, a programmer must learn and manipulate two sets of design/programming patterns.
 While this distinction can be hidden away in library code, effective use of the library still has to take both paradigms into account.
+In contrast, approaches based on shared-state models more closely resemble the standard call/return programming model, resulting in a single programming paradigm.
+Finally, a newer approach for restricting non-determinism is transactional memory~\cite{Herlihy93}.
+While this approach is pursued in hardware~\cite{Nakaike15} and system languages, like \CC~\cite{Cpp-Transactions}, the performance and feature set is still too restrictive~\cite{Cascaval08,Boehm09} to be the main concurrency paradigm for system languages.
+In contrast, approaches based on stateful models more closely resemble the standard call/return programming model, resulting in a single programming paradigm.
+At the lowest level, concurrent control is implemented by atomic operations, upon which different kinds of locking mechanisms are constructed, \eg semaphores~\cite{Dijkstra68b}, barriers, and path expressions~\cite{Campbell74}.
+However, for productivity it is always desirable to use the highest-level construct that provides the necessary efficiency~\cite{Hochstein05}.
+A newer approach for restricting non-determinism is transactional memory~\cite{Herlihy93}.
+While this approach is pursued in hardware~\cite{Nakaike15} and system languages, like \CC~\cite{Cpp-Transactions}, the performance and feature set is still too restrictive to be the main concurrency paradigm for system languages, which is why it is rejected as the core paradigm for concurrency in \CFA.
+One of the most natural, elegant, and efficient mechanisms for mutual exclusion and synchronization for shared-memory systems is the \emph{monitor}.
+First proposed by Brinch Hansen~\cite{Hansen73} and later described and extended by C.A.R.~Hoare~\cite{Hoare74}, many concurrent programming languages provide monitors as an explicit language construct: \eg Concurrent Pascal~\cite{ConcurrentPascal}, Mesa~\cite{Mesa}, Modula~\cite{Modula-2}, Turing~\cite{Turing:old}, Modula-3~\cite{Modula-3}, NeWS~\cite{NeWS}, Emerald~\cite{Emerald}, \uC~\cite{Buhr92a} and Java~\cite{Java}.
+In addition, operating-system kernels and device drivers have a monitor-like structure, although they often use lower-level primitives such as mutex locks or semaphores to simulate monitors.
+For these reasons, \CFA selected monitors as the core high-level concurrency construct, upon which higher-level approaches can be easily constructed.
+\subsection{Mutual Exclusion}
+A group of instructions manipulating a specific instance of shared data that must be performed atomically is called a \newterm{critical section}~\cite{Dijkstra65}, which is enforced by \newterm{simple mutual-exclusion}.
+The generalization is called a \newterm{group critical-section}~\cite{Joung00}, where multiple tasks with the same session use the resource simultaneously and different sessions are segregated, which is enforced by \newterm{complex mutual-exclusion} providing the correct kind and number of threads using a group critical-section.
+The readers/writer problem~\cite{Courtois71} is an instance of a group critical-section, where readers share a session but writers have a unique session.
+However, many solutions exist for mutual exclusion, which vary in terms of performance, flexibility and ease of use.
+Methods range from low-level locks, which are fast and flexible but require significant attention for correctness, to higher-level concurrency techniques, which sacrifice some performance to improve ease of use.
+Ease of use comes by either guaranteeing some problems cannot occur, \eg deadlock free, or by offering a more explicit coupling between shared data and critical section.
+For example, the \CC @std::atomic<T>@ offers an easy way to express mutual-exclusion on a restricted set of operations, \eg reading/writing, for numerical types.
+However, a significant challenge with locks is composability because it takes careful organization for multiple locks to be used while preventing deadlock.
+Easing composability is another feature higher-level mutual-exclusion mechanisms can offer.
+\subsection{Synchronization}
+Synchronization enforces relative ordering of execution, and synchronization tools provide numerous mechanisms to establish these timing relationships.
+Low-level synchronization primitives offer good performance and flexibility at the cost of ease of use;
+higher-level mechanisms often simplify usage by adding better coupling between synchronization and data, \eg receive-specific versus receive-any thread in message passing or offering specialized solutions, \eg barrier lock.
+Often synchronization is used to order access to a critical section, \eg ensuring a waiting writer thread enters the critical section before a calling reader thread.
+If the calling reader is scheduled before the waiting writer, the reader has barged.
+Barging can result in staleness/freshness problems, where a reader barges ahead of a writer and reads temporally stale data, or a writer barges ahead of another writer overwriting data with a fresh value preventing the previous value from ever being read (lost computation).
+Preventing or detecting barging is an involved challenge with low-level locks, which is made easier through higher-level constructs.
+This challenge is often split into two different approaches: barging avoidance and prevention.
+Algorithms that unconditionally releasing a lock for competing threads to acquire use barging avoidance during synchronization to force a barging thread to wait;
+algorithms that conditionally hold locks during synchronization, \eg baton-passing~\cite{Andrews89}, prevent barging completely.
 …
 \label{s:Monitor}
+One of the most natural, elegant, efficient, high-level mechanisms for mutual exclusion and synchronization for shared-memory systems is the \emph{monitor} (Table~\ref{t:ExecutionPropertyComposition} case 2).
+First proposed by Brinch Hansen~\cite{Hansen73} and later described and extended by C.A.R.~Hoare~\cite{Hoare74}, many concurrent programming languages provide monitors as an explicit language construct: \eg Concurrent Pascal~\cite{ConcurrentPascal}, Mesa~\cite{Mesa}, Modula~\cite{Modula-2}, Turing~\cite{Turing:old}, Modula-3~\cite{Modula-3}, NeWS~\cite{NeWS}, Emerald~\cite{Emerald}, \uC~\cite{Buhr92a} and Java~\cite{Java}.
+In addition, operating-system kernels and device drivers have a monitor-like structure, although they often use lower-level primitives such as mutex locks or semaphores to manually implement a monitor.
+For these reasons, \CFA selected monitors as the core high-level concurrency construct, upon which higher-level approaches can be easily constructed.
+Specifically, a \textbf{monitor} is a set of functions that ensure mutual exclusion when accessing shared state.
+More precisely, a monitor is a programming technique that implicitly binds mutual exclusion to static function scope by call/return, as opposed to locks, where mutual-exclusion is defined by acquire/release calls, independent of lexical context (analogous to block and heap storage allocation).
+A \textbf{monitor} is a set of functions that ensure mutual exclusion when accessing shared state.
+More precisely, a monitor is a programming technique that implicitly binds mutual exclusion to static function scope, as opposed to locks, where mutual-exclusion is defined by acquire/release calls, independent of lexical context (analogous to block and heap storage allocation).
 Restricting acquire/release points eases programming, comprehension, and maintenance, at a slight cost in flexibility and efficiency.
 \CFA uses a custom @monitor@ type and leverages declaration semantics (deallocation) to protect active or waiting threads in a monitor.
 The following is a \CFA monitor implementation of an atomic counter.
 \begin{cfa}
+\begin{cfa}[morekeywords=nomutex]
 `monitor` Aint { int cnt; }; $\C[4.25in]{// atomic integer counter}$
+int ++?( Aint & `mutex` this ) with( this ) { return ++cnt; } $\C{// increment}$
+int ?=?( Aint & `mutex` lhs, int rhs ) with( lhs ) { cnt = rhs; } $\C{// conversions with int, mutex optional}\CRT$
+int ?=?( int & lhs, Aint & `mutex` rhs ) with( rhs ) { lhs = cnt; }
+\end{cfa}
+The operators use the parameter-only declaration type-qualifier @mutex@ to mark which parameters require locking during function execution to protect from race conditions.
+The assignment operators provide bidirectional conversion between an atomic and normal integer without accessing field @cnt@.
+(These operations only need @mutex@, if reading/writing the implementation type is not atomic.)
+The atomic counter is used without any explicit mutual-exclusion and provides thread-safe semantics.
+int ++?( Aint & `mutex`$\(_{opt}\)$ this ) with( this ) { return ++cnt; } $\C{// increment}$
+int ?=?( Aint & `mutex`$\(_{opt}\)$ lhs, int rhs ) with( lhs ) { cnt = rhs; } $\C{// conversions with int}\CRT$
+int ?=?( int & lhs, Aint & `mutex`$\(_{opt}\)$ rhs ) with( rhs ) { lhs = cnt; }
+\end{cfa}
+% The @Aint@ constructor, @?{}@, uses the \lstinline[morekeywords=nomutex]@nomutex@ qualifier indicating mutual exclusion is unnecessary during construction because an object is inaccessible (private) until after it is initialized.
+% (While a constructor may publish its address into a global variable, doing so generates a race-condition.)
+The prefix increment operation, @++?@, is normally @mutex@, indicating mutual exclusion is necessary during function execution, to protect the incrementing from race conditions, unless there is an atomic increment instruction for the implementation type.
+The assignment operators provide bidirectional conversion between an atomic and normal integer without accessing field @cnt@;
+these operations only need @mutex@, if reading/writing the implementation type is not atomic.
+The atomic counter is used without any explicit mutual-exclusion and provides thread-safe semantics, which is similar to the \CC template @std::atomic@.
 \begin{cfa}
 int i = 0, j = 0, k = 5;
 …
 i = x; j = y; k = z;
 \end{cfa}
-Note, like other concurrent programming languages, \CFA has specializations for the basic types using atomic instructions for performance and a general trait similar to the \CC template @std::atomic@.
 \CFA monitors have \newterm{multi-acquire} semantics so the thread in the monitor may acquire it multiple times without deadlock, allowing recursion and calling other interface functions.
-\newpage
 \begin{cfa}
 monitor M { ... } m;
 …
 \end{cfa}
 \CFA monitors also ensure the monitor lock is released regardless of how an acquiring function ends (normal or exceptional), and returning a shared variable is safe via copying before the lock is released.
+Similar safety is offered by \emph{explicit} opt-in disciplines like \CC RAII versus the monitor \emph{implicit} language-enforced safety guarantee ensuring no programmer usage errors.
+Similar safety is offered by \emph{explicit} mechanisms like \CC RAII;
+monitor \emph{implicit} safety ensures no programmer usage errors.
 Furthermore, RAII mechanisms cannot handle complex synchronization within a monitor, where the monitor lock may not be released on function exit because it is passed to an unblocking thread;
 RAII is purely a mutual-exclusion mechanism (see Section~\ref{s:Scheduling}).
 …
 \end{cquote}
 The @dtype@ property prevents \emph{implicit} copy operations and the @is_monitor@ trait provides no \emph{explicit} copy operations, so monitors must be passed by reference (pointer).
+% Copying a lock is insecure because it is possible to copy an open lock and then use the open copy when the original lock is closed to simultaneously access the shared data.
+% Copying a monitor is secure because both the lock and shared data are copies, but copying the shared data is meaningless because it no longer represents a unique entity.
 Similarly, the function definitions ensures there is a mechanism to get (read) the monitor descriptor from its handle, and a special destructor to prevent deallocation if a thread using the shared data.
 The custom monitor type also inserts any locks needed to implement the mutual exclusion semantics.
 …
 For example, a monitor may be passed through multiple helper functions before it is necessary to acquire the monitor's mutual exclusion.
+\CFA requires programmers to identify the kind of parameter with the @mutex@ keyword and uses no keyword to mean \lstinline[morekeywords=nomutex]@nomutex@, because @mutex@ parameters are rare and no keyword is the \emph{normal} parameter semantics.
+Hence, @mutex@ parameters are documentation, at the function and its prototype, to both programmer and compiler, without other redundant keywords.
+Furthermore, \CFA relies heavily on traits as an abstraction mechanism, so the @mutex@ qualifier prevents coincidentally matching of a monitor trait with a type that is not a monitor, similar to coincidental inheritance where a shape and playing card can both be drawable.
+The benefit of mandatory monitor qualifiers is self-documentation, but requiring both @mutex@ and \lstinline[morekeywords=nomutex]@nomutex@ for all monitor parameters is redundant.
+Instead, the semantics has one qualifier as the default and the other required.
+For example, make the safe @mutex@ qualifier the default because assuming \lstinline[morekeywords=nomutex]@nomutex@ may cause subtle errors.
+Alternatively, make the unsafe \lstinline[morekeywords=nomutex]@nomutex@ qualifier the default because it is the \emph{normal} parameter semantics while @mutex@ parameters are rare.
+Providing a default qualifier implies knowing whether a parameter is a monitor.
+Since \CFA relies heavily on traits as an abstraction mechanism, types can coincidentally match the monitor trait but not be a monitor, similar to inheritance where a shape and playing card can both be drawable.
+For this reason, \CFA requires programmers to identify the kind of parameter with the @mutex@ keyword and uses no keyword to mean \lstinline[morekeywords=nomutex]@nomutex@.
 The next semantic decision is establishing which parameter \emph{types} may be qualified with @mutex@.
 …
 Function @f3@ has a multiple object matrix, and @f4@ a multiple object data structure.
 While shown shortly, multiple object acquisition is possible, but the number of objects must be statically known.
 Therefore, \CFA only acquires one monitor per parameter with exactly one level of indirection, and exclude pointer types to unknown sized arrays.
+Therefore, \CFA only acquires one monitor per parameter with at most one level of indirection, excluding pointers as it is impossible to statically determine the size.
 For object-oriented monitors, \eg Java, calling a mutex member \emph{implicitly} acquires mutual exclusion of the receiver object, @`rec`.foo(...)@.
 …
 While object-oriented monitors can be extended with a mutex qualifier for multiple-monitor members, no prior example of this feature could be found.}
 called \newterm{bulk acquire}.
 \CFA guarantees bulk acquisition order is consistent across calls to @mutex@ functions using the same monitors as arguments, so acquiring multiple monitors in a bulk acquire is safe from deadlock.
+\CFA guarantees acquisition order is consistent across calls to @mutex@ functions using the same monitors as arguments, so acquiring multiple monitors is safe from deadlock.
 Figure~\ref{f:BankTransfer} shows a trivial solution to the bank transfer problem~\cite{BankTransfer}, where two resources must be locked simultaneously, using \CFA monitors with implicit locking and \CC with explicit locking.
 A \CFA programmer only has to manage when to acquire mutual exclusion;
 …
 void transfer( BankAccount & `mutex` my,
         BankAccount & `mutex` your, int me2you ) {
+        // bulk acquire
         deposit( my, -me2you ); // debit
         deposit( your, me2you ); // credit
 …
 void transfer( BankAccount & my,
                         BankAccount & your, int me2you ) {
         `scoped_lock lock( my.m, your.m );` // bulk acquire
+        `scoped_lock lock( my.m, your.m );`
         deposit( my, -me2you ); // debit
         deposit( your, me2you ); // credit
 …
 \end{figure}
 Users can still force the acquiring order by using or not using @mutex@.
+Users can still force the acquiring order by using @mutex@/\lstinline[morekeywords=nomutex]@nomutex@.
 \begin{cfa}
 void foo( M & mutex m1, M & mutex m2 ); $\C{// acquire m1 and m2}$
 void bar( M & mutex m1, M & m2 ) { $\C{// only acquire m1}$
+void bar( M & mutex m1, M & /* nomutex */ m2 ) { $\C{// acquire m1}$
         ... foo( m1, m2 ); ... $\C{// acquire m2}$
+}
 void baz( M & m1, M & mutex m2 ) { $\C{// only acquire m2}$
+void baz( M & /* nomutex */ m1, M & mutex m2 ) { $\C{// acquire m2}$
         ... foo( m1, m2 ); ... $\C{// acquire m1}$
+}
 …
 % There are many aspects of scheduling in a concurrency system, all related to resource utilization by waiting threads, \ie which thread gets the resource next.
 % Different forms of scheduling include access to processors by threads (see Section~\ref{s:RuntimeStructureCluster}), another is access to a shared resource by a lock or monitor.
+This section discusses scheduling for waiting threads eligible for monitor entry, \ie which user thread gets the shared resource next. (See Section~\ref{s:RuntimeStructureCluster} for scheduling kernel threads on virtual processors.)
+While monitor mutual-exclusion provides safe access to its shared data, the data may indicate a thread cannot proceed, \eg a bounded buffer may be full/\-empty so produce/consumer threads must block.
+Leaving the monitor and retrying (busy waiting) is impractical for high-level programming.
+Monitors eliminate busy waiting by providing synchronization within the monitor critical-section to schedule threads needing access to the shared data, where threads block versus spin.
+This section discusses monitor scheduling for waiting threads eligible for entry, \ie which thread gets the shared resource next. (See Section~\ref{s:RuntimeStructureCluster} for scheduling threads on virtual processors.)
+While monitor mutual-exclusion provides safe access to shared data, the monitor data may indicate that a thread accessing it cannot proceed, \eg a bounded buffer may be full/empty so produce/consumer threads must block.
+Leaving the monitor and trying again (busy waiting) is impractical for high-level programming.
+Monitors eliminate busy waiting by providing synchronization to schedule threads needing access to the shared data, where threads block versus spinning.
 Synchronization is generally achieved with internal~\cite{Hoare74} or external~\cite[\S~2.9.2]{uC++} scheduling.
+\newterm{Internal} (largely) schedules threads located \emph{inside} the monitor and is accomplished using condition variables with signal and wait.
+\newterm{External} (largely) schedules threads located \emph{outside} the monitor and is accomplished with the @waitfor@ statement.
+Note, internal scheduling has a small amount of external scheduling and vice versus, so the naming denotes where the majority of the block threads reside (inside or outside) for scheduling.
+For complex scheduling, the approaches can be combined, so there can be an equal number of threads waiting inside and outside.
+\CFA monitors do not allow calling threads to barge ahead of signalled threads (via barging prevention), which simplifies synchronization among threads in the monitor and increases correctness.
+A direct consequence of this semantics is that unblocked waiting threads are not required to recheck the waiting condition, \ie waits are not in a starvation-prone busy-loop as required by the signals-as-hints style with barging.
+Preventing barging comes directly from Hoare's semantics in the seminal paper on monitors~\cite[p.~550]{Hoare74}.
+\newterm{Internal scheduling} is characterized by each thread entering the monitor and making an individual decision about proceeding or blocking, while \newterm{external scheduling} is characterized by an entering thread making a decision about proceeding for itself and on behalf of other threads attempting entry.
+Finally, \CFA monitors do not allow calling threads to barge ahead of signalled threads, which simplifies synchronization among threads in the monitor and increases correctness.
+If barging is allowed, synchronization between a signaller and signallee is difficult, often requiring additional flags and multiple unblock/block cycles.
+In fact, signals-as-hints is completely opposite from that proposed by Hoare in the seminal paper on monitors~\cite[p.~550]{Hoare74}.
 % \begin{cquote}
 % However, we decree that a signal operation be followed immediately by resumption of a waiting program, without possibility of an intervening procedure call from yet a third program.
 % It is only in this way that a waiting program has an absolute guarantee that it can acquire the resource just released by the signalling program without any danger that a third program will interpose a monitor entry and seize the resource instead.~\cite[p.~550]{Hoare74}
 % \end{cquote}
+Furthermore, \CFA concurrency has no spurious wakeup~\cite[\S~9]{Buhr05a}, which eliminates an implicit self barging.
+Monitor mutual-exclusion means signalling cannot have the signaller and signalled thread in the monitor simultaneously, so only the signaller or signallee can proceed.
+Figure~\ref{f:MonitorScheduling} shows internal/external scheduling for the bounded-buffer examples in Figure~\ref{f:GenericBoundedBuffer}.
+For internal scheduling in Figure~\ref{f:BBInt}, the @signal@ moves the signallee (front thread of the specified condition queue) to urgent and the signaller continues (solid line).
+Multiple signals move multiple signallees to urgent until the condition queue is empty.
+When the signaller exits or waits, a thread is implicitly unblocked from urgent (if available) before unblocking a calling thread to prevent barging.
+Furthermore, \CFA concurrency has no spurious wakeup~\cite[\S~9]{Buhr05a}, which eliminates an implicit form of self barging.
+Hence, a \CFA @wait@ statement is not enclosed in a @while@ loop retesting a blocking predicate, which can cause thread starvation due to barging.
+Figure~\ref{f:MonitorScheduling} shows general internal/external scheduling (for the bounded-buffer example in Figure~\ref{f:InternalExternalScheduling}).
+External calling threads block on the calling queue, if the monitor is occupied, otherwise they enter in FIFO order.
+Internal threads block on condition queues via @wait@ and reenter from the condition in FIFO order.
+Alternatively, internal threads block on urgent from the @signal_block@ or @waitfor@, and reenter implicitly when the monitor becomes empty, \ie, the thread in the monitor exits or waits.
+There are three signalling mechanisms to unblock waiting threads to enter the monitor.
+Note, signalling cannot have the signaller and signalled thread in the monitor simultaneously because of the mutual exclusion, so either the signaller or signallee can proceed.
+For internal scheduling, threads are unblocked from condition queues using @signal@, where the signallee is moved to urgent and the signaller continues (solid line).
+Multiple signals move multiple signallees to urgent until the condition is empty.
+When the signaller exits or waits, a thread blocked on urgent is processed before calling threads to prevent barging.
 (Java conceptually moves the signalled thread to the calling queue, and hence, allows barging.)
+Signal is used when the signaller is providing the cooperation needed by the signallee (\eg creating an empty slot in a buffer for a producer) and the signaller immediately exits the monitor to run concurrently (consume the buffer element) and passes control of the monitor to the signalled thread, which can immediately take advantage of the state change.
+Specifically, the @wait@ function atomically blocks the calling thread and implicitly releases the monitor lock(s) for all monitors in the function's parameter list.
+Signalling is unconditional because signalling an empty condition queue does nothing.
+It is common to declare condition queues as monitor fields to prevent shared access, hence no locking is required for access as the queues are protected by the monitor lock.
+In \CFA, a condition queue can be created/stored independently.
+The alternative unblock is in the opposite order using @signal_block@, where the signaller is moved to urgent and the signallee continues (dashed line), and is implicitly unblocked from urgent when the signallee exits or waits.
+For external scheduling, the condition queues are not used;
+instead threads are unblocked directly from the calling queue using @waitfor@ based on function names requesting mutual exclusion.
+(The linear search through the calling queue to locate a particular call can be reduced to $O(1)$.)
+The @waitfor@ has the same semantics as @signal_block@, where the signalled thread executes before the signallee, which waits on urgent.
+Executing multiple @waitfor@s from different signalled functions causes the calling threads to move to urgent.
+External scheduling requires urgent to be a stack, because the signaller expects to execute immediately after the specified monitor call has exited or waited.
+Internal scheduling behaves the same for an urgent stack or queue, except for multiple signalling, where the threads unblock from urgent in reverse order from signalling.
+If the restart order is important, multiple signalling by a signal thread can be transformed into daisy-chain signalling among threads, where each thread signals the next thread.
+We tried both a stack for @waitfor@ and queue for signalling, but that resulted in complex semantics about which thread enters next.
+Hence, \CFA uses a single urgent stack to correctly handle @waitfor@ and adequately support both forms of signalling.
 \begin{figure}
 …
 \end{figure}
+Figure~\ref{f:BBInt} shows a \CFA generic bounded-buffer with internal scheduling, where producers/consumers enter the monitor, detect the buffer is full/empty, and block on an appropriate condition variable, @full@/@empty@.
+The @wait@ function atomically blocks the calling thread and implicitly releases the monitor lock(s) for all monitors in the function's parameter list.
+The appropriate condition variable is signalled to unblock an opposite kind of thread after an element is inserted/removed from the buffer.
+Signalling is unconditional, because signalling an empty condition variable does nothing.
+It is common to declare condition variables as monitor fields to prevent shared access, hence no locking is required for access as the conditions are protected by the monitor lock.
+In \CFA, a condition variable can be created/stored independently.
+% To still prevent expensive locking on access, a condition variable is tied to a \emph{group} of monitors on first use, called \newterm{branding}, resulting in a low-cost boolean test to detect sharing from other monitors.
+% Signalling semantics cannot have the signaller and signalled thread in the monitor simultaneously, which means:
+% \begin{enumerate}
+% \item
+% The signalling thread returns immediately and the signalled thread continues.
+% \item
+% The signalling thread continues and the signalled thread is marked for urgent unblocking at the next scheduling point (exit/wait).
+% \item
+% The signalling thread blocks but is marked for urgent unblocking at the next scheduling point and the signalled thread continues.
+% \end{enumerate}
+% The first approach is too restrictive, as it precludes solving a reasonable class of problems, \eg dating service (see Figure~\ref{f:DatingService}).
+% \CFA supports the next two semantics as both are useful.
 \begin{figure}
 \centering
 …
                 T elements[10];
         };
         void ?{}( Buffer(T) & buf ) with(buf) {
+        void ?{}( Buffer(T) & buffer ) with(buffer) {
                 front = back = count = 0;
+        }
         void insert(Buffer(T) & mutex buf, T elm) with(buf){
                 if ( count == 10 ) `wait( empty )`; // full ?
                 // insert elm into buf
+        void insert( Buffer(T) & mutex buffer, T elem )
+                                with(buffer) {
+                if ( count == 10 ) `wait( empty )`;
+                // insert elem into buffer
                 `signal( full )`;
+        }
         T remove( Buffer(T) & mutex buf ) with(buf) {
                 if ( count == 0 ) `wait( full )`; // empty ?
                 // remove elm from buf
+        T remove( Buffer(T) & mutex buffer ) with(buffer) {
+                if ( count == 0 ) `wait( full )`;
+                // remove elem from buffer
                 `signal( empty )`;
                 return elm;
+                return elem;
+        }
+}
 \end{cfa}
 \end{lrbox}
+% \newbox\myboxB
+% \begin{lrbox}{\myboxB}
+% \begin{cfa}[aboveskip=0pt,belowskip=0pt]
+% forall( otype T ) { // distribute forall
+%       monitor Buffer {
+%
+%               int front, back, count;
+%               T elements[10];
+%       };
+%       void ?{}( Buffer(T) & buffer ) with(buffer) {
+%               [front, back, count] = 0;
+%       }
+%       T remove( Buffer(T) & mutex buffer ); // forward
+%       void insert( Buffer(T) & mutex buffer, T elem )
+%                               with(buffer) {
+%               if ( count == 10 ) `waitfor( remove, buffer )`;
+%               // insert elem into buffer
+%
+%       }
+%       T remove( Buffer(T) & mutex buffer ) with(buffer) {
+%               if ( count == 0 ) `waitfor( insert, buffer )`;
+%               // remove elem from buffer
+%
+%               return elem;
+%       }
+% }
+% \end{cfa}
+% \end{lrbox}
 \newbox\myboxB
 \begin{lrbox}{\myboxB}
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
-forall( otype T ) { // distribute forall
-        monitor Buffer {
-                int front, back, count;
-                T elements[10];
-        };
-        void ?{}( Buffer(T) & buf ) with(buf) {
-                front = back = count = 0;
+        }
-        T remove( Buffer(T) & mutex buf ); // forward
-        void insert(Buffer(T) & mutex buf, T elm) with(buf){
-                if ( count == 10 ) `waitfor( remove : buf )`;
-                // insert elm into buf
+        }
-        T remove( Buffer(T) & mutex buf ) with(buf) {
-                if ( count == 0 ) `waitfor( insert : buf )`;
-                // remove elm from buf
-                return elm;
+        }
+}
-\end{cfa}
-\end{lrbox}
-\subfloat[Internal scheduling]{\label{f:BBInt}\usebox\myboxA}
-\hspace{1pt}
-\vrule
-\hspace{3pt}
-\subfloat[External scheduling]{\label{f:BBExt}\usebox\myboxB}
-\caption{Generic bounded buffer}
-\label{f:GenericBoundedBuffer}
-\end{figure}
-The @signal_block@ provides the opposite unblocking order, where the signaller is moved to urgent and the signallee continues and a thread is implicitly unblocked from urgent when the signallee exits or waits (dashed line).
-Signal block is used when the signallee is providing the cooperation needed by the signaller (\eg if the buffer is removed and a producer hands off an item to a consumer, as in Figure~\ref{f:DatingSignalBlock}) so the signaller must wait until the signallee unblocks, provides the cooperation, exits the monitor to run concurrently, and passes control of the monitor to the signaller, which can immediately take advantage of the state change.
-Using @signal@ or @signal_block@ can be a dynamic decision based on whether the thread providing the cooperation arrives before or after the thread needing the cooperation.
-External scheduling in Figure~\ref{f:BBExt} simplifies internal scheduling by eliminating condition queues and @signal@/@wait@ (cases where it cannot are discussed shortly), and has existed in the programming language Ada for almost 40 years with variants in other languages~\cite{SR,ConcurrentC++,uC++}.
-While prior languages use external scheduling solely for thread interaction, \CFA generalizes it to both monitors and threads.
-External scheduling allows waiting for events from other threads while restricting unrelated events, that would otherwise have to wait on condition queues in the monitor.
-Scheduling is controlled by the @waitfor@ statement, which atomically blocks the calling thread, releases the monitor lock, and restricts the function calls that can next acquire mutual exclusion.
-Specifically, a thread calling the monitor is unblocked directly from the calling queue based on function names that can fulfill the cooperation required by the signaller.
-(The linear search through the calling queue to locate a particular call can be reduced to $O(1)$.)
-Hence, the @waitfor@ has the same semantics as @signal_block@, where the signallee thread from the calling queue executes before the signaller, which waits on urgent.
-Now when a producer/consumer detects a full/empty buffer, the necessary cooperation for continuation is specified by indicating the next function call that can occur.
-For example, a producer detecting a full buffer must have cooperation from a consumer to remove an item so function @remove@ is accepted, which prevents producers from entering the monitor, and after a consumer calls @remove@, the producer waiting on urgent is \emph{implicitly} unblocked because it can now continue its insert operation.
-Hence, this mechanism is done in terms of control flow, next call, versus in terms of data, channels, as in Go/Rust @select@.
-While both mechanisms have strengths and weaknesses, \CFA uses the control-flow mechanism to be consistent with other language features.
-Figure~\ref{f:ReadersWriterLock} shows internal/external scheduling for a readers/writer lock with no barging and threads are serviced in FIFO order to eliminate staleness/freshness among the reader/writer threads.
-For internal scheduling in Figure~\ref{f:RWInt}, the readers and writers wait on the same condition queue in FIFO order, making it impossible to tell if a waiting thread is a reader or writer.
-To clawback the kind of thread, a \CFA condition can store user data in the node for a blocking thread at the @wait@, \ie whether the thread is a @READER@ or @WRITER@.
-An unblocked reader thread checks if the thread at the front of the queue is a reader and unblock it, \ie the readers daisy-chain signal the next group of readers demarcated by the next writer or end of the queue.
-For external scheduling in Figure~\ref{f:RWExt}, a waiting reader checks if a writer is using the resource, and if so, restricts further calls until the writer exits by calling @EndWrite@.
-The writer does a similar action for each reader or writer using the resource.
-Note, no new calls to @StartRead@/@StartWrite@ may occur when waiting for the call to @EndRead@/@EndWrite@.
-\begin{figure}
-\centering
-\newbox\myboxA
-\begin{lrbox}{\myboxA}
-\begin{cfa}[aboveskip=0pt,belowskip=0pt]
-enum RW { READER, WRITER };
 monitor ReadersWriter {
+        int rcnt, wcnt; // readers/writer using resource
+        `condition RWers;`
+        int rcnt, wcnt; // readers/writer using resource
 };
 void ?{}( ReadersWriter & rw ) with(rw) {
 …
 void EndRead( ReadersWriter & mutex rw ) with(rw) {
         rcnt -= 1;
-        if ( rcnt == 0 ) `signal( RWers )`;
+}
 void EndWrite( ReadersWriter & mutex rw ) with(rw) {
         wcnt = 0;
-        `signal( RWers );`
+}
 void StartRead( ReadersWriter & mutex rw ) with(rw) {
+        if ( wcnt !=0 || ! empty( RWers ) )
+                `wait( RWers, READER )`;
+        if ( wcnt > 0 ) `waitfor( EndWrite, rw );`
         rcnt += 1;
-        if ( ! empty(RWers) && `front(RWers) == READER` )
-                `signal( RWers )`;  // daisy-chain signalling
+}
 void StartWrite( ReadersWriter & mutex rw ) with(rw) {
         if ( wcnt != 0 || rcnt != 0 ) `wait( RWers, WRITER )`;
+        if ( wcnt > 0 ) `waitfor( EndWrite, rw );`
+        else while ( rcnt > 0 ) `waitfor( EndRead, rw );`
         wcnt = 1;
+}
 \end{cfa}
 \end{lrbox}
+\newbox\myboxB
+\begin{lrbox}{\myboxB}
+\begin{cfa}[aboveskip=0pt,belowskip=0pt]
+monitor ReadersWriter {
+        int rcnt, wcnt; // readers/writer using resource
+};
+void ?{}( ReadersWriter & rw ) with(rw) {
+        rcnt = wcnt = 0;
+}
+void EndRead( ReadersWriter & mutex rw ) with(rw) {
+        rcnt -= 1;
+}
+void EndWrite( ReadersWriter & mutex rw ) with(rw) {
+        wcnt = 0;
+}
+void StartRead( ReadersWriter & mutex rw ) with(rw) {
+        if ( wcnt > 0 ) `waitfor( EndWrite : rw );`
+        rcnt += 1;
+}
+void StartWrite( ReadersWriter & mutex rw ) with(rw) {
+        if ( wcnt > 0 ) `waitfor( EndWrite : rw );`
+        else while ( rcnt > 0 ) `waitfor( EndRead : rw );`
+        wcnt = 1;
+}
+\end{cfa}
+\end{lrbox}
+\subfloat[Internal scheduling]{\label{f:RWInt}\usebox\myboxA}
+\hspace{1pt}
+\subfloat[Generic bounded buffer, internal scheduling]{\label{f:BBInt}\usebox\myboxA}
+\hspace{3pt}
 \vrule
 \hspace{3pt}
 \subfloat[External scheduling]{\label{f:RWExt}\usebox\myboxB}
 \caption{Readers / writer lock}
 \label{f:ReadersWriterLock}
+\subfloat[Readers / writer lock, external scheduling]{\label{f:RWExt}\usebox\myboxB}
+\caption{Internal / external scheduling}
+\label{f:InternalExternalScheduling}
 \end{figure}
+Finally, external scheduling requires urgent to be a stack, because the signaller expects to execute immediately after the specified monitor call has exited or waited.
+Internal schedulling performing multiple signalling results in unblocking from urgent in the reverse order from signalling.
+It is rare for the unblocking order to be important as an unblocked thread can be time-sliced immediately after leaving the monitor.
+If the unblocking order is important, multiple signalling can be restructured into daisy-chain signalling, where each thread signals the next thread.
+Hence, \CFA uses a single urgent stack to correctly handle @waitfor@ and adequately support both forms of signalling.
+(Advanced @waitfor@ features are discussed in Section~\ref{s:ExtendedWaitfor}.)
+Figure~\ref{f:BBInt} can be transformed into external scheduling by removing the condition variables and signals/waits, and adding the following lines at the locations of the current @wait@s in @insert@/@remove@, respectively.
+\begin{cfa}[aboveskip=2pt,belowskip=1pt]
+if ( count == 10 ) `waitfor( remove, buffer )`;       |      if ( count == 0 ) `waitfor( insert, buffer )`;
+\end{cfa}
+Here, the producers/consumers detects a full/\-empty buffer and prevents more producers/consumers from entering the monitor until there is a free/empty slot in the buffer.
+External scheduling is controlled by the @waitfor@ statement, which atomically blocks the calling thread, releases the monitor lock, and restricts the function calls that can next acquire mutual exclusion.
+If the buffer is full, only calls to @remove@ can acquire the buffer, and if the buffer is empty, only calls to @insert@ can acquire the buffer.
+Threads calling excluded functions block outside of (external to) the monitor on the calling queue, versus blocking on condition queues inside of (internal to) the monitor.
+Figure~\ref{f:RWExt} shows a readers/writer lock written using external scheduling, where a waiting reader detects a writer using the resource and restricts further calls until the writer exits by calling @EndWrite@.
+The writer does a similar action for each reader or writer using the resource.
+Note, no new calls to @StarRead@/@StartWrite@ may occur when waiting for the call to @EndRead@/@EndWrite@.
+External scheduling allows waiting for events from other threads while restricting unrelated events, that would otherwise have to wait on conditions in the monitor.
+The mechnaism can be done in terms of control flow, \eg Ada @accept@ or \uC @_Accept@, or in terms of data, \eg Go @select@ on channels.
+While both mechanisms have strengths and weaknesses, this project uses the control-flow mechanism to be consistent with other language features.
+% Two challenges specific to \CFA for external scheduling are loose object-definitions (see Section~\ref{s:LooseObjectDefinitions}) and multiple-monitor functions (see Section~\ref{s:Multi-MonitorScheduling}).
+Figure~\ref{f:DatingService} shows a dating service demonstrating non-blocking and blocking signalling.
+The dating service matches girl and boy threads with matching compatibility codes so they can exchange phone numbers.
+A thread blocks until an appropriate partner arrives.
+The complexity is exchanging phone numbers in the monitor because of the mutual-exclusion property.
+For signal scheduling, the @exchange@ condition is necessary to block the thread finding the match, while the matcher unblocks to take the opposite number, post its phone number, and unblock the partner.
+For signal-block scheduling, the implicit urgent-queue replaces the explict @exchange@-condition and @signal_block@ puts the finding thread on the urgent condition and unblocks the matcher.
+The dating service is an example of a monitor that cannot be written using external scheduling because it requires knowledge of calling parameters to make scheduling decisions, and parameters of waiting threads are unavailable;
+as well, an arriving thread may not find a partner and must wait, which requires a condition variable, and condition variables imply internal scheduling.
+Furthermore, barging corrupts the dating service during an exchange because a barger may also match and change the phone numbers, invalidating the previous exchange phone number.
+Putting loops around the @wait@s does not correct the problem;
+the simple solution must be restructured to account for barging.
 \begin{figure}
 …
 };
 int girl( DS & mutex ds, int phNo, int ccode ) {
         if ( empty( Boys[ccode] ) ) {
+        if ( is_empty( Boys[ccode] ) ) {
                 wait( Girls[ccode] );
                 GirlPhNo = phNo;
 …
 };
 int girl( DS & mutex ds, int phNo, int ccode ) {
         if ( empty( Boys[ccode] ) ) { // no compatible
+        if ( is_empty( Boys[ccode] ) ) { // no compatible
                 wait( Girls[ccode] ); // wait for boy
                 GirlPhNo = phNo; // make phone number available
 …
 \qquad
 \subfloat[\lstinline@signal_block@]{\label{f:DatingSignalBlock}\usebox\myboxB}
 \caption{Dating service Monitor}
 \label{f:DatingServiceMonitor}
+\caption{Dating service}
+\label{f:DatingService}
 \end{figure}
+Figure~\ref{f:DatingServiceMonitor} shows a dating service demonstrating non-blocking and blocking signalling.
+The dating service matches girl and boy threads with matching compatibility codes so they can exchange phone numbers.
+A thread blocks until an appropriate partner arrives.
+The complexity is exchanging phone numbers in the monitor because of the mutual-exclusion property.
+For signal scheduling, the @exchange@ condition is necessary to block the thread finding the match, while the matcher unblocks to take the opposite number, post its phone number, and unblock the partner.
+For signal-block scheduling, the implicit urgent-queue replaces the explicit @exchange@-condition and @signal_block@ puts the finding thread on the urgent stack and unblocks the matcher.
+The dating service is an important example of a monitor that cannot be written using external scheduling.
+First, because scheduling requires knowledge of calling parameters to make matching decisions, and parameters of calling threads are unavailable within the monitor.
+For example, a girl thread within the monitor cannot examine the @ccode@ of boy threads waiting on the calling queue to determine if there is a matching partner.
+Second, because a scheduling decision may be delayed when there is no immediate match, which requires a condition queue for waiting, and condition queues imply internal scheduling.
+For example, if a girl thread could determine there is no calling boy with the same @ccode@, it must wait until a matching boy arrives.
+Finally, barging corrupts the dating service during an exchange because a barger may also match and change the phone numbers, invalidating the previous exchange phone number.
+This situation shows rechecking the waiting condition and waiting again (signals-as-hints) fails, requiring significant restructured to account for barging.
+In summation, for internal scheduling, non-blocking signalling (as in the producer/consumer example) is used when the signaller is providing the cooperation for a waiting thread;
+the signaller enters the monitor and changes state, detects a waiting threads that can use the state, performs a non-blocking signal on the condition queue for the waiting thread, and exits the monitor to run concurrently.
+The waiter unblocks next from the urgent queue, uses/takes the state, and exits the monitor.
+Blocking signal is the reverse, where the waiter is providing the cooperation for the signalling thread;
+the signaller enters the monitor, detects a waiting thread providing the necessary state, performs a blocking signal to place it on the urgent queue and unblock the waiter.
+The waiter changes state and exits the monitor, and the signaller unblocks next from the urgent queue to use/take the state.
 Both internal and external scheduling extend to multiple monitors in a natural way.
 \begin{cquote}
 \begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
+\begin{tabular}{@{}l@{\hspace{3\parindentlnth}}l@{}}
 \begin{cfa}
 monitor M { `condition e`; ... };
 …
+&
 \begin{cfa}
 void rtn$\(_1\)$( M & mutex m1, M & mutex m2 ); // overload rtn
+void rtn$\(_1\)$( M & mutex m1, M & mutex m2 );
 void rtn$\(_2\)$( M & mutex m1 );
 void bar( M & mutex m1, M & mutex m2 ) {
         ... waitfor( `rtn`${\color{red}\(_1\)}$ ); ...       // $\LstCommentStyle{waitfor( rtn\(_1\) : m1, m2 )}$
         ... waitfor( `rtn${\color{red}\(_2\)}$ : m1` ); ...
+        ... waitfor( `rtn` ); ...       // $\LstCommentStyle{waitfor( rtn\(_1\), m1, m2 )}$
+        ... waitfor( `rtn, m1` ); ... // $\LstCommentStyle{waitfor( rtn\(_2\), m1 )}$
+}
 \end{cfa}
 …
 For @wait( e )@, the default semantics is to atomically block the signaller and release all acquired mutex parameters, \ie @wait( e, m1, m2 )@.
 To override the implicit multi-monitor wait, specific mutex parameter(s) can be specified, \eg @wait( e, m1 )@.
 Wait cannot statically verify the released monitors are the acquired mutex-parameters without disallowing separately compiled helper functions calling @wait@.
 While \CC supports bulk locking, @wait@ only accepts a single lock for a condition queue, so bulk locking with condition queues is asymmetric.
+Wait cannot statically verifies the released monitors are the acquired mutex-parameters without disallowing separately compiled helper functions calling @wait@.
+While \CC supports bulk locking, @wait@ only accepts a single lock for a condition variable, so bulk locking with condition variables is asymmetric.
 Finally, a signaller,
 \begin{cfa}
 …
 must have acquired at least the same locks as the waiting thread signalled from a condition queue to allow the locks to be passed, and hence, prevent barging.
 Similarly, for @waitfor( rtn )@, the default semantics is to atomically block the acceptor and release all acquired mutex parameters, \ie @waitfor( rtn : m1, m2 )@.
 To override the implicit multi-monitor wait, specific mutex parameter(s) can be specified, \eg @waitfor( rtn : m1 )@.
+Similarly, for @waitfor( rtn )@, the default semantics is to atomically block the acceptor and release all acquired mutex parameters, \ie @waitfor( rtn, m1, m2 )@.
+To override the implicit multi-monitor wait, specific mutex parameter(s) can be specified, \eg @waitfor( rtn, m1 )@.
 @waitfor@ does statically verify the monitor types passed are the same as the acquired mutex-parameters of the given function or function pointer, hence the function (pointer) prototype must be accessible.
 % When an overloaded function appears in an @waitfor@ statement, calls to any function with that name are accepted.
 …
 void rtn( M & mutex m );
 `int` rtn( M & mutex m );
+waitfor( (`int` (*)( M & mutex ))rtn : m );
+\end{cfa}
+The ability to release a subset of acquired monitors can result in a \newterm{nested monitor}~\cite{Lister77} deadlock (see Section~\ref{s:MutexAcquisition}).
+\newpage
+waitfor( (`int` (*)( M & mutex ))rtn, m );
+\end{cfa}
+The ability to release a subset of acquired monitors can result in a \newterm{nested monitor}~\cite{Lister77} deadlock.
 \begin{cfa}
 void foo( M & mutex m1, M & mutex m2 ) {
         ... wait( `e, m1` ); ...                                $\C{// release m1, keeping m2 acquired}$
 void bar( M & mutex m1, M & mutex m2 ) {        $\C{// must acquire m1 and m2}$
+        ... wait( `e, m1` ); ...                                $\C{// release m1, keeping m2 acquired )}$
+void bar( M & mutex m1, M & mutex m2 ) {        $\C{// must acquire m1 and m2 )}$
         ... signal( `e` ); ...
 \end{cfa}
 The @wait@ only releases @m1@ so the signalling thread cannot acquire @m1@ and @m2@ to enter @bar@ and @signal@ the condition.
+While deadlock can occur with multiple/nesting acquisition, this is a consequence of locks, and by extension monitor locking is not perfectly composable.
+While deadlock can occur with multiple/nesting acquisition, this is a consequence of locks, and by extension monitors, not being perfectly composable.
 \subsection{\texorpdfstring{Extended \protect\lstinline@waitfor@}{Extended waitfor}}
-\label{s:ExtendedWaitfor}
 Figure~\ref{f:ExtendedWaitfor} shows the extended form of the @waitfor@ statement to conditionally accept one of a group of mutex functions, with an optional statement to be performed \emph{after} the mutex function finishes.
 …
 Hence, the terminating @else@ clause allows a conditional attempt to accept a call without blocking.
 If both @timeout@ and @else@ clause are present, the @else@ must be conditional, or the @timeout@ is never triggered.
+There is also a traditional future wait queue (not shown) (\eg Microsoft @WaitForMultipleObjects@), to wait for a specified number of future elements in the queue.
+Finally, there is a shorthand for specifying multiple functions using the same set of monitors: @waitfor( f, g, h : m1, m2, m3 )@.
+There is also a traditional future wait queue (not shown) (\eg Microsoft (@WaitForMultipleObjects@)), to wait for a specified number of future elements in the queue.
 \begin{figure}
 …
 The right example accepts either @mem1@ or @mem2@ if @C1@ and @C2@ are true.
 An interesting use of @waitfor@ is accepting the @mutex@ destructor to know when an object is deallocated, \eg assume the bounded buffer is restructured from a monitor to a thread with the following @main@.
+An interesting use of @waitfor@ is accepting the @mutex@ destructor to know when an object is deallocated, \eg assume the bounded buffer is restructred from a monitor to a thread with the following @main@.
 \begin{cfa}
 void main( Buffer(T) & buffer ) with(buffer) {
         for () {
                 `waitfor( ^?{} : buffer )` break;
                 or when ( count != 20 ) waitfor( insert : buffer ) { ... }
                 or when ( count != 0 ) waitfor( remove : buffer ) { ... }
+                `waitfor( ^?{}, buffer )` break;
+                or when ( count != 20 ) waitfor( insert, buffer ) { ... }
+                or when ( count != 0 ) waitfor( remove, buffer ) { ... }
+        }
         // clean up
 …
 To support this efficient semantics (and prevent barging), the implementation maintains a list of monitors acquired for each blocked thread.
 When a signaller exits or waits in a monitor function/statement, the front waiter on urgent is unblocked if all its monitors are released.
 Implementing a fast subset check for the necessary released monitors is important and discussed in the following sections.
+Implementing a fast subset check for the necessary released monitors is important.
 % The benefit is encapsulating complexity into only two actions: passing monitors to the next owner when they should be released and conditionally waking threads if all conditions are met.
+\subsection{\texorpdfstring{\protect\lstinline@waitfor@ Implementation}{waitfor Implementation}}
+\label{s:waitforImplementation}
+In a statically-typed object-oriented programming language, a class has an exhaustive list of members, even when members are added via static inheritance (see Figure~\ref{f:uCinheritance}).
+Knowing all members at compilation (even separate compilation) allows uniquely numbered them so the accept-statement implementation can use a fast/compact bit mask with $O(1)$ compare.
+\begin{figure}
+\centering
+\begin{lrbox}{\myboxA}
+\begin{uC++}[aboveskip=0pt,belowskip=0pt]
+$\emph{translation unit 1}$
+_Monitor B { // common type in .h file
+        _Mutex virtual void `f`( ... );
+        _Mutex virtual void `g`( ... );
+        _Mutex virtual void w1( ... ) { ... _Accept(`f`, `g`); ... }
+};
+$\emph{translation unit 2}$
+// include B
+_Monitor D : public B { // inherit
+        _Mutex void `h`( ... ); // add
+        _Mutex void w2( ... ) { ... _Accept(`f`, `h`); ... }
+};
+\end{uC++}
+\end{lrbox}
+\begin{lrbox}{\myboxB}
+\begin{cfa}[aboveskip=0pt,belowskip=0pt]
+$\emph{translation unit 1}$
+monitor M { ... }; // common type in .h file
+void `f`( M & mutex m, ... );
+void `g`( M & mutex m, ... );
+void w1( M & mutex m, ... ) { ... waitfor(`f`, `g` : m); ... }
+$\emph{translation unit 2}$
+// include M
+extern void `f`( M & mutex m, ... ); // import f but not g
+void `h`( M & mutex m ); // add
+void w2( M & mutex m, ... ) { ... waitfor(`f`, `h` : m); ... }
+\end{cfa}
+\end{lrbox}
+\subfloat[\uC]{\label{f:uCinheritance}\usebox\myboxA}
+\hspace{3pt}
+\vrule
+\hspace{3pt}
+\subfloat[\CFA]{\label{f:CFinheritance}\usebox\myboxB}
+\caption{Member / Function visibility}
+\label{f:MemberFunctionVisibility}
+\end{figure}
+However, the @waitfor@ statement in translation unit 2 (see Figure~\ref{f:CFinheritance}) cannot see function @g@ in translation unit 1 precluding a unique numbering for a bit-mask because the monitor type only carries the protected shared-data.
+\subsection{Loose Object Definitions}
+\label{s:LooseObjectDefinitions}
+In an object-oriented programming language, a class includes an exhaustive list of operations.
+A new class can add members via static inheritance but the subclass still has an exhaustive list of operations.
+(Dynamic member adding, \eg JavaScript~\cite{JavaScript}, is not considered.)
+In the object-oriented scenario, the type and all its operators are always present at compilation (even separate compilation), so it is possible to number the operations in a bit mask and use an $O(1)$ compare with a similar bit mask created for the operations specified in a @waitfor@.
+However, in \CFA, monitor functions can be statically added/removed in translation units, making a fast subset check difficult.
+\begin{cfa}
+        monitor M { ... }; // common type, included in .h file
+translation unit 1
+        void `f`( M & mutex m );
+        void g( M & mutex m ) { waitfor( `f`, m ); }
+translation unit 2
+        void `f`( M & mutex m ); $\C{// replacing f and g for type M in this translation unit}$
+        void `g`( M & mutex m );
+        void h( M & mutex m ) { waitfor( `f`, m ) or waitfor( `g`, m ); } $\C{// extending type M in this translation unit}$
+\end{cfa}
+The @waitfor@ statements in each translation unit cannot form a unique bit-mask because the monitor type does not carry that information.
+Hence, function pointers are used to identify the functions listed in the @waitfor@ statement, stored in a variable-sized array.
+Then, the same implementation approach used for the urgent stack is used for the calling queue.
+Each caller has a list of monitors acquired, and the @waitfor@ statement performs a (usually short) linear search matching functions in the @waitfor@ list with called functions, and then verifying the associated mutex locks can be transfers.
 (A possible way to construct a dense mapping is at link or load-time.)
-Hence, function pointers are used to identify the functions listed in the @waitfor@ statement, stored in a variable-sized array.
-Then, the same implementation approach used for the urgent stack (see Section~\ref{s:Scheduling}) is used for the calling queue.
-Each caller has a list of monitors acquired, and the @waitfor@ statement performs a (short) linear search matching functions in the @waitfor@ list with called functions, and then verifying the associated mutex locks can be transfers.
 …
 The solution is for the programmer to disambiguate:
 \begin{cfa}
 waitfor( f : `m2` ); $\C{// wait for call to f with argument m2}$
+waitfor( f, `m2` ); $\C{// wait for call to f with argument m2}$
 \end{cfa}
 Both locks are acquired by function @g@, so when function @f@ is called, the lock for monitor @m2@ is passed from @g@ to @f@, while @g@ still holds lock @m1@.
 …
 monitor M { ... };
 void f( M & mutex m1, M & mutex m2 );
 void g( M & mutex m1, M & mutex m2 ) { waitfor( f : `m1, m2` ); $\C{// wait for call to f with arguments m1 and m2}$
+void g( M & mutex m1, M & mutex m2 ) { waitfor( f, `m1, m2` ); $\C{// wait for call to f with arguments m1 and m2}$
 \end{cfa}
 Again, the set of monitors passed to the @waitfor@ statement must be entirely contained in the set of monitors already acquired by the accepting function.
 % Also, the order of the monitors in a @waitfor@ statement must match the order of the mutex parameters.
 Figure~\ref{f:UnmatchedMutexSets} shows internal and external scheduling with multiple monitors that must match exactly with a signalling or accepting thread, \ie partial matching results in waiting.
 In both cases, the set of monitors is disjoint so unblocking is impossible.
+Also, the order of the monitors in a @waitfor@ statement is unimportant.
+Figure~\ref{f:UnmatchedMutexSets} shows an example where, for internal and external scheduling with multiple monitors, a signalling or accepting thread must match exactly, \ie partial matching results in waiting.
+For both examples, the set of monitors is disjoint so unblocking is impossible.
 \begin{figure}
 …
+}
 void g( M1 & mutex m1, M2 & mutex m2 ) {
         waitfor( f : m1, m2 );
+        waitfor( f, m1, m2 );
+}
 g( `m11`, m2 ); // block on accept
 …
 \end{figure}
+\subsection{\texorpdfstring{\protect\lstinline@mutex@ Threads}{mutex Threads}}
+Threads in \CFA can also be monitors to allow \emph{direct communication} among threads, \ie threads can have mutex functions that are called by other threads.
+Hence, all monitor features are available when using threads.
+Figure~\ref{f:DirectCommunication} shows a comparison of direct call communication in \CFA with direct channel communication in Go.
+(Ada provides a similar mechanism to the \CFA direct communication.)
+The program main in both programs communicates directly with the other thread versus indirect communication where two threads interact through a passive monitor.
+Both direct and indirection thread communication are valuable tools in structuring concurrent programs.
 \begin{figure}
 \centering
 …
 struct Msg { int i, j; };
 monitor thread GoRtn { int i;  float f;  Msg m; };
+thread GoRtn { int i;  float f;  Msg m; };
 void mem1( GoRtn & mutex gortn, int i ) { gortn.i = i; }
 void mem2( GoRtn & mutex gortn, float f ) { gortn.f = f; }
 …
         for () {
                 `waitfor( mem1 : gortn )` sout | i;  // wait for calls
                 or `waitfor( mem2 : gortn )` sout | f;
                 or `waitfor( mem3 : gortn )` sout | m.i | m.j;
                 or `waitfor( ^?{} : gortn )` break; // low priority
+                `waitfor( mem1, gortn )` sout | i;  // wait for calls
+                or `waitfor( mem2, gortn )` sout | f;
+                or `waitfor( mem3, gortn )` sout | m.i | m.j;
+                or `waitfor( ^?{}, gortn )` break;
+        }
 …
 \hspace{3pt}
 \subfloat[Go]{\label{f:Gochannel}\usebox\myboxB}
+\caption{Direct versus indirect communication}
+\label{f:DirectCommunicationComparison}
+\medskip
+\begin{cfa}
+monitor thread DatingService {
+        condition Girls[CompCodes], Boys[CompCodes];
+        int girlPhoneNo, boyPhoneNo, ccode;
+};
+int girl( DatingService & mutex ds, int phoneno, int code ) with( ds ) {
+        girlPhoneNo = phoneno;  ccode = code;
+        `wait( Girls[ccode] );`                                                         $\C{// wait for boy}$
+        girlPhoneNo = phoneno;  return boyPhoneNo;
+}
+int boy( DatingService & mutex ds, int phoneno, int code ) with( ds ) {
+        boyPhoneNo = phoneno;  ccode = code;
+        `wait( Boys[ccode] );`                                                          $\C{// wait for girl}$
+        boyPhoneNo = phoneno;  return girlPhoneNo;
+}
+void main( DatingService & ds ) with( ds ) {                    $\C{// thread starts, ds defaults to mutex}$
+        for () {
+                waitfor( ^?{} ) break;                                                  $\C{// high priority}$
+                or waitfor( girl )                                                              $\C{// girl called, compatible boy ? restart boy then girl}$
+                        if ( ! is_empty( Boys[ccode] ) ) { `signal_block( Boys[ccode] );  signal_block( Girls[ccode] );` }
+                or waitfor( boy ) {                                                             $\C{// boy called, compatible girl ? restart girl then boy}$
+                        if ( ! is_empty( Girls[ccode] ) ) { `signal_block( Girls[ccode] );  signal_block( Boys[ccode] );` }
+        }
+}
+\end{cfa}
+\caption{Direct communication dating service}
+\label{f:DirectCommunicationDatingService}
+\caption{Direct communication}
+\label{f:DirectCommunication}
 \end{figure}
 …
 void main( Ping & pi ) {
         for ( 10 ) {
                 `waitfor( ping : pi );`
+                `waitfor( ping, pi );`
                 `pong( po );`
+        }
 …
         for ( 10 ) {
                 `ping( pi );`
                 `waitfor( pong : po );`
+                `waitfor( pong, po );`
+        }
+}
 …
+\subsection{\texorpdfstring{\protect\lstinline@monitor@ Generators / Coroutines / Threads}{monitor Generators / Coroutines / Threads}}
+\CFA generators, coroutines, and threads can also be monitors (Table~\ref{t:ExecutionPropertyComposition} cases 4, 6, 12) allowing safe \emph{direct communication} with threads, \ie the custom types can have mutex functions that are called by other threads.
+All monitor features are available within these mutex functions.
+For example, if the formatter generator (or coroutine equivalent) in Figure~\ref{f:CFAFormatGen} is extended with the monitor property and this interface function is used to communicate with the formatter:
+\begin{cfa}
+void fmt( Fmt & mutex fmt, char ch ) { fmt.ch = ch; resume( fmt ) }
+\end{cfa}
+multiple threads can safely pass characters for formatting.
+Figure~\ref{f:DirectCommunicationComparison} shows a comparison of direct call-communication in \CFA versus indirect channel-communication in Go.
+(Ada has a similar mechanism to \CFA direct communication.)
+The program thread in \CFA @main@ uses the call/return paradigm to directly communicate with the @GoRtn main@, whereas Go switches to the channel paradigm to indirectly communicate with the goroutine.
+Communication by multiple threads is safe for the @gortn@ thread via mutex calls in \CFA or channel assignment in Go.
+Figure~\ref{f:DirectCommunicationDatingService} shows the dating-service problem in Figure~\ref{f:DatingServiceMonitor} extended from indirect monitor communication to direct thread communication.
+When converting a monitor to a thread (server), the coding pattern is to move as much code as possible from the accepted members into the thread main so it does an much work as possible.
+Notice, the dating server is postponing requests for an unspecified time while continuing to accept new requests.
+For complex servers (web-servers), there can be hundreds of lines of code in the thread main and safe interaction with clients can be complex.
+\subsection{Execution Properties}
+Table~\ref{t:ObjectPropertyComposition} shows how the \CFA high-level constructs cover 3 fundamental execution properties: thread, stateful function, and mutual exclusion.
+Case 1 is a basic object, with none of the new execution properties.
+Case 2 allows @mutex@ calls to Case 1 to protect shared data.
+Case 3 allows stateful functions to suspend/resume but restricts operations because the state is stackless.
+Case 4 allows @mutex@ calls to Case 3 to protect shared data.
+Cases 5 and 6 are the same as 3 and 4 without restriction because the state is stackful.
+Cases 7 and 8 are rejected because a thread cannot execute without a stackful state in a preemptive environment when context switching from the signal handler.
+Cases 9 and 10 have a stackful thread without and with @mutex@ calls.
+For situations where threads do not require direct communication, case 9 provides faster creation/destruction by eliminating @mutex@ setup.
+\begin{table}
+\caption{Object property composition}
+\centering
+\label{t:ObjectPropertyComposition}
+\renewcommand{\arraystretch}{1.25}
+%\setlength{\tabcolsep}{5pt}
+\begin{tabular}{c|c||l|l}
+\multicolumn{2}{c||}{object properties} & \multicolumn{2}{c}{mutual exclusion} \\
+\hline
+thread  & stateful                              & \multicolumn{1}{c|}{No} & \multicolumn{1}{c}{Yes} \\
+\hline
+\hline
+No              & No                                    & \textbf{1}\ \ \ aggregate type                & \textbf{2}\ \ \ @monitor@ aggregate type \\
+\hline
+No              & Yes (stackless)               & \textbf{3}\ \ \ @generator@                   & \textbf{4}\ \ \ @monitor@ @generator@ \\
+\hline
+No              & Yes (stackful)                & \textbf{5}\ \ \ @coroutine@                   & \textbf{6}\ \ \ @monitor@ @coroutine@ \\
+\hline
+Yes             & No / Yes (stackless)  & \textbf{7}\ \ \ {\color{red}rejected} & \textbf{8}\ \ \ {\color{red}rejected} \\
+\hline
+Yes             & Yes (stackful)                & \textbf{9}\ \ \ @thread@                              & \textbf{10}\ \ @monitor@ @thread@ \\
+\end{tabular}
+\end{table}
 …
 For completeness and efficiency, \CFA provides a standard set of low-level locks: recursive mutex, condition, semaphore, barrier, \etc, and atomic instructions: @fetchAssign@, @fetchAdd@, @testSet@, @compareSet@, \etc.
 Some of these low-level mechanism are used to build the \CFA runtime, but we always advocate using high-level mechanisms whenever possible.
+Some of these low-level mechanism are used in the \CFA runtime, but we strongly advocate using high-level mechanisms whenever possible.
 …
 \begin{cfa}
 struct Adder {
         int * row, cols;
+    int * row, cols;
 };
 int operator()() {
 …
 \label{s:RuntimeStructureCluster}
+A \newterm{cluster} is a collection of user and kernel threads, where the kernel threads run the user threads from the cluster's ready queue, and the operating system runs the kernel threads on the processors from its ready queue.
+The term \newterm{virtual processor} is introduced as a synonym for kernel thread to disambiguate between user and kernel thread.
+From the language perspective, a virtual processor is an actual processor (core).
+A \newterm{cluster} is a collection of threads and virtual processors (abstract kernel-thread) that execute the (user) threads from its own ready queue (like an OS executing kernel threads).
 The purpose of a cluster is to control the amount of parallelism that is possible among threads, plus scheduling and other execution defaults.
 The default cluster-scheduler is single-queue multi-server, which provides automatic load-balancing of threads on processors.
 …
 Programs may use more virtual processors than hardware processors.
 On a multiprocessor, kernel threads are distributed across the hardware processors resulting in virtual processors executing in parallel.
 (It is possible to use affinity to lock a virtual processor onto a particular hardware processor~\cite{affinityLinux,affinityWindows}, which is used when caching issues occur or for heterogeneous hardware processors.) %, affinityFreebsd, affinityNetbsd, affinityMacosx
+(It is possible to use affinity to lock a virtual processor onto a particular hardware processor~\cite{affinityLinux, affinityWindows, affinityFreebsd, affinityNetbsd, affinityMacosx}, which is used when caching issues occur or for heterogeneous hardware processors.)
 The \CFA runtime attempts to block unused processors and unblock processors as the system load increases;
 balancing the workload with processors is difficult because it requires future knowledge, \ie what will the application workload do next.
+balancing the workload with processors is difficult because it requires future knowledge, \ie what will the applicaton workload do next.
 Preemption occurs on virtual processors rather than user threads, via operating-system interrupts.
 Thus virtual processors execute user threads, where preemption frequency applies to a virtual processor, so preemption occurs randomly across the executed user threads.
 …
 Nondeterministic preemption provides fairness from long-running threads, and forces concurrent programmers to write more robust programs, rather than relying on code between cooperative scheduling to be atomic.
 This atomic reliance can fail on multi-core machines, because execution across cores is nondeterministic.
 A different reason for not supporting preemption is that it significantly complicates the runtime system, \eg Windows runtime does not support interrupts and on Linux systems, interrupts are complex (see below).
+A different reason for not supporting preemption is that it significantly complicates the runtime system, \eg Microsoft runtime does not support interrupts and on Linux systems, interrupts are complex (see below).
 Preemption is normally handled by setting a countdown timer on each virtual processor.
 When the timer expires, an interrupt is delivered, and its signal handler resets the countdown timer, and if the virtual processor is executing in user code, the signal handler performs a user-level context-switch, or if executing in the language runtime kernel, the preemption is ignored or rolled forward to the point where the runtime kernel context switches back to user code.
+When the timer expires, an interrupt is delivered, and the interrupt handler resets the countdown timer, and if the virtual processor is executing in user code, the signal handler performs a user-level context-switch, or if executing in the language runtime kernel, the preemption is ignored or rolled forward to the point where the runtime kernel context switches back to user code.
 Multiple signal handlers may be pending.
 When control eventually switches back to the signal handler, it returns normally, and execution continues in the interrupted user thread, even though the return from the signal handler may be on a different kernel thread than the one where the signal is delivered.
 The only issue with this approach is that signal masks from one kernel thread may be restored on another as part of returning from the signal handler;
 therefore, the same signal mask is required for all virtual processors in a cluster.
+Because preemption interval is usually long (1 millisecond) performance cost is negligible.
+Linux switched a decade ago from specific to arbitrary virtual-processor signal-delivery for applications with multiple kernel threads.
+In the new semantics, a virtual-processor directed signal may be delivered to any virtual processor created by the application that does not have the signal blocked.
+Because preemption frequency is usually long (1 millisecond) performance cost is negligible.
+Linux switched a decade ago from specific to arbitrary process signal-delivery for applications with multiple kernel threads.
+\begin{cquote}
+A process-directed signal may be delivered to any one of the threads that does not currently have the signal blocked.
+If more than one of the threads has the signal unblocked, then the kernel chooses an arbitrary thread to which it will deliver the signal.
+SIGNAL(7) - Linux Programmer's Manual
+\end{cquote}
 Hence, the timer-expiry signal, which is generated \emph{externally} by the Linux kernel to an application, is delivered to any of its Linux subprocesses (kernel threads).
 To ensure each virtual processor receives a preemption signal, a discrete-event simulation is run on a special virtual processor, and only it sets and receives timer events.
 …
 \label{s:Performance}
 To test the performance of the \CFA runtime, a series of microbenchmarks are used to compare \CFA with pthreads, Java 11.0.6, Go 1.12.6, Rust 1.37.0, Python 3.7.6, Node.js 12.14.1, and \uC 7.0.0.
+To verify the implementation of the \CFA runtime, a series of microbenchmarks are performed comparing \CFA with pthreads, Java OpenJDK-9, Go 1.12.6 and \uC 7.0.0.
 For comparison, the package must be multi-processor (M:N), which excludes libdill/libmil~\cite{libdill} (M:1)), and use a shared-memory programming model, \eg not message passing.
 The benchmark computer is an AMD Opteron\texttrademark\ 6380 NUMA 64-core, 8 socket, 2.5 GHz processor, running Ubuntu 16.04.6 LTS, and pthreads/\CFA/\uC are compiled with gcc 9.2.1.
+The benchmark computer is an AMD Opteron\texttrademark\ 6380 NUMA 64-core, 8 socket, 2.5 GHz processor, running Ubuntu 16.04.6 LTS, and \CFA/\uC are compiled with gcc 6.5.
 All benchmarks are run using the following harness. (The Java harness is augmented to circumvent JIT issues.)
 \begin{cfa}
+#define BENCH( `run` ) uint64_t start = cputime_ns();  `run;`  double result = (double)(cputime_ns() - start) / N;
+\end{cfa}
+where CPU time in nanoseconds is from the appropriate language clock.
+Each benchmark is performed @N@ times, where @N@ is selected so the benchmark runs in the range of 2--20 seconds for the specific programming language.
+The total time is divided by @N@ to obtain the average time for a benchmark.
+Each benchmark experiment is run 13 times and the average appears in the table.
+unsigned int N = 10_000_000;
+#define BENCH( `run` ) Time before = getTimeNsec();  `run;`  Duration result = (getTimeNsec() - before) / N;
+\end{cfa}
+The method used to get time is @clock_gettime( CLOCK_REALTIME )@.
+Each benchmark is performed @N@ times, where @N@ varies depending on the benchmark;
+the total time is divided by @N@ to obtain the average time for a benchmark.
+Each benchmark experiment is run 31 times.
 All omitted tests for other languages are functionally identical to the \CFA tests and available online~\cite{CforallBenchMarks}.
+% tar --exclude-ignore=exclude -cvhf benchmark.tar benchmark
+\paragraph{Context Switching}
+% tar --exclude=.deps --exclude=Makefile --exclude=Makefile.in --exclude=c.c --exclude=cxx.cpp --exclude=fetch_add.c -cvhf benchmark.tar benchmark
+\paragraph{Object Creation}
+Object creation is measured by creating/deleting the specific kind of concurrent object.
+Figure~\ref{f:creation} shows the code for \CFA, with results in Table~\ref{tab:creation}.
+The only note here is that the call stacks of \CFA coroutines are lazily created, therefore without priming the coroutine to force stack creation, the creation cost is artificially low.
+\begin{multicols}{2}
+\lstset{language=CFA,moredelim=**[is][\color{red}]{@}{@},deletedelim=**[is][]{`}{`}}
+\begin{cfa}
+@thread@ MyThread {};
+void @main@( MyThread & ) {}
+int main() {
+        BENCH( for ( N ) { @MyThread m;@ } )
+        sout | result`ns;
+}
+\end{cfa}
+\captionof{figure}{\CFA object-creation benchmark}
+\label{f:creation}
+\columnbreak
+\vspace*{-16pt}
+\captionof{table}{Object creation comparison (nanoseconds)}
+\label{tab:creation}
+\begin{tabular}[t]{@{}r*{3}{D{.}{.}{5.2}}@{}}
+\multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} & \multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
+\CFA Coroutine Lazy             & 13.2          & 13.1          & 0.44          \\
+\CFA Coroutine Eager    & 531.3         & 536.0         & 26.54         \\
+\CFA Thread                             & 2074.9        & 2066.5        & 170.76        \\
+\uC Coroutine                   & 89.6          & 90.5          & 1.83          \\
+\uC Thread                              & 528.2         & 528.5         & 4.94          \\
+Goroutine                               & 4068.0        & 4113.1        & 414.55        \\
+Java Thread                             & 103848.5      & 104295.4      & 2637.57       \\
+Pthreads                                & 33112.6       & 33127.1       & 165.90
+\end{tabular}
+\end{multicols}
+\paragraph{Context-Switching}
 In procedural programming, the cost of a function call is important as modularization (refactoring) increases.
 (In many cases, a compiler inlines function calls to increase the size and number of basic blocks for optimizing.)
 Similarly, when modularization extends to coroutines/threads, the time for a context switch becomes a relevant factor.
+(In many cases, a compiler inlines function calls to eliminate this cost.)
+Similarly, when modularization extends to coroutines/tasks, the time for a context switch becomes a relevant factor.
 The coroutine test is from resumer to suspender and from suspender to resumer, which is two context switches.
-%For async-await systems, the test is scheduling and fulfilling @N@ empty promises, where all promises are allocated before versus interleaved with fulfillment to avoid garbage collection.
-For async-await systems, the test measures the cost of the @await@ expression entering the event engine by awaiting @N@ promises, where each created promise is resolved by an immediate event in the engine (using Node.js @setImmediate@).
 The thread test is using yield to enter and return from the runtime kernel, which is two context switches.
 The difference in performance between coroutine and thread context-switch is the cost of scheduling for threads, whereas coroutines are self-scheduling.
+Figure~\ref{f:ctx-switch} shows the \CFA code for a coroutine/thread with results in Table~\ref{t:ctx-switch}.
+% From: Gregor Richards <gregor.richards@uwaterloo.ca>
+% To: "Peter A. Buhr" <pabuhr@plg2.cs.uwaterloo.ca>
+% Date: Fri, 24 Jan 2020 13:49:18 -0500
+%
+% I can also verify that the previous version, which just tied a bunch of promises together, *does not* go back to the
+% event loop at all in the current version of Node. Presumably they're taking advantage of the fact that the ordering of
+% events is intentionally undefined to just jump right to the next 'then' in the chain, bypassing event queueing
+% entirely. That's perfectly correct behavior insofar as its difference from the specified behavior isn't observable, but
+% it isn't typical or representative of much anything useful, because most programs wouldn't have whole chains of eager
+% promises. Also, it's not representative of *anything* you can do with async/await, as there's no way to encode such an
+% eager chain that way.
+Figure~\ref{f:ctx-switch} only shows the \CFA code for coroutines/threads (other systems are similar) with all results in Table~\ref{tab:ctx-switch}.
 \begin{multicols}{2}
 \lstset{language=CFA,moredelim=**[is][\color{red}]{@}{@},deletedelim=**[is][]{`}{`}}
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
 @coroutine@ C {};
 void main( C & ) { for () { @suspend;@ } }
+@coroutine@ C {} c;
+void main( C & ) { for ( ;; ) { @suspend;@ } }
 int main() { // coroutine test
-        C c;
         BENCH( for ( N ) { @resume( c );@ } )
         sout | result;
+}
 int main() { // thread test
+        sout | result`ns;
+}
+int main() { // task test
         BENCH( for ( N ) { @yield();@ } )
         sout | result;
+        sout | result`ns;
+}
 \end{cfa}
 …
 \vspace*{-16pt}
 \captionof{table}{Context switch comparison (nanoseconds)}
 \label{t:ctx-switch}
+\label{tab:ctx-switch}
 \begin{tabular}{@{}r*{3}{D{.}{.}{3.2}}@{}}
 \multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} &\multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
+C function                      & 1.8           & 1.8           & 0.0   \\
+\CFA generator          & 1.8           & 2.0           & 0.3   \\
+\CFA coroutine          & 32.5          & 32.9          & 0.8   \\
+\CFA thread                     & 93.8          & 93.6          & 2.2   \\
+\uC coroutine           & 50.3          & 50.3          & 0.2   \\
+\uC thread                      & 97.3          & 97.4          & 1.0   \\
+Python generator        & 40.9          & 41.3          & 1.5   \\
+Node.js generator       & 32.6          & 32.2          & 1.0   \\
+Node.js await           & 1852.2        & 1854.7        & 16.4  \\
+Goroutine thread        & 143.0         & 143.3         & 1.1   \\
+Rust thread                     & 332.0         & 331.4         & 2.4   \\
+Java thread                     & 405.0         & 415.0         & 17.6  \\
+Pthreads thread         & 334.3         & 335.2         & 3.9
+C function              & 1.8   & 1.8   & 0.01  \\
+\CFA generator  & 2.4   & 2.2   & 0.25  \\
+\CFA Coroutine  & 36.2  & 36.2  & 0.25  \\
+\CFA Thread             & 93.2  & 93.5  & 2.09  \\
+\uC Coroutine   & 52.0  & 52.1  & 0.51  \\
+\uC Thread              & 96.2  & 96.3  & 0.58  \\
+Goroutine               & 141.0 & 141.3 & 3.39  \\
+Java Thread             & 374.0 & 375.8 & 10.38 \\
+Pthreads Thread & 361.0 & 365.3 & 13.19
 \end{tabular}
 \end{multicols}
+\paragraph{Internal Scheduling}
+Internal scheduling is measured using a cycle of two threads signalling and waiting.
+Figure~\ref{f:schedint} shows the code for \CFA, with results in Table~\ref{t:schedint}.
+\paragraph{Mutual-Exclusion}
+Uncontented mutual exclusion, which frequently occurs, is measured by entering/leaving a critical section.
+For monitors, entering and leaving a monitor function is measured.
+To put the results in context, the cost of entering a non-inline function and the cost of acquiring and releasing a @pthread_mutex@ lock is also measured.
+Figure~\ref{f:mutex} shows the code for \CFA with all results in Table~\ref{tab:mutex}.
 Note, the incremental cost of bulk acquire for \CFA, which is largely a fixed cost for small numbers of mutex objects.
-Java scheduling is significantly greater because the benchmark explicitly creates multiple thread in order to prevent the JIT from making the program sequential, \ie removing all locking.
 \begin{multicols}{2}
 \lstset{language=CFA,moredelim=**[is][\color{red}]{@}{@},deletedelim=**[is][]{`}{`}}
 \begin{cfa}
-volatile int go = 0;
-@condition c;@
 @monitor@ M {} m1/*, m2, m3, m4*/;
+void call( M & @mutex p1/*, p2, p3, p4*/@ ) {
+        @signal( c );@
+}
+void wait( M & @mutex p1/*, p2, p3, p4*/@ ) {
+        go = 1; // continue other thread
+        for ( N ) { @wait( c );@ } );
+}
+thread T {};
+void main( T & ) {
+        while ( go == 0 ) { yield(); } // waiter must start first
+        BENCH( for ( N ) { call( m1/*, m2, m3, m4*/ ); } )
+        sout | result;
+}
+void __attribute__((noinline))
+do_call( M & @mutex m/*, m2, m3, m4*/@ ) {}
 int main() {
+        T t;
+        wait( m1/*, m2, m3, m4*/ );
+}
+\end{cfa}
+\captionof{figure}{\CFA Internal-scheduling benchmark}
+\label{f:schedint}
+        BENCH(
+                for( N ) do_call( m1/*, m2, m3, m4*/ );
+        )
+        sout | result`ns;
+}
+\end{cfa}
+\captionof{figure}{\CFA acquire/release mutex benchmark}
+\label{f:mutex}
 \columnbreak
 \vspace*{-16pt}
+\captionof{table}{Internal-scheduling comparison (nanoseconds)}
+\label{t:schedint}
+\bigskip
+\begin{tabular}{@{}r*{3}{D{.}{.}{5.2}}@{}}
+\multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} & \multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
+\CFA @signal@, 1 monitor        & 364.4         & 364.2         & 4.4           \\
+\CFA @signal@, 2 monitor        & 484.4         & 483.9         & 8.8           \\
+\CFA @signal@, 4 monitor        & 709.1         & 707.7         & 15.0          \\
+\uC @signal@ monitor            & 328.3         & 327.4         & 2.4           \\
+Rust cond. variable                     & 7514.0        & 7437.4        & 397.2         \\
+Java @notify@ monitor           & 9623.0        & 9654.6        & 236.2         \\
+Pthreads cond. variable         & 5553.7        & 5576.1        & 345.6
+\captionof{table}{Mutex comparison (nanoseconds)}
+\label{tab:mutex}
+\begin{tabular}{@{}r*{3}{D{.}{.}{3.2}}@{}}
+\multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} &\multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
+test and test-and-test lock             & 19.1  & 18.9  & 0.40  \\
+\CFA @mutex@ function, 1 arg.   & 45.9  & 46.6  & 1.45  \\
+\CFA @mutex@ function, 2 arg.   & 105.0 & 104.7 & 3.08  \\
+\CFA @mutex@ function, 4 arg.   & 165.0 & 167.6 & 5.65  \\
+\uC @monitor@ member rtn.               & 54.0  & 53.7  & 0.82  \\
+Java synchronized method                & 31.0  & 31.1  & 0.50  \\
+Pthreads Mutex Lock                             & 33.6  & 32.6  & 1.14
 \end{tabular}
 \end{multicols}
 …
 External scheduling is measured using a cycle of two threads calling and accepting the call using the @waitfor@ statement.
 Figure~\ref{f:schedext} shows the code for \CFA with results in Table~\ref{t:schedext}.
+Figure~\ref{f:ext-sched} shows the code for \CFA, with results in Table~\ref{tab:ext-sched}.
 Note, the incremental cost of bulk acquire for \CFA, which is largely a fixed cost for small numbers of mutex objects.
 …
 \vspace*{-16pt}
 \begin{cfa}
+@monitor@ M {} m1/*, m2, m3, m4*/;
+void call( M & @mutex p1/*, p2, p3, p4*/@ ) {}
+void wait( M & @mutex p1/*, p2, p3, p4*/@ ) {
+        for ( N ) { @waitfor( call : p1/*, p2, p3, p4*/ );@ }
+}
+volatile int go = 0;
+@monitor@ M {} m;
 thread T {};
+void __attribute__((noinline))
+do_call( M & @mutex@ ) {}
 void main( T & ) {
+        BENCH( for ( N ) { call( m1/*, m2, m3, m4*/ ); } )
+        sout | result;
+        while ( go == 0 ) { yield(); }
+        while ( go == 1 ) { do_call( m ); }
+}
+int __attribute__((noinline))
+do_wait( M & @mutex@ m ) {
+        go = 1; // continue other thread
+        BENCH( for ( N ) { @waitfor( do_call, m );@ } )
+        go = 0; // stop other thread
+        sout | result`ns;
+}
 int main() {
         T t;
         wait( m1/*, m2, m3, m4*/ );
+        do_wait( m );
+}
 \end{cfa}
 \captionof{figure}{\CFA external-scheduling benchmark}
 \label{f:schedext}
+\label{f:ext-sched}
 \columnbreak
 …
 \vspace*{-16pt}
 \captionof{table}{External-scheduling comparison (nanoseconds)}
 \label{t:schedext}
+\label{tab:ext-sched}
 \begin{tabular}{@{}r*{3}{D{.}{.}{3.2}}@{}}
 \multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} &\multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
+\CFA @waitfor@, 1 monitor       & 367.1 & 365.3 & 5.0   \\
+\CFA @waitfor@, 2 monitor       & 463.0 & 464.6 & 7.1   \\
+\CFA @waitfor@, 4 monitor       & 689.6 & 696.2 & 21.5  \\
+\uC \lstinline[language=uC++]|_Accept| monitor  & 328.2 & 329.1 & 3.4   \\
+Go \lstinline[language=Golang]|select| channel  & 365.0 & 365.5 & 1.2
+\CFA @waitfor@, 1 @monitor@     & 376.4 & 376.8 & 7.63  \\
+\CFA @waitfor@, 2 @monitor@     & 491.4 & 492.0 & 13.31 \\
+\CFA @waitfor@, 4 @monitor@     & 681.0 & 681.7 & 19.10 \\
+\uC @_Accept@                           & 331.1 & 331.4 & 2.66
 \end{tabular}
 \end{multicols}
+\paragraph{Mutual-Exclusion}
+Uncontented mutual exclusion, which frequently occurs, is measured by entering/leaving a critical section.
 For monitors, entering and leaving a monitor function is measured, otherwise the language-appropriate mutex-lock is measured.
 For comparison, a spinning (versus blocking) test-and-test-set lock is presented.
 Figure~\ref{f:mutex} shows the code for \CFA with results in Table~\ref{t:mutex}.
 Note the incremental cost of bulk acquire for \CFA, which is largely a fixed cost for small numbers of mutex objects.
+\paragraph{Internal Scheduling}
+Internal scheduling is measured using a cycle of two threads signalling and waiting.
+Figure~\ref{f:int-sched} shows the code for \CFA, with results in Table~\ref{tab:int-sched}.
+Note, the incremental cost of bulk acquire for \CFA, which is largely a fixed cost for small numbers of mutex objects.
+Java scheduling is significantly greater because the benchmark explicitly creates multiple thread in order to prevent the JIT from making the program sequential, \ie removing all locking.
 \begin{multicols}{2}
 \lstset{language=CFA,moredelim=**[is][\color{red}]{@}{@},deletedelim=**[is][]{`}{`}}
 \begin{cfa}
+@monitor@ M {} m1/*, m2, m3, m4*/;
+call( M & @mutex p1/*, p2, p3, p4*/@ ) {}
+volatile int go = 0;
+@monitor@ M { @condition c;@ } m;
+void __attribute__((noinline))
+do_call( M & @mutex@ a1 ) { @signal( c );@ }
+thread T {};
+void main( T & this ) {
+        while ( go == 0 ) { yield(); }
+        while ( go == 1 ) { do_call( m ); }
+}
+int  __attribute__((noinline))
+do_wait( M & mutex m ) with(m) {
+        go = 1; // continue other thread
+        BENCH( for ( N ) { @wait( c );@ } );
+        go = 0; // stop other thread
+        sout | result`ns;
+}
 int main() {
         BENCH( for( N ) call( m1/*, m2, m3, m4*/ ); )
         sout | result;
+}
 \end{cfa}
 \captionof{figure}{\CFA acquire/release mutex benchmark}
 \label{f:mutex}
+        T t;
+        do_wait( m );
+}
+\end{cfa}
+\captionof{figure}{\CFA Internal-scheduling benchmark}
+\label{f:int-sched}
 \columnbreak
 \vspace*{-16pt}
+\captionof{table}{Mutex comparison (nanoseconds)}
+\label{t:mutex}
+\begin{tabular}{@{}r*{3}{D{.}{.}{3.2}}@{}}
+\multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} &\multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
+test-and-test-set lock                  & 19.1  & 18.9  & 0.4   \\
+\CFA @mutex@ function, 1 arg.   & 48.3  & 47.8  & 0.9   \\
+\CFA @mutex@ function, 2 arg.   & 86.7  & 87.6  & 1.9   \\
+\CFA @mutex@ function, 4 arg.   & 173.4 & 169.4 & 5.9   \\
+\uC @monitor@ member rtn.               & 54.8  & 54.8  & 0.1   \\
+Goroutine mutex lock                    & 34.0  & 34.0  & 0.0   \\
+Rust mutex lock                                 & 33.0  & 33.2  & 0.8   \\
+Java synchronized method                & 31.0  & 31.0  & 0.0   \\
+Pthreads mutex Lock                             & 31.0  & 31.1  & 0.4
+\captionof{table}{Internal-scheduling comparison (nanoseconds)}
+\label{tab:int-sched}
+\bigskip
+\begin{tabular}{@{}r*{3}{D{.}{.}{5.2}}@{}}
+\multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} & \multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
+\CFA @signal@, 1 @monitor@      & 372.6         & 374.3         & 14.17         \\
+\CFA @signal@, 2 @monitor@      & 492.7         & 494.1         & 12.99         \\
+\CFA @signal@, 4 @monitor@      & 749.4         & 750.4         & 24.74         \\
+\uC @signal@                            & 320.5         & 321.0         & 3.36          \\
+Java @notify@                           & 10160.5       & 10169.4       & 267.71        \\
+Pthreads Cond. Variable         & 4949.6        & 5065.2        & 363
 \end{tabular}
 \end{multicols}
-\paragraph{Creation}
-Creation is measured by creating/deleting a specific kind of control-flow object.
-Figure~\ref{f:creation} shows the code for \CFA with results in Table~\ref{t:creation}.
-Note, the call stacks of \CFA coroutines are lazily created on the first resume, therefore the cost of creation with and without a stack are presented.
-\begin{multicols}{2}
-\lstset{language=CFA,moredelim=**[is][\color{red}]{@}{@},deletedelim=**[is][]{`}{`}}
-\begin{cfa}
-@coroutine@ MyCoroutine {};
-void ?{}( MyCoroutine & this ) {
-#ifdef EAGER
-        resume( this );
-#endif
+}
-void main( MyCoroutine & ) {}
-int main() {
-        BENCH( for ( N ) { @MyCoroutine c;@ } )
-        sout | result;
+}
-\end{cfa}
-\captionof{figure}{\CFA creation benchmark}
-\label{f:creation}
-\columnbreak
-\vspace*{-16pt}
-\captionof{table}{Creation comparison (nanoseconds)}
-\label{t:creation}
-\begin{tabular}[t]{@{}r*{3}{D{.}{.}{5.2}}@{}}
-\multicolumn{1}{@{}c}{} & \multicolumn{1}{c}{Median} & \multicolumn{1}{c}{Average} & \multicolumn{1}{c@{}}{Std Dev} \\
-\CFA generator                  & 0.6           & 0.6           & 0.0           \\
-\CFA coroutine lazy             & 13.4          & 13.1          & 0.5           \\
-\CFA coroutine eager    & 144.7         & 143.9         & 1.5           \\
-\CFA thread                             & 466.4         & 468.0         & 11.3          \\
-\uC coroutine                   & 155.6         & 155.7         & 1.7           \\
-\uC thread                              & 523.4         & 523.9         & 7.7           \\
-Python generator                & 123.2         & 124.3         & 4.1           \\
-Node.js generator               & 32.3          & 32.2          & 0.3           \\
-Goroutine thread                & 751.0         & 750.5         & 3.1           \\
-Rust thread                             & 53801.0       & 53896.8       & 274.9         \\
-Java thread                             & 120274.0      & 120722.9      & 2356.7        \\
-Pthreads thread                 & 31465.5       & 31419.5       & 140.4
-\end{tabular}
-\end{multicols}
-\subsection{Discussion}
-Languages using 1:1 threading based on pthreads can at best meet or exceed (due to language overhead) the pthread results.
-Note, pthreads has a fast zero-contention mutex lock checked in user space.
-Languages with M:N threading have better performance than 1:1 because there is no operating-system interactions.
-Languages with stackful coroutines have higher cost than stackless coroutines because of stack allocation and context switching;
-however, stackful \uC and \CFA coroutines have approximately the same performance as stackless Python and Node.js generators.
-The \CFA stackless generator is approximately 25 times faster for suspend/resume and 200 times faster for creation than stackless Python and Node.js generators.
 \section{Conclusion}
 …
 Advanced control-flow will always be difficult, especially when there is temporal ordering and nondeterminism.
 However, many systems exacerbate the difficulty through their presentation mechanisms.
+This paper shows it is possible to understand high-level control-flow using three properties: statefulness, thread, mutual-exclusion/synchronization.
+Combining these properties creates a number of high-level, efficient, and maintainable control-flow types: generator, coroutine, thread, each of which can be a monitor.
+Eliminated from \CFA are barging and spurious wakeup, which are nonintuitive and lead to errors, and having to work with a bewildering set of low-level locks and acquisition techniques.
+\CFA high-level race-free monitors and threads provide the core mechanisms for mutual exclusion and synchronization, without having to resort to magic qualifiers like @volatile@/@atomic@.
+This paper shows it is possible to present a hierarchy of control-flow features, generator, coroutine, thread, and monitor, providing an integrated set of high-level, efficient, and maintainable control-flow features.
+Eliminated from \CFA are spurious wakeup and barging, which are nonintuitive and lead to errors, and having to work with a bewildering set of low-level locks and acquisition techniques.
+\CFA high-level race-free monitors and tasks provide the core mechanisms for mutual exclusion and synchronization, without having to resort to magic qualifiers like @volatile@/@atomic@.
 Extending these mechanisms to handle high-level deadlock-free bulk acquire across both mutual exclusion and synchronization is a unique contribution.
 The \CFA runtime provides concurrency based on a preemptive M:N user-level threading-system, executing in clusters, which encapsulate scheduling of work on multiple kernel threads providing parallelism.
 The M:N model is judged to be efficient and provide greater flexibility than a 1:1 threading model.
 These concepts and the \CFA runtime-system are written in the \CFA language, extensively leveraging the \CFA type-system, which demonstrates the expressiveness of the \CFA language.
 Performance comparisons with other concurrent systems/languages show the \CFA approach is competitive across all basic operations, which translates directly into good performance in well-written applications with advanced control-flow.
 C programmers should feel comfortable using these mechanisms for developing complex control-flow in applications, with the ability to obtain maximum available performance by selecting mechanisms at the appropriate level of need using only calling communication.
+Performance comparisons with other concurrent systems/languages show the \CFA approach is competitive across all low-level operations, which translates directly into good performance in well-written concurrent applications.
+C programmers should feel comfortable using these mechanisms for developing complex control-flow in applications, with the ability to obtain maximum available performance by selecting mechanisms at the appropriate level of need.
 …
 \label{futur:nbio}
 Many modern workloads are not bound by computation but IO operations, common cases being web servers and XaaS~\cite{XaaS} (anything as a service).
+Many modern workloads are not bound by computation but IO operations, a common case being web servers and XaaS~\cite{XaaS} (anything as a service).
 These types of workloads require significant engineering to amortizing costs of blocking IO-operations.
 At its core, non-blocking I/O is an operating-system level feature queuing IO operations, \eg network operations, and registering for notifications instead of waiting for requests to complete.
 …
 \section{Acknowledgements}
 The authors recognize the design assistance of Aaron Moss, Rob Schluntz, Andrew Beach, and Michael Brooks; David Dice for commenting and helping with the Java benchmarks; and Gregor Richards for helping with the Node.js benchmarks.
 This research is funded by a grant from Waterloo-Huawei (\url{http://www.huawei.com}) Joint Innovation Lab. %, and Peter Buhr is partially funded by the Natural Sciences and Engineering Research Council of Canada.
+The authors would like to recognize the design assistance of Aaron Moss, Rob Schluntz, Andrew Beach and Michael Brooks on the features described in this paper.
+Funding for this project has been provided by Huawei Ltd.\ (\url{http://www.huawei.com}). %, and Peter Buhr is partially funded by the Natural Sciences and Engineering Research Council of Canada.
 {%
 \fontsize{9bp}{11.5bp}\selectfont%
+\fontsize{9bp}{12bp}\selectfont%
 \bibliography{pl,local}
 }%

doc/papers/concurrency/examples/Fib.py

-              r7030dab
+              r71d6bd8
         while True:
                 fn = fn1 + fn2; fn2 = fn1; fn1 = fn; yield fn
 f1 = Fib()
 …
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.7 Fib.py" #
+# compile-command: "python3.5 Fib.py" #
 # End: #

doc/papers/concurrency/examples/Fib2.c

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
+void mary() {
+        printf( "MARY\n" );
+}
 #define FIB_INIT { 0 }
 typedef struct { int restart; int fn1, fn2; } Fib;
+typedef struct { int next; int fn1, fn2; } Fib;
 int fib( Fib * f ) {
+        static void * states[] = { &&s0, &&s1, &&s2 };
+        goto *states[f->restart];
+  s0:
+        static void * states[] = { &&s1, &&s2, &&s3 };
+        goto *states[f->next];
+  s1:
+        mary();
         f->fn1 = 0;
         f->restart = 1;
+        f->next = 1;
         return f->fn1;
+  s1:
+  s2:
+        mary();
         f->fn2 = f->fn1;
         f->fn1 = 1;
         f->restart = 2;
+        f->next = 2;
         return f->fn1;
+  s2:;
+  s3:;
+        mary();
         int fn = f->fn1 + f->fn2;
         f->fn2 = f->fn1;

doc/papers/concurrency/examples/Fib2.py

-              r7030dab
+              r71d6bd8
 def Fib():
     fn1, fn = 1, 0
+    fn1, fn = 0, 1
     while True:
         yield fn
+        yield fn1
         fn1, fn = fn, fn1 + fn
 …
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.7 Fib2.py" #
+# compile-command: "python3.5 Fib2.py" #
 # End: #

doc/papers/concurrency/examples/Fib3.c

-              r7030dab
+              r71d6bd8
 typedef struct {
+        int restart, fn1, fn;
+        int fn1, fn;
+        void * next;
 } Fib;
 #define FibCtor { 0, 1, 0 }
+#define FibCtor { 1, 0, NULL }
 Fib * comain( Fib * f ) {
+        static void * states[] = {&&s0, &&s1};
+        goto *states[f->restart];
+  s0: f->restart = 1;
+        if ( __builtin_expect(f->next != 0, 1) ) goto *f->next;
+        f->next = &&s1;
         for ( ;; ) {
                 return f;

doc/papers/concurrency/examples/FibRefactor.py

r7030dab	r71d6bd8
22	22	# Local Variables: #
23	23	# tab-width: 4 #
24		# compile-command: "python3.7 FibRefactor.py" #
	24	# compile-command: "python3.5 FibRefactor.py" #
25	25	# End: #

doc/papers/concurrency/examples/Format.c

-              r7030dab
+              r71d6bd8
 typedef struct {
         int restart, g, b;
+        void * next;
         char ch;
+        int g, b;
 } Fmt;
 void comain( Fmt * f ) {
+        static void * states[] = {&&s0, &&s1};
+        goto *states[f->restart];
+  s0: f->restart = 1;
+        if ( __builtin_expect(f->next != 0, 1) ) goto *f->next;
+        f->next = &&s1;
         for ( ;; ) {
                 for ( f->g = 0; f->g < 5; f->g += 1 ) {                 // groups
                         for ( f->b = 0; f->b < 4; f->b += 1 ) {         // blocks
+                                do {
+                                        return;  s1: ;
+                                } while ( f->ch == '\n' );                              // ignore
+                                return;
+                          s1:;  while ( f->ch == '\n' ) return;         // ignore
                                 printf( "%c", f->ch );                                  // print character
+                        }
 …
 int main() {
         Fmt fmt = { 0 };
+        Fmt fmt = { NULL };
         comain( &fmt );                                                                         // prime
         for ( ;; ) {

doc/papers/concurrency/examples/Format.cc

-              r7030dab
+              r71d6bd8
                         for ( g = 0; g < 5; g += 1 ) { // groups of 5 blocks
                                 for ( b = 0; b < 4; b += 1 ) { // blocks of 4 characters
                                         for ( ;; ) { // for newline characters
+//                                      for ( ;; ) { // for newline characters
                                                 suspend();
                                                 if ( ch != '\n' ) break; // ignore newline
+                                        }
+//                                              if ( ch != '\n' ) break; // ignore newline
+//                                      }
 //                                      cout << ch; // print character
+                                }
 …
 // Local Variables: //
 // tab-width: 4 //
 // compile-command: "u++-work -O2 -nodebug Format.cc" //
+// compile-command: "u++-work -O2 -nodebubg Format.cc" //
 // End: //

doc/papers/concurrency/examples/Format.cfa

-              r7030dab
+              r71d6bd8
                 for ( g = 0; g < 5; g += 1 ) {          // groups of 5 blocks
                         for ( b = 0; b < 4; b += 1 ) {  // blocks of 4 characters
                                 do {
+//                              do {
                                         suspend();
                                 } while ( ch == '\n' || ch == '\t' );
+//                              } while ( ch == '\n' || ch == '\t' );
                                 sout | ch;                                      // print character
+                        }

doc/papers/concurrency/examples/Format.data

r7030dab	r71d6bd8
1		abcdefghijklmnop
2		qrstuvwxyzx
3		xxxxxxxxxxxxx
	1	abcdefghijklmnopqrstuvwxyzxxxxxxxxxxxxxx

doc/papers/concurrency/examples/Format.py

-              r7030dab
+              r71d6bd8
                         for g in range( 5 ):    # groups of 5 blocks
                                 for b in range( 4 ): # blocks of 4 characters
+                                        while True:
+                                                ch = (yield) # receive from send
+                                                if '\n' not in ch:
+                                                        break
+                                        print( ch, end='' ) # receive from send
+                                        print( (yield), end='' ) # receive from send
                                 print( '  ', end='' ) # block separator
                         print()                                 # group separator
 …
                         print()
-input = "abcdefghijklmnop\nqrstuvwx\nyzxxxxxxxxxxxxxx\n"
 fmt = Format()
 next( fmt )                                                     # prime generator
 for i in input:
         fmt.send( i );                          # send to yield
+for i in range( 41 ):
+        fmt.send( 'a' );                                # send to yield
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.7 Format.py" #
+# compile-command: "python3.5 Format.py" #
 # End: #

doc/papers/concurrency/examples/Format1.c

-              r7030dab
+              r71d6bd8
 typedef struct {
         int restart, g, b;
+        void * next;
         char ch;
+        int g, b;
 } Fmt;
 void format( Fmt * f ) {
+        static void * states[] = {&&s0, &&s1};
+        goto *states[f->restart];
+  s0: f->restart = 1;
+        if ( __builtin_expect(f->next != 0, 1) ) goto *f->next;
+        f->next = &&s1;
         for ( ;; ) {
                 for ( f->g = 0; f->g < 5; f->g += 1 ) {                 // groups
                         for ( f->b = 0; f->b < 4; f->b += 1 ) {         // blocks
                                 return;
+                          s1: if ( f->ch == '\0' ) goto fini;           // EOF ?
+                          s1: ;
+                                if ( f->ch == '\0' ) goto fini;                 // EOF ?
                                 while ( f->ch == '\n' ) return;                 // ignore
 //                              printf( "%c", f->ch );                                  // print character
+                                printf( "%c", f->ch );                                  // print character
+                        }
 //                      printf( " " );                                                          // block separator
+                        printf( " " );                                                          // block separator
+                }
 //              printf( "\n" );                                                                 // group separator
+                printf( "\n" );                                                                 // group separator
+        }
   fini:;
 //      if ( f->g != 0 || f->b != 0 ) printf( "\n" );
+  fini:
+        if ( f->g != 0 || f->b != 0 ) printf( "\n" );
+}
 int main() {
         Fmt fmt = { 0 };
+        Fmt fmt = { NULL };
         format( &fmt );                                                                         // prime
+        fmt.ch = 'a';
+        for ( long int i = 0; i < 1000000000; i += 1 ) {
+//              scanf( "%c", &fmt.ch );                                                 // direct read into communication variable
+//        if ( feof( stdin ) ) break;
+        for ( ;; ) {
+                scanf( "%c", &fmt.ch );                                                 // direct read into communication variable
+          if ( feof( stdin ) ) break;
                 format( &fmt );
+        }
         fmt.ch = '\0';                                                                          // sentential (EOF)
+        fmt.ch = '\0';
         format( &fmt );
+}

doc/papers/concurrency/examples/PingPong.c

-              r7030dab
+              r71d6bd8
 typedef struct PingPong {
         int restart;                                                                            // style 1
+        const char * name;
         int N, i;
-        const char * name;
         struct PingPong * partner;
         void * next;                                                                            // style 2
+        void * next;
 } PingPong;
+#define PPCtor( name, N ) { 0, N, 0, name, NULL, NULL }
+#define PPCtor( name, N ) { name, N, 0, NULL, NULL }
 void comain( PingPong * pp ) __attribute__(( noinline ));
 void comain( PingPong * pp ) {
+        if ( __builtin_expect(pp->next != 0, 1) ) goto *pp->next;
 #if 0
+        if ( __builtin_expect(pp->next != 0, 1) ) goto *pp->next;
+        pp->next = &&here;
+                asm( "mov  %0,%%rdi" : "=m" (pp) );
+                asm( "mov  %rdi,%rax" );
+#ifndef OPT
+#ifdef PRINT
+                asm( "add  $16, %rsp" );
+#endif // PRINT
+                asm( "popq %rbp" );
+#endif // ! OPT
+#ifdef OPT
+#ifdef PRINT
+                asm( "popq %rbx" );
+#endif // PRINT
+#endif // OPT
+                asm( "jmp  comain" );
+  here: ;
+#endif // 0
         pp->next = &&cycle;
         for ( ; pp->i < pp->N; pp->i += 1 ) {
 …
           cycle: ;
         } // for
-#endif // 0
-#if 1
-        static void * states[] = {&&s0, &&s1};
-        goto *states[pp->restart];
-  s0: pp->restart = 1;
-        for ( ; pp->i < pp->N; pp->i += 1 ) {
-#ifdef PRINT
-                printf( "%s %d\n", pp->name, pp->i );
-#endif // PRINT
-                asm( "mov  %0,%%rdi" : "=m" (pp->partner) );
-                asm( "mov  %rdi,%rax" );
-#ifndef OPT
-#ifdef PRINT
-                asm( "add  $16, %rsp" );
-#endif // PRINT
-                asm( "popq %rbp" );
-#endif // ! OPT
-#ifdef OPT
-#ifdef PRINT
-                asm( "popq %rbx" );
-#endif // PRINT
-#endif // OPT
-                asm( "jmp  comain" );
-          s1: ;
-        } // for
-#endif // 0
+}
 …
 // Local Variables: //
 // tab-width: 4 //
 // compile-command: "gcc-9 -g -DPRINT PingPong.c" //
+// compile-command: "gcc-8 -g -DPRINT PingPong.c" //
 // End: //

doc/papers/concurrency/examples/Pingpong.py

-              r7030dab
+              r71d6bd8
 def PingPong( name, N ):
         partner = yield                         # get partner
         yield                                           # resume scheduler
+        partner = (yield)           # get partner
+        yield                       # resume scheduler
         for i in range( N ):
                 print( name )
                 yield partner                   # execute next
+                yield partner           # execute next
         print( "end", name )
 def Scheduler():
+        n = yield                                       # starting coroutine
+        try:
+                while True:
+                        n = next( n )           # schedule coroutine
+        except StopIteration:
+                pass
+        n = (yield)                 # starting coroutine
+        while True:
+                n = next( n )           # schedule coroutine
 pi = PingPong( "ping", 5 )
 po = PingPong( "pong", 5 )
 next( pi )                                              # prime
 pi.send( po )                                   # send partner
 next( po )                                              # prime
 po.send( pi )                                   # send partner
+next( pi )                      # prime
+pi.send( po )                   # send partner
+next( po )                      # prime
+po.send( pi )                   # send partner
 s = Scheduler();
 next( s )                                               # prime
+next( s )                       # prime
 try:
         s.send( pi )                            # start cycle
 except StopIteration:                   # scheduler stopped
         pass
+except StopIteration:
+        print( "scheduler stop" )
 print( "stop" )
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.7 Pingpong.py" #
+# compile-command: "python3.5 Pingpong.py" #
 # End: #

doc/papers/concurrency/examples/ProdCons.py

-              r7030dab
+              r71d6bd8
 def Prod( N ):
         cons = yield                            # get cons
         yield                                           # resume scheduler
+        cons = (yield)              # get cons
+        yield                       # resume scheduler
         for i in range( N ):
                 print( "prod" )
                 yield cons                              # execute next
+                yield cons              # execute next
         print( "end", "prod" )
 def Cons( N ):
         prod = yield                            # get prod
         yield                                           # resume scheduler
+        prod = (yield)              # get prod
+        yield                       # resume scheduler
         for i in range( N ):
                 print( "cons" )
                 yield prod                              # execute next
+                yield prod              # execute next
         print( "end", "cons" )
 def Scheduler():
+        n = yield                                       # starting coroutine
+        try:
+                while True:
+                        n = next( n )           # schedule coroutine
+        except StopIteration:
+                pass
+        n = (yield)                 # starting coroutine
+        while True:
+                n = next( n )           # schedule coroutine
 prod = Prod( 5 )
 cons = Cons( 5 )
 next( prod )                                    # prime
 prod.send( cons )                               # send cons
 next( cons )                                    # prime
 cons.send( prod )                               # send prod
+next( prod )                    # prime
+prod.send( cons )               # send cons
+next( cons )                    # prime
+cons.send( prod )               # send prod
 s = Scheduler();
 next( s )                                               # prime
+next( s )                       # prime
 try:
         s.send( prod )                          # start cycle
 except StopIteration:                   # scheduler stopped
         pass
+except StopIteration:
+        print( "scheduler stop" )
 print( "stop" )
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.7 ProdCons.py" #
+# compile-command: "python3.5 ProdCons.py" #
 # End: #

doc/papers/concurrency/examples/Refactor.py

r7030dab	r71d6bd8
26	26	# Local Variables: #
27	27	# tab-width: 4 #
28		# compile-command: "python3.7 Refactor.py" #
	28	# compile-command: "python3.5 Refactor.py" #
29	29	# End: #

doc/papers/concurrency/figures/FullCoroutinePhases.fig

-              r7030dab
+              r71d6bd8
 -2
 2
 1 0 1 0 7 100 0 -1 0.000 0 0 1 0 5175.000 2437.500 4875 1875 5175 1800 5475 1875
+1 0 1 0 7 100 0 -1 0.000 0 0 1 0 4575.000 2437.500 4275 1875 4575 1800 4875 1875
 1 1.00 45.00 90.00
 1 0 1 0 7 100 0 -1 0.000 0 0 1 0 5175.000 1537.500 5475 2100 5175 2175 4875 2100
+1 0 1 0 7 100 0 -1 0.000 0 0 1 0 4575.000 1537.500 4875 2100 4575 2175 4275 2100
 1 1.00 45.00 90.00
 1 0 1 0 7 50 -1 -1 0.000 0 1 1 0 4807.500 1642.500 4725 1425 4575 1650 4800 1875
+1 0 1 0 7 50 -1 -1 0.000 0 1 1 0 4207.500 1642.500 4125 1425 3975 1650 4200 1875
 1 1.00 45.00 90.00
-1575 1575 2700 2025
 1 0 1 0 7 100 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 …
 1 1.00 45.00 90.00
 1575 2400 1800
+1 0 1 0 7 100 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+1575 3300 1800
+1 0 1 0 7 100 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+2025 3300 2250
+1 0 100 0 0 10 0.0000 2 105 555 2100 1200 creation\001
 1 0 100 0 4 10 0.0000 2 165 300 1725 1950 ping\001
 1 0 100 0 4 10 0.0000 2 135 360 2475 1950 pong\001
+-6
+3075 1575 4200 2025
+3075 1575 4200 2025
+1 0 1 0 7 100 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+1575 3300 1800
+1 0 1 0 7 100 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+1575 3900 1800
+1 0 100 0 4 10 0.0000 2 165 300 3225 1950 ping\001
+1 0 100 0 4 10 0.0000 2 135 360 3975 1950 pong\001
+-6
+-6
+1 0 100 0 4 10 0.0000 2 165 300 3300 1950 ping\001
+1 0 100 0 4 10 0.0000 2 135 360 3300 2400 pong\001
+1 0 100 0 0 10 0.0000 2 105 675 4575 1200 execution\001
+1 0 100 0 4 10 0.0000 2 165 300 4275 2025 ping\001
+1 0 100 0 4 10 0.0000 2 135 360 4875 2025 pong\001
+1 0 100 0 0 10 0.0000 2 90 420 3300 1200 starter\001
 1 0 100 0 4 10 0.0000 2 165 705 2100 1500 pgm main\001
+1 0 100 0 4 10 0.0000 2 165 705 3600 1500 pgm main\001
+1 0 100 0 4 10 0.0000 2 165 300 4875 2025 ping\001
+1 0 100 0 4 10 0.0000 2 135 360 5475 2025 pong\001
+1 0 100 0 4 10 0.0000 2 165 705 5100 1500 pgm main\001
+1 0 100 0 2 10 0.0000 2 105 540 2100 1275 creator\001
+1 0 100 0 2 10 0.0000 2 105 495 3600 1275 starter\001
+1 0 100 0 2 10 0.0000 2 105 690 5175 1275 execution\001
+1 0 100 0 4 10 0.0000 2 165 705 3300 1500 pgm main\001
+1 0 100 0 4 10 0.0000 2 165 705 4500 1500 pgm main\001

doc/papers/concurrency/figures/RunTimeStructure.fig

-              r7030dab
+              r71d6bd8
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4500 3600 15 15 4500 3600 4515 3615
 -6
+3225 4125 4650 4425
+4350 4200 4650 4350
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4425 4275 15 15 4425 4275 4440 4290
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4500 4275 15 15 4500 4275 4515 4290
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4575 4275 15 15 4575 4275 4590 4290
+2175 4650 7050 4950
+3 0 1 0 0 0 0 0 0.000 1 0.0000 2250 4830 30 30 2250 4830 2280 4860
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4200 4800 150 75 4200 4800 4350 4875
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3275 4800 100 100 3275 4800 3375 4800
+2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
+4950 5400 4725 5175 4725 5175 4950 5400 4950
+2 1 1 -1 -1 0 0 -1 3.000 0 0 0 0 0 5
+4950 6300 4950 6300 4725 6525 4725 6525 4950
+0 -1 0 0 0 10 0.0000 2 105 450 6600 4875 cluster\001
+0 -1 0 0 0 10 0.0000 2 105 660 5475 4875 processor\001
+0 -1 0 0 0 10 0.0000 2 105 555 4425 4875 monitor\001
+0 -1 0 0 0 10 0.0000 2 120 270 3450 4875 task\001
+0 -1 0 0 0 10 0.0000 2 105 660 2325 4875 coroutine\001
 -6
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3450 4275 225 150 3450 4275 3675 4425
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4050 4275 225 150 4050 4275 4275 4425
+3450 1275 3750 1425
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3525 1350 15 15 3525 1350 3540 1365
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3600 1350 15 15 3600 1350 3615 1365
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3675 1350 15 15 3675 1350 3690 1365
 -6
+6675 4125 7500 4425
+7200 4200 7500 4350
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 7275 4275 15 15 7275 4275 7290 4290
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 7350 4275 15 15 7350 4275 7365 4290
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 7425 4275 15 15 7425 4275 7440 4290
+-6
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 6900 4275 225 150 6900 4275 7125 4425
+-6
+6675 3525 8025 3975
+1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+3750 6975 3750
+1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+3750 7350 3750
+2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
+3975 7800 3525 7350 3525 7350 3975 7800 3975
+1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+3750 8025 3750
+5550 1275 5850 1425
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 5625 1350 15 15 5625 1350 5640 1365
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 5700 1350 15 15 5700 1350 5715 1365
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 5775 1350 15 15 5775 1350 5790 1365
 -6
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 5550 2625 150 150 5550 2625 5700 2625
 …
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4425 2850 150 150 4425 2850 4575 2850
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4650 2475 150 150 4650 2475 4800 2475
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3525 3600 150 150 3525 3600 3675 3600
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3975 3600 150 150 3975 3600 4125 3600
 3 0 1 0 0 0 0 0 0.000 1 0.0000 3525 3600 30 30 3525 3600 3555 3630
 …
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3975 2850 150 150 3975 2850 4125 2850
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 7200 2775 150 150 7200 2775 7350 2775
+3 0 1 0 0 0 0 0 0.000 1 0.0000 2250 4830 30 30 2250 4830 2280 4860
+3 0 1 0 0 0 0 0 0.000 1 0.0000 7200 2775 30 30 7200 2775 7230 2805
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3525 3600 150 150 3525 3600 3675 3600
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3875 4800 100 100 3875 4800 3975 4800
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4650 4800 150 75 4650 4800 4800 4875
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4650 1350 225 150 4650 1350 4875 1500
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 5250 1350 225 150 5250 1350 5475 1500
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4050 1350 225 150 4050 1350 4275 1500
 2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
 4200 2400 3750 1950 3750 1950 4200 2400 4200
 …
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
+3975 6975 3975
+1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
 2775 6825 2775
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 0 0 2
+2775 6825 3750
+2775 6825 3975
+1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+3975 7350 3975
+2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
+4200 7800 3750 7350 3750 7350 4200 7800 4200
+1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
+1 1.00 45.00 90.00
+3975 8025 3975
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 4
 1 1.00 45.00 90.00
+3750 7875 2325 7200 2325 7200 2550
+2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
+4950 5850 4725 5625 4725 5625 4950 5850 4950
+2 1 1 -1 -1 0 0 -1 3.000 0 0 0 0 0 5
+4950 6750 4950 6750 4725 6975 4725 6975 4950
+3975 7875 2325 7200 2325 7200 2550
 1 -1 0 0 0 10 0.0000 2 105 720 5550 4425 Processors\001
 1 -1 0 0 0 10 0.0000 2 120 1005 4200 3225 Blocked Tasks\001
 …
 1 -1 0 0 0 10 0.0000 2 105 990 2175 3525 Discrete-event\001
 1 -1 0 0 0 10 0.0000 2 135 795 2175 4350 preemption\001
-0 -1 0 0 0 10 0.0000 2 150 1290 2325 4875 genrator/coroutine\001
-0 -1 0 0 0 10 0.0000 2 120 270 4050 4875 task\001
-0 -1 0 0 0 10 0.0000 2 105 450 7050 4875 cluster\001
-0 -1 0 0 0 10 0.0000 2 105 660 5925 4875 processor\001
-0 -1 0 0 0 10 0.0000 2 105 555 4875 4875 monitor\001

doc/papers/concurrency/mail2

-              r7030dab
+              r71d6bd8
 Software: Practice and Experience Editorial Office
-Date: Tue, 12 Nov 2019 22:25:17 +0000
-From: Richard Jones <onbehalfof@manuscriptcentral.com>
-Reply-To: R.E.Jones@kent.ac.uk
-To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
-Subject: Software: Practice and Experience - Decision on Manuscript ID
- SPE-19-0219
--Nov-2019
-Dear Dr Buhr,
-Many thanks for submitting SPE-19-0219 entitled "Advanced Control-flow and Concurrency in Cforall" to Software: Practice and Experience. The paper has now been reviewed and the comments of the referees are included at the bottom of this letter.
-The decision on this paper is that it requires substantial further work is required. The referees have a number of substantial concerns. All the reviewers found the submission very hard to read; two of the reviewers state that it needs very substantial restructuring. These concerns must be addressed before your submission can be considered further.
-A revised version of your manuscript that takes into account the comments of the referees will be reconsidered for publication.
-Please note that submitting a revision of your manuscript does not guarantee eventual acceptance, and that your revision will be subject to re-review by the referees before a decision is rendered.
-You have 90 days from the date of this email to submit your revision. If you are unable to complete the revision within this time, please contact me to request an extension.
-You can upload your revised manuscript and submit it through your Author Center. Log into https://mc.manuscriptcentral.com/spe  and enter your Author Center, where you will find your manuscript title listed under "Manuscripts with Decisions".
-When submitting your revised manuscript, you will be able to respond to the comments made by the referee(s) in the space provided.  You can use this space to document any changes you make to the original manuscript.
-If you feel that your paper could benefit from English language polishing, you may wish to consider having your paper professionally edited for English language by a service such as Wiley's at http://wileyeditingservices.com. Please note that while this service will greatly improve the readability of your paper, it does not guarantee acceptance of your paper by the journal.
-Once again, thank you for submitting your manuscript to Software: Practice and Experience and I look forward to receiving your revision.
-Sincerely,
-Prof. Richard Jones
-Software: Practice and Experience
-R.E.Jones@kent.ac.uk
-Referee(s)' Comments to Author:
-Reviewing: 1
-Comments to the Author
-This article presents the design and rationale behind the various
-threading and synchronization mechanisms of C-forall, a new low-level
-programming language.  This paper is very similar to a companion paper
-which I have also received: as the papers are similar, so will these
-reviews be --- in particular any general comments from the other
-review apply to this paper also.
-As far as I can tell, the article contains three main ideas: an
-asynchronous execution / threading model; a model for monitors to
-provide mutual exclusion; and an implementation.  The first two ideas
-are drawn together in Table 1: unfortunately this is on page 25 of 30
-pages of text. Implementation choices and descriptions are scattered
-throughout the paper - and the sectioning of the paper seems almost
-arbitrary.
-The article is about its contributions.  Simply adding feature X to
-language Y isn't by itself a contribution, (when feature X isn't
-already a contribution).  The contribution can be in the design: the
-motivation, the space of potential design options, the particular
-design chosen and the rationale for that choice, or the resulting
-performance.  For example: why support two kinds of generators as well
-as user-level threads?  Why support both low and high level
-synchronization constructs?  Similarly I would have found the article
-easier to follow if it was written top down, presenting the design
-principles, present the space of language features, justify chosen
-language features (and rationale) and those excluded, and then present
-implementation, and performance.
-Then the writing of the article is often hard to follow, to say the
-least. Two examples: section 3 "stateful functions" - I've some idea
-what that is (a function with Algol's "own" or C's "static" variables?
-but in fact the paper has a rather more specific idea than that. The
-top of page 3 throws a whole lot of defintions at the reader
-"generator" "coroutine" "stackful" "stackless" "symmetric"
-"asymmetric" without every stopping to define each one --- but then in
-footnote "C" takes the time to explain what C's "main" function is?  I
-cannot imagine a reader of this paper who doesn't know what "main" is
-in C; especially if they understand the other concepts already
-presented in the paper.  The start of section 3 then does the same
-thing: putting up a whole lot of definitions, making distinctions and
-comparisons, even talking about some runtime details, but the critical
-definition of a monitor doesn't appear until three pages later, at the
-start of section 5 on p15, lines 29-34 are a good, clear, description
-of what a monitor actually is.  That needs to come first, rather than
-being buried again after two sections of comparisons, discussions,
-implementations, and options that are ungrounded because they haven't
-told the reader what they are actually talking about.  First tell the
-reader what something is, then how they might use it (as programmers:
-what are the rules and restrictions) and only then start comparison
-with other things, other approaches, other languages, or
-implementations.
-The description of the implementation is similarly lost in the trees
-without ever really seeing the wood. Figure 19 is crucial here, but
-it's pretty much at the end of the paper, and comments about
-implementations are threaded throughout the paper without the context
-(fig 19) to understand what's going on.   The protocol for performance
-testing may just about suffice for C (although is N constantly ten
-million, or does it vary for each benchmark) but such evaluation isn't
-appropriate for garbage-collected or JITTed languages like Java or Go.
-other comments working through the paper - these are mostly low level
-and are certainly not comprehensive.
-p1 only a subset of C-forall extensions?
-p1 "has features often associated with object-oriented programming
-languages, such as constructors, destructors, virtuals and simple
-inheritance."   There's no need to quibble about this. Once a language
-has inheritance, it's hard to claim it's not object-oriented.
-p2 barging? signals-as-hints?
-p3 start your discussion of generations with a simple example of a
-C-forall generator.  Fig 1(b) might do: but put it inline instead of
-the python example - and explain the key rules and restrictions on the
-construct.  Then don't even start to compare with coroutines until
-you've presented, described and explained your coroutines...
-p3 I'd probably leave out the various "C" versions unless there are
-key points to make you can't make in C-forall. All the alternatives
-are just confusing.
-p4 but what's that "with" in Fig 1(B)
-p5 start with the high level features of C-forall generators...
-p5 why is the paper explaining networking protocols?
-p7 lines 1-9 (transforming generator to coroutine - why would I do any
-of this? Why would I want one instead of the other (do not use "stack"
-in your answer!)
-p10 last para "A coroutine must retain its last resumer to suspend
-back because the resumer is on a different stack. These reverse
-pointers allow suspend to cycle backwards, "  I've no idea what is
-going on here?  why should I care?  Shouldn't I just be using threads
-instead?  why not?
-p16 for the same reasons - what reasons?
-p17 if the multiple-monitor entry procedure really is novel, write a
-paper about that, and only about that.
-p23 "Loose Object Definitions" - no idea what that means.  in that
-section: you can't leave out JS-style dynamic properties.  Even in
-OOLs that (one way or another) allow separate definitions of methods
-(like Objective-C, Swift, Ruby, C#) at any time a runtime class has a
-fixed definition.  Quite why the detail about bit mask implementation
-is here anyway, I've no idea.
-p25 this cluster isn't a CLU cluster then?
-* conclusion should conclude the paper, not the related.
-Reviewing: 2
-Comments to the Author
-This paper describes the concurrency features of an extension of C (whose name I will write as "C\/" here, for convenience), including much design-level discussion of the coroutine- and monitor-based features and some microbenchmarks exploring the current implementation's performance. The key message of the latter is that the system's concurrency abstractions are much lighter-weight than the threading found in mainstream C or Java implementations.
-There is much description of the system and its details, but nothing about (non-artificial) uses of it. Although the microbenchmark data is encouraging, arguably not enough practical experience with the system has been reported here to say much about either its usability advantages or its performance.
-As such, the main contribution of the paper seem to be to document the existence of the described system and to provide a detailed design rationale and (partial) tutorial. I believe that could be of interest to some readers, so an acceptable manuscript is lurking in here somewhere.
-Unfortunately, at present the writing style is somewhere between unclear and infuriating. It omits to define terms; it uses needlessly many terms for what are apparently (but not clearly) the same things; it interrupts itself rather than deliver the natural consequent of whatever it has just said; and so on. Section 5 is particularly bad in these regards -- see my detailed comments below. Fairly major additional efforts will be needed to turn the present text into a digestible design-and-tutorial document. I suspect that a shorter paper could do this job better than the present manuscript, which is overwrought in parts.
-p2: lines 4--9 are a little sloppy. It is not the languages but their popular implementations which "adopt" the 1:1 kernel threading model.
-line 10: "medium work" -- "medium-sized work"?
-line 18: "is all sequential to the compiler" -- not true in modern compilers, and in 2004 H-J Boehm wrote a tech report describing exactly why ("Threads cannot be implemented as a library", HP Labs).
-line 20: "knows the optimization boundaries" -- I found this vague. What's an example?
-line 31: this paragraph has made a lot of claims. Perhaps forward-reference to the parts of the paper that discuss each one.
-line 33: "so the reader can judge if" -- this reads rather passive-aggressively. Perhaps better: "... to support our argument that..."
-line 41: "a dynamic partitioning mechanism" -- I couldn't tell what this meant
-p3. Presenting concept of a "stateful function" as a new language feature seems odd. In C, functions often have local state thanks to static local variables (or globals, indeed). Of course, that has several limitations. Can you perhaps present your contributions by enumerating these limitations? See also my suggestion below about a possible framing centred on a strawman.
-line 2: "an old idea that is new again" -- this is too oblique
-lines 2--15: I found this to be a word/concept soup. Stacks, closures, generators, stackless stackful, coroutine, symmetric, asymmetric, resume/suspend versus resume/resume... there needs to be a more gradual and structured way to introduce all this, and ideally one that minimises redundancy. Maybe present it as a series of "definitions" each with its own heading, e.g. "A closure is stackless if its local state has statically known fixed size"; "A generator simply means a stackless closure." And so on. Perhaps also strongly introduce the word "activate" as a direct contrast with resume and suspend. These are just a flavour of the sort of changes that might make this paragraph into something readable.
-Continuing the thought: I found it confusing that by these definitinos, a stackful closure is not a stack, even though logically the stack *is* a kind of closure (it is a representation of the current thread's continuation).
-lines 24--27: without explaining what the boost functor types mean, I don't think the point here comes across.
-line 34: "semantically coupled" -- I wasn't surew hat this meant
-p4: the point of Figure 1 (C) was not immediately clear. It seem to be showing how one might "compile down" Figure 1 (B). Or is that Figure 1 (A)?
-It's right that the incidental language features of the system are not front-and-centre, but I'd appreciate some brief glossing of non-C languages features as they appear. Examples are the square bracket notation, the pipe notation and the constructor syntax. These explanations could go in the caption of the figure which first uses them, perhaps. Overall I found the figure captions to be terse, and a missed opportunity to explain clearly what was going on.
-p5 line 23: "This restriction is removed..." -- give us some up-front summary of your contributions and the elements of the language design that will be talked about, so that this isn't an aside. This will reduce the "twisty passages" feeling that characterises much of the paper.
-line 40: "a killer asymmetric generator" -- this is stylistically odd, and the sentence about failures doesn't convincigly argue that C\/ will help with them. Have you any experience writing device drivers using C\/? Or any argument that the kinds of failures can be traced to the "stack-ripping" style that one is forced to use without coroutines? Also, a typo on line 41: "device drives". And saying "Windows/Linux" is sloppy... what does the cited paper actually say?
-p6 lines 13--23: this paragraph is difficult to understand. It seems to be talking about a control-flow pattern roughly equivalent to tail recursion. What is the high-level point, other than that this is possible?
-line 34: "which they call coroutines" -- a better way to make this point is presumably that the C++20 proposal only provides a specialised kind of coroutine, namely generators, despite its use of the more general word.
-line 47: "... due to dynamic stack allocation, execution..." -- this sentence doesn't scan. I suggest adding "and for" in the relevant places where currently there are only commas.
-p8 / Figure 5 (B) -- the GNU C extension of unary "&&" needs to be explained. The whole figure needs a better explanation, in fact.
-p9, lines 1--10: I wasn't sure this stepping-through really added much value. What are the truly important points to note about this code?
-p10: similarly, lines 3--27 again are somewhere between tedious and confusing. I'm sure the motivation and details of "starter semantics" can both be stated much more pithily.
-line 32: "a self-resume does not overwrite the last resumer" -- is this a hack or a defensible principled decision?
-p11: "a common source of errors" -- among beginners or among production code? Presumably the former.
-line 23: "with builtin and library" -- not sure what this means
-lines 31--36: these can be much briefer. The only important point here seems to be that coroutines cannot be copied.
-p12: line 1: what is a "task"? Does it matter?
-line 7: calling it "heap stack" seems to be a recipe for confusion. "Stack-and-heap" might be better, and contrast with "stack-and-VLS" perhaps. When "VLS" is glossed, suggest actually expanding its initials: say "length" not "size".
-line 21: are you saying "cooperative threading" is the same as "non-preemptive scheduling", or that one is a special case (kind) of the other? Both are defensible, but be clear.
-line 27: "mutual exclusion and synchronization" -- the former is a kind of the latter, so I suggest "and other forms of synchronization".
-line 30: "can either be a stackless or stackful" -- stray "a", but also, this seems to be switching from generic/background terminology to C\/-specific terminology.
-An expositional idea occurs: start the paper with a strawman naive/limited realisation of coroutines -- say, Simon Tatham's popular "Coroutines in C" web page -- and identify point by point what the limitations are and how C\/ overcomes them. Currently the presentation is often flat (lacking motivating contrasts) and backwards (stating solutions before problems). The foregoing approach might fix both of these.
-page 13: line 23: it seems a distraction to mention the Python feature here.
-p14 line 5: it seems odd to describe these as "stateless" just because they lack shared mutable state. It means the code itself is even more stateful. Maybe the "stack ripping" argument could usefully be given here.
-line 16: "too restrictive" -- would be good to have a reference to justify this, or at least give a sense of what the state-of-the-art performance in transactional memory systems is (both software and hardware)
-line 22: "simulate monitors" -- what about just *implementing* monitors? isn't that what these systems do? or is the point more about refining them somehow into something more specialised?
-p15: sections 4.1 and 4.2 seem adrift and misplaced. Split them into basic parts (which go earlier) and more advanced parts (e.g. barging, which can be explained later).
-line 31: "acquire/release" -- misses an opportunity to contrast the monitor's "enter/exit" abstraction with the less structured acquire/release of locks.
-p16 line 12: the "implicit" versus "explicit" point is unclear. Is it perhaps about the contract between an opt-in *discipline* and a language-enforced *guarantee*?
-line 28: no need to spend ages dithering about which one is default and which one is the explicit qualifier. Tell us what you decided, briefly justify it, and move on.
-p17: Figure 11: since the main point seems to be to highlight bulk acquire, include a comment which identifies the line where this is happening.
-line 2: "impossible to statically..." -- or dynamically. Doing it dynamically would be perfectly acceptable (locking is a dynamic operation after all)
-"guarantees acquisition order is consistent" -- assuming it's done in a single bulk acquire.
-p18: section 5.3: the text here is a mess. The explanations of "internal" versus "external" scheduling are unclear, and "signals as hints" is not explained. "... can cause thread starvation" -- means including a while loop, or not doing so? "There are three signalling mechanisms.." but the text does not follow that by telling us what they are. My own scribbled attempt at unpicking the internal/external thing: "threads already in the monitor, albeit waiting, have priority over those trying to enter".
-p19: line 3: "empty condition" -- explain that condition variables don't store anything. So being "empty" means that the queue of waiting threads (threads waiting to be signalled that the condition has become true) is empty.
-line 6: "... can be transformed into external scheduling..." -- OK, but give some motivation.
-p20: line 6: "mechnaism"
-lines 16--20: this is dense and can probably only be made clear with an example
-p21 line 21: clarify that nested monitor deadlock was describe earlier (in 5.2). (Is the repetition necessary?)
-line 27: "locks, and by extension monitors" -- this is true but the "by extension" argument is faulty. It is perfectly possible to use locks as a primitive and build a compositional mechanism out of them, e.g. transactions.
-p22 line 2: should say "restructured"
-line 33: "Implementing a fast subset check..." -- make clear that the following section explains how to do this. Restructuring the sections themselves could do this, or noting in the text.
-p23: line 3: "dynamic member adding, eg, JavaScript" -- needs to say "as permitted in JavaScript", and "dynamically adding members" is stylistically better
-p23: line 18: "urgent stack" -- back-reference to where this was explained before
-p24 line 7: I did not understand what was more "direct" about "direct communication". Also, what is a "passive monitor" -- just a monitor, given that monitors are passive by design?
-line 14 / section 5.9: this table was useful and it (or something like it) could be used much earlier on to set the structure of the rest of the paper. The explanation at present is too brief, e.g. I did not really understand the point about cases 7 and 8.
-p25 line 2: instead of casually dropping in a terse explanation for the newly intrdouced term "virtual processor", introduce it properly. Presumably the point is to give a less ambiguous meaning to "thread" by reserving it only for C\/'s green threads.
-Table 1: what does "No / Yes" mean?
-p26 line 15: "transforms user threads into fibres" -- a reference is needed to explain what "fibres" means... guessing it's in the sense of Adya et al.
-line 20: "Microsoft runtime" -- means Windows?
-lines 21--26: don't say "interrupt" to mean "signal", especially not without clear introduction. You can use "POSIX signal" to disambiguate from condition variables' "signal".
-p27 line 3: "frequency is usually long" -- that's a "time period" or "interval", not a frequency
-line 5: the lengthy quotation is not really necessary; just paraphrase the first sentence and move on.
-line 20: "to verify the implementation" -- I don't think that means what is intended
-Tables in section 7 -- too many significant figures. How many overall runs are described? What is N in each case?
-p29 line 2: "to eliminate this cost" -- arguably confusing since nowadays on commodity CPUs most of the benefits of inlining are not to do with call overheads, but from later optimizations enabled as a consequence of the inlining
-line 41: "a hierarchy" -- are they a hierarchy? If so, this could be explained earlier. Also, to say these make up "an integrated set... of control-flow features" verges on the tautologous.
-p30 line 15: "a common case being web servers and XaaS" -- that's two cases
-Reviewing: 3
-Comments to the Author
-# Cforall review
-Overall, I quite enjoyed reading the paper. Cforall has some very interesting ideas. I did have some suggestions that I think would be helpful before final publication. I also left notes on various parts of the paper that I find confusing when reading, in hopes that it may be useful to you.
-## Summary
-* Expand on the motivations for including both generator and coroutines, vs trying to build one atop the other
-* Expand on the motivations for having Why both symmetric and asymettric coroutines?
-* Comparison to async-await model adopted by other languages
-    * C#, JS
-    * Rust and its async/await model
-* Consider performance comparisons against node.js and Rust frameworks
-* Discuss performance of monitors vs finer-grained memory models and atomic operations found in other languages
-* Why both internal/external scheduling for synchronization?
-## Generator/coroutines
-In general, this section was clear, but I thought it would be useful to provide a somewhat deeper look into why Cforall opted for the particular combination of features that it offers. I see three main differences from other languages:
-* Generators are not exposed as a "function" that returns a generator object, but rather as a kind of struct, with communication happening via mutable state instead of "return values". That is, the generator must be manually resumed and (if I understood) it is expected to store values that can then later be read (perhaps via methods), instead of having a `yield <Expr>` statement that yields up a value explicitly.
-* Both "symmetric" and "asymmetric" generators are supported, instead of only asymmetric.
-* Coroutines (multi-frame generators) are an explicit mechanism.
-In most other languages, coroutines are rather built by layering single-frame generators atop one another (e.g., using a mechanism like async-await), and symmetric coroutines are basically not supported. I'd like to see a bit more justification for Cforall including all the above mechanisms -- it seemed like symmetric coroutines were a useful building block for some of the user-space threading and custom scheduler mechanisms that were briefly mentioned later in the paper.
-In the discussion of coroutines, I would have expected a bit more of a comparison to the async-await mechanism offered in other languages. Certainly the semantics of async-await in JavaScript implies significantly more overhead (because each async fn is a distinct heap object). [Rust's approach avoids this overhead][zc], however, and might be worthy of a comparison (see the Performance section).
-## Locks and threading
-### Comparison to atomics overlooks performance
-There are several sections in the paper that compare against atomics -- for example, on page 15, the paper shows a simple monitor that encapsulates an integer and compares that to C++ atomics. Later, the paper compares the simplicity of monitors against the `volatile` quantifier from Java. The conclusion in section 8 also revisits this point.
-While I agree that monitors are simpler, they are obviously also significantly different from a performance perspective -- the paper doesn't seem to address this at all. It's plausible that (e.g.) the `Aint` monitor type described in the paper can be compiled and mapped to the specialized instructions offered by hardware, but I didn't see any mention of how this would be done. There is also no mention of the more nuanced memory ordering relations offered by C++11 and how one might achieve similar performance characteristics in Cforall (perhaps the answer is that one simply doesn't need to; I think that's defensible, but worth stating explicitly).
-### Justification for external scheduling feels lacking
-Cforall includes both internal and external scheduling; I found the explanation for the external scheduling mechanism to be lacking in justification. Why include both mechanisms when most languages seem to make do with only internal scheduling? It would be useful to show some scenarios where external scheduling is truly more powerful.
-I would have liked to see some more discussion of external scheduling and how it  interacts with software engineering best practices. It seems somewhat similar to AOP in certain regards. It seems to add a bit of "extra semantics" to monitor methods, in that any method may now also become a kind of synchronization point. The "open-ended" nature of this feels like it could easily lead to subtle bugs, particularly when code refactoring occurs (which may e.g. split an existing method into two). This seems particularly true if external scheduling can occur across compilation units -- the paper suggested that this is true, but I wasn't entirely clear.
-I would have also appreciated a few more details on how external scheduling is implemented. It seems to me that there must be some sort of "hooks" on mutex methods so that they can detect whether some other function is waiting on them and awaken those blocked threads. I'm not sure how such hooks are inserted, particularly across compilation units. The material in Section 5.6 didn't quite clarify the matter for me. For example, it left me somewhat confused about whether the `f` and `g` functions declared were meant to be local to a translation unit, or shared with other unit.
-### Presentation of monitors is somewhat confusing
-I found myself confused fairly often in the section on monitors. I'm just going to leave some notes here on places that I got confused in how that it could be useful to you as feedback on writing that might want to be clarified.
-To start, I did not realize that the `mutex_opt` notation was a keyword, I thought it was a type annotation. I think this could be called out more explicitly.
-Later, in section 5.2, the paper discusses `nomutex` annotations, which initially threw me, as they had not been introduced (now I realize that this paragraph is there to justify why there is no such keyword). The paragraph might be rearranged to make that clearer, perhaps by leading with the choice that Cforall made.
-On page 17, the paper states that "acquiring multiple monitors is safe from deadlock", but this could be stated a bit more precisely: acquiring multiple monitors in a bulk-acquire is safe from deadlock (deadlock can still result from nested acquires).
-On page 18, the paper states that wait states do not have to be enclosed in loops, as there is no concern of barging. This seems true but there are also other reasons to use loops (e.g., if there are multiple reasons to notify on the same condition). Thus the statement initially surprised me, as barging is only one of many reasons that I typically employ loops around waits.
-I did not understand the diagram in Figure 12 for some time. Initially, I thought that it was generic to all monitors, and I could not understand the state space. It was only later that I realized it was specific to your example. Updating the caption from "Monitor scheduling to "Monitor scheduling in the example from Fig 13" might have helped me quite a bit.
-I spent quite some time reading the boy/girl dating example (\*) and I admit I found it somewhat confusing. For example, I couldn't tell whether there were supposed to be many "girl" threads executing at once, or if there was only supposed to be one girl and one boy thread executing in a loop. Are the girl/boy threads supposed to invoke the girl/boy methods or vice versa? Surely there is some easier way to set this up? I believe that when reading the paper I convinced myself of how it was supposed to be working, but I'm writing this review some days later, and I find myself confused all over again and not able to easily figure it out.
-(\*) as an aside, I would consider modifying the example to some other form of matching, like customers and support personnel.
-## Related work
-The paper offered a number of comparisons to Go, C#, Scala, and so forth, but seems to have overlooked another recent language, Rust. In many ways, Rust seems to be closest in philosophy to Cforall, so it seems like an odd omission. I already mentioned above that Rust is in the process of shipping [async-await syntax][aa], which is definitely an alternative to the generator/coroutine approach in Cforall (though one with clear pros/cons).
-## Performance
-In the performance section in particular, you might consider comparing against some of the Rust web servers and threading systems. For example, actix is top of the [single query TechEmpower Framework benchmarks], and tokio is near the top of the [plainthreading benchmarks][pt] (hyper, the top, is more of an HTTP framework, though it is also written in Rust). It would seem worth trying to compare their "context switching" costs as well -- I believe both actix and tokio have a notion of threads that could be readily compared.
-Another addition that might be worth considering is to compare against node.js promises, although I think the comparison to process creation is not as clean.
-That said, I think that the performance comparison is not a big focus of the paper, so it may not be necessary to add anything to it.
-## Authorship of this review
-I'm going to sign this review. This review was authored by Nicholas D. Matsakis. In the intrerest of full disclosure, I'm heavily involved in the Rust project, although I dont' think that influenced this review in particular. Feel free to reach out to me for clarifying questions.
-## Links
-[aa]: https://blog.rust-lang.org/2019/09/30/Async-await-hits-beta.html
-[zc]: https://aturon.github.io/blog/2016/08/11/futures/
-[sq]: https://www.techempower.com/benchmarks/#section=data-r18&hw=ph&test=db
-[pt]: https://www.techempower.com/benchmarks/#section=data-r18&hw=ph&test=plaintext
-Subject: Re: manuscript SPE-19-0219
-To: "Peter A. Buhr" <pabuhr@uwaterloo.ca>
-From: Richard Jones <R.E.Jones@kent.ac.uk>
-Date: Tue, 12 Nov 2019 22:43:55 +0000
-Dear Dr Buhr
-Your should have received a decision letter on this today. I am sorry that this
-has taken so long. Unfortunately SP&E receives a lot of submissions and getting
-reviewers is a perennial problem.
-Regards
-Richard
-Peter A. Buhr wrote on 11/11/2019 13:10:
->     26-Jun-2019
->     Your manuscript entitled "Advanced Control-flow and Concurrency in Cforall"
->     has been received by Software: Practice and Experience. It will be given
->     full consideration for publication in the journal.
+>
-> Hi, it has been over 4 months since submission of our manuscript SPE-19-0219
-> with no response.
+>
-> Currently, I am refereeing a paper for IEEE that already cites our prior SP&E
-> paper and the Master's thesis forming the bases of the SP&E paper under
-> review. Hence our work is apropos and we want to get it disseminates as soon as
-> possible.
+>
-> [3] A. Moss, R. Schluntz, and P. A. Buhr, "Cforall: Adding modern programming
->      language features to C," Software - Practice and Experience, vol. 48,
->      no. 12, pp. 2111-2146, 2018.
+>
-> [4] T. Delisle, "Concurrency in C for all," Master's thesis, University of
->      Waterloo, 2018.  [Online].  Available:
->      https://uwspace.uwaterloo.ca/bitstream/handle/10012/12888
-Date: Mon, 13 Jan 2020 05:33:15 +0000
-From: Richard Jones <onbehalfof@manuscriptcentral.com>
-Reply-To: R.E.Jones@kent.ac.uk
-To: pabuhr@uwaterloo.ca
-Subject: Revision reminder - SPE-19-0219
--Jan-2020
-Dear Dr Buhr
-SPE-19-0219
-This is a reminder that your opportunity to revise and re-submit your
-manuscript will expire 28 days from now. If you require more time please
-contact me directly and I may grant an extension to this deadline, otherwise
-the option to submit a revision online, will not be available.
-I look forward to receiving your revision.
-Sincerely,
-Prof. Richard Jones
-Editor, Software: Practice and Experience
-https://mc.manuscriptcentral.com/spe
-Date: Wed, 5 Feb 2020 04:22:18 +0000
-From: Aaron Thomas <onbehalfof@manuscriptcentral.com>
-Reply-To: speoffice@wiley.com
-To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
-Subject: SPE-19-0219.R1 successfully submitted
--Feb-2020
-Dear Dr Buhr,
-Your manuscript entitled "Advanced Control-flow and Concurrency in Cforall" has
-been successfully submitted online and is presently being given full
-consideration for publication in Software: Practice and Experience.
-Your manuscript number is SPE-19-0219.R1.  Please mention this number in all
-future correspondence regarding this submission.
-You can view the status of your manuscript at any time by checking your Author
-Center after logging into https://mc.manuscriptcentral.com/spe.  If you have
-difficulty using this site, please click the 'Get Help Now' link at the top
-right corner of the site.
-Thank you for submitting your manuscript to Software: Practice and Experience.
-Sincerely,
-Software: Practice and Experience Editorial Office

doc/theses/thierry_delisle_PhD/code/relaxed_list.cpp

-              r7030dab
+              r71d6bd8
 #include <vector>
-#include <getopt.h>
 #include <unistd.h>
 #include <sys/sysinfo.h>
 …
         int value;
+        int id;
+        Node() { creates++; }
+        Node(int value): value(value) { creates++; }
+        ~Node() { destroys++; }
+        Node(int value): value(value) {
+                creates++;
+        }
+        ~Node() {
+                destroys++;
+        }
 };
 …
 std::atomic_size_t Node::destroys = { 0 };
+static const constexpr int nodes_per_threads = 128;
+struct NodeArray {
+        __attribute__((aligned(64))) Node * array[nodes_per_threads];
+        __attribute__((aligned(64))) char pad;
+};
 bool enable_stats = false;
-template<>
-thread_local relaxed_list<Node>::TLS relaxed_list<Node>::tls = {};
-template<>
-relaxed_list<Node> * relaxed_list<Node>::head = nullptr;
-#ifndef NO_STATS
-template<>
-relaxed_list<Node>::GlobalStats relaxed_list<Node>::global_stats = {};
-#endif
-// ================================================================================================
-//                        UTILS
-// ================================================================================================
 struct local_stat_t {
 …
         size_t crc_in  = 0;
         size_t crc_out = 0;
-        size_t valmax = 0;
-        size_t valmin = 100000000ul;
 };
+struct global_stat_t {
+        std::atomic_size_t in  = { 0 };
+        std::atomic_size_t out = { 0 };
+        std::atomic_size_t empty = { 0 };
+        std::atomic_size_t crc_in  = { 0 };
+        std::atomic_size_t crc_out = { 0 };
+        std::atomic_size_t valmax = { 0 };
+        std::atomic_size_t valmin = { 100000000ul };
+};
+void atomic_max(std::atomic_size_t & target, size_t value) {
+        for(;;) {
+                size_t expect = target.load(std::memory_order_relaxed);
+                if(value <= expect) return;
+                bool success = target.compare_exchange_strong(expect, value);
+                if(success) return;
+        }
+}
+void atomic_min(std::atomic_size_t & target, size_t value) {
+        for(;;) {
+                size_t expect = target.load(std::memory_order_relaxed);
+                if(value >= expect) return;
+                bool success = target.compare_exchange_strong(expect, value);
+                if(success) return;
+        }
+}
+void tally_stats(global_stat_t & global, local_stat_t & local) {
+        global.in    += local.in;
+        global.out   += local.out;
+        global.empty += local.empty;
+        global.crc_in  += local.crc_in;
+        global.crc_out += local.crc_out;
+        atomic_max(global.valmax, local.valmax);
+        atomic_min(global.valmin, local.valmin);
+        relaxed_list<Node>::stats_tls_tally();
+}
+void waitfor(double & duration, barrier_t & barrier, std::atomic_bool & done) {
+__attribute__((noinline)) void run_body(
+        std::atomic<bool>& done,
+        Random & rand,
+        Node * (&my_nodes)[128],
+        local_stat_t & local,
+        relaxed_list<Node> & list
+) {
+        while(__builtin_expect(!done.load(std::memory_order_relaxed), true)) {
+                int idx = rand.next() % nodes_per_threads;
+                if (auto node = my_nodes[idx]) {
+                        local.crc_in += node->value;
+                        list.push(node);
+                        my_nodes[idx] = nullptr;
+                        local.in++;
+                }
+                else if(auto node = list.pop()) {
+                        local.crc_out += node->value;
+                        my_nodes[idx] = node;
+                        local.out++;
+                }
+                else {
+                        local.empty++;
+                }
+        }
+}
+void run(unsigned nthread, unsigned nqueues, unsigned fill, double duration) {
+        // List being tested
+        relaxed_list<Node> list = { nthread * nqueues };
+        // Barrier for synchronization
+        barrier_t barrier(nthread + 1);
+        // Data to check everything is OK
+        struct {
+                std::atomic_size_t in  = { 0 };
+                std::atomic_size_t out = { 0 };
+                std::atomic_size_t empty = { 0 };
+                std::atomic_size_t crc_in  = { 0 };
+                std::atomic_size_t crc_out = { 0 };
+                struct {
+                        struct {
+                                std::atomic_size_t attempt = { 0 };
+                                std::atomic_size_t success = { 0 };
+                        } push;
+                        struct {
+                                std::atomic_size_t attempt = { 0 };
+                                std::atomic_size_t success = { 0 };
+                        } pop;
+                } pick;
+        } global;
+        // Flag to signal termination
+        std::atomic_bool done  = { false };
+        // Prep nodes
+        std::cout << "Initializing ";
+        size_t nnodes  = 0;
+        size_t npushed = 0;
+        NodeArray all_nodes[nthread];
+        for(auto & nodes : all_nodes) {
+                Random rand(rdtscl());
+                for(auto & node : nodes.array) {
+                        auto r = rand.next() % 100;
+                        if(r < fill) {
+                                node = new Node(rand.next() % 100);
+                                nnodes++;
+                        } else {
+                                node = nullptr;
+                        }
+                }
+                for(int i = 0; i < 10; i++) {
+                        int idx = rand.next() % nodes_per_threads;
+                        if (auto node = nodes.array[idx]) {
+                                global.crc_in += node->value;
+                                list.push(node);
+                                npushed++;
+                                nodes.array[idx] = nullptr;
+                        }
+                }
+        }
+        std::cout << nnodes << " nodes " << fill << "% (" << npushed << " pushed)" << std::endl;
+        enable_stats = true;
+        std::thread * threads[nthread];
+        unsigned i = 1;
+        for(auto & t : threads) {
+                auto & my_nodes = all_nodes[i - 1].array;
+                t = new std::thread([&done, &list, &barrier, &global, &my_nodes](unsigned tid) {
+                        Random rand(tid + rdtscl());
+                        local_stat_t local;
+                        // affinity(tid);
+                        barrier.wait(tid);
+                        // EXPERIMENT START
+                        run_body(done, rand, my_nodes, local, list);
+                        // EXPERIMENT END
+                        barrier.wait(tid);
+                        global.in    += local.in;
+                        global.out   += local.out;
+                        global.empty += local.empty;
+                        for(auto node : my_nodes) {
+                                delete node;
+                        }
+                        global.crc_in  += local.crc_in;
+                        global.crc_out += local.crc_out;
+                        global.pick.push.attempt += relaxed_list<Node>::tls.pick.push.attempt;
+                        global.pick.push.success += relaxed_list<Node>::tls.pick.push.success;
+                        global.pick.pop .attempt += relaxed_list<Node>::tls.pick.pop.attempt;
+                        global.pick.pop .success += relaxed_list<Node>::tls.pick.pop.success;
+                }, i++);
+        }
         std::cout << "Starting" << std::endl;
         auto before = Clock::now();
 …
         duration = durr.count();
         std::cout << "\rClosing down" << std::endl;
+}
+void waitfor(double & duration, barrier_t & barrier, const std::atomic_size_t & count) {
+        std::cout << "Starting" << std::endl;
+        auto before = Clock::now();
+        barrier.wait(0);
+        while(true) {
+                usleep(100000);
+                size_t c = count.load();
+                if( c == 0 ) {
+                        break;
+                }
+                std::cout << "\r" << c;
+                std::cout.flush();
+        }
+        barrier.wait(0);
+        auto after = Clock::now();
+        duration_t durr = after - before;
+        duration = durr.count();
+        std::cout << "\rClosing down" << std::endl;
+}
+void print_stats(double duration, unsigned nthread, global_stat_t & global) {
+        for(auto t : threads) {
+                t->join();
+                delete t;
+        }
+        enable_stats = false;
+        while(auto node = list.pop()) {
+                global.crc_out += node->value;
+                delete node;
+        }
         assert(Node::creates == Node::destroys);
         assert(global.crc_in == global.crc_out);
 …
         std::cout << "Ops/sec       : " << ops_sec << "\n";
         std::cout << "Total ops     : " << ops << "(" << global.in << "i, " << global.out << "o, " << global.empty << "e)\n";
-        if(global.valmax != 0) {
-                std::cout << "Max runs      : " << global.valmax << "\n";
-                std::cout << "Min runs      : " << global.valmin << "\n";
+        }
         #ifndef NO_STATS
+                relaxed_list<Node>::stats_print(std::cout);
+                double push_sur = (100.0 * double(global.pick.push.success) / global.pick.push.attempt);
+                double pop_sur  = (100.0 * double(global.pick.pop .success) / global.pick.pop .attempt);
+                std::cout << "Push Pick %   : " << push_sur << "(" << global.pick.push.success << " / " << global.pick.push.attempt << ")\n";
+                std::cout << "Pop  Pick %   : " << pop_sur  << "(" << global.pick.pop .success << " / " << global.pick.pop .attempt << ")\n";
         #endif
+}
+void save_fairness(const int data[], int factor, unsigned nthreads, size_t columns, size_t rows, const std::string & output);
+// ================================================================================================
+//                        EXPERIMENTS
+// ================================================================================================
+// ================================================================================================
+__attribute__((noinline)) void runChurn_body(
+        std::atomic<bool>& done,
+        Random & rand,
+        Node * my_nodes[],
+        unsigned nslots,
+        local_stat_t & local,
+        relaxed_list<Node> & list
+) {
+        while(__builtin_expect(!done.load(std::memory_order_relaxed), true)) {
+                int idx = rand.next() % nslots;
+                if (auto node = my_nodes[idx]) {
+                        local.crc_in += node->value;
+                        list.push(node);
+                        my_nodes[idx] = nullptr;
+                        local.in++;
+                }
+                else if(auto node = list.pop()) {
+                        local.crc_out += node->value;
+                        my_nodes[idx] = node;
+                        local.out++;
+                }
+                else {
+                        local.empty++;
+                }
+        }
+}
+void runChurn(unsigned nthread, unsigned nqueues, double duration, unsigned nnodes, const unsigned nslots) {
+        std::cout << "Churn Benchmark" << std::endl;
+        assert(nnodes <= nslots);
+        // List being tested
+        // Barrier for synchronization
+        barrier_t barrier(nthread + 1);
+        // Data to check everything is OK
+        global_stat_t global;
+        // Flag to signal termination
+        std::atomic_bool done  = { false };
+        // Prep nodes
+        std::cout << "Initializing ";
+        size_t npushed = 0;
+        relaxed_list<Node> list = { nthread * nqueues };
+        {
+                Node** all_nodes[nthread];
+                for(auto & nodes : all_nodes) {
+                        nodes = new __attribute__((aligned(64))) Node*[nslots + 8];
+                        Random rand(rdtscl());
+                        for(unsigned i = 0; i < nnodes; i++) {
+                                nodes[i] = new Node(rand.next() % 100);
+                        }
+                        for(unsigned i = nnodes; i < nslots; i++) {
+                                nodes[i] = nullptr;
+                        }
+                        for(int i = 0; i < 10 && i < (int)nslots; i++) {
+                                int idx = rand.next() % nslots;
+                                if (auto node = nodes[idx]) {
+                                        global.crc_in += node->value;
+                                        list.push(node);
+                                        npushed++;
+                                        nodes[idx] = nullptr;
+                                }
+                        }
+                }
+                std::cout << nnodes << " nodes (" << nslots << " slots)" << std::endl;
+                enable_stats = true;
+                std::thread * threads[nthread];
+                unsigned i = 1;
+                for(auto & t : threads) {
+                        auto & my_nodes = all_nodes[i - 1];
+                        t = new std::thread([&done, &list, &barrier, &global, &my_nodes, nslots](unsigned tid) {
+                                Random rand(tid + rdtscl());
+                                local_stat_t local;
+                                // affinity(tid);
+                                barrier.wait(tid);
+                                // EXPERIMENT START
+                                runChurn_body(done, rand, my_nodes, nslots, local, list);
+                                // EXPERIMENT END
+                                barrier.wait(tid);
+                                tally_stats(global, local);
+                                for(unsigned i = 0; i < nslots; i++) {
+                                        delete my_nodes[i];
+                                }
+                        }, i++);
+                }
+                waitfor(duration, barrier, done);
+                for(auto t : threads) {
+                        t->join();
+                        delete t;
+                }
+                enable_stats = false;
+                while(auto node = list.pop()) {
+                        global.crc_out += node->value;
+                        delete node;
+                }
+                for(auto nodes : all_nodes) {
+                        delete[] nodes;
+                }
+        }
+        print_stats(duration, nthread, global);
+}
+// ================================================================================================
+__attribute__((noinline)) void runPingPong_body(
+        std::atomic<bool>& done,
+        Node initial_nodes[],
+        unsigned nnodes,
+        local_stat_t & local,
+        relaxed_list<Node> & list
+) {
+        Node * nodes[nnodes];
+        {
+                unsigned i = 0;
+                for(auto & n : nodes) {
+                        n = &initial_nodes[i++];
+                }
+        }
+        while(__builtin_expect(!done.load(std::memory_order_relaxed), true)) {
+                for(Node * & node : nodes) {
+                        local.crc_in += node->value;
+                        list.push(node);
+                        local.in++;
+                }
+                // -----
+                for(Node * & node : nodes) {
+                        node = list.pop();
+                        assert(node);
+                        local.crc_out += node->value;
+                        local.out++;
+                }
+        }
+}
+void runPingPong(unsigned nthread, unsigned nqueues, double duration, unsigned nnodes) {
+        std::cout << "PingPong Benchmark" << std::endl;
+        // Barrier for synchronization
+        barrier_t barrier(nthread + 1);
+        // Data to check everything is OK
+        global_stat_t global;
+        // Flag to signal termination
+        std::atomic_bool done  = { false };
+        std::cout << "Initializing ";
+        // List being tested
+        relaxed_list<Node> list = { nthread * nqueues };
+        {
+                enable_stats = true;
+                std::thread * threads[nthread];
+                unsigned i = 1;
+                for(auto & t : threads) {
+                        t = new std::thread([&done, &list, &barrier, &global, nnodes](unsigned tid) {
+                                Random rand(tid + rdtscl());
+                                Node nodes[nnodes];
+                                for(auto & n : nodes) {
+                                        n.value = (int)rand.next() % 100;
+                                }
+                                local_stat_t local;
+                                // affinity(tid);
+                                barrier.wait(tid);
+                                // EXPERIMENT START
+                                runPingPong_body(done, nodes, nnodes, local, list);
+                                // EXPERIMENT END
+                                barrier.wait(tid);
+                                tally_stats(global, local);
+                        }, i++);
+                }
+                waitfor(duration, barrier, done);
+                for(auto t : threads) {
+                        t->join();
+                        delete t;
+                }
+                enable_stats = false;
+        }
+        print_stats(duration, nthread, global);
+}
+// ================================================================================================
+__attribute__((noinline)) void runFairness_body(
+        unsigned tid,
+        size_t width,
+        size_t length,
+        int output[],
+        std::atomic_size_t & count,
+        Node initial_nodes[],
+        unsigned nnodes,
+        local_stat_t & local,
+        relaxed_list<Node> & list
+) {
+        Node * nodes[nnodes];
+        {
+                unsigned i = 0;
+                for(auto & n : nodes) {
+                        n = &initial_nodes[i++];
+                }
+        }
+        while(__builtin_expect(0 != count.load(std::memory_order_relaxed), true)) {
+                for(Node * & node : nodes) {
+                        local.crc_in += node->id;
+                        list.push(node);
+                        local.in++;
+                }
+                // -----
+                for(Node * & node : nodes) {
+                        node = list.pop();
+                        assert(node);
+                        if (unsigned(node->value) < length) {
+                                size_t idx = (node->value * width) + node->id;
+                                assert(idx < (width * length));
+                                output[idx] = tid;
+                        }
+                        node->value++;
+                        if(unsigned(node->value) == length) count--;
+                        local.crc_out += node->id;
+                        local.out++;
+                }
+        }
+}
+void runFairness(unsigned nthread, unsigned nqueues, double duration, unsigned nnodes, const std::string & output) {
+        std::cout << "Fairness Benchmark, outputing to : " << output << std::endl;
+        // Barrier for synchronization
+        barrier_t barrier(nthread + 1);
+        // Data to check everything is OK
+        global_stat_t global;
+        std::cout << "Initializing ";
+        // Check fairness by creating a png of where the threads ran
+        size_t width = nthread * nnodes;
+        size_t length = 100000;
+        std::unique_ptr<int[]> data_out { new int[width * length] };
+        // Flag to signal termination
+        std::atomic_size_t count = width;
+        // List being tested
+        relaxed_list<Node> list = { nthread * nqueues };
+        {
+                enable_stats = true;
+                std::thread * threads[nthread];
+                unsigned i = 1;
+                for(auto & t : threads) {
+                        t = new std::thread([&count, &list, &barrier, &global, nnodes, width, length, data_out = data_out.get()](unsigned tid) {
+                                unsigned int start = (tid - 1) * nnodes;
+                                Node nodes[nnodes];
+                                for(auto & n : nodes) {
+                                        n.id = start;
+                                        n.value = 0;
+                                        start++;
+                                }
+                                local_stat_t local;
+                                // affinity(tid);
+                                barrier.wait(tid);
+                                // EXPERIMENT START
+                                runFairness_body(tid, width, length, data_out, count, nodes, nnodes, local, list);
+                                // EXPERIMENT END
+                                barrier.wait(tid);
+                                for(const auto & n : nodes) {
+                                        local.valmax = max(local.valmax, size_t(n.value));
+                                        local.valmin = min(local.valmin, size_t(n.value));
+                                }
+                                tally_stats(global, local);
+                        }, i++);
+                }
+                waitfor(duration, barrier, count);
+                for(auto t : threads) {
+                        t->join();
+                        delete t;
+                }
+                enable_stats = false;
+        }
+        print_stats(duration, nthread, global);
+        save_fairness(data_out.get(), 100, nthread, width, length, output);
+}
+// ================================================================================================
+bool iequals(const std::string& a, const std::string& b)
+{
+    return std::equal(a.begin(), a.end(),
+                      b.begin(), b.end(),
+                      [](char a, char b) {
+                          return std::tolower(a) == std::tolower(b);
+                      });
+void usage(char * argv[]) {
+        std::cerr << argv[0] << ": [DURATION (FLOAT:SEC)] [NTHREADS] [NQUEUES] [FILL]" << std::endl;;
+        std::exit(1);
+}
 …
         double duration   = 5.0;
         unsigned nthreads = 2;
+        unsigned nqueues  = 4;
+        unsigned nnodes   = 100;
+        unsigned nslots   = 100;
+        std::string out   = "fairness.png";
+        enum {
+                Churn,
+                PingPong,
+                Fairness,
+                NONE
+        } benchmark = NONE;
+        unsigned nqueues  = 2;
+        unsigned fill     = 100;
         std::cout.imbue(std::locale(""));
+        for(;;) {
+                static struct option options[] = {
+                        {"duration",  required_argument, 0, 'd'},
+                        {"nthreads",  required_argument, 0, 't'},
+                        {"nqueues",   required_argument, 0, 'q'},
+                        {"benchmark", required_argument, 0, 'b'},
+                        {0, 0, 0, 0}
+                };
+                int idx = 0;
+                int opt = getopt_long(argc, argv, "d:t:q:b:", options, &idx);
+                std::string arg = optarg ? optarg : "";
+                size_t len = 0;
+                switch(opt) {
+                        // Exit Case
+                        case -1:
+                                /* paranoid */ assert(optind <= argc);
+                                switch(benchmark) {
+                                case NONE:
+                                        std::cerr << "Must specify a benchmark" << std::endl;
+                                        goto usage;
+                                case PingPong:
+                                        nnodes = 1;
+                                        nslots = 1;
+                                        switch(argc - optind) {
+                                        case 0: break;
+                                        case 1:
+                                                try {
+                                                        arg = optarg = argv[optind];
+                                                        nnodes = stoul(optarg, &len);
+                                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                                } catch(std::invalid_argument &) {
+                                                        std::cerr << "Number of nodes must be a positive integer, was " << arg << std::endl;
+                                                        goto usage;
+                                                }
+                                                break;
+                                        default:
+                                                std::cerr << "'PingPong' benchmark doesn't accept more than 2 extra arguments" << std::endl;
+                                                goto usage;
+                                        }
+                                        break;
+                                case Churn:
+                                        nnodes = 100;
+                                        nslots = 100;
+                                        switch(argc - optind) {
+                                        case 0: break;
+                                        case 1:
+                                                try {
+                                                        arg = optarg = argv[optind];
+                                                        nnodes = stoul(optarg, &len);
+                                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                                        nslots = nnodes;
+                                                } catch(std::invalid_argument &) {
+                                                        std::cerr << "Number of nodes must be a positive integer, was " << arg << std::endl;
+                                                        goto usage;
+                                                }
+                                                break;
+                                        case 2:
+                                                try {
+                                                        arg = optarg = argv[optind];
+                                                        nnodes = stoul(optarg, &len);
+                                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                                } catch(std::invalid_argument &) {
+                                                        std::cerr << "Number of nodes must be a positive integer, was " << arg << std::endl;
+                                                        goto usage;
+                                                }
+                                                try {
+                                                        arg = optarg = argv[optind + 1];
+                                                        nslots = stoul(optarg, &len);
+                                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                                } catch(std::invalid_argument &) {
+                                                        std::cerr << "Number of slots must be a positive integer, was " << arg << std::endl;
+                                                        goto usage;
+                                                }
+                                                break;
+                                        default:
+                                                std::cerr << "'Churn' benchmark doesn't accept more than 2 extra arguments" << std::endl;
+                                                goto usage;
+                                        }
+                                        break;
+                                case Fairness:
+                                        nnodes = 1;
+                                        switch(argc - optind) {
+                                        case 0: break;
+                                        case 1:
+                                                arg = optarg = argv[optind];
+                                                out = arg;
+                                                break;
+                                        default:
+                                                std::cerr << "'Churn' benchmark doesn't accept more than 2 extra arguments" << std::endl;
+                                                goto usage;
+                                        }
+                                }
+                                goto run;
+                        // Benchmarks
+                        case 'b':
+                                if(benchmark != NONE) {
+                                        std::cerr << "Only when benchmark can be run" << std::endl;
+                                        goto usage;
+                                }
+                                if(iequals(arg, "churn")) {
+                                        benchmark = Churn;
+                                        break;
+                                }
+                                if(iequals(arg, "pingpong")) {
+                                        benchmark = PingPong;
+                                        break;
+                                }
+                                if(iequals(arg, "fairness")) {
+                                        benchmark = Fairness;
+                                        break;
+                                }
+                                std::cerr << "Unkown benchmark " << arg << std::endl;
+                                goto usage;
+                        // Numeric Arguments
+                        case 'd':
+                                try {
+                                        duration = stod(optarg, &len);
+                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                } catch(std::invalid_argument &) {
+                                        std::cerr << "Duration must be a valid double, was " << arg << std::endl;
+                                        goto usage;
+                                }
+                                break;
+                        case 't':
+                                try {
+                                        nthreads = stoul(optarg, &len);
+                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                } catch(std::invalid_argument &) {
+                                        std::cerr << "Number of threads must be a positive integer, was " << arg << std::endl;
+                                        goto usage;
+                                }
+                                break;
+                        case 'q':
+                                try {
+                                        nqueues = stoul(optarg, &len);
+                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                } catch(std::invalid_argument &) {
+                                        std::cerr << "Number of queues must be a positive integer, was " << arg << std::endl;
+                                        goto usage;
+                                }
+                                break;
+                        // Other cases
+                        default: /* ? */
+                                std::cerr << opt << std::endl;
+                        usage:
+                                std::cerr << "Usage: " << argv[0] << ": [options] -b churn [NNODES] [NSLOTS = NNODES]" << std::endl;
+                                std::cerr << "  or:  " << argv[0] << ": [options] -b pingpong [NNODES]" << std::endl;
+                                std::cerr << std::endl;
+                                std::cerr << "  -d, --duration=DURATION  Duration of the experiment, in seconds" << std::endl;
+                                std::cerr << "  -t, --nthreads=NTHREADS  Number of kernel threads" << std::endl;
+                                std::cerr << "  -q, --nqueues=NQUEUES    Number of queues per threads" << std::endl;
+                                std::exit(1);
+                }
+        }
+        run:
+        switch (argc)
+        {
+        case 5:
+                fill = std::stoul(argv[4]);
+                [[fallthrough]];
+        case 4:
+                nqueues = std::stoul(argv[3]);
+                [[fallthrough]];
+        case 3:
+                nthreads = std::stoul(argv[2]);
+                [[fallthrough]];
+        case 2:
+                duration = std::stod(argv[1]);
+                if( duration <= 0.0 ) {
+                        std::cerr << "Duration must be positive, was " << argv[1] << "(" << duration << ")" << std::endl;
+                        usage(argv);
+                }
+                [[fallthrough]];
+        case 1:
+                break;
+        default:
+                usage(argv);
+                break;
+        }
         check_cache_line_size();
         std::cout << "Running " << nthreads << " threads (" << (nthreads * nqueues) << " queues) for " << duration << " seconds" << std::endl;
+        switch(benchmark) {
+                case Churn:
+                        runChurn(nthreads, nqueues, duration, nnodes, nslots);
+                        break;
+                case PingPong:
+                        runPingPong(nthreads, nqueues, duration, nnodes);
+                        break;
+                case Fairness:
+                        runFairness(nthreads, nqueues, duration, nnodes, out);
+                        break;
+                default:
+                        abort();
+        }
+        run(nthreads, nqueues, fill, duration);
         return 0;
+}
+template<>
+thread_local relaxed_list<Node>::TLS relaxed_list<Node>::tls = {};
+template<>
+relaxed_list<Node>::intrusive_queue_t::stat::Dif relaxed_list<Node>::intrusive_queue_t::stat::dif = {};
 const char * __my_progname = "Relaxed List";
-struct rgb_t {
-    double r;       // a fraction between 0 and 1
-    double g;       // a fraction between 0 and 1
-    double b;       // a fraction between 0 and 1
-};
-struct hsv_t {
-    double h;       // angle in degrees
-    double s;       // a fraction between 0 and 1
-    double v;       // a fraction between 0 and 1
-};
-rgb_t hsv2rgb(hsv_t in) {
-        double hh, p, q, t, ff;
-        long   i;
-        rgb_t  out;
-        if(in.s <= 0.0) {       // < is bogus, just shuts up warnings
-                out.r = in.v;
-                out.g = in.v;
-                out.b = in.v;
-                return out;
+        }
-        hh = in.h;
-        if(hh >= 360.0) hh = 0.0;
-        hh /= 60.0;
-        i = (long)hh;
-        ff = hh - i;
-        p = in.v * (1.0 - in.s);
-        q = in.v * (1.0 - (in.s * ff));
-        t = in.v * (1.0 - (in.s * (1.0 - ff)));
-        switch(i) {
-        case 0:
-                out.r = in.v;
-                out.g = t;
-                out.b = p;
-                break;
-        case 1:
-                out.r = q;
-                out.g = in.v;
-                out.b = p;
-                break;
-        case 2:
-                out.r = p;
-                out.g = in.v;
-                out.b = t;
-                break;
-        case 3:
-                out.r = p;
-                out.g = q;
-                out.b = in.v;
-                break;
-        case 4:
-                out.r = t;
-                out.g = p;
-                out.b = in.v;
-                break;
-        case 5:
-        default:
-                out.r = in.v;
-                out.g = p;
-                out.b = q;
-                break;
+        }
-        return out;
+}
-void save_fairness(const int data[], int factor, unsigned nthreads, size_t columns, size_t rows, const std::string & output) {
-        std::ofstream os(output);
-        os << "<html>\n";
-        os << "<head>\n";
-        os << "<style>\n";
-        os << "</style>\n";
-        os << "</head>\n";
-        os << "<body>\n";
-        os << "<table style=\"width=100%\">\n";
-        size_t idx = 0;
-        for(size_t r = 0ul; r < rows; r++) {
-                os << "<tr>\n";
-                for(size_t c = 0ul; c < columns; c++) {
-                        os << "<td class=\"custom custom" << data[idx] << "\"></td>\n";
-                        idx++;
+                }
-                os << "</tr>\n";
+        }
-        os << "</table>\n";
-        os << "</body>\n";
-        os << "</html>\n";
-        os << std::endl;
+}
-#include <png.h>
-#include <setjmp.h>
-/*
-void save_fairness(const int data[], int factor, unsigned nthreads, size_t columns, size_t rows, const std::string & output) {
-        int width  = columns * factor;
-        int height = rows / factor;
-        int code = 0;
-        int idx = 0;
-        FILE *fp = NULL;
-        png_structp png_ptr = NULL;
-        png_infop info_ptr = NULL;
-        png_bytep row = NULL;
-        // Open file for writing (binary mode)
-        fp = fopen(output.c_str(), "wb");
-        if (fp == NULL) {
-                fprintf(stderr, "Could not open file %s for writing\n", output.c_str());
-                code = 1;
-                goto finalise;
+        }
-           // Initialize write structure
-        png_ptr = png_create_write_struct(PNG_LIBPNG_VER_STRING, NULL, NULL, NULL);
-        if (png_ptr == NULL) {
-                fprintf(stderr, "Could not allocate write struct\n");
-                code = 1;
-                goto finalise;
+        }
-        // Initialize info structure
-        info_ptr = png_create_info_struct(png_ptr);
-        if (info_ptr == NULL) {
-                fprintf(stderr, "Could not allocate info struct\n");
-                code = 1;
-                goto finalise;
+        }
-        // Setup Exception handling
-        if (setjmp(png_jmpbuf(png_ptr))) {
-                fprintf(stderr, "Error during png creation\n");
-                code = 1;
-                goto finalise;
+        }
-        png_init_io(png_ptr, fp);
-        // Write header (8 bit colour depth)
-        png_set_IHDR(png_ptr, info_ptr, width, height,
-, PNG_COLOR_TYPE_RGB, PNG_INTERLACE_NONE,
-                PNG_COMPRESSION_TYPE_BASE, PNG_FILTER_TYPE_BASE);
-        png_write_info(png_ptr, info_ptr);
-        // Allocate memory for one row (3 bytes per pixel - RGB)
-        row = (png_bytep) malloc(3 * width * sizeof(png_byte));
-        // Write image data
-        int x, y;
-        for (y=0 ; y<height ; y++) {
-                for (x=0 ; x<width ; x++) {
-                        auto & r = row[(x * 3) + 0];
-                        auto & g = row[(x * 3) + 1];
-                        auto & b = row[(x * 3) + 2];
-                        assert(idx < (rows * columns));
-                        int color = data[idx] - 1;
-                        assert(color < nthreads);
-                        assert(color >= 0);
-                        idx++;
-                        double angle = double(color) / double(nthreads);
-                        auto c = hsv2rgb({ 360.0 * angle, 0.8, 0.8 });
-                        r = char(c.r * 255.0);
-                        g = char(c.g * 255.0);
-                        b = char(c.b * 255.0);
+                }
-                png_write_row(png_ptr, row);
+        }
-        assert(idx == (rows * columns));
-        // End write
-        png_write_end(png_ptr, NULL);
-        finalise:
-        if (fp != NULL) fclose(fp);
-        if (info_ptr != NULL) png_free_data(png_ptr, info_ptr, PNG_FREE_ALL, -1);
-        if (png_ptr != NULL) png_destroy_write_struct(&png_ptr, (png_infopp)NULL);
-        if (row != NULL) free(row);
+}
-*/

doc/theses/thierry_delisle_PhD/code/relaxed_list.hpp

-              r7030dab
+              r71d6bd8
 };
-static inline bool bts(std::atomic_size_t & target, size_t bit ) {
-        //*
-        int result = 0;
-        asm volatile(
-                "LOCK btsq %[bit], %[target]\n\t"
-                :"=@ccc" (result)
-                : [target] "m" (target), [bit] "r" (bit)
-        );
-        return result != 0;
-        /*/
-        size_t mask = 1ul << bit;
-        size_t ret = target.fetch_or(mask, std::memory_order_relaxed);
-        return (ret & mask) != 0;
-        //*/
+}
-static inline bool btr(std::atomic_size_t & target, size_t bit ) {
-        //*
-        int result = 0;
-        asm volatile(
-                "LOCK btrq %[bit], %[target]\n\t"
-                :"=@ccc" (result)
-                : [target] "m" (target), [bit] "r" (bit)
-        );
-        return result != 0;
-        /*/
-        size_t mask = 1ul << bit;
-        size_t ret = target.fetch_and(~mask, std::memory_order_relaxed);
-        return (ret & mask) != 0;
-        //*/
+}
 extern bool enable_stats;
 …
                 size_t attempt = 0;
                 size_t success = 0;
-                size_t mask_attempt = 0;
-        } pop;
-};
-struct empty_stat {
-        struct {
-                size_t value = 0;
-                size_t count = 0;
-        } push;
-        struct {
-                size_t value = 0;
-                size_t count = 0;
         } pop;
 };
 …
         static_assert(std::is_same<decltype(node_t::_links), _LinksFields_t<node_t>>::value, "Node must have a links field");
 public:
         relaxed_list(unsigned numLists)
+                : lists(new intrusive_queue_t[numLists])
+                : numNonEmpty{0}
+                , lists(new intrusive_queue_t[numLists])
                 , numLists(numLists)
+        {
+                assertf(7 * 8 * 8 >= numLists, "List currently only supports 448 sublists");
+                // assert(sizeof(*this) == 128);
+                std::cout << "Constructing Relaxed List with " << numLists << std::endl;
+        {}
+        ~relaxed_list() {
+                lists.reset();
                 #ifndef NO_STATS
+                        if(head) this->next = head;
+                        head = this;
+                        std::cout << "Difference   : "
+                                << ssize_t(double(intrusive_queue_t::stat::dif.value) / intrusive_queue_t::stat::dif.num  ) << " avg\t"
+                                << intrusive_queue_t::stat::dif.max << "max" << std::endl;
                 #endif
+        }
-        ~relaxed_list() {
-                std::cout << "Destroying Relaxed List" << std::endl;
-                lists.reset();
+        }
 …
                 while(true) {
                         // Pick a random list
                         unsigned i = tls.rng.next() % numLists;
+                        int i = tls.rng.next() % numLists;
                         #ifndef NO_STATS
 …
                         if( !lists[i].lock.try_lock() ) continue;
-                        __attribute__((unused)) int num = numNonEmpty;
                         // Actually push it
+                        if(lists[i].push(node)) {
+                                numNonEmpty++;
+                                size_t qword = i >> 6ull;
+                                size_t bit   = i & 63ull;
+                                assertf((list_mask[qword] & (1ul << bit)) == 0, "Before set %zu:%zu (%u), %zx & %zx", qword, bit, i, list_mask[qword].load(), (1ul << bit));
+                                __attribute__((unused)) bool ret = bts(list_mask[qword], bit);
+                                assert(!ret);
+                                assertf((list_mask[qword] & (1ul << bit)) != 0, "After set %zu:%zu (%u), %zx & %zx", qword, bit, i, list_mask[qword].load(), (1ul << bit));
+                        }
+                        lists[i].push(node, numNonEmpty);
                         assert(numNonEmpty <= (int)numLists);
 …
                         #ifndef NO_STATS
                                 tls.pick.push.success++;
-                                tls.empty.push.value += num;
-                                tls.empty.push.count += 1;
                         #endif
                         return;
 …
         __attribute__((noinline, hot)) node_t * pop() {
+                #if !defined(NO_BITMASK)
+                        // for(int r = 0; r < 10 && numNonEmpty != 0; r++) {
+                        //      // Pick two lists at random
+                        //      unsigned i = tls.rng.next() % numLists;
+                        //      unsigned j = tls.rng.next() % numLists;
+                        //      if(auto node = try_pop(i, j)) return node;
+                        // }
+                        int nnempty;
+                        while(0 != (nnempty = numNonEmpty)) {
+                                tls.pick.pop.mask_attempt++;
+                                unsigned i, j;
+                                // if( numLists < 4 || (numLists / nnempty) < 4 ) {
+                                //      // Pick two lists at random
+                                //      i = tls.rng.next() % numLists;
+                                //      j = tls.rng.next() % numLists;
+                                // } else
+                                {
+                                        #ifndef NO_STATS
+                                                // tls.pick.push.mask_attempt++;
+                                        #endif
+                                        // Pick two lists at random
+                                        unsigned num = ((numLists - 1) >> 6) + 1;
+                                        unsigned ri = tls.rng.next();
+                                        unsigned rj = tls.rng.next();
+                                        unsigned wdxi = (ri >> 6u) % num;
+                                        unsigned wdxj = (rj >> 6u) % num;
+                                        size_t maski = list_mask[wdxi].load(std::memory_order_relaxed);
+                                        size_t maskj = list_mask[wdxj].load(std::memory_order_relaxed);
+                                        if(maski == 0 && maskj == 0) continue;
+                                        unsigned bi = rand_bit(ri, maski);
+                                        unsigned bj = rand_bit(rj, maskj);
+                                        assertf(bi < 64, "%zu %u", maski, bi);
+                                        assertf(bj < 64, "%zu %u", maskj, bj);
+                                        i = bi | (wdxi << 6);
+                                        j = bj | (wdxj << 6);
+                                        assertf(i < numLists, "%u", wdxi << 6);
+                                        assertf(j < numLists, "%u", wdxj << 6);
+                                }
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                #else
+                        while(numNonEmpty != 0) {
+                                // Pick two lists at random
+                                int i = tls.rng.next() % numLists;
+                                int j = tls.rng.next() % numLists;
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                #endif
+                while(numNonEmpty != 0) {
+                        // Pick two lists at random
+                        int i = tls.rng.next() % numLists;
+                        int j = tls.rng.next() % numLists;
+                        #ifndef NO_STATS
+                                tls.pick.pop.attempt++;
+                        #endif
+                        // Pick the bet list
+                        int w = i;
+                        if( __builtin_expect(lists[j].ts() != 0, true) ) {
+                                w = (lists[i].ts() < lists[j].ts()) ? i : j;
+                        }
+                        auto & list = lists[w];
+                        // If list looks empty retry
+                        if( list.ts() == 0 ) continue;
+                        // If we can't get the lock retry
+                        if( !list.lock.try_lock() ) continue;
+                        // If list is empty, unlock and retry
+                        if( list.ts() == 0 ) {
+                                list.lock.unlock();
+                                continue;
+                        }
+                        // Actually pop the list
+                        auto node = list.pop(numNonEmpty);
+                        assert(node);
+                        // Unlock and return
+                        list.lock.unlock();
+                        assert(numNonEmpty >= 0);
+                        #ifndef NO_STATS
+                                tls.pick.pop.success++;
+                        #endif
+                        return node;
+                }
                 return nullptr;
+        }
-private:
-        node_t * try_pop(unsigned i, unsigned j) {
-                #ifndef NO_STATS
-                        tls.pick.pop.attempt++;
-                #endif
-                // Pick the bet list
-                int w = i;
-                if( __builtin_expect(lists[j].ts() != 0, true) ) {
-                        w = (lists[i].ts() < lists[j].ts()) ? i : j;
+                }
-                auto & list = lists[w];
-                // If list looks empty retry
-                if( list.ts() == 0 ) return nullptr;
-                // If we can't get the lock retry
-                if( !list.lock.try_lock() ) return nullptr;
-                __attribute__((unused)) int num = numNonEmpty;
-                // If list is empty, unlock and retry
-                if( list.ts() == 0 ) {
-                        list.lock.unlock();
-                        return nullptr;
+                }
-                // Actually pop the list
-                node_t * node;
-                bool emptied;
-                std::tie(node, emptied) = list.pop();
-                assert(node);
-                if(emptied) {
-                        numNonEmpty--;
-                        size_t qword = w >> 6ull;
-                        size_t bit   = w & 63ull;
-                        assert((list_mask[qword] & (1ul << bit)) != 0);
-                        __attribute__((unused)) bool ret = btr(list_mask[qword], bit);
-                        assert(ret);
-                        assert((list_mask[qword] & (1ul << bit)) == 0);
+                }
-                // Unlock and return
-                list.lock.unlock();
-                assert(numNonEmpty >= 0);
-                #ifndef NO_STATS
-                        tls.pick.pop.success++;
-                        tls.empty.pop.value += num;
-                        tls.empty.pop.count += 1;
-                #endif
-                return node;
+        }
 private:
 …
                 struct stat {
                         ssize_t diff = 0;
+                        size_t  push = 0;
+                        size_t  pop  = 0;
+                        // size_t value = 0;
+                        // size_t count = 0;
+                        static struct Dif {
+                                ssize_t value = 0;
+                                size_t  num   = 0;
+                                ssize_t max   = 0;
+                        } dif;
                 };
 …
                 sentinel_t before;
                 sentinel_t after;
+                #ifndef NO_STATS
+                        stat s;
+                #endif
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Winvalid-offsetof"
+                stat s;
                 static constexpr auto fields_offset = offsetof( node_t, _links );
-#pragma GCC diagnostic pop
         public:
                 intrusive_queue_t()
 …
                         , after {{ head(), nullptr }}
+                {
+                        /* paranoid */ assert((reinterpret_cast<uintptr_t>( head() ) + fields_offset) == reinterpret_cast<uintptr_t>(&before));
+                        /* paranoid */ assert((reinterpret_cast<uintptr_t>( tail() ) + fields_offset) == reinterpret_cast<uintptr_t>(&after ));
+                        /* paranoid */ assert(head()->_links.prev == nullptr);
+                        /* paranoid */ assert(head()->_links.next == tail() );
+                        /* paranoid */ assert(tail()->_links.next == nullptr);
+                        /* paranoid */ assert(tail()->_links.prev == head() );
+                        /* paranoid */ assert(sizeof(*this) == 128);
+                        /* paranoid */ assert((intptr_t(this) % 128) == 0);
+                }
+                ~intrusive_queue_t() = default;
+                        assert((reinterpret_cast<uintptr_t>( head() ) + fields_offset) == reinterpret_cast<uintptr_t>(&before));
+                        assert((reinterpret_cast<uintptr_t>( tail() ) + fields_offset) == reinterpret_cast<uintptr_t>(&after ));
+                        assert(head()->_links.prev == nullptr);
+                        assert(head()->_links.next == tail() );
+                        assert(tail()->_links.next == nullptr);
+                        assert(tail()->_links.prev == head() );
+                        assert(sizeof(*this) == 128);
+                        assert((intptr_t(this) % 128) == 0);
+                }
+                ~intrusive_queue_t() {
+                        #ifndef NO_STATS
+                                stat::dif.value+= s.diff;
+                                stat::dif.num  ++;
+                                stat::dif.max  = std::abs(stat::dif.max) > std::abs(s.diff) ? stat::dif.max : s.diff;
+                        #endif
+                }
                 inline node_t * head() const {
 …
+                }
                 inline bool push(node_t * node) {
+                inline void push(node_t * node, std::atomic_int & nonEmpty) {
                         assert(lock);
                         assert(node->_links.ts != 0);
 …
                         prev->_links.next = node;
                         tail->_links.prev = node;
-                        #ifndef NO_STATS
-                                if(enable_stats) {
-                                        s.diff++;
-                                        s.push++;
+                                }
-                        #endif
                         if(before._links.ts == 0l) {
+                                nonEmpty += 1;
                                 before._links.ts = node->_links.ts;
                                 assert(node->_links.prev == this->head());
                                 return true;
+                        }
                         return false;
+                }
                 inline std::pair<node_t *, bool> pop() {
+                        }
+                        #ifndef NO_STATS
+                                if(enable_stats) s.diff++;
+                        #endif
+                }
+                inline node_t * pop(std::atomic_int & nonEmpty) {
                         assert(lock);
                         node_t * head = this->head();
 …
                         node_t * node = head->_links.next;
                         node_t * next = node->_links.next;
                         if(node == tail) return {nullptr, false};
+                        if(node == tail) return nullptr;
                         head->_links.next = next;
                         next->_links.prev = head;
-                        #ifndef NO_STATS
-                                if(enable_stats) {
-                                        s.diff--;
-                                        s.pop ++;
+                                }
-                        #endif
                         if(next == tail) {
                                 before._links.ts = 0l;
                                 return {node, true};
+                                nonEmpty -= 1;
+                        }
                         else {
 …
                                 before._links.ts = next->_links.ts;
                                 assert(before._links.ts != 0);
+                                return {node, false};
+                        }
+                        }
+                        #ifndef NO_STATS
+                                if(enable_stats) s.diff--;
+                        #endif
+                        return node;
+                }
 …
         static __attribute__((aligned(128))) thread_local struct TLS {
+                Random     rng = { int(rdtscl()) };
+                pick_stat  pick;
+                empty_stat empty;
+                Random    rng = { int(rdtscl()) };
+                pick_stat pick;
         } tls;
-public:
-        std::atomic_int numNonEmpty  = { 0 };  // number of non-empty lists
-        std::atomic_size_t list_mask[7] = { {0}, {0}, {0}, {0}, {0}, {0}, {0} }; // which queues are empty
 private:
+        std::atomic_int numNonEmpty; // number of non-empty lists
         __attribute__((aligned(64))) std::unique_ptr<intrusive_queue_t []> lists;
         const unsigned numLists;
 …
 public:
         static const constexpr size_t sizeof_queue = sizeof(intrusive_queue_t);
+#ifndef NO_STATS
+        static void stats_print(std::ostream & os) {
+                auto it = head;
+                while(it) {
+                        it->stats_print_local(os);
+                        it = it->next;
+                }
+        }
+        static void stats_tls_tally() {
+                global_stats.pick.push.attempt += tls.pick.push.attempt;
+                global_stats.pick.push.success += tls.pick.push.success;
+                global_stats.pick.pop .attempt += tls.pick.pop.attempt;
+                global_stats.pick.pop .success += tls.pick.pop.success;
+                global_stats.pick.pop .mask_attempt += tls.pick.pop.mask_attempt;
+                global_stats.qstat.push.value += tls.empty.push.value;
+                global_stats.qstat.push.count += tls.empty.push.count;
+                global_stats.qstat.pop .value += tls.empty.pop .value;
+                global_stats.qstat.pop .count += tls.empty.pop .count;
+        }
+private:
+        static struct GlobalStats {
+                struct {
+                        struct {
+                                std::atomic_size_t attempt = { 0 };
+                                std::atomic_size_t success = { 0 };
+                        } push;
+                        struct {
+                                std::atomic_size_t attempt = { 0 };
+                                std::atomic_size_t success = { 0 };
+                                std::atomic_size_t mask_attempt = { 0 };
+                        } pop;
+                } pick;
+                struct {
+                        struct {
+                                std::atomic_size_t value = { 0 };
+                                std::atomic_size_t count = { 0 };
+                        } push;
+                        struct {
+                                std::atomic_size_t value = { 0 };
+                                std::atomic_size_t count = { 0 };
+                        } pop;
+                } qstat;
+        } global_stats;
+        // Link list of all lists for stats
+        __attribute__((aligned(64))) relaxed_list<node_t> * next = nullptr;
+        static relaxed_list<node_t> * head;
+        void stats_print_local(std::ostream & os ) {
+                std::cout << "----- Relaxed List Stats -----" << std::endl;
+                {
+                        ssize_t diff = 0;
+                        size_t  num  = 0;
+                        ssize_t max  = 0;
+                        for(size_t i = 0; i < numLists; i++) {
+                                const auto & list = lists[i];
+                                diff+= list.s.diff;
+                                num ++;
+                                max  = std::abs(max) > std::abs(list.s.diff) ? max : list.s.diff;
+                                os << "Local Q ops   : " << (list.s.push + list.s.pop) << "(" << list.s.push << "i, " << list.s.pop << "o)\n";
+                        }
+                        os << "Difference   : " << ssize_t(double(diff) / num  ) << " avg\t" << max << "max" << std::endl;
+                }
+                const auto & global = global_stats;
+                double push_sur = (100.0 * double(global.pick.push.success) / global.pick.push.attempt);
+                double pop_sur  = (100.0 * double(global.pick.pop .success) / global.pick.pop .attempt);
+                double mpop_sur = (100.0 * double(global.pick.pop .success) / global.pick.pop .mask_attempt);
+                os << "Push   Pick % : " << push_sur << "(" << global.pick.push.success << " / " << global.pick.push.attempt << ")\n";
+                os << "Pop    Pick % : " << pop_sur  << "(" << global.pick.pop .success << " / " << global.pick.pop .attempt << ")\n";
+                os << "TryPop Pick % : " << mpop_sur << "(" << global.pick.pop .success << " / " << global.pick.pop .mask_attempt << ")\n";
+                double avgQ_push = double(global.qstat.push.value) / global.qstat.push.count;
+                double avgQ_pop  = double(global.qstat.pop .value) / global.qstat.pop .count;
+                double avgQ      = double(global.qstat.push.value + global.qstat.pop .value) / (global.qstat.push.count + global.qstat.pop .count);
+                os << "Push   Avg Qs : " << avgQ_push << " (" << global.qstat.push.count << "ops)\n";
+                os << "Pop    Avg Qs : " << avgQ_pop  << " (" << global.qstat.pop .count << "ops)\n";
+                os << "Global Avg Qs : " << avgQ      << " (" << (global.qstat.push.count + global.qstat.pop .count) << "ops)\n";
+        }
+#endif
+};
+};

doc/theses/thierry_delisle_PhD/code/utils.hpp

-              r7030dab
+              r71d6bd8
 #include <unistd.h>
 #include <sys/sysinfo.h>
-#include <x86intrin.h>
 // Barrier from
 …
+}
 static inline void affinity(int tid) {
+void affinity(int tid) {
         static int cpus = get_nprocs();
 …
 static const constexpr std::size_t cache_line_size = 64;
 static inline void check_cache_line_size() {
+void check_cache_line_size() {
         std::cout << "Checking cache line size" << std::endl;
         const std::string cache_file = "/sys/devices/system/cpu/cpu0/cache/index0/coherency_line_size";
 …
         return std::chrono::duration_cast<std::chrono::duration<T, Ratio>>(std::chrono::duration<T>(seconds)).count();
+}
-static inline unsigned rand_bit(unsigned rnum, size_t mask) {
-        unsigned bit = mask ? rnum % __builtin_popcountl(mask) : 0;
-#if !defined(__BMI2__)
-        uint64_t v = mask;   // Input value to find position with rank r.
-        unsigned int r = bit + 1;// Input: bit's desired rank [1-64].
-        unsigned int s;      // Output: Resulting position of bit with rank r [1-64]
-        uint64_t a, b, c, d; // Intermediate temporaries for bit count.
-        unsigned int t;      // Bit count temporary.
-        // Do a normal parallel bit count for a 64-bit integer,
-        // but store all intermediate steps.
-        a =  v - ((v >> 1) & ~0UL/3);
-        b = (a & ~0UL/5) + ((a >> 2) & ~0UL/5);
-        c = (b + (b >> 4)) & ~0UL/0x11;
-        d = (c + (c >> 8)) & ~0UL/0x101;
-        t = (d >> 32) + (d >> 48);
-        // Now do branchless select!
-        s  = 64;
-        s -= ((t - r) & 256) >> 3; r -= (t & ((t - r) >> 8));
-        t  = (d >> (s - 16)) & 0xff;
-        s -= ((t - r) & 256) >> 4; r -= (t & ((t - r) >> 8));
-        t  = (c >> (s - 8)) & 0xf;
-        s -= ((t - r) & 256) >> 5; r -= (t & ((t - r) >> 8));
-        t  = (b >> (s - 4)) & 0x7;
-        s -= ((t - r) & 256) >> 6; r -= (t & ((t - r) >> 8));
-        t  = (a >> (s - 2)) & 0x3;
-        s -= ((t - r) & 256) >> 7; r -= (t & ((t - r) >> 8));
-        t  = (v >> (s - 1)) & 0x1;
-        s -= ((t - r) & 256) >> 8;
-        return s - 1;
-#else
-        uint64_t picked = _pdep_u64(1ul << bit, mask);
-        return picked ? __builtin_ctzl(picked) : 0;
-#endif
+}

doc/user/user.tex

-              r7030dab
+              r71d6bd8
 %% Created On       : Wed Apr  6 14:53:29 2016
 %% Last Modified By : Peter A. Buhr
 %% Last Modified On : Fri Mar  6 13:34:52 2020
 %% Update Count     : 3924
+%% Last Modified On : Sat Jul 13 18:36:18 2019
+%% Update Count     : 3876
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 …
 Even with all its problems, C continues to be popular because it allows writing software at virtually any level in a computer system without restriction.
 For system programming, where direct access to hardware, storage management, and real-time issues are a requirement, C is usually the only language of choice.
 The TIOBE index~\cite{TIOBE} for February 2020 ranks the top six most \emph{popular} programming languages as \Index*{Java} 17.4\%, C 16.8\%, Python 9.3\%, \Index*[C++]{\CC{}} 6.2\%, \Csharp 5.9\%, Visual Basic 5.9\% = 61.5\%, where the next 50 languages are less than 2\% each, with a long tail.
 The top 4 rankings over the past 35 years are:
+The TIOBE index~\cite{TIOBE} for July 2018 ranks the top five most \emph{popular} programming languages as \Index*{Java} 16\%, C 14\%, \Index*[C++]{\CC{}} 7.5\%, Python 6\%, Visual Basic 4\% = 47.5\%, where the next 50 languages are less than 4\% each, with a long tail.
+The top 3 rankings over the past 30 years are:
 \begin{center}
 \setlength{\tabcolsep}{10pt}
+\begin{tabular}{@{}rcccccccc@{}}
+                & 2020  & 2015  & 2010  & 2005  & 2000  & 1995  & 1990  & 1985  \\ \hline
+Java    & 1             & 2             & 1             & 2             & 3             & -             & -             & -             \\
+\R{C}   & \R{2} & \R{1} & \R{2} & \R{1} & \R{1} & \R{2} & \R{1} & \R{1} \\
+Python  & 3             & 7             & 6             & 6             & 22    & 21    & -             & -             \\
+\CC             & 4             & 4             & 4             & 3             & 2             & 1             & 2             & 12    \\
+\begin{tabular}{@{}rccccccc@{}}
+                & 2018  & 2013  & 2008  & 2003  & 1998  & 1993  & 1988  \\ \hline
+Java    & 1             & 2             & 1             & 1             & 16    & -             & -             \\
+\R{C}   & \R{2} & \R{1} & \R{2} & \R{2} & \R{1} & \R{1} & \R{1} \\
+\CC             & 3             & 4             & 3             & 3             & 2             & 2             & 5             \\
 \end{tabular}
 \end{center}
 …
 Keyword clashes are accommodated by syntactic transformations using the \CFA backquote escape-mechanism:
 \begin{cfa}
 int ®``®otype = 3; §\C{// make keyword an identifier}§
 double ®``®forall = 3.5;
+int ®`®otype®`® = 3; §\C{// make keyword an identifier}§
+double ®`®forall®`® = 3.5;
 \end{cfa}
 …
 // include file uses the CFA keyword "with".
 #if ! defined( with ) §\C{// nesting ?}§
 #define with ®``®with §\C{// make keyword an identifier}§
+#define with ®`®with®`® §\C{// make keyword an identifier}§
 #define __CFA_BFD_H__
 #endif
+§{\color{red}\#\textbf{include\_next} <bfdlink.h>}§ §\C{// must have internal check for multiple expansion}§
+®#include_next <bfdlink.h> §\C{// must have internal check for multiple expansion}§
+®
 #if defined( with ) && defined( __CFA_BFD_H__ ) §\C{// reset only if set}§
 #undef with
 …
 \section{Exponentiation Operator}
 C, \CC, and Java (and many other programming languages) have no exponentiation operator\index{exponentiation!operator}\index{operator!exponentiation}, \ie $x^y$, and instead use a routine, like \Indexc{pow(x,y)}, to perform the exponentiation operation.
 \CFA extends the basic operators with the exponentiation operator ©?®\®?©\index{?\\?@©?®\®?©} and ©?\=?©\index{?\\=?@©®\®=?©}, as in, ©x ®\® y© and ©x ®\®= y©, which means $x^y$ and $x \leftarrow x^y$.
+C, \CC, and Java (and many other programming languages) have no exponentiation operator\index{exponentiation!operator}\index{operator!exponentiation}, \ie $x^y$, and instead use a routine, like \Indexc{pow}, to perform the exponentiation operation.
+\CFA extends the basic operators with the exponentiation operator ©?\?©\index{?\\?@©?\?©} and ©?\=?©\index{?\\=?@©\=?©}, as in, ©x \ y© and ©x \= y©, which means $x^y$ and $x \leftarrow x^y$.
 The priority of the exponentiation operator is between the cast and multiplicative operators, so that ©w * (int)x \ (int)y * z© is parenthesized as ©((w * (((int)x) \ ((int)y))) * z)©.
 There are exponentiation operators for integral and floating types, including the builtin \Index{complex} types.
+As for \Index{division}, there are exponentiation operators for integral and floating types, including the builtin \Index{complex} types.
 Integral exponentiation\index{exponentiation!unsigned integral} is performed with repeated multiplication\footnote{The multiplication computation is $O(\log y)$.} (or shifting if the exponent is 2).
 Overflow for a large exponent or negative exponent returns zero.
+Overflow from large exponents or negative exponents return zero.
 Floating exponentiation\index{exponentiation!floating} is performed using \Index{logarithm}s\index{exponentiation!logarithm}, so the exponent cannot be negative.
 \begin{cfa}
 …
 1 256 -64 125 ®0® 3273344365508751233 ®0® ®0® -0.015625 18.3791736799526 0.264715-1.1922i
 \end{cfa}
 Note, ©5 \ 32© and ©5L \ 64© overflow, and ©-4 \ -3© is a fraction but stored in an integer so all three computations generate an integral zero.
+Note, ©5 ®\® 32© and ©5L ®\® 64© overflow, and ©-4 ®\® -3© is a fraction but stored in an integer so all three computations generate an integral zero.
 Parenthesis are necessary for complex constants or the expression is parsed as ©1.0f+®(®2.0fi \ 3.0f®)®+2.0fi©.
 The exponentiation operator is available for all the basic types, but for user-defined types, only the integral-computation version is available.
 …
 OT ?®\®?( OT ep, unsigned long int y );
 \end{cfa}
 The user type ©T© must define multiplication, one (©1©), and ©*©.
+The user type ©T© must define multiplication, one, ©1©, and, ©*©.
 …
+%\section{\texorpdfstring{\protect\lstinline@case@ Clause}{case Clause}}
+\subsection{\texorpdfstring{\LstKeywordStyle{case} Clause}{case Clause}}
+C restricts the ©case© clause of a ©switch© statement to a single value.
+For multiple ©case© clauses associated with the same statement, it is necessary to have multiple ©case© clauses rather than multiple values.
+Requiring a ©case© clause for each value does not seem to be in the spirit of brevity normally associated with C.
+Therefore, the ©case© clause is extended with a list of values, as in:
+\subsection{Loop Control}
+The ©for©/©while©/©do-while© loop-control allows empty or simplified ranges (see Figure~\ref{f:LoopControlExamples}).
+\begin{itemize}
+\item
+An empty conditional implies ©1©.
+\item
+The up-to range ©~©\index{~@©~©} means exclusive range [M,N).
+\item
+The up-to range ©~=©\index{~=@©~=©} means inclusive range [M,N].
+\item
+The down-to range ©-~©\index{-~@©-~©} means exclusive range [N,M).
+\item
+The down-to range ©-~=©\index{-~=@©-~=©} means inclusive range [N,M].
+\item
+©@© means put nothing in this field.
+\item
+©0© is the implicit start value;
+\item
+©1© is the implicit increment value.
+\item
+The up-to range uses ©+=© for increment;
+\item
+The down-to range uses ©-=© for decrement.
+\item
+The loop index is polymorphic in the type of the start value or comparison value when start is implicitly ©0©.
+\end{itemize}
+\begin{figure}
 \begin{cquote}
+\begin{tabular}{@{}l@{\hspace{3em}}l@{\hspace{2em}}l@{}}
+\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c@{\hspace{2em}}}{\textbf{C}} \\
+\begin{cfa}
+switch ( i ) {
+  case ®1, 3, 5®:
+        ...
+  case ®2, 4, 6®:
+        ...
+}
+\begin{tabular}{@{}l|l@{}}
+\multicolumn{1}{c|}{loop control} & \multicolumn{1}{c}{output} \\
+\hline
+\begin{cfa}
+sout | nlOff;
+while ®()® { sout | "empty"; break; } sout | nl;
+do { sout | "empty"; break; } while ®()®; sout | nl;
+for ®()® { sout | "empty"; break; } sout | nl;
+for ( ®0® ) { sout | "A"; } sout | "zero" | nl;
+for ( ®1® ) { sout | "A"; } sout | nl;
+for ( ®10® ) { sout | "A"; } sout | nl;
+for ( ®1 ~= 10 ~ 2® ) { sout | "B"; } sout | nl;
+for ( ®10 -~= 1 ~ 2® ) { sout | "C"; } sout | nl;
+for ( ®0.5 ~ 5.5® ) { sout | "D"; } sout | nl;
+for ( ®5.5 -~ 0.5® ) { sout | "E"; } sout | nl;
+for ( ®i; 10® ) { sout | i; } sout | nl;
+for ( ®i; 1 ~= 10 ~ 2® ) { sout | i; } sout | nl;
+for ( ®i; 10 -~= 1 ~ 2® ) { sout | i; } sout | nl;
+for ( ®i; 0.5 ~ 5.5® ) { sout | i; } sout | nl;
+for ( ®i; 5.5 -~ 0.5® ) { sout | i; } sout | nl;
+for ( ®ui; 2u ~= 10u ~ 2u® ) { sout | ui; } sout | nl;
+for ( ®ui; 10u -~= 2u ~ 2u® ) { sout | ui; } sout | nl;
+enum { N = 10 };
+for ( ®N® ) { sout | "N"; } sout | nl;
+for ( ®i; N® ) { sout | i; } sout | nl;
+for ( ®i; N -~ 0® ) { sout | i; } sout | nl;
+const int start = 3, comp = 10, inc = 2;
+for ( ®i; start ~ comp ~ inc + 1® ) { sout | i; } sout | nl;
+for ( ®i; 1 ~ @® ) { if ( i > 10 ) break;
+        sout | i; } sout | nl;
+for ( ®i; 10 -~ @® ) { if ( i < 0 ) break;
+        sout | i; } sout | nl;
+for ( ®i; 2 ~ @ ~ 2® ) { if ( i > 10 ) break;
+        sout | i; } sout | nl;
+for ( ®i; 2.1 ~ @ ~ @® ) { if ( i > 10.5 ) break;
+        sout | i; i += 1.7; } sout | nl;
+for ( ®i; 10 -~ @ ~ 2® ) { if ( i < 0 ) break;
+        sout | i; } sout | nl;
+for ( ®i; 12.1 ~ @ ~ @® ) { if ( i < 2.5 ) break;
+        sout | i; i -= 1.7; } sout | nl;
+for ( ®i; 5 : j; -5 ~ @® ) { sout | i | j; } sout | nl;
+for ( ®i; 5 : j; -5 -~ @® ) { sout | i | j; } sout | nl;
+for ( ®i; 5 : j; -5 ~ @ ~ 2® ) { sout | i | j; } sout | nl;
+for ( ®i; 5 : j; -5 -~ @ ~ 2® ) { sout | i | j; } sout | nl;
+for ( ®j; -5 ~ @ : i; 5® ) { sout | i | j; } sout | nl;
+for ( ®j; -5 -~ @ : i; 5® ) { sout | i | j; } sout | nl;
+for ( ®j; -5 ~ @ ~ 2 : i; 5® ) { sout | i | j; } sout | nl;
+for ( ®j; -5 -~ @ ~ 2 : i; 5® ) { sout | i | j; } sout | nl;
+for ( ®j; -5 -~ @ ~ 2 : i; 5 : k; 1.5 ~ @® ) {
+        sout | i | j | k; } sout | nl;
+for ( ®j; -5 -~ @ ~ 2 : k; 1.5 ~ @ : i; 5® ) {
+        sout | i | j | k; } sout | nl;
+for ( ®k; 1.5 ~ @ : j; -5 -~ @ ~ 2 : i; 5® ) {
+        sout | i | j | k; } sout | nl;
 \end{cfa}
+&
 \begin{cfa}
+switch ( i ) {
+  case 1: case 3 : case 5:
+        ...
+  case 2: case 4 : case 6:
+        ...
+}
+\end{cfa}
+&
+\begin{cfa}
+// odd values
+// even values
+empty
+empty
+empty
+zero
+A
+A A A A A A A A A A
+B B B B B
+C C C C C
+D D D D D
+E E E E E
+1 2 3 4 5 6 7 8 9
+3 5 7 9
+8 6 4 2
+.5 1.5 2.5 3.5 4.5
+.5 4.5 3.5 2.5 1.5
+4 6 8 10
+8 6 4 2
+N N N N N N N N N N
+1 2 3 4 5 6 7 8 9
+9 8 7 6 5 4 3 2 1
+6 9
+2 3 4 5 6 7 8 9 10
+9 8 7 6 5 4 3 2 1 0
+4 6 8 10
+.1 3.8 5.5 7.2 8.9
+8 6 4 2 0
+.1 10.4 8.7 7 5.3 3.6
+-5 1 -4 2 -3 3 -2 4 -1
+-5 1 -6 2 -7 3 -8 4 -9
+-5 1 -3 2 -1 3 1 4 3
+-5 1 -7 2 -9 3 -11 4 -13
+-5 1 -4 2 -3 3 -2 4 -1
+-5 1 -6 2 -7 3 -8 4 -9
+-5 1 -3 2 -1 3 1 4 3
+-5 1 -7 2 -9 3 -11 4 -13
+-5 1.5 1 -7 2.5 2 -9 3.5 3 -11 4.5 4 -13 5.5
+-5 1.5 1 -7 2.5 2 -9 3.5 3 -11 4.5 4 -13 5.5
+-5 1.5 1 -7 2.5 2 -9 3.5 3 -11 4.5 4 -13 5.5
 \end{cfa}
 \end{tabular}
 \end{cquote}
+In addition, subranges are allowed to specify case values.\footnote{
+gcc has the same mechanism but awkward syntax, \lstinline@2 ...42@, because a space is required after a number, otherwise the period is a decimal point.}
+\begin{cfa}
+switch ( i ) {
+  case ®1~5:® §\C{// 1, 2, 3, 4, 5}§
+        ...
+  case ®10~15:® §\C{// 10, 11, 12, 13, 14, 15}§
+        ...
+}
+\end{cfa}
+Lists of subranges are also allowed.
+\begin{cfa}
+case ®1~5, 12~21, 35~42®:
+\end{cfa}
+\caption{Loop Control Examples}
+\label{f:LoopControlExamples}
+\end{figure}
 …
+\subsection{Non-terminating and Labelled \texorpdfstring{\LstKeywordStyle{fallthrough}}{Non-terminating and Labelled fallthrough}}
+The ©fallthrough© clause may be non-terminating within a ©case© clause or have a target label to common code from multiple case clauses.
+\begin{center}
+\begin{tabular}{@{}lll@{}}
+\begin{cfa}
+choose ( ... ) {
+  case 3:
+        if ( ... ) {
+                ... ®fallthru;® // goto case 4
+        } else {
+                ...
+        }
+        // implicit break
+  case 4:
+%\section{\texorpdfstring{\protect\lstinline@case@ Clause}{case Clause}}
+\subsection{\texorpdfstring{\LstKeywordStyle{case} Statement}{case Statement}}
+C restricts the ©case© clause of a ©switch© statement to a single value.
+For multiple ©case© clauses associated with the same statement, it is necessary to have multiple ©case© clauses rather than multiple values.
+Requiring a ©case© clause for each value does not seem to be in the spirit of brevity normally associated with C.
+Therefore, the ©case© clause is extended with a list of values, as in:
+\begin{cquote}
+\begin{tabular}{@{}l@{\hspace{3em}}l@{\hspace{2em}}l@{}}
+\multicolumn{1}{c@{\hspace{3em}}}{\textbf{\CFA}}        & \multicolumn{1}{c@{\hspace{2em}}}{\textbf{C}} \\
+\begin{cfa}
+switch ( i ) {
+  case ®1, 3, 5®:
+        ...
+  case ®2, 4, 6®:
+        ...
+}
 \end{cfa}
+&
 \begin{cfa}
+choose ( ... ) {
+  case 3:
+        ... ®fallthrough common;®
+  case 4:
+        ... ®fallthrough common;®
+  ®common:® // below fallthrough
+                          // at case-clause level
+        ...     // common code for cases 3/4
+        // implicit break
+  case 4:
+switch ( i ) {
+  case 1: case 3 : case 5:
+        ...
+  case 2: case 4 : case 6:
+        ...
+}
 \end{cfa}
+&
 \begin{cfa}
+choose ( ... ) {
+  case 3:
+        choose ( ... ) {
+          case 4:
+                for ( ... ) {
+                        // multi-level transfer
+                        ... ®fallthru common;®
+                }
+                ...
+        }
+// odd values
+// even values
+\end{cfa}
+\end{tabular}
+\end{cquote}
+In addition, subranges are allowed to specify case values.\footnote{
+gcc has the same mechanism but awkward syntax, \lstinline@2 ...42@, because a space is required after a number, otherwise the period is a decimal point.}
+\begin{cfa}
+switch ( i ) {
+  case ®1~5:® §\C{// 1, 2, 3, 4, 5}§
         ...
+  ®common:® // below fallthrough
+                          // at case-clause level
+\end{cfa}
+\end{tabular}
+\end{center}
+The target label must be below the ©fallthrough© and may not be nested in a control structure, and
+the target label must be at the same or higher level as the containing ©case© clause and located at
+the same level as a ©case© clause; the target label may be case ©default©, but only associated
+with the current ©switch©/©choose© statement.
+\subsection{Loop Control}
+The ©for©/©while©/©do-while© loop-control allows empty or simplified ranges (see Figure~\ref{f:LoopControlExamples}).
+\begin{itemize}
+\item
+The loop index is polymorphic in the type of the comparison value N (when the start value is implicit) or the start value M.
+\item
+An empty conditional implies comparison value of ©1© (true).
+\item
+A comparison N is implicit up-to exclusive range [0,N©®)®©.
+\item
+A comparison ©=© N is implicit up-to inclusive range [0,N©®]®©.
+\item
+The up-to range M ©~©\index{~@©~©} N means exclusive range [M,N©®)®©.
+\item
+The up-to range M ©~=©\index{~=@©~=©} N means inclusive range [M,N©®]®©.
+\item
+The down-to range M ©-~©\index{-~@©-~©} N means exclusive range [N,M©®)®©.
+\item
+The down-to range M ©-~=©\index{-~=@©-~=©} N means inclusive range [N,M©®]®©.
+\item
+©0© is the implicit start value;
+\item
+©1© is the implicit increment value.
+\item
+The up-to range uses operator ©+=© for increment;
+\item
+The down-to range uses operator ©-=© for decrement.
+\item
+©@© means put nothing in this field.
+\item
+©:© means start another index.
+\end{itemize}
+\begin{figure}
+\begin{tabular}{@{}l|l@{}}
+\multicolumn{1}{c|}{loop control} & \multicolumn{1}{c}{output} \\
+\hline
+\begin{cfa}[xleftmargin=0pt]
+while ®()® { sout | "empty"; break; }
+do { sout | "empty"; break; } while ®()®;
+for ®()® { sout | "empty"; break; }
+for ( ®0® ) { sout | "A"; } sout | "zero";
+for ( ®1® ) { sout | "A"; }
+for ( ®10® ) { sout | "A"; }
+for ( ®= 10® ) { sout | "A"; }
+for ( ®1 ~= 10 ~ 2® ) { sout | "B"; }
+for ( ®10 -~= 1 ~ 2® ) { sout | "C"; }
+for ( ®0.5 ~ 5.5® ) { sout | "D"; }
+for ( ®5.5 -~ 0.5® ) { sout | "E"; }
+for ( ®i; 10® ) { sout | i; }
+for ( ®i; = 10® ) { sout | i; }
+for ( ®i; 1 ~= 10 ~ 2® ) { sout | i; }
+for ( ®i; 10 -~= 1 ~ 2® ) { sout | i; }
+for ( ®i; 0.5 ~ 5.5® ) { sout | i; }
+for ( ®i; 5.5 -~ 0.5® ) { sout | i; }
+for ( ®ui; 2u ~= 10u ~ 2u® ) { sout | ui; }
+for ( ®ui; 10u -~= 2u ~ 2u® ) { sout | ui; }
+enum { N = 10 };
+for ( ®N® ) { sout | "N"; }
+for ( ®i; N® ) { sout | i; }
+for ( ®i; N -~ 0® ) { sout | i; }
+const int start = 3, comp = 10, inc = 2;
+for ( ®i; start ~ comp ~ inc + 1® ) { sout | i; }
+for ( i; 1 ~ ®@® ) { if ( i > 10 ) break; sout | i; }
+for ( i; 10 -~ ®@® ) { if ( i < 0 ) break; sout | i; }
+for ( i; 2 ~ ®@® ~ 2 ) { if ( i > 10 ) break; sout | i; }
+for ( i; 2.1 ~ ®@® ~ ®@® ) { if ( i > 10.5 ) break; sout | i; i += 1.7; }
+for ( i; 10 -~ ®@® ~ 2 ) { if ( i < 0 ) break; sout | i; }
+for ( i; 12.1 ~ ®@® ~ ®@® ) { if ( i < 2.5 ) break; sout | i; i -= 1.7; }
+for ( i; 5 ®:® j; -5 ~ @ ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 -~ @ ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 ~ @ ~ 2 ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 -~ @ ~ 2 ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 ~ @ ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 -~ @ ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 ~ @ ~ 2 ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 -~ @ ~ 2 ) { sout | i | j; }
+for ( i; 5 ®:® j; -5 -~ @ ~ 2 ®:® k; 1.5 ~ @ ) { sout | i | j | k; }
+for ( i; 5 ®:® j; -5 -~ @ ~ 2 ®:® k; 1.5 ~ @ ) { sout | i | j | k; }
+for ( i; 5 ®:® k; 1.5 ~ @ ®:® j; -5 -~ @ ~ 2 ) { sout | i | j | k; }
+\end{cfa}
+&
+\begin{cfa}
+empty
+empty
+empty
+zero
+A
+A A A A A A A A A A
+A A A A A A A A A A A
+B B B B B
+C C C C C
+D D D D D
+E E E E E
+1 2 3 4 5 6 7 8 9
+1 2 3 4 5 6 7 8 9 10
+3 5 7 9
+8 6 4 2
+.5 1.5 2.5 3.5 4.5
+.5 4.5 3.5 2.5 1.5
+4 6 8 10
+8 6 4 2
+N N N N N N N N N N
+1 2 3 4 5 6 7 8 9
+9 8 7 6 5 4 3 2 1
+6 9
+2 3 4 5 6 7 8 9 10
+9 8 7 6 5 4 3 2 1 0
+4 6 8 10
+.1 3.8 5.5 7.2 8.9
+8 6 4 2 0
+.1 10.4 8.7 7. 5.3 3.6
+-5 1 -4 2 -3 3 -2 4 -1
+-5 1 -6 2 -7 3 -8 4 -9
+-5 1 -3 2 -1 3 1 4 3
+-5 1 -7 2 -9 3 -11 4 -13
+-5 1 -4 2 -3 3 -2 4 -1
+-5 1 -6 2 -7 3 -8 4 -9
+-5 1 -3 2 -1 3 1 4 3
+-5 1 -7 2 -9 3 -11 4 -13
+-5 1.5 1 -7 2.5 2 -9 3.5 3 -11 4.5 4 -13 5.5
+-5 1.5 1 -7 2.5 2 -9 3.5 3 -11 4.5 4 -13 5.5
+-5 1.5 1 -7 2.5 2 -9 3.5 3 -11 4.5 4 -13 5.5
+\end{cfa}
+\end{tabular}
+\caption{Loop Control Examples}
+\label{f:LoopControlExamples}
+\end{figure}
+  case ®10~15:® §\C{// 10, 11, 12, 13, 14, 15}§
+        ...
+}
+\end{cfa}
+Lists of subranges are also allowed.
+\begin{cfa}
+case ®1~5, 12~21, 35~42®:
+\end{cfa}
 % for ()  => for ( ;; )
 …
 hence, names in these include files are not mangled\index{mangling!name} (see~\VRef{s:Interoperability}).
 All other C header files must be explicitly wrapped in ©extern "C"© to prevent name mangling.
 This approach is different from \Index*[C++]{\CC{}} where the name-mangling issue is handled internally in C header-files through checks for preprocessor variable ©__cplusplus©, which adds appropriate ©extern "C"© qualifiers.
+For \Index*[C++]{\CC{}}, the name-mangling issue is often handled internally in many C header-files through checks for preprocessor variable ©__cplusplus©, which adds appropriate ©extern "C"© qualifiers.
 …
 The storage-management routines extend their C equivalents by overloading, alternate names, providing shallow type-safety, and removing the need to specify the allocation size for non-array types.
 C storage management provides the following capabilities:
+Storage management provides the following capabilities:
 \begin{description}
 \item[filled]
 after allocation with a specified character or value.
+\item[fill]
+after allocation the storage is filled with a specified character.
 \item[resize]
 an existing allocation to decreased or increased its size.
 In either case, new storage may or may not be allocated and, if there is a new allocation, as much data from the existing allocation is copied into the new allocation.
+an existing allocation is decreased or increased in size.
+In either case, new storage may or may not be allocated and, if there is a new allocation, as much data from the existing allocation is copied.
 For an increase in storage size, new storage after the copied data may be filled.
 \item[align]
 an allocation on a specified memory boundary, \eg, an address multiple of 64 or 128 for cache-line purposes.
+\item[alignment]
+an allocation starts on a specified memory boundary, \eg, an address multiple of 64 or 128 for cache-line purposes.
 \item[array]
 the allocation size is scaled to the specified number of array elements.
 An array may be filled, resized, or aligned.
 \end{description}
+\VRef[Table]{t:AllocationVersusCapabilities} shows allocation routines supporting different combinations of storage-management capabilities.
+\begin{table}
+\centering
+\begin{minipage}{0.75\textwidth}
+\begin{tabular}{@{}r|l|l|l|l|l@{}}
+The table shows allocation routines supporting different combinations of storage-management capabilities:
+\begin{center}
+\begin{tabular}{@{}r|r|l|l|l|l@{}}
 \multicolumn{1}{c}{}&           & \multicolumn{1}{c|}{fill}     & resize        & alignment     & array \\
 \hline
 C               & ©malloc©                      & no                    & no            & no            & no    \\
                 & ©calloc©                      & yes (0 only)  & no            & no            & yes   \\
                 & ©realloc©                     & copy                  & yes           & no            & no    \\
+                & ©realloc©                     & no/copy               & yes           & no            & no    \\
                 & ©memalign©            & no                    & no            & yes           & no    \\
-                & ©aligned_alloc©\footnote{Same as ©memalign© but size is an integral multiple of alignment, which is universally ignored.}
-                                                        & no                    & no            & yes           & no    \\
                 & ©posix_memalign©      & no                    & no            & yes           & no    \\
-                & ©valloc©                      & no                    & no            & yes (page size)& no   \\
-                & ©pvalloc©\footnote{Same as ©valloc© but rounds size to multiple of page size.}
-                                                        & no                    & no            & yes (page size)& no   \\
 \hline
+\CFA    & ©cmemalign©           & yes (0 only)  & no            & yes           & yes   \\
+                & ©realloc©                     & copy                  & yes           & yes           & no    \\
+                & ©alloc©                       & no                    & yes           & no            & yes   \\
+                & ©alloc_set©           & yes                   & yes           & no            & yes   \\
+                & ©alloc_align©         & no                    & yes           & yes           & yes   \\
+                & ©alloc_align_set©     & yes                   & yes           & yes           & yes   \\
+C11             & ©aligned_alloc©       & no                    & no            & yes           & no    \\
+\hline
+\CFA    & ©alloc©                       & no/copy/yes   & no/yes        & no            & yes   \\
+                & ©align_alloc©         & no/yes                & no            & yes           & yes   \\
 \end{tabular}
+\end{minipage}
+\caption{Allocation Routines versus Storage-Management Capabilities}
+\label{t:AllocationVersusCapabilities}
+\end{table}
+\CFA memory management extends the type safety of all allocations by using the type of the left-hand-side type to determine the allocation size and return a matching type for the new storage.
+Type-safe allocation is provided for all C allocation routines and new \CFA allocation routines, \eg in
+\begin{cfa}
+int * ip = (int *)malloc( sizeof(int) );                §\C{// C}§
+int * ip = malloc();                                                    §\C{// \CFA type-safe version of C malloc}§
+int * ip = alloc();                                                             §\C{// \CFA type-safe uniform alloc}§
+\end{cfa}
+the latter two allocations determine the allocation size from the type of ©p© (©int©) and cast the pointer to the allocated storage to ©int *©.
+\CFA memory management extends allocation safety by implicitly honouring all alignment requirements, \eg in
+\begin{cfa}
+struct S { int i; } __attribute__(( aligned( 128 ) )); // cache-line alignment
+S * sp = malloc();                                                              §\C{// honour type alignment}§
+\end{cfa}
+the storage allocation is implicitly aligned to 128 rather than the default 16.
+The alignment check is performed at compile time so there is no runtime cost.
+\CFA memory management extends the resize capability with the notion of \newterm{sticky properties}.
+Hence, initial allocation capabilities are remembered and maintained when resize requires copying.
+For example, an initial alignment and fill capability are preserved during a resize copy so the copy has the same alignment and extended storage is filled.
+Without sticky properties it is dangerous to use ©realloc©, resulting in an idiom of manually performing the reallocation to maintain correctness.
+\CFA memory management extends allocation to support constructors for initialization of allocated storage, \eg in
+\begin{cfa}
+struct S { int i; };                                                    §\C{// cache-line aglinment}§
+void ?{}( S & s, int i ) { s.i = i; }
+// assume ?|? operator for printing an S
+S & sp = *®new®( 3 );                                                   §\C{// call constructor after allocation}§
+sout | sp.i;
+®delete®( &sp );
+S * spa = ®anew®( 10, 5 );                                              §\C{// allocate array and initialize each array element}§
+for ( i; 10 ) sout | spa[i] | nonl;
+sout | nl;
+®adelete®( 10, spa );
+\end{cfa}
+Allocation routines ©new©/©anew© allocate a variable/array and initialize storage using the allocated type's constructor.
+Note, the matching deallocation routines ©delete©/©adelete©.
+\end{center}
+It is impossible to resize with alignment because the underlying ©realloc© allocates storage if more space is needed, and it does not honour alignment from the original allocation.
 \leavevmode
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
+// C unsafe allocation
 extern "C" {
+        // C unsafe allocation
+        void * malloc( size_t size );§\indexc{malloc}§
+        void * calloc( size_t dim, size_t size );§\indexc{calloc}§
+        void * realloc( void * ptr, size_t size );§\indexc{realloc}§
+        void * memalign( size_t align, size_t size );§\indexc{memalign}§
+        void * aligned_alloc( size_t align, size_t size );§\indexc{aligned_alloc}§
+        int posix_memalign( void ** ptr, size_t align, size_t size );§\indexc{posix_memalign}§
+        void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize );§\indexc{cmemalign}§ // CFA
+        // C unsafe initialization/copy
+        void * memset( void * dest, int c, size_t size );§\indexc{memset}§
+        void * memcpy( void * dest, const void * src, size_t size );§\indexc{memcpy}§
+}
+void * realloc( void * oaddr, size_t nalign, size_t size ); // CFA heap
+void * malloc( size_t size );§\indexc{memset}§
+void * calloc( size_t dim, size_t size );§\indexc{calloc}§
+void * realloc( void * ptr, size_t size );§\indexc{realloc}§
+void * memalign( size_t align, size_t size );§\indexc{memalign}§
+int posix_memalign( void ** ptr, size_t align, size_t size );§\indexc{posix_memalign}§
+// C unsafe initialization/copy
+void * memset( void * dest, int c, size_t size );
+void * memcpy( void * dest, const void * src, size_t size );
+}
 forall( dtype T | sized(T) ) {
         // §\CFA§ safe equivalents, i.e., implicit size specification
+// §\CFA§ safe equivalents, i.e., implicit size specification
         T * malloc( void );
         T * calloc( size_t dim );
         T * realloc( T * ptr, size_t size );
         T * memalign( size_t align );
-        T * cmemalign( size_t align, size_t dim  );
         T * aligned_alloc( size_t align );
         int posix_memalign( T ** ptr, size_t align );
         // §\CFA§ safe general allocation, fill, resize, alignment, array
+// §\CFA§ safe general allocation, fill, resize, array
         T * alloc( void );§\indexc{alloc}§
+        T * alloc( char fill );
         T * alloc( size_t dim );
+        T * alloc( size_t dim, char fill );
         T * alloc( T ptr[], size_t dim );
+        T * alloc_set( char fill );§\indexc{alloc_set}§
+        T * alloc_set( T fill );
+        T * alloc_set( size_t dim, char fill );
+        T * alloc_set( size_t dim, T fill );
+        T * alloc_set( size_t dim, const T fill[] );
+        T * alloc_set( T ptr[], size_t dim, char fill );
+        T * alloc_align( size_t align );
+        T * alloc_align( size_t align, size_t dim );
+        T * alloc_align( T ptr[], size_t align ); // aligned realloc array
+        T * alloc_align( T ptr[], size_t align, size_t dim ); // aligned realloc array
+        T * alloc_align_set( size_t align, char fill );
+        T * alloc_align_set( size_t align, T fill );
+        T * alloc_align_set( size_t align, size_t dim, char fill );
+        T * alloc_align_set( size_t align, size_t dim, T fill );
+        T * alloc_align_set( size_t align, size_t dim, const T fill[] );
+        T * alloc_align_set( T ptr[], size_t align, size_t dim, char fill );
+        // §\CFA§ safe initialization/copy, i.e., implicit size specification
+        T * memset( T * dest, char fill );§\indexc{memset}§
+        T * alloc( T ptr[], size_t dim, char fill );
+// §\CFA§ safe general allocation, align, fill, array
+        T * align_alloc( size_t align );
+        T * align_alloc( size_t align, char fill );
+        T * align_alloc( size_t align, size_t dim );
+        T * align_alloc( size_t align, size_t dim, char fill );
+// §\CFA§ safe initialization/copy, i.e., implicit size specification
+        T * memset( T * dest, char c );§\indexc{memset}§
         T * memcpy( T * dest, const T * src );§\indexc{memcpy}§
+        // §\CFA§ safe initialization/copy, i.e., implicit size specification, array types
         T * amemset( T dest[], char fill, size_t dim );
+// §\CFA§ safe initialization/copy array
+        T * amemset( T dest[], char c, size_t dim );
         T * amemcpy( T dest[], const T src[], size_t dim );
+}
 // §\CFA§ allocation/deallocation and constructor/destructor, non-array types
 forall( dtype T | sized(T), ttype Params | { void ?{}( T &, Params ); } ) T * new( Params p );§\indexc{new}§
 forall( dtype T | sized(T) | { void ^?{}( T & ); } ) void delete( T * ptr );§\indexc{delete}§
 forall( dtype T, ttype Params | sized(T) | { void ^?{}( T & ); void delete( Params ); } )
+// §\CFA§ allocation/deallocation and constructor/destructor
+forall( dtype T | sized(T), ttype Params | { void ?{}( T *, Params ); } ) T * new( Params p );§\indexc{new}§
+forall( dtype T | { void ^?{}( T * ); } ) void delete( T * ptr );§\indexc{delete}§
+forall( dtype T, ttype Params | { void ^?{}( T * ); void delete( Params ); } )
   void delete( T * ptr, Params rest );
 // §\CFA§ allocation/deallocation and constructor/destructor, array types
 forall( dtype T | sized(T), ttype Params | { void ?{}( T &, Params ); } ) T * anew( size_t dim, Params p );§\indexc{anew}§
 forall( dtype T | sized(T) | { void ^?{}( T & ); } ) void adelete( size_t dim, T arr[] );§\indexc{adelete}§
 forall( dtype T | sized(T) | { void ^?{}( T & ); }, ttype Params | { void adelete( Params ); } )
+// §\CFA§ allocation/deallocation and constructor/destructor, array
+forall( dtype T | sized(T), ttype Params | { void ?{}( T *, Params ); } ) T * anew( size_t dim, Params p );§\indexc{anew}§
+forall( dtype T | sized(T) | { void ^?{}( T * ); } ) void adelete( size_t dim, T arr[] );§\indexc{adelete}§
+forall( dtype T | sized(T) | { void ^?{}( T * ); }, ttype Params | { void adelete( Params ); } )
   void adelete( size_t dim, T arr[], Params rest );
 \end{cfa}

driver/cc1.cc

-              r7030dab
+              r71d6bd8
         #endif // __DEBUG_H__
-        enum {
-                Color_Auto   = 0,
-                Color_Always = 1,
-                Color_Never  = 2,
-        } color_arg = Color_Auto;
-        const char * color_names[3] = { "--colors=auto", "--colors=always", "--colors=never" };
         // process all the arguments
 …
                 if ( prefix( arg, "-" ) ) {
                         // strip inappropriate flags
-                        if ( prefix( arg, "-fdiagnostics-color=" ) ) {
-                                string choice = arg.substr(20);
-                                     if(choice == "always") color_arg = Color_Always;
-                                else if(choice == "never" ) color_arg = Color_Never;
-                                else if(choice == "auto"  ) color_arg = Color_Auto;
-                        } else if ( arg == "-fno-diagnostics-color" ) {
-                                color_arg = Color_Auto;
+                        }
                         if ( arg == "-quiet" || arg == "-version" || arg == "-fpreprocessed" ||
 …
                         cargs[ncargs++] = cfa_cpp_out.c_str();
                 } // if
-                cargs[ncargs++] = color_names[color_arg];
                 cargs[ncargs] = nullptr;                                                // terminate argument list

driver/cfa.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Aug 20 13:44:49 2002
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Jan 31 16:48:03 2020
 // Update Count     : 421
+// Last Modified On : Tue Sep 10 17:00:15 2019
+// Update Count     : 420
 //
 …
                         } else if ( arg == "-XCFA" ) {                          // CFA pass through
                                 i += 1;
-                                if ( i == argc ) continue;                              // next argument available ?
                                 Putenv( argv, argv[i] );
 …
                 args[nargs++] = "-Xlinker";
                 args[nargs++] = "--undefined=__cfaabi_appready_startup";
-                args[nargs++] = "-z";
-                args[nargs++] = "execstack";
                 // include the cfa library in case it is needed
 …
                 args[nargs++] = "-lcfathread";
                 args[nargs++] = "-Wl,--pop-state";
-                args[nargs++] = "-Wl,--push-state,--no-as-needed";
                 args[nargs++] = "-lcfa";
+                args[nargs++] = "-Wl,--pop-state";
+                args[nargs++] = "-pthread";
+                args[nargs++] = "-lpthread";
                 args[nargs++] = "-ldl";
                 args[nargs++] = "-lrt";

libcfa/configure

r7030dab	r71d6bd8
3000	3000	case $CONFIGURATION in
3001	3001	"debug" )
3002		CONFIG_CFLAGS="-O0 -g"
	3002	CONFIG_CFLAGS="-Og -g"
3003	3003	CONFIG_CFAFLAGS="-debug"
3004	3004	CONFIG_BUILDLIB="yes"

libcfa/configure.ac

r7030dab	r71d6bd8
68	68	case $CONFIGURATION in
69	69	"debug" )
70		CONFIG_CFLAGS="-O0 -g"
	70	CONFIG_CFLAGS="-Og -g"
71	71	CONFIG_CFAFLAGS="-debug"
72	72	CONFIG_BUILDLIB="yes"

libcfa/prelude/Makefile.am

-              r7030dab
+              r71d6bd8
 ## Created On       : Sun May 31 08:54:01 2015
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Mon Feb  3 21:27:18 2020
 ## Update Count     : 208
+## Last Modified On : Wed Dec 14 15:00:35 2016
+## Update Count     : 205
 ###############################################################################
 …
 extras.cf : ${srcdir}/extras.regx ${srcdir}/extras.c
         ${AM_V_GEN}gcc ${AM_CFLAGS} -E ${srcdir}/extras.c | grep -f ${srcdir}/extras.regx > extras.cf
-        ${AM_V_GEN}gcc ${AM_CFLAGS} -E ${srcdir}/extras.c | grep -zo -f ${srcdir}/extras.regx2 | tr '\0' '\n' >> extras.cf
 # create forward declarations for gcc builtins

libcfa/prelude/Makefile.in

-              r7030dab
+              r71d6bd8
 # Makefile.in generated by automake 1.16.1 from Makefile.am.
+# Makefile.in generated by automake 1.15 from Makefile.am.
 # @configure_input@
 # Copyright (C) 1994-2018 Free Software Foundation, Inc.
+# Copyright (C) 1994-2014 Free Software Foundation, Inc.
 # This Makefile.in is free software; the Free Software Foundation
 …
             cd $(top_builddir) && $(MAKE) $(AM_MAKEFLAGS) am--refresh;; \
           *) \
             echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__maybe_remake_depfiles)'; \
             cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__maybe_remake_depfiles);; \
+            echo ' cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe)'; \
+            cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe);; \
         esac;
 …
+distdir: $(BUILT_SOURCES)
+        $(MAKE) $(AM_MAKEFLAGS) distdir-am
+distdir-am: $(DISTFILES)
+distdir: $(DISTFILES)
         @srcdirstrip=`echo "$(srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \
         topsrcdirstrip=`echo "$(top_srcdir)" | sed 's/[].[^$$\\*]/\\\\&/g'`; \
 …
 extras.cf : ${srcdir}/extras.regx ${srcdir}/extras.c
         ${AM_V_GEN}gcc ${AM_CFLAGS} -E ${srcdir}/extras.c | grep -f ${srcdir}/extras.regx > extras.cf
-        ${AM_V_GEN}gcc ${AM_CFLAGS} -E ${srcdir}/extras.c | grep -zo -f ${srcdir}/extras.regx2 | tr '\0' '\n' >> extras.cf
 # create forward declarations for gcc builtins

libcfa/prelude/builtins.c

-              r7030dab
+              r71d6bd8
 // Created On       : Fri Jul 21 16:21:03 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Nov 21 16:31:39 2019
 // Update Count     : 101
+// Last Modified On : Tue Jun 25 18:06:52 2019
+// Update Count     : 97
 //
 …
 void abort( const char fmt[], ... ) __attribute__ (( format(printf, 1, 2), __nothrow__, __leaf__, __noreturn__ ));
-forall(dtype T)
-static inline T & identity(T & i) {
-        return i;
+}
-// generator support
-struct $generator {
-        inline int;
-};
-static inline void  ?{}($generator & this) { ((int&)this) = 0; }
-static inline void ^?{}($generator &) {}
-trait is_generator(dtype T) {
-      void main(T & this);
-      $generator * get_generator(T & this);
-};
-forall(dtype T | is_generator(T))
-static inline T & resume(T & gen) {
-        main(gen);
-        return gen;
+}
 // implicit increment, decrement if += defined, and implicit not if != defined
 …
 // universal typed pointer constant
+static inline forall( dtype DT ) DT * intptr( uintptr_t addr ) { return (DT *)addr; }
+// Compiler issue: there is a problem with anonymous types that do not have a size.
+static inline forall( dtype DT | sized(DT) ) DT * intptr( uintptr_t addr ) { return (DT *)addr; }
 // exponentiation operator implementation

libcfa/prelude/extras.regx

-              r7030dab
+              r71d6bd8
 typedef.* char32_t;
 typedef.* wchar_t;
+extern.*\*malloc\(.*\).*
+extern.* free\(.*\).*
+extern.* exit\(.*\).*
+extern.* atexit\(.*\).*
+extern.* abort\(.*\).*
+extern.* printf\(.*\).*

libcfa/prelude/prototypes.awk

-              r7030dab
+              r71d6bd8
 # Created On       : Sat May 16 07:57:37 2015
 # Last Modified By : Peter A. Buhr
 # Last Modified On : Sat Feb  8 09:46:58 2020
 # Update Count     : 36
+# Last Modified On : Thu Jun  6 20:46:28 2019
+# Update Count     : 34
+#
 …
 BEGIN {
         FS = "[( )]"
+  FS = "[( )]"
         # order so string search is longest string
         i=-1
 …
 /BT_FN/ {
         for (i = 1; i <= NF; i += 1 ) {
                 if ( match($i, "BT_FN") != 0 ) {
                         prototypes[$i] = $i
+                }
+        for (i = 1; i <= NF; i++) {
+          if( match($i, "BT_FN") != 0 ) {
+                prototypes[$i] = $i
+          }
+        }
+}
+  }
 END {
 …
         for ( prototype in prototypes ) {
+                # printf( "//\"%s\"\n", prototype )
+                if ( index( "BT_LAST", prototype ) == 1 ) {
+                        continue
+          # printf( "//\"%s\"\n", prototype )
+          if ( index( "BT_LAST", prototype ) == 1 ) {
+                continue
+          } # if
+          printf( "#define %s(NAME) FUNC_SIMPLE(", prototype )
+          if ( sub( "BT_FN_", "", prototype ) == 0 ) {
+                printf( "\n********** BAD MACRO NAME \"%s\" **********\n", prototype )
+                exit 0
+          } # if
+          # generate function return type as macro
+          for ( t = 0; t < N; t += 1 ) {                                        # find longest match
+                type = types[t];
+                if ( index( prototype, type ) == 1 ) {          # found match
+                  printf( "BT_%s, NAME", type )
+                  sub( type, "", prototype )
+                  break;
                 } # if
+          } # for
+                printf( "#define %s(NAME) FUNC_SIMPLE(", prototype )
+                if ( sub( "BT_FN_", "", prototype ) == 0 ) {
+                        printf( "\n********** BAD MACRO NAME \"%s\" **********\n", prototype )
+          # generate function parameter types as macro
+          if ( index( prototype, "VAR" ) != 2 ) {                       # C-style empty parameters ?
+                for ( p = 0; length( prototype ) > 0; p += 1 ) { # until all parameters types are removed
+                  sub( "_", "", prototype)                              # remove "_"
+                  printf( ", ", type )
+                  temp = prototype
+                  for ( t = 0; t < N; t += 1 ) {                        # find longest match
+                        type = types[t];
+                        if ( index( prototype, type ) == 1 ) { # found match
+                          printf( "BT_%s", type )
+                          sub( type, "", prototype )
+                          break;
+                        } # if
+                  } # for
+                  if ( temp == prototype ) {                            # no match found for parameter in macro table
+                        printf( "\n********** MISSING TYPE \"%s\" **********\n", prototype )
                         exit 0
+                } # if
+                # generate function return type as macro
+                for ( t = 0; t < N; t += 1 ) {                                  # find longest match
+                        type = types[t];
+                        if ( index( prototype, type ) == 1 ) {          # found match
+                                printf( "BT_%s, NAME", type )
+                                sub( type, "", prototype )
+                                break;
+                        } # if
+                  } # if
                 } # for
+                # generate function parameter types as macro
+                if ( index( prototype, "VAR" ) != 2 ) {                 # C-style empty parameters ?
+                        for ( p = 0; length( prototype ) > 0; p += 1 ) { # until all parameters types are removed
+                                sub( "_", "", prototype)                                # remove "_"
+                                printf( ", ", type )
+                                temp = prototype
+                                for ( t = 0; t < N; t += 1 ) {                  # find longest match
+                                        type = types[t];
+                                        if ( index( prototype, type ) == 1 ) { # found match
+                                                printf( "BT_%s", type )
+                                                sub( type, "", prototype )
+                                                break;
+                                        } # if
+                                } # for
+                                if ( temp == prototype ) {                              # no match found for parameter in macro table
+                                        printf( "\n********** MISSING TYPE \"%s\" **********\n", prototype )
+                                        exit 0
+                                } # if
+                        } # for
+                } # if
+                printf( ")\n" )
+          } # if
+          printf( ")\n" )
         } # for

libcfa/prelude/sync-builtins.cf

-              r7030dab
+              r71d6bd8
 char __sync_fetch_and_add(volatile char *, char,...);
+char __sync_fetch_and_add_1(volatile char *, char,...);
 signed char __sync_fetch_and_add(volatile signed char *, signed char,...);
+signed char __sync_fetch_and_add_1(volatile signed char *, signed char,...);
 unsigned char __sync_fetch_and_add(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_fetch_and_add_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_fetch_and_add(volatile signed short *, signed short,...);
+signed short __sync_fetch_and_add_2(volatile signed short *, signed short,...);
 unsigned short __sync_fetch_and_add(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_fetch_and_add_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_fetch_and_add(volatile signed int *, signed int,...);
+signed int __sync_fetch_and_add_4(volatile signed int *, signed int,...);
 unsigned int __sync_fetch_and_add(volatile unsigned int *, unsigned int,...);
+signed long int __sync_fetch_and_add(volatile signed long int *, signed long int,...);
+unsigned long int __sync_fetch_and_add(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_fetch_and_add_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_fetch_and_add(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_fetch_and_add_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_fetch_and_add(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_fetch_and_add_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_fetch_and_add(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_fetch_and_add_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_fetch_and_add(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_fetch_and_add_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_fetch_and_sub(volatile char *, char,...);
+char __sync_fetch_and_sub_1(volatile char *, char,...);
 signed char __sync_fetch_and_sub(volatile signed char *, signed char,...);
+signed char __sync_fetch_and_sub_1(volatile signed char *, signed char,...);
 unsigned char __sync_fetch_and_sub(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_fetch_and_sub_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_fetch_and_sub(volatile signed short *, signed short,...);
+signed short __sync_fetch_and_sub_2(volatile signed short *, signed short,...);
 unsigned short __sync_fetch_and_sub(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_fetch_and_sub_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_fetch_and_sub(volatile signed int *, signed int,...);
+signed int __sync_fetch_and_sub_4(volatile signed int *, signed int,...);
 unsigned int __sync_fetch_and_sub(volatile unsigned int *, unsigned int,...);
+signed long int __sync_fetch_and_sub(volatile signed long int *, signed long int,...);
+unsigned long int __sync_fetch_and_sub(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_fetch_and_sub_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_fetch_and_sub(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_fetch_and_sub_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_fetch_and_sub(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_fetch_and_sub_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_fetch_and_sub(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_fetch_and_sub_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_fetch_and_sub(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_fetch_and_sub_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_fetch_and_or(volatile char *, char,...);
+char __sync_fetch_and_or_1(volatile char *, char,...);
 signed char __sync_fetch_and_or(volatile signed char *, signed char,...);
+signed char __sync_fetch_and_or_1(volatile signed char *, signed char,...);
 unsigned char __sync_fetch_and_or(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_fetch_and_or_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_fetch_and_or(volatile signed short *, signed short,...);
+signed short __sync_fetch_and_or_2(volatile signed short *, signed short,...);
 unsigned short __sync_fetch_and_or(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_fetch_and_or_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_fetch_and_or(volatile signed int *, signed int,...);
+signed int __sync_fetch_and_or_4(volatile signed int *, signed int,...);
 unsigned int __sync_fetch_and_or(volatile unsigned int *, unsigned int,...);
+signed long int __sync_fetch_and_or(volatile signed long int *, signed long int,...);
+unsigned long int __sync_fetch_and_or(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_fetch_and_or_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_fetch_and_or(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_fetch_and_or_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_fetch_and_or(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_fetch_and_or_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_fetch_and_or(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_fetch_and_or_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_fetch_and_or(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_fetch_and_or_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_fetch_and_and(volatile char *, char,...);
+char __sync_fetch_and_and_1(volatile char *, char,...);
 signed char __sync_fetch_and_and(volatile signed char *, signed char,...);
+signed char __sync_fetch_and_and_1(volatile signed char *, signed char,...);
 unsigned char __sync_fetch_and_and(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_fetch_and_and_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_fetch_and_and(volatile signed short *, signed short,...);
+signed short __sync_fetch_and_and_2(volatile signed short *, signed short,...);
 unsigned short __sync_fetch_and_and(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_fetch_and_and_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_fetch_and_and(volatile signed int *, signed int,...);
+signed int __sync_fetch_and_and_4(volatile signed int *, signed int,...);
 unsigned int __sync_fetch_and_and(volatile unsigned int *, unsigned int,...);
+signed long int __sync_fetch_and_and(volatile signed long int *, signed long int,...);
+unsigned long int __sync_fetch_and_and(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_fetch_and_and_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_fetch_and_and(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_fetch_and_and_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_fetch_and_and(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_fetch_and_and_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_fetch_and_and(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_fetch_and_and_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_fetch_and_and(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_fetch_and_and_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_fetch_and_xor(volatile char *, char,...);
+char __sync_fetch_and_xor_1(volatile char *, char,...);
 signed char __sync_fetch_and_xor(volatile signed char *, signed char,...);
+signed char __sync_fetch_and_xor_1(volatile signed char *, signed char,...);
 unsigned char __sync_fetch_and_xor(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_fetch_and_xor_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_fetch_and_xor(volatile signed short *, signed short,...);
+signed short __sync_fetch_and_xor_2(volatile signed short *, signed short,...);
 unsigned short __sync_fetch_and_xor(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_fetch_and_xor_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_fetch_and_xor(volatile signed int *, signed int,...);
+signed int __sync_fetch_and_xor_4(volatile signed int *, signed int,...);
 unsigned int __sync_fetch_and_xor(volatile unsigned int *, unsigned int,...);
+signed long int __sync_fetch_and_xor(volatile signed long int *, signed long int,...);
+unsigned long int __sync_fetch_and_xor(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_fetch_and_xor_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_fetch_and_xor(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_fetch_and_xor_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_fetch_and_xor(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_fetch_and_xor_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_fetch_and_xor(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_fetch_and_xor_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_fetch_and_xor(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_fetch_and_xor_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_fetch_and_nand(volatile char *, char,...);
+char __sync_fetch_and_nand_1(volatile char *, char,...);
 signed char __sync_fetch_and_nand(volatile signed char *, signed char,...);
+signed char __sync_fetch_and_nand_1(volatile signed char *, signed char,...);
 unsigned char __sync_fetch_and_nand(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_fetch_and_nand_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_fetch_and_nand(volatile signed short *, signed short,...);
+signed short __sync_fetch_and_nand_2(volatile signed short *, signed short,...);
 unsigned short __sync_fetch_and_nand(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_fetch_and_nand_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_fetch_and_nand(volatile signed int *, signed int,...);
+signed int __sync_fetch_and_nand_4(volatile signed int *, signed int,...);
 unsigned int __sync_fetch_and_nand(volatile unsigned int *, unsigned int,...);
+signed long int __sync_fetch_and_nand(volatile signed long int *, signed long int,...);
+unsigned long int __sync_fetch_and_nand(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_fetch_and_nand_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_fetch_and_nand(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_fetch_and_nand_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_fetch_and_nand(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_fetch_and_nand_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_fetch_and_nand(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_fetch_and_nand_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_fetch_and_nand(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_fetch_and_nand_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_add_and_fetch(volatile char *, char,...);
+char __sync_add_and_fetch_1(volatile char *, char,...);
 signed char __sync_add_and_fetch(volatile signed char *, signed char,...);
+signed char __sync_add_and_fetch_1(volatile signed char *, signed char,...);
 unsigned char __sync_add_and_fetch(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_add_and_fetch_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_add_and_fetch(volatile signed short *, signed short,...);
+signed short __sync_add_and_fetch_2(volatile signed short *, signed short,...);
 unsigned short __sync_add_and_fetch(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_add_and_fetch_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_add_and_fetch(volatile signed int *, signed int,...);
+signed int __sync_add_and_fetch_4(volatile signed int *, signed int,...);
 signed int __sync_add_and_fetch(volatile signed int *, signed int,...);
+signed long int __sync_add_and_fetch(volatile signed long int *, signed long int,...);
+unsigned long int __sync_add_and_fetch(volatile unsigned long int *, unsigned long int,...);
+signed int __sync_add_and_fetch_4(volatile signed int *, signed int,...);
 signed long long int __sync_add_and_fetch(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_add_and_fetch_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_add_and_fetch(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_add_and_fetch_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_add_and_fetch(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_add_and_fetch_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_add_and_fetch(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_add_and_fetch_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_sub_and_fetch(volatile char *, char,...);
+char __sync_sub_and_fetch_1(volatile char *, char,...);
 signed char __sync_sub_and_fetch(volatile signed char *, signed char,...);
+signed char __sync_sub_and_fetch_1(volatile signed char *, signed char,...);
 unsigned char __sync_sub_and_fetch(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_sub_and_fetch_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_sub_and_fetch(volatile signed short *, signed short,...);
+signed short __sync_sub_and_fetch_2(volatile signed short *, signed short,...);
 unsigned short __sync_sub_and_fetch(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_sub_and_fetch_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_sub_and_fetch(volatile signed int *, signed int,...);
+signed int __sync_sub_and_fetch_4(volatile signed int *, signed int,...);
 unsigned int __sync_sub_and_fetch(volatile unsigned int *, unsigned int,...);
+signed long int __sync_sub_and_fetch(volatile signed long int *, signed long int,...);
+unsigned long int __sync_sub_and_fetch(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_sub_and_fetch_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_sub_and_fetch(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_sub_and_fetch_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_sub_and_fetch(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_sub_and_fetch_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_sub_and_fetch(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_sub_and_fetch_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_sub_and_fetch(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_sub_and_fetch_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_or_and_fetch(volatile char *, char,...);
+char __sync_or_and_fetch_1(volatile char *, char,...);
 signed char __sync_or_and_fetch(volatile signed char *, signed char,...);
+signed char __sync_or_and_fetch_1(volatile signed char *, signed char,...);
 unsigned char __sync_or_and_fetch(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_or_and_fetch_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_or_and_fetch(volatile signed short *, signed short,...);
+signed short __sync_or_and_fetch_2(volatile signed short *, signed short,...);
 unsigned short __sync_or_and_fetch(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_or_and_fetch_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_or_and_fetch(volatile signed int *, signed int,...);
+signed int __sync_or_and_fetch_4(volatile signed int *, signed int,...);
 unsigned int __sync_or_and_fetch(volatile unsigned int *, unsigned int,...);
+signed long int __sync_or_and_fetch(volatile signed long int *, signed long int,...);
+unsigned long int __sync_or_and_fetch(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_or_and_fetch_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_or_and_fetch(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_or_and_fetch_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_or_and_fetch(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_or_and_fetch_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_or_and_fetch(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_or_and_fetch_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_or_and_fetch(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_or_and_fetch_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_and_and_fetch(volatile char *, char,...);
+char __sync_and_and_fetch_1(volatile char *, char,...);
 signed char __sync_and_and_fetch(volatile signed char *, signed char,...);
+signed char __sync_and_and_fetch_1(volatile signed char *, signed char,...);
 unsigned char __sync_and_and_fetch(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_and_and_fetch_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_and_and_fetch(volatile signed short *, signed short,...);
+signed short __sync_and_and_fetch_2(volatile signed short *, signed short,...);
 unsigned short __sync_and_and_fetch(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_and_and_fetch_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_and_and_fetch(volatile signed int *, signed int,...);
+signed int __sync_and_and_fetch_4(volatile signed int *, signed int,...);
 unsigned int __sync_and_and_fetch(volatile unsigned int *, unsigned int,...);
+signed long int __sync_and_and_fetch(volatile signed long int *, signed long int,...);
+unsigned long int __sync_and_and_fetch(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_and_and_fetch_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_and_and_fetch(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_and_and_fetch_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_and_and_fetch(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_and_and_fetch_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_and_and_fetch(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_and_and_fetch_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_and_and_fetch(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_and_and_fetch_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_xor_and_fetch(volatile char *, char,...);
+char __sync_xor_and_fetch_1(volatile char *, char,...);
 signed char __sync_xor_and_fetch(volatile signed char *, signed char,...);
+signed char __sync_xor_and_fetch_1(volatile signed char *, signed char,...);
 unsigned char __sync_xor_and_fetch(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_xor_and_fetch_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_xor_and_fetch(volatile signed short *, signed short,...);
+signed short __sync_xor_and_fetch_2(volatile signed short *, signed short,...);
 unsigned short __sync_xor_and_fetch(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_xor_and_fetch_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_xor_and_fetch(volatile signed int *, signed int,...);
+signed int __sync_xor_and_fetch_4(volatile signed int *, signed int,...);
 unsigned int __sync_xor_and_fetch(volatile unsigned int *, unsigned int,...);
+signed long int __sync_xor_and_fetch(volatile signed long int *, signed long int,...);
+unsigned long int __sync_xor_and_fetch(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_xor_and_fetch_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_xor_and_fetch(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_xor_and_fetch_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_xor_and_fetch(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_xor_and_fetch_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_xor_and_fetch(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_xor_and_fetch_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_xor_and_fetch(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_xor_and_fetch_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 char __sync_nand_and_fetch(volatile char *, char,...);
+char __sync_nand_and_fetch_1(volatile char *, char,...);
 signed char __sync_nand_and_fetch(volatile signed char *, signed char,...);
+signed char __sync_nand_and_fetch_1(volatile signed char *, signed char,...);
 unsigned char __sync_nand_and_fetch(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_nand_and_fetch_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_nand_and_fetch(volatile signed short *, signed short,...);
+signed short __sync_nand_and_fetch_2(volatile signed short *, signed short,...);
 unsigned short __sync_nand_and_fetch(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_nand_and_fetch_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_nand_and_fetch(volatile signed int *, signed int,...);
+signed int __sync_nand_and_fetch_4(volatile signed int *, signed int,...);
 unsigned int __sync_nand_and_fetch(volatile unsigned int *, unsigned int,...);
+signed long int __sync_nand_and_fetch(volatile signed long int *, signed long int,...);
+unsigned long int __sync_nand_and_fetch(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_nand_and_fetch_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_nand_and_fetch(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_nand_and_fetch_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_nand_and_fetch(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_nand_and_fetch_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_nand_and_fetch(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_nand_and_fetch_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_nand_and_fetch(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_nand_and_fetch_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 _Bool __sync_bool_compare_and_swap(volatile char *, char, char,...);
+_Bool __sync_bool_compare_and_swap_1(volatile char *, char, char,...);
 _Bool __sync_bool_compare_and_swap(volatile signed char *, signed char, signed char,...);
+_Bool __sync_bool_compare_and_swap_1(volatile signed char *, signed char, signed char,...);
 _Bool __sync_bool_compare_and_swap(volatile unsigned char *, unsigned char, unsigned char,...);
+_Bool __sync_bool_compare_and_swap_1(volatile unsigned char *, unsigned char, unsigned char,...);
 _Bool __sync_bool_compare_and_swap(volatile short *, signed short, signed short,...);
+_Bool __sync_bool_compare_and_swap_2(volatile short *, signed short, signed short,...);
 _Bool __sync_bool_compare_and_swap(volatile short *, unsigned short, unsigned short,...);
+_Bool __sync_bool_compare_and_swap_2(volatile short *, unsigned short, unsigned short,...);
 _Bool __sync_bool_compare_and_swap(volatile signed int *, signed int, signed int,...);
+_Bool __sync_bool_compare_and_swap_4(volatile signed int *, signed int, signed int,...);
 _Bool __sync_bool_compare_and_swap(volatile unsigned int *, unsigned int, unsigned int,...);
+_Bool __sync_bool_compare_and_swap(volatile signed long int *, signed long int, signed long int,...);
+_Bool __sync_bool_compare_and_swap(volatile unsigned long int *, unsigned long int, unsigned long int,...);
+_Bool __sync_bool_compare_and_swap_4(volatile unsigned int *, unsigned int, unsigned int,...);
 _Bool __sync_bool_compare_and_swap(volatile signed long long int *, signed long long int, signed long long int,...);
+_Bool __sync_bool_compare_and_swap_8(volatile signed long long int *, signed long long int, signed long long int,...);
 _Bool __sync_bool_compare_and_swap(volatile unsigned long long int *, unsigned long long int, unsigned long long int,...);
+_Bool __sync_bool_compare_and_swap_8(volatile unsigned long long int *, unsigned long long int, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 _Bool __sync_bool_compare_and_swap(volatile signed __int128 *, signed __int128, signed __int128,...);
+_Bool __sync_bool_compare_and_swap_16(volatile signed __int128 *, signed __int128, signed __int128,...);
 _Bool __sync_bool_compare_and_swap(volatile unsigned __int128 *, unsigned __int128, unsigned __int128,...);
+_Bool __sync_bool_compare_and_swap_16(volatile unsigned __int128 *, unsigned __int128, unsigned __int128,...);
 #endif
 forall(dtype T) _Bool __sync_bool_compare_and_swap(T * volatile *, T *, T*, ...);
 char __sync_val_compare_and_swap(volatile char *, char, char,...);
+char __sync_val_compare_and_swap_1(volatile char *, char, char,...);
 signed char __sync_val_compare_and_swap(volatile signed char *, signed char, signed char,...);
+signed char __sync_val_compare_and_swap_1(volatile signed char *, signed char, signed char,...);
 unsigned char __sync_val_compare_and_swap(volatile unsigned char *, unsigned char, unsigned char,...);
+unsigned char __sync_val_compare_and_swap_1(volatile unsigned char *, unsigned char, unsigned char,...);
 signed short __sync_val_compare_and_swap(volatile signed short *, signed short, signed short,...);
+signed short __sync_val_compare_and_swap_2(volatile signed short *, signed short, signed short,...);
 unsigned short __sync_val_compare_and_swap(volatile unsigned short *, unsigned short, unsigned short,...);
+unsigned short __sync_val_compare_and_swap_2(volatile unsigned short *, unsigned short, unsigned short,...);
 signed int __sync_val_compare_and_swap(volatile signed int *, signed int, signed int,...);
+signed int __sync_val_compare_and_swap_4(volatile signed int *, signed int, signed int,...);
 unsigned int __sync_val_compare_and_swap(volatile unsigned int *, unsigned int, unsigned int,...);
+signed long int __sync_val_compare_and_swap(volatile signed long int *, signed long int, signed long int,...);
+unsigned long int __sync_val_compare_and_swap(volatile unsigned long int *, unsigned long int, unsigned long int,...);
+unsigned int __sync_val_compare_and_swap_4(volatile unsigned int *, unsigned int, unsigned int,...);
 signed long long int __sync_val_compare_and_swap(volatile signed long long int *, signed long long int, signed long long int,...);
+signed long long int __sync_val_compare_and_swap_8(volatile signed long long int *, signed long long int, signed long long int,...);
 unsigned long long int __sync_val_compare_and_swap(volatile unsigned long long int *, unsigned long long int, unsigned long long int,...);
+unsigned long long int __sync_val_compare_and_swap_8(volatile unsigned long long int *, unsigned long long int, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_val_compare_and_swap(volatile signed __int128 *, signed __int128, signed __int128,...);
+signed __int128 __sync_val_compare_and_swap_16(volatile signed __int128 *, signed __int128, signed __int128,...);
 unsigned __int128 __sync_val_compare_and_swap(volatile unsigned __int128 *, unsigned __int128, unsigned __int128,...);
+unsigned __int128 __sync_val_compare_and_swap_16(volatile unsigned __int128 *, unsigned __int128, unsigned __int128,...);
 #endif
 forall(dtype T) T * __sync_val_compare_and_swap(T * volatile *, T *, T*,...);
 char __sync_lock_test_and_set(volatile char *, char,...);
+char __sync_lock_test_and_set_1(volatile char *, char,...);
 signed char __sync_lock_test_and_set(volatile signed char *, signed char,...);
+signed char __sync_lock_test_and_set_1(volatile signed char *, signed char,...);
 unsigned char __sync_lock_test_and_set(volatile unsigned char *, unsigned char,...);
+unsigned char __sync_lock_test_and_set_1(volatile unsigned char *, unsigned char,...);
 signed short __sync_lock_test_and_set(volatile signed short *, signed short,...);
+signed short __sync_lock_test_and_set_2(volatile signed short *, signed short,...);
 unsigned short __sync_lock_test_and_set(volatile unsigned short *, unsigned short,...);
+unsigned short __sync_lock_test_and_set_2(volatile unsigned short *, unsigned short,...);
 signed int __sync_lock_test_and_set(volatile signed int *, signed int,...);
+signed int __sync_lock_test_and_set_4(volatile signed int *, signed int,...);
 unsigned int __sync_lock_test_and_set(volatile unsigned int *, unsigned int,...);
+signed long int __sync_lock_test_and_set(volatile signed long int *, signed long int,...);
+unsigned long int __sync_lock_test_and_set(volatile unsigned long int *, unsigned long int,...);
+unsigned int __sync_lock_test_and_set_4(volatile unsigned int *, unsigned int,...);
 signed long long int __sync_lock_test_and_set(volatile signed long long int *, signed long long int,...);
+signed long long int __sync_lock_test_and_set_8(volatile signed long long int *, signed long long int,...);
 unsigned long long int __sync_lock_test_and_set(volatile unsigned long long int *, unsigned long long int,...);
+unsigned long long int __sync_lock_test_and_set_8(volatile unsigned long long int *, unsigned long long int,...);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __sync_lock_test_and_set(volatile signed __int128 *, signed __int128,...);
+signed __int128 __sync_lock_test_and_set_16(volatile signed __int128 *, signed __int128,...);
 unsigned __int128 __sync_lock_test_and_set(volatile unsigned __int128 *, unsigned __int128,...);
+unsigned __int128 __sync_lock_test_and_set_16(volatile unsigned __int128 *, unsigned __int128,...);
 #endif
 void __sync_lock_release(volatile char *,...);
+void __sync_lock_release_1(volatile char *,...);
 void __sync_lock_release(volatile signed char *,...);
+void __sync_lock_release_1(volatile signed char *,...);
 void __sync_lock_release(volatile unsigned char *,...);
+void __sync_lock_release_1(volatile unsigned char *,...);
 void __sync_lock_release(volatile signed short *,...);
+void __sync_lock_release_2(volatile signed short *,...);
 void __sync_lock_release(volatile unsigned short *,...);
+void __sync_lock_release_2(volatile unsigned short *,...);
 void __sync_lock_release(volatile signed int *,...);
+void __sync_lock_release_4(volatile signed int *,...);
 void __sync_lock_release(volatile unsigned int *,...);
+void __sync_lock_release(volatile signed long int *,...);
+void __sync_lock_release(volatile unsigned long int *,...);
+void __sync_lock_release_4(volatile unsigned int *,...);
 void __sync_lock_release(volatile signed long long int *,...);
+void __sync_lock_release_8(volatile signed long long int *,...);
 void __sync_lock_release(volatile unsigned long long int *,...);
+void __sync_lock_release_8(volatile unsigned long long int *,...);
 #if defined(__SIZEOF_INT128__)
 void __sync_lock_release(volatile signed __int128 *,...);
+void __sync_lock_release_16(volatile signed __int128 *,...);
 void __sync_lock_release(volatile unsigned __int128 *,...);
+void __sync_lock_release_16(volatile unsigned __int128 *,...);
 #endif
 …
 _Bool __atomic_test_and_set(volatile signed int *, int);
 _Bool __atomic_test_and_set(volatile unsigned int *, int);
-_Bool __atomic_test_and_set(volatile signed long int *, int);
-_Bool __atomic_test_and_set(volatile unsigned long int *, int);
 _Bool __atomic_test_and_set(volatile signed long long int *, int);
 _Bool __atomic_test_and_set(volatile unsigned long long int *, int);
 …
 void __atomic_clear(volatile signed int *, int);
 void __atomic_clear(volatile unsigned int *, int);
-void __atomic_clear(volatile signed long int *, int);
-void __atomic_clear(volatile unsigned long int *, int);
 void __atomic_clear(volatile signed long long int *, int);
 void __atomic_clear(volatile unsigned long long int *, int);
 …
 #endif
-_Bool __atomic_exchange_n(volatile _Bool *, _Bool, int);
-void __atomic_exchange(volatile _Bool *, volatile _Bool *, volatile _Bool *, int);
 char __atomic_exchange_n(volatile char *, char, int);
+char __atomic_exchange_1(volatile char *, char, int);
 void __atomic_exchange(volatile char *, volatile char *, volatile char *, int);
 signed char __atomic_exchange_n(volatile signed char *, signed char, int);
+signed char __atomic_exchange_1(volatile signed char *, signed char, int);
 void __atomic_exchange(volatile signed char *, volatile signed char *, volatile signed char *, int);
 unsigned char __atomic_exchange_n(volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_exchange_1(volatile unsigned char *, unsigned char, int);
 void __atomic_exchange(volatile unsigned char *, volatile unsigned char *, volatile unsigned char *, int);
 signed short __atomic_exchange_n(volatile signed short *, signed short, int);
+signed short __atomic_exchange_2(volatile signed short *, signed short, int);
 void __atomic_exchange(volatile signed short *, volatile signed short *, volatile signed short *, int);
 unsigned short __atomic_exchange_n(volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_exchange_2(volatile unsigned short *, unsigned short, int);
 void __atomic_exchange(volatile unsigned short *, volatile unsigned short *, volatile unsigned short *, int);
 signed int __atomic_exchange_n(volatile signed int *, signed int, int);
+signed int __atomic_exchange_4(volatile signed int *, signed int, int);
 void __atomic_exchange(volatile signed int *, volatile signed int *, volatile signed int *, int);
 unsigned int __atomic_exchange_n(volatile unsigned int *, unsigned int, int);
+unsigned int __atomic_exchange_4(volatile unsigned int *, unsigned int, int);
 void __atomic_exchange(volatile unsigned int *, volatile unsigned int *, volatile unsigned int *, int);
-signed long int __atomic_exchange_n(volatile signed long int *, signed long int, int);
-void __atomic_exchange(volatile signed long int *, volatile signed long int *, volatile signed long int *, int);
-unsigned long int __atomic_exchange_n(volatile unsigned long int *, unsigned long int, int);
-void __atomic_exchange(volatile unsigned long int *, volatile unsigned long int *, volatile unsigned long int *, int);
 signed long long int __atomic_exchange_n(volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_exchange_8(volatile signed long long int *, signed long long int, int);
 void __atomic_exchange(volatile signed long long int *, volatile signed long long int *, volatile signed long long int *, int);
 unsigned long long int __atomic_exchange_n(volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_exchange_8(volatile unsigned long long int *, unsigned long long int, int);
 void __atomic_exchange(volatile unsigned long long int *, volatile unsigned long long int *, volatile unsigned long long int *, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_exchange_n(volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_exchange_16(volatile signed __int128 *, signed __int128, int);
 void __atomic_exchange(volatile signed __int128 *, volatile signed __int128 *, volatile signed __int128 *, int);
 unsigned __int128 __atomic_exchange_n(volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_exchange_16(volatile unsigned __int128 *, unsigned __int128, int);
 void __atomic_exchange(volatile unsigned __int128 *, volatile unsigned __int128 *, volatile unsigned __int128 *, int);
 #endif
 …
 void __atomic_load(const volatile _Bool *, volatile _Bool *, int);
 char __atomic_load_n(const volatile char *, int);
+char __atomic_load_1(const volatile char *, int);
 void __atomic_load(const volatile char *, volatile char *, int);
 signed char __atomic_load_n(const volatile signed char *, int);
+signed char __atomic_load_1(const volatile signed char *, int);
 void __atomic_load(const volatile signed char *, volatile signed char *, int);
 unsigned char __atomic_load_n(const volatile unsigned char *, int);
+unsigned char __atomic_load_1(const volatile unsigned char *, int);
 void __atomic_load(const volatile unsigned char *, volatile unsigned char *, int);
 signed short __atomic_load_n(const volatile signed short *, int);
+signed short __atomic_load_2(const volatile signed short *, int);
 void __atomic_load(const volatile signed short *, volatile signed short *, int);
 unsigned short __atomic_load_n(const volatile unsigned short *, int);
+unsigned short __atomic_load_2(const volatile unsigned short *, int);
 void __atomic_load(const volatile unsigned short *, volatile unsigned short *, int);
 signed int __atomic_load_n(const volatile signed int *, int);
+signed int __atomic_load_4(const volatile signed int *, int);
 void __atomic_load(const volatile signed int *, volatile signed int *, int);
 unsigned int __atomic_load_n(const volatile unsigned int *, int);
+unsigned int __atomic_load_4(const volatile unsigned int *, int);
 void __atomic_load(const volatile unsigned int *, volatile unsigned int *, int);
-signed long int __atomic_load_n(const volatile signed long int *, int);
-void __atomic_load(const volatile signed long int *, volatile signed long int *, int);
-unsigned long int __atomic_load_n(const volatile unsigned long int *, int);
-void __atomic_load(const volatile unsigned long int *, volatile unsigned long int *, int);
 signed long long int __atomic_load_n(const volatile signed long long int *, int);
+signed long long int __atomic_load_8(const volatile signed long long int *, int);
 void __atomic_load(const volatile signed long long int *, volatile signed long long int *, int);
 unsigned long long int __atomic_load_n(const volatile unsigned long long int *, int);
+unsigned long long int __atomic_load_8(const volatile unsigned long long int *, int);
 void __atomic_load(const volatile unsigned long long int *, volatile unsigned long long int *, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_load_n(const volatile signed __int128 *, int);
+signed __int128 __atomic_load_16(const volatile signed __int128 *, int);
 void __atomic_load(const volatile signed __int128 *, volatile signed __int128 *, int);
 unsigned __int128 __atomic_load_n(const volatile unsigned __int128 *, int);
+unsigned __int128 __atomic_load_16(const volatile unsigned __int128 *, int);
 void __atomic_load(const volatile unsigned __int128 *, volatile unsigned __int128 *, int);
 #endif
 …
 _Bool __atomic_compare_exchange_n(volatile char *, char *, char, _Bool, int, int);
+_Bool __atomic_compare_exchange_1(volatile char *, char *, char, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile char *, char *, char *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile signed char *, signed char *, signed char, _Bool, int, int);
+_Bool __atomic_compare_exchange_1(volatile signed char *, signed char *, signed char, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile signed char *, signed char *, signed char *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile unsigned char *, unsigned char *, unsigned char, _Bool, int, int);
+_Bool __atomic_compare_exchange_1(volatile unsigned char *, unsigned char *, unsigned char, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile unsigned char *, unsigned char *, unsigned char *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile signed short *, signed short *, signed short, _Bool, int, int);
+_Bool __atomic_compare_exchange_2(volatile signed short *, signed short *, signed short, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile signed short *, signed short *, signed short *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile unsigned short *, unsigned short *, unsigned short, _Bool, int, int);
+_Bool __atomic_compare_exchange_2(volatile unsigned short *, unsigned short *, unsigned short, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile unsigned short *, unsigned short *, unsigned short *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile signed int *, signed int *, signed int, _Bool, int, int);
+_Bool __atomic_compare_exchange_4(volatile signed int *, signed int *, signed int, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile signed int *, signed int *, signed int *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile unsigned int *, unsigned int *, unsigned int, _Bool, int, int);
+_Bool __atomic_compare_exchange_4(volatile unsigned int *, unsigned int *, unsigned int, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile unsigned int *, unsigned int *, unsigned int *, _Bool, int, int);
-_Bool __atomic_compare_exchange_n(volatile signed long int *, signed long int *, signed long int, _Bool, int, int);
-_Bool __atomic_compare_exchange  (volatile signed long int *, signed long int *, signed long int *, _Bool, int, int);
-_Bool __atomic_compare_exchange_n(volatile unsigned long int *, unsigned long int *, unsigned long int, _Bool, int, int);
-_Bool __atomic_compare_exchange  (volatile unsigned long int *, unsigned long int *, unsigned long int *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile signed long long int *, signed long long int *, signed long long int, _Bool, int, int);
+_Bool __atomic_compare_exchange_8(volatile signed long long int *, signed long long int *, signed long long int, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile signed long long int *, signed long long int *, signed long long int *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n(volatile unsigned long long int *, unsigned long long int *, unsigned long long int, _Bool, int, int);
+_Bool __atomic_compare_exchange_8(volatile unsigned long long int *, unsigned long long int *, unsigned long long int, _Bool, int, int);
 _Bool __atomic_compare_exchange  (volatile unsigned long long int *, unsigned long long int *, unsigned long long int *, _Bool, int, int);
 #if defined(__SIZEOF_INT128__)
 _Bool __atomic_compare_exchange_n (volatile signed __int128 *, signed __int128 *, signed __int128, _Bool, int, int);
+_Bool __atomic_compare_exchange_16(volatile signed __int128 *, signed __int128 *, signed __int128, _Bool, int, int);
 _Bool __atomic_compare_exchange   (volatile signed __int128 *, signed __int128 *, signed __int128 *, _Bool, int, int);
 _Bool __atomic_compare_exchange_n (volatile unsigned __int128 *, unsigned __int128 *, unsigned __int128, _Bool, int, int);
+_Bool __atomic_compare_exchange_16(volatile unsigned __int128 *, unsigned __int128 *, unsigned __int128, _Bool, int, int);
 _Bool __atomic_compare_exchange   (volatile unsigned __int128 *, unsigned __int128 *, unsigned __int128 *, _Bool, int, int);
 #endif
 …
 void __atomic_store(volatile _Bool *, _Bool *, int);
 void __atomic_store_n(volatile char *, char, int);
+void __atomic_store_1(volatile char *, char, int);
 void __atomic_store(volatile char *, char *, int);
 void __atomic_store_n(volatile signed char *, signed char, int);
+void __atomic_store_1(volatile signed char *, signed char, int);
 void __atomic_store(volatile signed char *, signed char *, int);
 void __atomic_store_n(volatile unsigned char *, unsigned char, int);
+void __atomic_store_1(volatile unsigned char *, unsigned char, int);
 void __atomic_store(volatile unsigned char *, unsigned char *, int);
 void __atomic_store_n(volatile signed short *, signed short, int);
+void __atomic_store_2(volatile signed short *, signed short, int);
 void __atomic_store(volatile signed short *, signed short *, int);
 void __atomic_store_n(volatile unsigned short *, unsigned short, int);
+void __atomic_store_2(volatile unsigned short *, unsigned short, int);
 void __atomic_store(volatile unsigned short *, unsigned short *, int);
 void __atomic_store_n(volatile signed int *, signed int, int);
+void __atomic_store_4(volatile signed int *, signed int, int);
 void __atomic_store(volatile signed int *, signed int *, int);
 void __atomic_store_n(volatile unsigned int *, unsigned int, int);
+void __atomic_store_4(volatile unsigned int *, unsigned int, int);
 void __atomic_store(volatile unsigned int *, unsigned int *, int);
-void __atomic_store_n(volatile signed long int *, signed long int, int);
-void __atomic_store(volatile signed long int *, signed long int *, int);
-void __atomic_store_n(volatile unsigned long int *, unsigned long int, int);
-void __atomic_store(volatile unsigned long int *, unsigned long int *, int);
 void __atomic_store_n(volatile signed long long int *, signed long long int, int);
+void __atomic_store_8(volatile signed long long int *, signed long long int, int);
 void __atomic_store(volatile signed long long int *, signed long long int *, int);
 void __atomic_store_n(volatile unsigned long long int *, unsigned long long int, int);
+void __atomic_store_8(volatile unsigned long long int *, unsigned long long int, int);
 void __atomic_store(volatile unsigned long long int *, unsigned long long int *, int);
 #if defined(__SIZEOF_INT128__)
 void __atomic_store_n(volatile signed __int128 *, signed __int128, int);
+void __atomic_store_16(volatile signed __int128 *, signed __int128, int);
 void __atomic_store(volatile signed __int128 *, signed __int128 *, int);
 void __atomic_store_n(volatile unsigned __int128 *, unsigned __int128, int);
+void __atomic_store_16(volatile unsigned __int128 *, unsigned __int128, int);
 void __atomic_store(volatile unsigned __int128 *, unsigned __int128 *, int);
 #endif
 …
 char __atomic_add_fetch  (volatile char *, char, int);
+char __atomic_add_fetch_1(volatile char *, char, int);
 signed char __atomic_add_fetch  (volatile signed char *, signed char, int);
+signed char __atomic_add_fetch_1(volatile signed char *, signed char, int);
 unsigned char __atomic_add_fetch  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_add_fetch_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_add_fetch  (volatile signed short *, signed short, int);
+signed short __atomic_add_fetch_2(volatile signed short *, signed short, int);
 unsigned short __atomic_add_fetch  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_add_fetch_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_add_fetch  (volatile signed int *, signed int, int);
+signed int __atomic_add_fetch_4(volatile signed int *, signed int, int);
 unsigned int __atomic_add_fetch  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_add_fetch  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_add_fetch  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_add_fetch_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_add_fetch  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_add_fetch_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_add_fetch  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_add_fetch_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_add_fetch   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_add_fetch_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_add_fetch   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_add_fetch_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_sub_fetch  (volatile char *, char, int);
+char __atomic_sub_fetch_1(volatile char *, char, int);
 signed char __atomic_sub_fetch  (volatile signed char *, signed char, int);
+signed char __atomic_sub_fetch_1(volatile signed char *, signed char, int);
 unsigned char __atomic_sub_fetch  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_sub_fetch_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_sub_fetch  (volatile signed short *, signed short, int);
+signed short __atomic_sub_fetch_2(volatile signed short *, signed short, int);
 unsigned short __atomic_sub_fetch  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_sub_fetch_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_sub_fetch  (volatile signed int *, signed int, int);
+signed int __atomic_sub_fetch_4(volatile signed int *, signed int, int);
 unsigned int __atomic_sub_fetch  (volatile unsigned int *, unsigned int, int);
+signed long long int __atomic_sub_fetch  (volatile signed long int *, signed long int, int);
+unsigned long long int __atomic_sub_fetch  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_sub_fetch_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_sub_fetch  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_sub_fetch_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_sub_fetch  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_sub_fetch_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_sub_fetch   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_sub_fetch_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_sub_fetch   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_sub_fetch_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_and_fetch  (volatile char *, char, int);
+char __atomic_and_fetch_1(volatile char *, char, int);
 signed char __atomic_and_fetch  (volatile signed char *, signed char, int);
+signed char __atomic_and_fetch_1(volatile signed char *, signed char, int);
 unsigned char __atomic_and_fetch  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_and_fetch_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_and_fetch  (volatile signed short *, signed short, int);
+signed short __atomic_and_fetch_2(volatile signed short *, signed short, int);
 unsigned short __atomic_and_fetch  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_and_fetch_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_and_fetch  (volatile signed int *, signed int, int);
+signed int __atomic_and_fetch_4(volatile signed int *, signed int, int);
 unsigned int __atomic_and_fetch  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_and_fetch  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_and_fetch  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_and_fetch_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_and_fetch  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_and_fetch_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_and_fetch  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_and_fetch_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_and_fetch   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_and_fetch_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_and_fetch   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_and_fetch_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_nand_fetch  (volatile char *, char, int);
+char __atomic_nand_fetch_1(volatile char *, char, int);
 signed char __atomic_nand_fetch  (volatile signed char *, signed char, int);
+signed char __atomic_nand_fetch_1(volatile signed char *, signed char, int);
 unsigned char __atomic_nand_fetch  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_nand_fetch_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_nand_fetch  (volatile signed short *, signed short, int);
+signed short __atomic_nand_fetch_2(volatile signed short *, signed short, int);
 unsigned short __atomic_nand_fetch  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_nand_fetch_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_nand_fetch  (volatile signed int *, signed int, int);
+signed int __atomic_nand_fetch_4(volatile signed int *, signed int, int);
 unsigned int __atomic_nand_fetch  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_nand_fetch  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_nand_fetch  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_nand_fetch_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_nand_fetch  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_nand_fetch_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_nand_fetch  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_nand_fetch_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_nand_fetch   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_nand_fetch_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_nand_fetch   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_nand_fetch_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_xor_fetch  (volatile char *, char, int);
+char __atomic_xor_fetch_1(volatile char *, char, int);
 signed char __atomic_xor_fetch  (volatile signed char *, signed char, int);
+signed char __atomic_xor_fetch_1(volatile signed char *, signed char, int);
 unsigned char __atomic_xor_fetch  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_xor_fetch_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_xor_fetch  (volatile signed short *, signed short, int);
+signed short __atomic_xor_fetch_2(volatile signed short *, signed short, int);
 unsigned short __atomic_xor_fetch  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_xor_fetch_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_xor_fetch  (volatile signed int *, signed int, int);
+signed int __atomic_xor_fetch_4(volatile signed int *, signed int, int);
 unsigned int __atomic_xor_fetch  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_xor_fetch  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_xor_fetch  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_xor_fetch_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_xor_fetch  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_xor_fetch_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_xor_fetch  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_xor_fetch_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_xor_fetch   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_xor_fetch_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_xor_fetch   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_xor_fetch_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_or_fetch  (volatile char *, char, int);
+char __atomic_or_fetch_1(volatile char *, char, int);
 signed char __atomic_or_fetch  (volatile signed char *, signed char, int);
+signed char __atomic_or_fetch_1(volatile signed char *, signed char, int);
 unsigned char __atomic_or_fetch  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_or_fetch_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_or_fetch  (volatile signed short *, signed short, int);
+signed short __atomic_or_fetch_2(volatile signed short *, signed short, int);
 unsigned short __atomic_or_fetch  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_or_fetch_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_or_fetch  (volatile signed int *, signed int, int);
+signed int __atomic_or_fetch_4(volatile signed int *, signed int, int);
 unsigned int __atomic_or_fetch  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_or_fetch  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_or_fetch  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_or_fetch_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_or_fetch  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_or_fetch_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_or_fetch  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_or_fetch_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_or_fetch   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_or_fetch_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_or_fetch   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_or_fetch_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_fetch_add  (volatile char *, char, int);
+char __atomic_fetch_add_1(volatile char *, char, int);
 signed char __atomic_fetch_add  (volatile signed char *, signed char, int);
+signed char __atomic_fetch_add_1(volatile signed char *, signed char, int);
 unsigned char __atomic_fetch_add  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_fetch_add_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_fetch_add  (volatile signed short *, signed short, int);
+signed short __atomic_fetch_add_2(volatile signed short *, signed short, int);
 unsigned short __atomic_fetch_add  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_fetch_add_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_fetch_add  (volatile signed int *, signed int, int);
+signed int __atomic_fetch_add_4(volatile signed int *, signed int, int);
 unsigned int __atomic_fetch_add  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_fetch_add  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_fetch_add  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_fetch_add_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_fetch_add  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_fetch_add_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_fetch_add  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_fetch_add_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_fetch_add   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_fetch_add_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_fetch_add   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_fetch_add_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_fetch_sub  (volatile char *, char, int);
+char __atomic_fetch_sub_1(volatile char *, char, int);
 signed char __atomic_fetch_sub  (volatile signed char *, signed char, int);
+signed char __atomic_fetch_sub_1(volatile signed char *, signed char, int);
 unsigned char __atomic_fetch_sub  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_fetch_sub_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_fetch_sub  (volatile signed short *, signed short, int);
+signed short __atomic_fetch_sub_2(volatile signed short *, signed short, int);
 unsigned short __atomic_fetch_sub  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_fetch_sub_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_fetch_sub  (volatile signed int *, signed int, int);
+signed int __atomic_fetch_sub_4(volatile signed int *, signed int, int);
 unsigned int __atomic_fetch_sub  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_fetch_sub  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_fetch_sub  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_fetch_sub_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_fetch_sub  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_fetch_sub_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_fetch_sub  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_fetch_sub_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_fetch_sub   (volatile signed  __int128 *, signed __int128, int);
+signed __int128 __atomic_fetch_sub_16(volatile signed  __int128 *, signed __int128, int);
 unsigned __int128 __atomic_fetch_sub   (volatile unsigned  __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_fetch_sub_16(volatile unsigned  __int128 *, unsigned __int128, int);
 #endif
 char __atomic_fetch_and  (volatile char *, char, int);
+char __atomic_fetch_and_1(volatile char *, char, int);
 signed char __atomic_fetch_and  (volatile signed char *, signed char, int);
+signed char __atomic_fetch_and_1(volatile signed char *, signed char, int);
 unsigned char __atomic_fetch_and  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_fetch_and_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_fetch_and  (volatile signed short *, signed short, int);
+signed short __atomic_fetch_and_2(volatile signed short *, signed short, int);
 unsigned short __atomic_fetch_and  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_fetch_and_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_fetch_and  (volatile signed int *, signed int, int);
+signed int __atomic_fetch_and_4(volatile signed int *, signed int, int);
 unsigned int __atomic_fetch_and  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_fetch_and  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_fetch_and  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_fetch_and_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_fetch_and  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_fetch_and_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_fetch_and  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_fetch_and_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_fetch_and   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_fetch_and_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_fetch_and   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_fetch_and_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_fetch_nand  (volatile char *, char, int);
+char __atomic_fetch_nand_1(volatile char *, char, int);
 signed char __atomic_fetch_nand  (volatile signed char *, signed char, int);
+signed char __atomic_fetch_nand_1(volatile signed char *, signed char, int);
 unsigned char __atomic_fetch_nand  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_fetch_nand_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_fetch_nand  (volatile signed short *, signed short, int);
+signed short __atomic_fetch_nand_2(volatile signed short *, signed short, int);
 unsigned short __atomic_fetch_nand  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_fetch_nand_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_fetch_nand  (volatile signed int *, signed int, int);
+signed int __atomic_fetch_nand_4(volatile signed int *, signed int, int);
 unsigned int __atomic_fetch_nand  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_fetch_nand  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_fetch_nand  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_fetch_nand_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_fetch_nand  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_fetch_nand_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_fetch_nand  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_fetch_nand_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_fetch_nand   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_fetch_nand_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_fetch_nand   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_fetch_nand_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_fetch_xor  (volatile char *, char, int);
+char __atomic_fetch_xor_1(volatile char *, char, int);
 signed char __atomic_fetch_xor  (volatile signed char *, signed char, int);
+signed char __atomic_fetch_xor_1(volatile signed char *, signed char, int);
 unsigned char __atomic_fetch_xor  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_fetch_xor_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_fetch_xor  (volatile signed short *, signed short, int);
+signed short __atomic_fetch_xor_2(volatile signed short *, signed short, int);
 unsigned short __atomic_fetch_xor  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_fetch_xor_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_fetch_xor  (volatile signed int *, signed int, int);
+signed int __atomic_fetch_xor_4(volatile signed int *, signed int, int);
 unsigned int __atomic_fetch_xor  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_fetch_xor  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_fetch_xor  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_fetch_xor_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_fetch_xor  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_fetch_xor_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_fetch_xor  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_fetch_xor_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_fetch_xor   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_fetch_xor_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_fetch_xor   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_fetch_xor_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif
 char __atomic_fetch_or  (volatile char *, char, int);
+char __atomic_fetch_or_1(volatile char *, char, int);
 signed char __atomic_fetch_or  (volatile signed char *, signed char, int);
+signed char __atomic_fetch_or_1(volatile signed char *, signed char, int);
 unsigned char __atomic_fetch_or  (volatile unsigned char *, unsigned char, int);
+unsigned char __atomic_fetch_or_1(volatile unsigned char *, unsigned char, int);
 signed short __atomic_fetch_or  (volatile signed short *, signed short, int);
+signed short __atomic_fetch_or_2(volatile signed short *, signed short, int);
 unsigned short __atomic_fetch_or  (volatile unsigned short *, unsigned short, int);
+unsigned short __atomic_fetch_or_2(volatile unsigned short *, unsigned short, int);
 signed int __atomic_fetch_or  (volatile signed int *, signed int, int);
+signed int __atomic_fetch_or_4(volatile signed int *, signed int, int);
 unsigned int __atomic_fetch_or  (volatile unsigned int *, unsigned int, int);
+signed long int __atomic_fetch_or  (volatile signed long int *, signed long int, int);
+unsigned long int __atomic_fetch_or  (volatile unsigned long int *, unsigned long int, int);
+unsigned int __atomic_fetch_or_4(volatile unsigned int *, unsigned int, int);
 signed long long int __atomic_fetch_or  (volatile signed long long int *, signed long long int, int);
+signed long long int __atomic_fetch_or_8(volatile signed long long int *, signed long long int, int);
 unsigned long long int __atomic_fetch_or  (volatile unsigned long long int *, unsigned long long int, int);
+unsigned long long int __atomic_fetch_or_8(volatile unsigned long long int *, unsigned long long int, int);
 #if defined(__SIZEOF_INT128__)
 signed __int128 __atomic_fetch_or   (volatile signed __int128 *, signed __int128, int);
+signed __int128 __atomic_fetch_or_16(volatile signed __int128 *, signed __int128, int);
 unsigned __int128 __atomic_fetch_or   (volatile unsigned __int128 *, unsigned __int128, int);
+unsigned __int128 __atomic_fetch_or_16(volatile unsigned __int128 *, unsigned __int128, int);
 #endif

libcfa/src/Makefile.am

-              r7030dab
+              r71d6bd8
 ## Created On       : Sun May 31 08:54:01 2015
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Mon Mar 16 18:07:59 2020
 ## Update Count     : 242
+## Last Modified On : Mon Jul 15 22:43:27 2019
+## Update Count     : 241
 ###############################################################################
 …
 # The built sources must not depend on the installed headers
 AM_CFAFLAGS = -quiet -cfalib -I$(srcdir)/stdhdr $(if $(findstring ${gdbwaittarget}, ${@}), -XCFA --gdb) @CONFIG_CFAFLAGS@
 AM_CFLAGS = -g -Wall -Wno-unused-function -fPIC -pthread @ARCH_FLAGS@ @CONFIG_CFLAGS@
+AM_CFLAGS = -g -Wall -Wno-unused-function -fPIC @ARCH_FLAGS@ @CONFIG_CFLAGS@
 AM_CCASFLAGS = -g -Wall -Wno-unused-function @ARCH_FLAGS@ @CONFIG_CFLAGS@
 CFACC = @CFACC@
 …
 #----------------------------------------------------------------------------------------------------------------
 if BUILDLIB
 headers_nosrc = bitmanip.hfa math.hfa gmp.hfa time_t.hfa bits/align.hfa bits/containers.hfa bits/defs.hfa bits/debug.hfa bits/locks.hfa
+headers_nosrc = math.hfa gmp.hfa time_t.hfa bits/align.hfa bits/containers.hfa bits/defs.hfa bits/debug.hfa bits/locks.hfa
 headers = fstream.hfa iostream.hfa iterator.hfa limits.hfa rational.hfa time.hfa stdlib.hfa common.hfa \
           containers/maybe.hfa containers/pair.hfa containers/result.hfa containers/vector.hfa

libcfa/src/Makefile.in

-              r7030dab
+              r71d6bd8
         limits.hfa rational.hfa time.hfa stdlib.hfa common.hfa \
         containers/maybe.hfa containers/pair.hfa containers/result.hfa \
         containers/vector.hfa bitmanip.hfa math.hfa gmp.hfa time_t.hfa \
+        containers/vector.hfa math.hfa gmp.hfa time_t.hfa \
         bits/align.hfa bits/containers.hfa bits/defs.hfa \
         bits/debug.hfa bits/locks.hfa concurrency/coroutine.hfa \
 …
 LTCFACOMPILE = $(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) \
         $(LIBTOOLFLAGS) --mode=compile $(CFACC) $(DEFS) \
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(AM_CFLAGS) $(CFAFLAGS) $(CFLAGS)
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) \
+        $(AM_CFLAGS) $(CFLAGS)
 AM_V_CFA = $(am__v_CFA_@AM_V@)
 …
 am__v_CFA_0 = @echo "  CFA     " $@;
 am__v_CFA_1 =
+AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
+am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
+am__v_JAVAC_0 = @echo "  JAVAC   " $@;
+am__v_JAVAC_1 =
+AM_V_GOC = $(am__v_GOC_@AM_V@)
+am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
+am__v_GOC_0 = @echo "  GOC     " $@;
+am__v_GOC_1 =
 UPPCC = u++
 UPPCOMPILE = $(UPPCC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_UPPFLAGS) $(UPPFLAGS) $(AM_CXXFLAGS) $(CXXFLAGS) $(AM_CFLAGS) $(CFLAGS)
 …
 am__v_UPP_0 = @echo "  UPP     " $@;
 am__v_UPP_1 =
-AM_V_GOC = $(am__v_GOC_@AM_V@)
-am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
-am__v_GOC_0 = @echo "  GOC     " $@;
-am__v_GOC_1 =
-AM_V_PY = $(am__v_PY_@AM_V@)
-am__v_PY_ = $(am__v_PY_@AM_DEFAULT_V@)
-am__v_PY_0 = @echo "  PYTHON  " $@;
-am__v_PY_1 =
-AM_V_RUST = $(am__v_RUST_@AM_V@)
-am__v_RUST_ = $(am__v_RUST_@AM_DEFAULT_V@)
-am__v_RUST_0 = @echo "  RUST    " $@;
-am__v_RUST_1 =
-AM_V_NODEJS = $(am__v_NODEJS_@AM_V@)
-am__v_NODEJS_ = $(am__v_NODEJS_@AM_DEFAULT_V@)
-am__v_NODEJS_0 = @echo "  NODEJS  " $@;
-am__v_NODEJS_1 =
-AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
-am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
-am__v_JAVAC_0 = @echo "  JAVAC   " $@;
-am__v_JAVAC_1 =
 lib_LTLIBRARIES = libcfa.la libcfathread.la
 gdbwaittarget = ""
 …
 # The built sources must not depend on the installed headers
 AM_CFAFLAGS = -quiet -cfalib -I$(srcdir)/stdhdr $(if $(findstring ${gdbwaittarget}, ${@}), -XCFA --gdb) @CONFIG_CFAFLAGS@
 AM_CFLAGS = -g -Wall -Wno-unused-function -fPIC -pthread @ARCH_FLAGS@ @CONFIG_CFLAGS@
+AM_CFLAGS = -g -Wall -Wno-unused-function -fPIC @ARCH_FLAGS@ @CONFIG_CFLAGS@
 AM_CCASFLAGS = -g -Wall -Wno-unused-function @ARCH_FLAGS@ @CONFIG_CFLAGS@
 @BUILDLIB_FALSE@headers_nosrc =
 #----------------------------------------------------------------------------------------------------------------
 @BUILDLIB_TRUE@headers_nosrc = bitmanip.hfa math.hfa gmp.hfa time_t.hfa bits/align.hfa bits/containers.hfa bits/defs.hfa bits/debug.hfa bits/locks.hfa
+@BUILDLIB_TRUE@headers_nosrc = math.hfa gmp.hfa time_t.hfa bits/align.hfa bits/containers.hfa bits/defs.hfa bits/debug.hfa bits/locks.hfa
 @BUILDLIB_FALSE@headers =
 @BUILDLIB_TRUE@headers = fstream.hfa iostream.hfa iterator.hfa limits.hfa rational.hfa time.hfa stdlib.hfa common.hfa \

libcfa/src/assert.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Nov 28 12:27:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 13:00:18 2020
 // Update Count     : 6
+// Last Modified On : Thu Jul 20 15:10:26 2017
+// Update Count     : 2
 //
 …
 #include <stdarg.h>                                                             // varargs
 #include <stdio.h>                                                              // fprintf
-#include <unistd.h>                                                             // STDERR_FILENO
 #include "bits/debug.hfa"
 …
         // called by macro assert in assert.h
         void __assert_fail( const char assertion[], const char file[], unsigned int line, const char function[] ) {
                 __cfaabi_bits_print_safe( STDERR_FILENO, CFA_ASSERT_FMT ".\n", assertion, __progname, function, line, file );
+        void __assert_fail( const char *assertion, const char *file, unsigned int line, const char *function ) {
+                __cfaabi_dbg_bits_print_safe( CFA_ASSERT_FMT ".\n", assertion, __progname, function, line, file );
                 abort();
+        }
         // called by macro assertf
         void __assert_fail_f( const char assertion[], const char file[], unsigned int line, const char function[], const char fmt[], ... ) {
                 __cfaabi_bits_acquire();
                 __cfaabi_bits_print_nolock( STDERR_FILENO, CFA_ASSERT_FMT ": ", assertion, __progname, function, line, file );
+        void __assert_fail_f( const char *assertion, const char *file, unsigned int line, const char *function, const char *fmt, ... ) {
+                __cfaabi_dbg_bits_acquire();
+                __cfaabi_dbg_bits_print_nolock( CFA_ASSERT_FMT ": ", assertion, __progname, function, line, file );
                 va_list args;
                 va_start( args, fmt );
                 __cfaabi_bits_print_vararg( STDERR_FILENO, fmt, args );
+                __cfaabi_dbg_bits_print_vararg( fmt, args );
                 va_end( args );
                 __cfaabi_bits_print_nolock( STDERR_FILENO, "\n" );
                 __cfaabi_bits_release();
+                __cfaabi_dbg_bits_print_nolock( "\n" );
+                __cfaabi_dbg_bits_release();
                 abort();
+        }

libcfa/src/bits/align.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Nov 28 12:27:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Nov 16 18:58:22 2019
 // Update Count     : 3
+// Last Modified On : Fri Jul 21 23:05:35 2017
+// Update Count     : 2
 //
 // This  library is free  software; you  can redistribute  it and/or  modify it
 …
 // Minimum size used to align memory boundaries for memory allocations.
+//#define libAlign() (sizeof(double))
+// gcc-7 uses xmms instructions, which require 16 byte alignment.
+#define libAlign() (16)
+#define libAlign() (sizeof(double))
 // Check for power of 2

libcfa/src/bits/containers.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Oct 31 16:38:50 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Jan 15 07:42:35 2020
 // Update Count     : 28
+// Last Modified On : Wed Jun 26 08:52:20 2019
+// Update Count     : 4
 #pragma once
 …
         forall(dtype T | sized(T))
         static inline T & ?[?]( __small_array(T) & this, __lock_size_t idx ) {
+        static inline T& ?[?]( __small_array(T) & this, __lock_size_t idx) {
                 return ((typeof(this.data))this.data)[idx];
+        }
         forall(dtype T | sized(T))
         static inline T & ?[?]( const __small_array(T) & this, __lock_size_t idx ) {
+        static inline T& ?[?]( const __small_array(T) & this, __lock_size_t idx) {
                 return ((typeof(this.data))this.data)[idx];
+        }
         forall(dtype T)
         static inline T * begin( const __small_array(T) & this ) {
+        forall(dtype T | sized(T))
+        static inline T* begin( const __small_array(T) & this ) {
                 return ((typeof(this.data))this.data);
+        }
         forall(dtype T | sized(T))
         static inline T * end( const __small_array(T) & this ) {
+        static inline T* end( const __small_array(T) & this ) {
                 return ((typeof(this.data))this.data) + this.size;
+        }
 …
 #ifdef __cforall
         trait is_node(dtype T) {
                 T *& get_next( T & );
+                T*& get_next( T& );
         };
 #endif
 …
         forall(dtype T)
         static inline void ?{}( __stack(T) & this ) {
+                (this.top){ 0p };
+        }
+        static inline forall( dtype T | is_node(T) ) {
+                void push( __stack(T) & this, T * val ) {
+                        verify( !get_next( *val ) );
+                        get_next( *val ) = this.top;
+                        this.top = val;
+                }
+                T * pop( __stack(T) & this ) {
+                        T * top = this.top;
+                        if( top ) {
+                                this.top = get_next( *top );
+                                get_next( *top ) = 0p;
+                        }
+                        return top;
+                }
+                int ?!=?( const __stack(T) & this, __attribute__((unused)) zero_t zero ) {
+                        return this.top != 0;
+                }
+                (this.top){ NULL };
+        }
+        forall(dtype T | is_node(T) | sized(T))
+        static inline void push( __stack(T) & this, T * val ) {
+                verify( !get_next( *val ) );
+                get_next( *val ) = this.top;
+                this.top = val;
+        }
+        forall(dtype T | is_node(T) | sized(T))
+        static inline T * pop( __stack(T) & this ) {
+                T * top = this.top;
+                if( top ) {
+                        this.top = get_next( *top );
+                        get_next( *top ) = NULL;
+                }
+                return top;
+        }
+        forall(dtype T | is_node(T))
+        static inline int ?!=?( const __stack(T) & this, __attribute__((unused)) zero_t zero ) {
+                return this.top != 0;
+        }
 #endif
 …
 #ifdef __cforall
+        static inline forall( dtype T | is_node(T) ) {
+                void ?{}( __queue(T) & this ) with( this ) {
+                        head{ 1p };
+                        tail{ &head };
+                        verify(*tail == 1p);
+                }
+                void append( __queue(T) & this, T * val ) with( this ) {
+                        verify(tail != 0p);
+                        verify(*tail == 1p);
+                        *tail = val;
+                        tail = &get_next( *val );
+                        *tail = 1p;
+                }
+                T * pop_head( __queue(T) & this ) {
+                        verify(*this.tail == 1p);
+                        T * head = this.head;
+                        if( head != 1p ) {
+                                this.head = get_next( *head );
+                                if( get_next( *head ) == 1p ) {
+                                        this.tail = &this.head;
+                                }
+                                get_next( *head ) = 0p;
+                                verify(*this.tail == 1p);
+                                verify( get_next(*head) == 0p );
+                                return head;
+        forall(dtype T)
+        static inline void ?{}( __queue(T) & this ) with( this ) {
+                head{ NULL };
+                tail{ &head };
+        }
+        forall(dtype T | is_node(T) | sized(T))
+        static inline void append( __queue(T) & this, T * val ) with( this ) {
+                verify(tail != NULL);
+                *tail = val;
+                tail = &get_next( *val );
+        }
+        forall(dtype T | is_node(T) | sized(T))
+        static inline T * pop_head( __queue(T) & this ) {
+                T * head = this.head;
+                if( head ) {
+                        this.head = get_next( *head );
+                        if( !get_next( *head ) ) {
+                                this.tail = &this.head;
+                        }
+                        verify(*this.tail == 1p);
+                        return 0p;
+                }
+                T * remove( __queue(T) & this, T ** it ) with( this ) {
+                        T * val = *it;
+                        verify( val );
+                        (*it) = get_next( *val );
+                        if( tail == &get_next( *val ) ) {
+                                tail = it;
+                        }
+                        get_next( *val ) = 0p;
+                        verify( (head == 1p) == (&head == tail) );
+                        verify( *tail == 1p );
+                        return val;
+                }
+                int ?!=?( const __queue(T) & this, __attribute__((unused)) zero_t zero ) {
+                        return this.head != 0;
+                }
+                        get_next( *head ) = NULL;
+                }
+                return head;
+        }
+        forall(dtype T | is_node(T) | sized(T))
+        static inline T * remove( __queue(T) & this, T ** it ) with( this ) {
+                T * val = *it;
+                verify( val );
+                (*it) = get_next( *val );
+                if( tail == &get_next( *val ) ) {
+                        tail = it;
+                }
+                get_next( *val ) = NULL;
+                verify( (head == NULL) == (&head == tail) );
+                verify( *tail == NULL );
+                return val;
+        }
+        forall(dtype T | is_node(T))
+        static inline int ?!=?( const __queue(T) & this, __attribute__((unused)) zero_t zero ) {
+                return this.head != 0;
+        }
 #endif
 …
 #ifdef __cforall
+        forall(dtype T )
+        forall(dtype T | sized(T))
         static inline [void] ?{}( __dllist(T) & this, * [T * & next, T * & prev] ( T & ) __get ) {
                 this.head{ 0p };
+                this.head{ NULL };
                 this.__get = __get;
+        }
 …
         #define next 0
         #define prev 1
+        static inline forall(dtype T) {
+                void push_front( __dllist(T) & this, T & node ) with( this ) {
+                        verify(__get);
+                        if ( head ) {
+                                __get( node ).next = head;
+                                __get( node ).prev = __get( *head ).prev;
+                                // inserted node must be consistent before it is seen
+                                // prevent code movement across barrier
+                                asm( "" : : : "memory" );
+                                __get( *head ).prev = &node;
+                                T & _prev = *__get( node ).prev;
+                                __get( _prev ).next = &node;
+                        } else {
+                                __get( node ).next = &node;
+                                __get( node ).prev = &node;
+                        }
+        forall(dtype T | sized(T))
+        static inline void push_front( __dllist(T) & this, T & node ) with( this ) {
+                verify(__get);
+                if ( head ) {
+                        __get( node ).next = head;
+                        __get( node ).prev = __get( *head ).prev;
+                        // inserted node must be consistent before it is seen
                         // prevent code movement across barrier
                         asm( "" : : : "memory" );
+                        head = &node;
+                }
+                void remove( __dllist(T) & this, T & node ) with( this ) {
+                        verify(__get);
+                        if ( &node == head ) {
+                                if ( __get( *head ).next == head ) {
+                                        head = 0p;
+                                } else {
+                                        head = __get( *head ).next;
+                                }
+                        __get( *head ).prev = &node;
+                        T & _prev = *__get( node ).prev;
+                        __get( _prev ).next = &node;
+                }
+                else {
+                        __get( node ).next = &node;
+                        __get( node ).prev = &node;
+                }
+                // prevent code movement across barrier
+                asm( "" : : : "memory" );
+                head = &node;
+        }
+        forall(dtype T | sized(T))
+        static inline void remove( __dllist(T) & this, T & node ) with( this ) {
+                verify(__get);
+                if ( &node == head ) {
+                        if ( __get( *head ).next == head ) {
+                                head = NULL;
+                        }
+                        __get( *__get( node ).next ).prev = __get( node ).prev;
+                        __get( *__get( node ).prev ).next = __get( node ).next;
+                        __get( node ).next = 0p;
+                        __get( node ).prev = 0p;
+                }
+                int ?!=?( const __dllist(T) & this, __attribute__((unused)) zero_t zero ) {
+                        return this.head != 0;
+                }
+                void move_to_front( __dllist(T) & src, __dllist(T) & dst, T & node ) {
+                        remove    (src, node);
+                        push_front(dst, node);
+                }
+                        else {
+                                head = __get( *head ).next;
+                        }
+                }
+                __get( *__get( node ).next ).prev = __get( node ).prev;
+                __get( *__get( node ).prev ).next = __get( node ).next;
+                __get( node ).next = NULL;
+                __get( node ).prev = NULL;
+        }
+        forall(dtype T | sized(T))
+        static inline int ?!=?( const __dllist(T) & this, __attribute__((unused)) zero_t zero ) {
+                return this.head != 0;
+        }
         #undef next
 …
 #endif
-// Local Variables: //
-// tab-width: 4 //
-// End: //

libcfa/src/bits/debug.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Mar 30 12:30:01 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 13:03:16 2020
 // Update Count     : 11
+// Last Modified On : Sun Jul 14 22:17:35 2019
+// Update Count     : 4
 //
 …
 extern "C" {
+        void __cfaabi_bits_write( int fd, const char in_buffer[], int len ) {
+        void __cfaabi_dbg_bits_write( const char *in_buffer, int len ) {
                 // ensure all data is written
                 for ( int count = 0, retcode; count < len; count += retcode ) {
 …
                         for ( ;; ) {
                                 retcode = write( fd, in_buffer, len - count );
+                                retcode = write( STDERR_FILENO, in_buffer, len - count );
                                 // not a timer interrupt ?
 …
+        }
         void __cfaabi_bits_acquire() __attribute__((__weak__)) {}
         void __cfaabi_bits_release() __attribute__((__weak__)) {}
+        void __cfaabi_dbg_bits_acquire() __attribute__((__weak__)) {}
+        void __cfaabi_dbg_bits_release() __attribute__((__weak__)) {}
         void __cfaabi_bits_print_safe  ( int fd, const char fmt[], ... ) __attribute__(( format(printf, 2, 3) )) {
+        void __cfaabi_dbg_bits_print_safe  ( const char fmt[], ... ) __attribute__(( format(printf, 1, 2) )) {
                 va_list args;
                 va_start( args, fmt );
                 __cfaabi_bits_acquire();
+                __cfaabi_dbg_bits_acquire();
                 int len = vsnprintf( buffer, buffer_size, fmt, args );
                 __cfaabi_bits_write( fd, buffer, len );
+                __cfaabi_dbg_bits_write( buffer, len );
                 __cfaabi_bits_release();
+                __cfaabi_dbg_bits_release();
                 va_end( args );
+        }
         void __cfaabi_bits_print_nolock( int fd, const char fmt[], ... ) __attribute__(( format(printf, 2, 3) )) {
+        void __cfaabi_dbg_bits_print_nolock( const char fmt[], ... ) __attribute__(( format(printf, 1, 2) )) {
                 va_list args;
 …
                 int len = vsnprintf( buffer, buffer_size, fmt, args );
                 __cfaabi_bits_write( fd, buffer, len );
+                __cfaabi_dbg_bits_write( buffer, len );
                 va_end( args );
+        }
         void __cfaabi_bits_print_vararg( int fd, const char fmt[], va_list args ) {
+        void __cfaabi_dbg_bits_print_vararg( const char fmt[], va_list args ) {
                 int len = vsnprintf( buffer, buffer_size, fmt, args );
                 __cfaabi_bits_write( fd, buffer, len );
+                __cfaabi_dbg_bits_write( buffer, len );
+        }
         void __cfaabi_bits_print_buffer( int fd, char in_buffer[], int in_buffer_size, const char fmt[], ... ) __attribute__(( format(printf, 4, 5) )) {
+        void __cfaabi_dbg_bits_print_buffer( char in_buffer[], int in_buffer_size, const char fmt[], ... ) __attribute__(( format(printf, 3, 4) )) {
                 va_list args;
 …
                 int len = vsnprintf( in_buffer, in_buffer_size, fmt, args );
                 __cfaabi_bits_write( fd, in_buffer, len );
+                __cfaabi_dbg_bits_write( in_buffer, len );
                 va_end( args );

libcfa/src/bits/debug.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Nov 28 12:27:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 12:29:21 2020
 // Update Count     : 9
+// Last Modified On : Thu Feb  8 12:35:19 2018
+// Update Count     : 2
 //
 …
         #define __cfaabi_dbg_ctx __PRETTY_FUNCTION__
         #define __cfaabi_dbg_ctx2 , __PRETTY_FUNCTION__
+        #define __cfaabi_dbg_ctx_param const char caller[]
+        #define __cfaabi_dbg_ctx_param2 , const char caller[]
+        #define __cfaabi_dbg_ctx_fwd caller
+        #define __cfaabi_dbg_ctx_fwd2 , caller
+        #define __cfaabi_dbg_ctx_param const char * caller
+        #define __cfaabi_dbg_ctx_param2 , const char * caller
 #else
         #define __cfaabi_dbg_debug_do(...)
 …
         #define __cfaabi_dbg_ctx_param
         #define __cfaabi_dbg_ctx_param2
-        #define __cfaabi_dbg_ctx_fwd
-        #define __cfaabi_dbg_ctx_fwd2
 #endif
 …
         #include <stdio.h>
         extern void __cfaabi_bits_write( int fd, const char buffer[], int len );
         extern void __cfaabi_bits_acquire();
         extern void __cfaabi_bits_release();
         extern void __cfaabi_bits_print_safe  ( int fd, const char fmt[], ... ) __attribute__(( format(printf, 2, 3) ));
         extern void __cfaabi_bits_print_nolock( int fd, const char fmt[], ... ) __attribute__(( format(printf, 2, 3) ));
         extern void __cfaabi_bits_print_vararg( int fd, const char fmt[], va_list arg );
         extern void __cfaabi_bits_print_buffer( int fd, char buffer[], int buffer_size, const char fmt[], ... ) __attribute__(( format(printf, 4, 5) ));
+      extern void __cfaabi_dbg_bits_write( const char *buffer, int len );
+      extern void __cfaabi_dbg_bits_acquire();
+      extern void __cfaabi_dbg_bits_release();
+      extern void __cfaabi_dbg_bits_print_safe  ( const char fmt[], ... ) __attribute__(( format(printf, 1, 2) ));
+      extern void __cfaabi_dbg_bits_print_nolock( const char fmt[], ... ) __attribute__(( format(printf, 1, 2) ));
+      extern void __cfaabi_dbg_bits_print_vararg( const char fmt[], va_list arg );
+      extern void __cfaabi_dbg_bits_print_buffer( char buffer[], int buffer_size, const char fmt[], ... ) __attribute__(( format(printf, 3, 4) ));
 #ifdef __cforall
+}
 …
 #ifdef __CFA_DEBUG_PRINT__
         #define __cfaabi_dbg_write( buffer, len )         __cfaabi_bits_write( STDERR_FILENO, buffer, len )
         #define __cfaabi_dbg_acquire()                    __cfaabi_bits_acquire()
         #define __cfaabi_dbg_release()                    __cfaabi_bits_release()
         #define __cfaabi_dbg_print_safe(...)              __cfaabi_bits_print_safe   (__VA_ARGS__)
         #define __cfaabi_dbg_print_nolock(...)            __cfaabi_bits_print_nolock (__VA_ARGS__)
         #define __cfaabi_dbg_print_buffer(...)            __cfaabi_bits_print_buffer (__VA_ARGS__)
         #define __cfaabi_dbg_print_buffer_decl(...)       char __dbg_text[256]; int __dbg_len = snprintf( __dbg_text, 256, __VA_ARGS__ ); __cfaabi_bits_write( __dbg_text, __dbg_len );
         #define __cfaabi_dbg_print_buffer_local(...)      __dbg_len = snprintf( __dbg_text, 256, __VA_ARGS__ ); __cfaabi_dbg_write( __dbg_text, __dbg_len );
+        #define __cfaabi_dbg_write( buffer, len )         __cfaabi_dbg_bits_write( buffer, len )
+        #define __cfaabi_dbg_acquire()                    __cfaabi_dbg_bits_acquire()
+        #define __cfaabi_dbg_release()                    __cfaabi_dbg_bits_release()
+        #define __cfaabi_dbg_print_safe(...)              __cfaabi_dbg_bits_print_safe   (__VA_ARGS__)
+        #define __cfaabi_dbg_print_nolock(...)            __cfaabi_dbg_bits_print_nolock (__VA_ARGS__)
+        #define __cfaabi_dbg_print_buffer(...)            __cfaabi_dbg_bits_print_buffer (__VA_ARGS__)
+        #define __cfaabi_dbg_print_buffer_decl(...)       char __dbg_text[256]; int __dbg_len = snprintf( __dbg_text, 256, __VA_ARGS__ ); __cfaabi_dbg_bits_write( __dbg_text, __dbg_len );
+        #define __cfaabi_dbg_print_buffer_local(...)      __dbg_len = snprintf( __dbg_text, 256, __VA_ARGS__ ); __cfaabi_dbg_bits_write( __dbg_text, __dbg_len );
 #else
         #define __cfaabi_dbg_write(...)               ((void)0)

libcfa/src/bits/defs.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Nov  9 13:24:10 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Jan 28 22:38:27 2020
 // Update Count     : 9
+// Last Modified On : Thu Feb  8 16:22:41 2018
+// Update Count     : 8
 //
 …
 #ifdef __cforall
+void abort( const char fmt[], ... ) __attribute__ (( format(printf, 1, 2), __nothrow__, __leaf__, __noreturn__ ));
+void abort( bool signalAbort, const char fmt[], ... ) __attribute__ (( format(printf, 2, 3), __nothrow__, __leaf__, __noreturn__ ));
+void abort ( const char fmt[], ... ) __attribute__ (( format(printf, 1, 2), __nothrow__, __leaf__, __noreturn__ ));
 extern "C" {
 #endif

libcfa/src/bits/locks.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Oct 31 15:14:38 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 13:03:19 2020
 // Update Count     : 11
+// Last Modified On : Sat Aug 11 15:42:24 2018
+// Update Count     : 10
 //
 …
                 #ifdef __CFA_DEBUG__
                         void __cfaabi_dbg_record_lock(__spinlock_t & this, const char prev_name[]);
+                        void __cfaabi_dbg_record(__spinlock_t & this, const char * prev_name);
                 #else
                         #define __cfaabi_dbg_record_lock(x, y)
+                        #define __cfaabi_dbg_record(x, y)
                 #endif
+        }
+        extern void yield( unsigned int );
         static inline void ?{}( __spinlock_t & this ) {
 …
         // Lock the spinlock, return false if already acquired
         static inline bool try_lock  ( __spinlock_t & this __cfaabi_dbg_ctx_param2 ) {
-                disable_interrupts();
                 bool result = (this.lock == 0) && (__atomic_test_and_set( &this.lock, __ATOMIC_ACQUIRE ) == 0);
                 if( result ) {
+                        __cfaabi_dbg_record_lock( this, caller );
+                } else {
+                        enable_interrupts_noPoll();
+                        disable_interrupts();
+                        __cfaabi_dbg_record( this, caller );
+                }
                 return result;
 …
                 #endif
-                disable_interrupts();
                 for ( unsigned int i = 1;; i += 1 ) {
                         if ( (this.lock == 0) && (__atomic_test_and_set( &this.lock, __ATOMIC_ACQUIRE ) == 0) ) break;
 …
                         #endif
+                }
+                __cfaabi_dbg_record_lock( this, caller );
+                disable_interrupts();
+                __cfaabi_dbg_record( this, caller );
+        }
         static inline void unlock( __spinlock_t & this ) {
+                enable_interrupts_noPoll();
                 __atomic_clear( &this.lock, __ATOMIC_RELEASE );
-                enable_interrupts_noPoll();
+        }
 …
+        }
+        static inline bool post(__bin_sem_t & this) with( this ) {
+        static inline void post(__bin_sem_t & this) with( this ) {
+                verify(__cfaabi_dbg_in_kernel());
                 pthread_mutex_lock(&lock);
                         bool needs_signal = !signaled;
 …
                 pthread_mutex_unlock(&lock);
+                if (needs_signal) pthread_cond_signal(&cond);
+                return needs_signal;
+                if (needs_signal)
+                        pthread_cond_signal(&cond);
+        }
 #endif

libcfa/src/bits/signal.hfa

-              r7030dab
+              r71d6bd8
         act.sa_sigaction = (void (*)(int, siginfo_t *, void *))handler;
-        sigemptyset( &act.sa_mask );
-        sigaddset( &act.sa_mask, SIGALRM );             // disabled during signal handler
-        sigaddset( &act.sa_mask, SIGUSR1 );
-        sigaddset( &act.sa_mask, SIGSEGV );
-        sigaddset( &act.sa_mask, SIGBUS );
-        sigaddset( &act.sa_mask, SIGILL );
-        sigaddset( &act.sa_mask, SIGFPE );
-        sigaddset( &act.sa_mask, SIGHUP );              // revert to default on second delivery
-        sigaddset( &act.sa_mask, SIGTERM );
-        sigaddset( &act.sa_mask, SIGINT );
         act.sa_flags = flags;
         if ( sigaction( sig, &act, 0p ) == -1 ) {
+        if ( sigaction( sig, &act, NULL ) == -1 ) {
                 __cfaabi_dbg_print_buffer_decl(
                         " __cfaabi_sigaction( sig:%d, handler:%p, flags:%d ), problem installing signal handler, error(%d) %s.\n",
 …
                 );
                 _exit( EXIT_FAILURE );
         } // if
+        }
+}
+// Sigaction wrapper : restore default handler
+static void __cfaabi_sigdefault( int sig ) {
+        struct sigaction act;
+        act.sa_handler = SIG_DFL;
+        act.sa_flags = 0;
+        sigemptyset( &act.sa_mask );
+        if ( sigaction( sig, &act, NULL ) == -1 ) {
+                __cfaabi_dbg_print_buffer_decl(
+                        " __cfaabi_sigdefault( sig:%d ), problem reseting signal handler, error(%d) %s.\n",
+                        sig, errno, strerror( errno )
+                );
+                _exit( EXIT_FAILURE );
+        }
+}

libcfa/src/clock.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Apr 12 14:36:06 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Jan  6 12:49:58 2020
 // Update Count     : 9
+// Last Modified On : Thu Jun 13 21:21:13 2019
+// Update Count     : 8
 //
 #include <time.hfa>
 //######################### C time #########################
 …
 static inline tm * localtime_r( time_t tp, tm * result ) { return localtime_r( &tp, result ); }
 //######################### Clock #########################
 struct Clock {                                                                                  // private
         Duration offset;                                                                        // for virtual clock: contains offset from real-time
+        int clocktype;                                                                          // implementation only -1 (virtual), CLOCK_REALTIME
 };
 static inline {
+        void resetClock( Clock & clk ) with( clk ) {
+                clocktype = CLOCK_REALTIME_COARSE;
+        } // Clock::resetClock
         void resetClock( Clock & clk, Duration adj ) with( clk ) {
+                clocktype = -1;
                 offset = adj + __timezone`s;                                    // timezone (global) is (UTC - local time) in seconds
         } // resetClock
+        void ?{}( Clock & clk ) { resetClock( clk ); }
         void ?{}( Clock & clk, Duration adj ) { resetClock( clk, adj ); }
 …
                 return ret;
         } // getTime
-        Time getCPUTime() {
-                timespec ts;
-                clock_gettime( CLOCK_THREAD_CPUTIME_ID, &ts );
-                return (Time){ ts };
-    } // getCPUTime
 } // distribution

libcfa/src/concurrency/CtxSwitch-arm.S

-              r7030dab
+              r71d6bd8
         .text
         .align  2
         .global __cfactx_switch
         .type   __cfactx_switch, %function
+        .global CtxSwitch
+        .type   CtxSwitch, %function
 __cfactx_switch:
+CtxSwitch:
         @ save callee-saved registers: r4-r8, r10, r11, r13(sp) (plus r9 depending on platform specification)
         @ I've seen reference to 31 registers on 64-bit, if this is the case, more need to be saved
 …
         mov r15, r14
         #endif // R9_SPECIAL
         .text
         .align  2
         .global __cfactx_invoke_stub
         .type   __cfactx_invoke_stub, %function
+        .global CtxInvokeStub
+        .type   CtxInvokeStub, %function
 __cfactx_invoke_stub:
+CtxInvokeStub:
         ldmfd r13!, {r0-r1}
         mov r15, r1

libcfa/src/concurrency/CtxSwitch-i386.S

-              r7030dab
+              r71d6bd8
         .text
         .align 2
         .globl __cfactx_switch
         .type  __cfactx_switch, @function
 __cfactx_switch:
+        .globl CtxSwitch
+        .type  CtxSwitch, @function
+CtxSwitch:
         // Copy the "from" context argument from the stack to register eax
 …
         ret
         .size  __cfactx_switch, .-__cfactx_switch
+        .size  CtxSwitch, .-CtxSwitch
 // Local Variables: //

libcfa/src/concurrency/CtxSwitch-x86_64.S

-              r7030dab
+              r71d6bd8
         .text
         .align 2
         .globl __cfactx_switch
         .type  __cfactx_switch, @function
 __cfactx_switch:
+        .globl CtxSwitch
+        .type  CtxSwitch, @function
+CtxSwitch:
         // Save volatile registers on the stack.
 …
         ret
         .size  __cfactx_switch, .-__cfactx_switch
+        .size  CtxSwitch, .-CtxSwitch
 //-----------------------------------------------------------------------------
 …
         .text
         .align 2
         .globl __cfactx_invoke_stub
         .type    __cfactx_invoke_stub, @function
 __cfactx_invoke_stub:
+        .globl CtxInvokeStub
+        .type    CtxInvokeStub, @function
+CtxInvokeStub:
         movq %rbx, %rdi
+        movq %r12, %rsi
+        jmp *%r13
+        .size  __cfactx_invoke_stub, .-__cfactx_invoke_stub
+        jmp *%r12
+        .size  CtxInvokeStub, .-CtxInvokeStub
 // Local Variables: //

libcfa/src/concurrency/alarm.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Fri Jun 2 11:31:25 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Jan  5 08:41:36 2020
 // Update Count     : 69
+// Last Modified On : Fri May 25 06:25:47 2018
+// Update Count     : 67
 //
 …
 void __kernel_set_timer( Duration alarm ) {
         verifyf(alarm >= 1`us || alarm == 0, "Setting timer to < 1us (%jins)", alarm`ns);
         setitimer( ITIMER_REAL, &(itimerval){ alarm }, 0p );
+        verifyf(alarm >= 1`us || alarm == 0, "Setting timer to < 1us (%jins)", alarm.tv);
+        setitimer( ITIMER_REAL, &(itimerval){ alarm }, NULL );
+}
 …
 //=============================================================================================
 void ?{}( alarm_node_t & this, $thread * thrd, Time alarm, Duration period ) with( this ) {
+void ?{}( alarm_node_t & this, thread_desc * thrd, Time alarm, Duration period ) with( this ) {
         this.thrd = thrd;
         this.alarm = alarm;
 …
                         this->tail = &this->head;
+                }
                 head->next = 0p;
+                head->next = NULL;
+        }
         verify( validate( this ) );
 …
                 this->tail = it;
+        }
         n->next = 0p;
+        n->next = NULL;
         verify( validate( this ) );

libcfa/src/concurrency/alarm.hfa

-              r7030dab
+              r71d6bd8
 #include "time.hfa"
 struct $thread;
+struct thread_desc;
 struct processor;
 …
         union {
                 $thread * thrd; // thrd who created event
+                thread_desc * thrd;     // thrd who created event
                 processor * proc;               // proc who created event
         };
 …
 typedef alarm_node_t ** __alarm_it_t;
 void ?{}( alarm_node_t & this, $thread * thrd, Time alarm, Duration period );
+void ?{}( alarm_node_t & this, thread_desc * thrd, Time alarm, Duration period );
 void ?{}( alarm_node_t & this, processor   * proc, Time alarm, Duration period );
 void ^?{}( alarm_node_t & this );

libcfa/src/concurrency/coroutine.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Nov 28 12:27:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 12:29:25 2020
 // Update Count     : 16
+// Last Modified On : Fri Mar 30 17:20:57 2018
+// Update Count     : 9
 //
 …
 extern "C" {
         void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct $coroutine *) __attribute__ ((__noreturn__));
+        void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct coroutine_desc *) __attribute__ ((__noreturn__));
         static void _CtxCoroutine_UnwindCleanup(_Unwind_Reason_Code, struct _Unwind_Exception *) __attribute__ ((__noreturn__));
         static void _CtxCoroutine_UnwindCleanup(_Unwind_Reason_Code, struct _Unwind_Exception *) {
 …
+}
 void ?{}( $coroutine & this, const char name[], void * storage, size_t storageSize ) with( this ) {
         (this.context){0p, 0p};
+void ?{}( coroutine_desc & this, const char * name, void * storage, size_t storageSize ) with( this ) {
+        (this.context){NULL, NULL};
         (this.stack){storage, storageSize};
         this.name = name;
         state = Start;
         starter = 0p;
         last = 0p;
         cancellation = 0p;
+}
 void ^?{}($coroutine& this) {
+        starter = NULL;
+        last = NULL;
+        cancellation = NULL;
+}
+void ^?{}(coroutine_desc& this) {
         if(this.state != Halted && this.state != Start && this.state != Primed) {
                 $coroutine * src = TL_GET( this_thread )->curr_cor;
                 $coroutine * dst = &this;
+                coroutine_desc * src = TL_GET( this_thread )->curr_cor;
+                coroutine_desc * dst = &this;
                 struct _Unwind_Exception storage;
 …
+                }
                 $ctx_switch( src, dst );
+                CoroutineCtxSwitch( src, dst );
+        }
+}
 …
 forall(dtype T | is_coroutine(T))
 void prime(T& cor) {
         $coroutine* this = get_coroutine(cor);
+        coroutine_desc* this = get_coroutine(cor);
         assert(this->state == Start);
 …
 [void *, size_t] __stack_alloc( size_t storageSize ) {
         const size_t stack_data_size = libCeiling( sizeof(__stack_t), 16 ); // minimum alignment
+        static const size_t stack_data_size = libCeiling( sizeof(__stack_t), 16 ); // minimum alignment
         assert(__page_size != 0l);
         size_t size = libCeiling( storageSize, 16 ) + stack_data_size;
 …
 void __stack_prepare( __stack_info_t * this, size_t create_size ) {
         const size_t stack_data_size = libCeiling( sizeof(__stack_t), 16 ); // minimum alignment
+        static const size_t stack_data_size = libCeiling( sizeof(__stack_t), 16 ); // minimum alignment
         bool userStack;
         void * storage;
 …
 // is not inline (We can't inline Cforall in C)
 extern "C" {
+        void __cfactx_cor_leave( struct $coroutine * src ) {
+                $coroutine * starter = src->cancellation != 0 ? src->last : src->starter;
+        void __suspend_internal(void) {
+                suspend();
+        }
+        void __leave_coroutine( coroutine_desc * src ) {
+                coroutine_desc * starter = src->cancellation != 0 ? src->last : src->starter;
                 src->state = Halted;
 …
                         src->name, src, starter->name, starter );
+                $ctx_switch( src, starter );
+        }
+        struct $coroutine * __cfactx_cor_finish(void) {
+                struct $coroutine * cor = kernelTLS.this_thread->curr_cor;
+                if(cor->state == Primed) {
+                        __cfactx_suspend();
+                }
+                cor->state = Active;
+                return cor;
+                CoroutineCtxSwitch( src, starter );
+        }
+}

libcfa/src/concurrency/coroutine.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Nov 28 12:27:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 12:29:26 2020
 // Update Count     : 11
+// Last Modified On : Fri Jun 21 17:49:39 2019
+// Update Count     : 9
 //
 …
 trait is_coroutine(dtype T) {
       void main(T & this);
       $coroutine * get_coroutine(T & this);
+      coroutine_desc * get_coroutine(T & this);
 };
 #define DECL_COROUTINE(X) static inline $coroutine* get_coroutine(X& this) { return &this.__cor; } void main(X& this)
+#define DECL_COROUTINE(X) static inline coroutine_desc* get_coroutine(X& this) { return &this.__cor; } void main(X& this)
 //-----------------------------------------------------------------------------
 …
 // void ^?{}( coStack_t & this );
 void  ?{}( $coroutine & this, const char name[], void * storage, size_t storageSize );
 void ^?{}( $coroutine & this );
+void ?{}( coroutine_desc & this, const char * name, void * storage, size_t storageSize );
+void ^?{}( coroutine_desc & this );
 static inline void ?{}( $coroutine & this)                                       { this{ "Anonymous Coroutine", 0p, 0 }; }
 static inline void ?{}( $coroutine & this, size_t stackSize)                     { this{ "Anonymous Coroutine", 0p, stackSize }; }
 static inline void ?{}( $coroutine & this, void * storage, size_t storageSize )  { this{ "Anonymous Coroutine", storage, storageSize }; }
 static inline void ?{}( $coroutine & this, const char name[])                    { this{ name, 0p, 0 }; }
 static inline void ?{}( $coroutine & this, const char name[], size_t stackSize ) { this{ name, 0p, stackSize }; }
+static inline void ?{}( coroutine_desc & this)                                       { this{ "Anonymous Coroutine", NULL, 0 }; }
+static inline void ?{}( coroutine_desc & this, size_t stackSize)                     { this{ "Anonymous Coroutine", NULL, stackSize }; }
+static inline void ?{}( coroutine_desc & this, void * storage, size_t storageSize )  { this{ "Anonymous Coroutine", storage, storageSize }; }
+static inline void ?{}( coroutine_desc & this, const char * name)                    { this{ name, NULL, 0 }; }
+static inline void ?{}( coroutine_desc & this, const char * name, size_t stackSize ) { this{ name, NULL, stackSize }; }
 //-----------------------------------------------------------------------------
 // Public coroutine API
+static inline void suspend(void);
+forall(dtype T | is_coroutine(T))
+static inline T & resume(T & cor);
 forall(dtype T | is_coroutine(T))
 void prime(T & cor);
 static inline struct $coroutine * active_coroutine() { return TL_GET( this_thread )->curr_cor; }
+static inline struct coroutine_desc * active_coroutine() { return TL_GET( this_thread )->curr_cor; }
 //-----------------------------------------------------------------------------
 …
 // Start coroutine routines
 extern "C" {
+        void __cfactx_invoke_coroutine(void (*main)(void *), void * this);
+      forall(dtype T | is_coroutine(T))
+      void CtxInvokeCoroutine(T * this);
         forall(dtype T)
         void __cfactx_start(void (*main)(T &), struct $coroutine * cor, T & this, void (*invoke)(void (*main)(void *), void *));
+      forall(dtype T | is_coroutine(T))
+      void CtxStart(T * this, void ( *invoke)(T *));
         extern void __cfactx_coroutine_unwind(struct _Unwind_Exception * storage, struct $coroutine *) __attribute__ ((__noreturn__));
+        extern void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct coroutine_desc *) __attribute__ ((__noreturn__));
         extern void __cfactx_switch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("__cfactx_switch");
+        extern void CtxSwitch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("CtxSwitch");
+}
 // Private wrappers for context switch and stack creation
 // Wrapper for co
 static inline void $ctx_switch( $coroutine * src, $coroutine * dst ) __attribute__((nonnull (1, 2))) {
+static inline void CoroutineCtxSwitch(coroutine_desc* src, coroutine_desc* dst) {
         // set state of current coroutine to inactive
         src->state = src->state == Halted ? Halted : Blocked;
+        src->state = src->state == Halted ? Halted : Inactive;
         // set new coroutine that task is executing
 …
         // context switch to specified coroutine
         verify( dst->context.SP );
         __cfactx_switch( &src->context, &dst->context );
         // when __cfactx_switch returns we are back in the src coroutine
+        CtxSwitch( &src->context, &dst->context );
+        // when CtxSwitch returns we are back in the src coroutine
         // set state of new coroutine to active
         src->state = Active;
         if( unlikely(src->cancellation != 0p) ) {
                 __cfactx_coroutine_unwind(src->cancellation, src);
+        if( unlikely(src->cancellation != NULL) ) {
+                _CtxCoroutine_Unwind(src->cancellation, src);
+        }
+}
 …
 // Suspend implementation inlined for performance
+extern "C" {
+        static inline void __cfactx_suspend(void) {
+                // optimization : read TLS once and reuse it
+                // Safety note: this is preemption safe since if
+                // preemption occurs after this line, the pointer
+                // will also migrate which means this value will
+                // stay in syn with the TLS
+                $coroutine * src = TL_GET( this_thread )->curr_cor;
+static inline void suspend(void) {
+        // optimization : read TLS once and reuse it
+        // Safety note: this is preemption safe since if
+        // preemption occurs after this line, the pointer
+        // will also migrate which means this value will
+        // stay in syn with the TLS
+        coroutine_desc * src = TL_GET( this_thread )->curr_cor;
                 assertf( src->last != 0,
                         "Attempt to suspend coroutine \"%.256s\" (%p) that has never been resumed.\n"
                         "Possible cause is a suspend executed in a member called by a coroutine user rather than by the coroutine main.",
                         src->name, src );
                 assertf( src->last->state != Halted,
                         "Attempt by coroutine \"%.256s\" (%p) to suspend back to terminated coroutine \"%.256s\" (%p).\n"
                         "Possible cause is terminated coroutine's main routine has already returned.",
                         src->name, src, src->last->name, src->last );
+        assertf( src->last != 0,
+                "Attempt to suspend coroutine \"%.256s\" (%p) that has never been resumed.\n"
+                "Possible cause is a suspend executed in a member called by a coroutine user rather than by the coroutine main.",
+                src->name, src );
+        assertf( src->last->state != Halted,
+                "Attempt by coroutine \"%.256s\" (%p) to suspend back to terminated coroutine \"%.256s\" (%p).\n"
+                "Possible cause is terminated coroutine's main routine has already returned.",
+                src->name, src, src->last->name, src->last );
+                $ctx_switch( src, src->last );
+        }
+        CoroutineCtxSwitch( src, src->last );
+}
 …
         // will also migrate which means this value will
         // stay in syn with the TLS
         $coroutine * src = TL_GET( this_thread )->curr_cor;
         $coroutine * dst = get_coroutine(cor);
+        coroutine_desc * src = TL_GET( this_thread )->curr_cor;
+        coroutine_desc * dst = get_coroutine(cor);
+        if( unlikely(dst->context.SP == 0p) ) {
+                TL_GET( this_thread )->curr_cor = dst;
+        if( unlikely(dst->context.SP == NULL) ) {
                 __stack_prepare(&dst->stack, 65000);
+                __cfactx_start(main, dst, cor, __cfactx_invoke_coroutine);
+                TL_GET( this_thread )->curr_cor = src;
+                CtxStart(&cor, CtxInvokeCoroutine);
+        }
 …
         // always done for performance testing
         $ctx_switch( src, dst );
+        CoroutineCtxSwitch( src, dst );
         return cor;
+}
 static inline void resume( $coroutine * dst ) __attribute__((nonnull (1))) {
+static inline void resume(coroutine_desc * dst) {
         // optimization : read TLS once and reuse it
         // Safety note: this is preemption safe since if
 …
         // will also migrate which means this value will
         // stay in syn with the TLS
         $coroutine * src = TL_GET( this_thread )->curr_cor;
+        coroutine_desc * src = TL_GET( this_thread )->curr_cor;
         // not resuming self ?
 …
         // always done for performance testing
         $ctx_switch( src, dst );
+        CoroutineCtxSwitch( src, dst );
+}

libcfa/src/concurrency/invoke.c

-              r7030dab
+              r71d6bd8
 // Called from the kernel when starting a coroutine or task so must switch back to user mode.
 extern struct $coroutine * __cfactx_cor_finish(void);
 extern void __cfactx_cor_leave ( struct $coroutine * );
 extern void __cfactx_thrd_leave();
+extern void __suspend_internal(void);
+extern void __leave_coroutine( struct coroutine_desc * );
+extern void __finish_creation( struct thread_desc * );
+extern void __leave_thread_monitor( struct thread_desc * this );
 extern void disable_interrupts() OPTIONAL_THREAD;
 extern void enable_interrupts( __cfaabi_dbg_ctx_param );
 void __cfactx_invoke_coroutine(
+void CtxInvokeCoroutine(
         void (*main)(void *),
+        struct coroutine_desc *(*get_coroutine)(void *),
         void *this
 ) {
+        // Finish setting up the coroutine by setting its state
+        struct $coroutine * cor = __cfactx_cor_finish();
+        struct coroutine_desc* cor = get_coroutine( this );
+        // Call the main of the coroutine
+        if(cor->state == Primed) {
+                __suspend_internal();
+        }
+        cor->state = Active;
         main( this );
         //Final suspend, should never return
         __cfactx_cor_leave( cor );
+        __leave_coroutine( cor );
         __cabi_abort( "Resumed dead coroutine" );
+}
 static _Unwind_Reason_Code __cfactx_coroutine_unwindstop(
+static _Unwind_Reason_Code _CtxCoroutine_UnwindStop(
         __attribute((__unused__)) int version,
         _Unwind_Action actions,
 …
                 // We finished unwinding the coroutine,
                 // leave it
                 __cfactx_cor_leave( param );
+                __leave_coroutine( param );
                 __cabi_abort( "Resumed dead coroutine" );
+        }
 …
+}
 void __cfactx_coroutine_unwind(struct _Unwind_Exception * storage, struct $coroutine * cor) __attribute__ ((__noreturn__));
 void __cfactx_coroutine_unwind(struct _Unwind_Exception * storage, struct $coroutine * cor) {
         _Unwind_Reason_Code ret = _Unwind_ForcedUnwind( storage, __cfactx_coroutine_unwindstop, cor );
+void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct coroutine_desc * cor) __attribute__ ((__noreturn__));
+void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct coroutine_desc * cor) {
+        _Unwind_Reason_Code ret = _Unwind_ForcedUnwind( storage, _CtxCoroutine_UnwindStop, cor );
         printf("UNWIND ERROR %d after force unwind\n", ret);
         abort();
+}
+void __cfactx_invoke_thread(
+void CtxInvokeThread(
+        void (*dtor)(void *),
         void (*main)(void *),
+        struct thread_desc *(*get_thread)(void *),
         void *this
 ) {
+        // Fetch the thread handle from the user defined thread structure
+        struct thread_desc* thrd = get_thread( this );
+        // First suspend, once the thread arrives here,
+        // the function pointer to main can be invalidated without risk
+        __finish_creation( thrd );
         // Officially start the thread by enabling preemption
         enable_interrupts( __cfaabi_dbg_ctx );
 …
         // The order of these 4 operations is very important
         //Final suspend, should never return
         __cfactx_thrd_leave();
+        __leave_thread_monitor( thrd );
         __cabi_abort( "Resumed dead thread" );
+}
+void __cfactx_start(
+void CtxStart(
         void (*main)(void *),
         struct $coroutine * cor,
+        struct coroutine_desc *(*get_coroutine)(void *),
         void *this,
         void (*invoke)(void *)
 ) {
+        struct coroutine_desc * cor = get_coroutine( this );
         struct __stack_t * stack = cor->stack.storage;
 …
         fs->dummyReturn = NULL;
+        fs->argument[0] = main;     // argument to invoke
+        fs->argument[1] = this;     // argument to invoke
+        fs->argument[0] = this;     // argument to invoke
         fs->rturn = invoke;
 …
         fs->dummyReturn = NULL;
+        fs->rturn = __cfactx_invoke_stub;
+        fs->fixedRegisters[0] = main;
+        fs->fixedRegisters[1] = this;
+        fs->fixedRegisters[2] = invoke;
+        fs->rturn = CtxInvokeStub;
+        fs->fixedRegisters[0] = this;
+        fs->fixedRegisters[1] = invoke;
 #elif defined( __ARM_ARCH )
+#error ARM needs to be upgrade to use to parameters like X86/X64 (A.K.A. : I broke this and do not know how to fix it)
         struct FakeStack {
                 float fpRegs[16];                       // floating point registers
 …
         struct FakeStack *fs = (struct FakeStack *)cor->context.SP;
         fs->intRegs[8] = __cfactx_invoke_stub;
+        fs->intRegs[8] = CtxInvokeStub;
         fs->arg[0] = this;
         fs->arg[1] = invoke;

libcfa/src/concurrency/invoke.h

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jan 17 12:27:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Dec  5 16:26:03 2019
 // Update Count     : 44
+// Last Modified On : Sat Jun 22 18:19:13 2019
+// Update Count     : 40
 //
 …
         extern "Cforall" {
                 extern __attribute__((aligned(128))) thread_local struct KernelThreadData {
                         struct $thread    * volatile this_thread;
+                        struct thread_desc    * volatile this_thread;
                         struct processor      * volatile this_processor;
 …
         };
+        enum coroutine_state { Halted, Start, Primed, Blocked, Ready, Active, Rerun };
+        enum __Preemption_Reason { __NO_PREEMPTION, __ALARM_PREEMPTION, __POLL_PREEMPTION, __MANUAL_PREEMPTION };
+        struct $coroutine {
+                // context that is switch during a __cfactx_switch
+        enum coroutine_state { Halted, Start, Inactive, Active, Primed };
+        struct coroutine_desc {
+                // context that is switch during a CtxSwitch
                 struct __stack_context_t context;
 …
                 // first coroutine to resume this one
                 struct $coroutine * starter;
+                struct coroutine_desc * starter;
                 // last coroutine to resume this one
                 struct $coroutine * last;
+                struct coroutine_desc * last;
                 // If non-null stack must be unwound with this exception
 …
         };
-        static inline struct __stack_t * __get_stack( struct $coroutine * cor ) { return (struct __stack_t*)(((uintptr_t)cor->stack.storage) & ((uintptr_t)-2)); }
         // struct which calls the monitor is accepting
 …
         };
         struct $monitor {
+        struct monitor_desc {
                 // spinlock to protect internal data
                 struct __spinlock_t lock;
                 // current owner of the monitor
                 struct $thread * owner;
+                struct thread_desc * owner;
                 // queue of threads that are blocked waiting for the monitor
                 __queue_t(struct $thread) entry_queue;
+                __queue_t(struct thread_desc) entry_queue;
                 // stack of conditions to run next once we exit the monitor
 …
         struct __monitor_group_t {
                 // currently held monitors
                 __cfa_anonymous_object( __small_array_t($monitor*) );
+                __cfa_anonymous_object( __small_array_t(monitor_desc*) );
                 // last function that acquired monitors
 …
         };
         struct $thread {
+        struct thread_desc {
                 // Core threading fields
                 // context that is switch during a __cfactx_switch
+                // context that is switch during a CtxSwitch
                 struct __stack_context_t context;
                 // current execution status for coroutine
+                volatile int state;
+                enum __Preemption_Reason preempted;
+                enum coroutine_state state;
                 //SKULLDUGGERY errno is not save in the thread data structure because returnToKernel appears to be the only function to require saving and restoring it
                 // coroutine body used to store context
                 struct $coroutine  self_cor;
+                struct coroutine_desc  self_cor;
                 // current active context
                 struct $coroutine * curr_cor;
+                struct coroutine_desc * curr_cor;
                 // monitor body used for mutual exclusion
                 struct $monitor    self_mon;
+                struct monitor_desc    self_mon;
                 // pointer to monitor with sufficient lifetime for current monitors
                 struct $monitor *  self_mon_p;
+                struct monitor_desc *  self_mon_p;
                 // pointer to the cluster on which the thread is running
 …
                 // Link lists fields
                 // instrusive link field for threads
                 struct $thread * next;
+                struct thread_desc * next;
                 struct {
                         struct $thread * next;
                         struct $thread * prev;
+                        struct thread_desc * next;
+                        struct thread_desc * prev;
                 } node;
+                #ifdef __CFA_DEBUG__
+                        // previous function to park/unpark the thread
+                        const char * park_caller;
+                        enum coroutine_state park_result;
+                        bool park_stale;
+                        const char * unpark_caller;
+                        enum coroutine_state unpark_result;
+                        bool unpark_stale;
+                #endif
+        };
+        #ifdef __CFA_DEBUG__
+                void __cfaabi_dbg_record_thrd($thread & this, bool park, const char prev_name[]);
+        #else
+                #define __cfaabi_dbg_record_thrd(x, y, z)
+        #endif
+        };
         #ifdef __cforall
         extern "Cforall" {
                 static inline $thread *& get_next( $thread & this ) __attribute__((const)) {
+                static inline thread_desc *& get_next( thread_desc & this ) {
                         return this.next;
+                }
                 static inline [$thread *&, $thread *& ] __get( $thread & this ) __attribute__((const)) {
+                static inline [thread_desc *&, thread_desc *& ] __get( thread_desc & this ) {
                         return this.node.[next, prev];
+                }
                 static inline void ?{}(__monitor_group_t & this) {
                         (this.data){0p};
+                        (this.data){NULL};
                         (this.size){0};
                         (this.func){NULL};
+                }
                 static inline void ?{}(__monitor_group_t & this, struct $monitor ** data, __lock_size_t size, fptr_t func) {
+                static inline void ?{}(__monitor_group_t & this, struct monitor_desc ** data, __lock_size_t size, fptr_t func) {
                         (this.data){data};
                         (this.size){size};
 …
+                }
                 static inline bool ?==?( const __monitor_group_t & lhs, const __monitor_group_t & rhs ) __attribute__((const)) {
+                static inline bool ?==?( const __monitor_group_t & lhs, const __monitor_group_t & rhs ) {
                         if( (lhs.data != 0) != (rhs.data != 0) ) return false;
                         if( lhs.size != rhs.size ) return false;
 …
         // assembler routines that performs the context switch
         extern void __cfactx_invoke_stub( void );
         extern void __cfactx_switch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("__cfactx_switch");
+        extern void CtxInvokeStub( void );
+        extern void CtxSwitch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("CtxSwitch");
         // void CtxStore ( void * this ) asm ("CtxStore");
         // void CtxRet   ( void * dst  ) asm ("CtxRet");

libcfa/src/concurrency/kernel.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jan 17 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 13:03:15 2020
 // Update Count     : 58
+// Last Modified On : Thu Jun 20 17:21:23 2019
+// Update Count     : 25
 //
 …
 #include <signal.h>
 #include <unistd.h>
-#include <limits.h>                                                                             // PTHREAD_STACK_MIN
-#include <sys/mman.h>                                                                   // mprotect
+}
 …
 //-----------------------------------------------------------------------------
 // Some assembly required
 #if defined( __i386 )
+#if   defined( __i386 )
         #define CtxGet( ctx )        \
                 __asm__ volatile (     \
 …
 //-----------------------------------------------------------------------------
 //Start and stop routine for the kernel, declared first to make sure they run first
+static void __kernel_startup (void) __attribute__(( constructor( STARTUP_PRIORITY_KERNEL ) ));
+static void __kernel_shutdown(void) __attribute__(( destructor ( STARTUP_PRIORITY_KERNEL ) ));
+//-----------------------------------------------------------------------------
+// Kernel Scheduling logic
+static $thread * __next_thread(cluster * this);
+static void __run_thread(processor * this, $thread * dst);
+static $thread * __halt(processor * this);
+static bool __wake_one(cluster * cltr, bool was_empty);
+static bool __wake_proc(processor *);
+static void kernel_startup(void)  __attribute__(( constructor( STARTUP_PRIORITY_KERNEL ) ));
+static void kernel_shutdown(void) __attribute__(( destructor ( STARTUP_PRIORITY_KERNEL ) ));
 //-----------------------------------------------------------------------------
 …
 KERNEL_STORAGE(cluster,         mainCluster);
 KERNEL_STORAGE(processor,       mainProcessor);
 KERNEL_STORAGE($thread, mainThread);
+KERNEL_STORAGE(thread_desc,     mainThread);
 KERNEL_STORAGE(__stack_t,       mainThreadCtx);
 cluster     * mainCluster;
 processor   * mainProcessor;
 $thread * mainThread;
+thread_desc * mainThread;
 extern "C" {
         struct { __dllist_t(cluster) list; __spinlock_t lock; } __cfa_dbg_global_clusters;
+struct { __dllist_t(cluster) list; __spinlock_t lock; } __cfa_dbg_global_clusters;
+}
 …
 // Global state
 thread_local struct KernelThreadData kernelTLS __attribute__ ((tls_model ( "initial-exec" ))) = {
         NULL,                                                                                           // cannot use 0p
+        NULL,
         NULL,
         { 1, false, false },
 …
 // Struct to steal stack
 struct current_stack_info_t {
         __stack_t * storage;                                                            // pointer to stack object
         void * base;                                                                            // base of stack
         void * limit;                                                                           // stack grows towards stack limit
         void * context;                                                                         // address of cfa_context_t
+        __stack_t * storage;            // pointer to stack object
+        void *base;                             // base of stack
+        void *limit;                    // stack grows towards stack limit
+        void *context;                  // address of cfa_context_t
 };
 …
 // Main thread construction
 void ?{}( $coroutine & this, current_stack_info_t * info) with( this ) {
+void ?{}( coroutine_desc & this, current_stack_info_t * info) with( this ) {
         stack.storage = info->storage;
         with(*stack.storage) {
 …
         name = "Main Thread";
         state = Start;
         starter = 0p;
         last = 0p;
         cancellation = 0p;
+}
 void ?{}( $thread & this, current_stack_info_t * info) with( this ) {
+        starter = NULL;
+        last = NULL;
+        cancellation = NULL;
+}
+void ?{}( thread_desc & this, current_stack_info_t * info) with( this ) {
         state = Start;
         self_cor{ info };
 …
         self_mon.recursion = 1;
         self_mon_p = &self_mon;
         next = 0p;
         node.next = 0p;
         node.prev = 0p;
+        next = NULL;
+        node.next = NULL;
+        node.prev = NULL;
         doregister(curr_cluster, this);
 …
+}
+static void * __invoke_processor(void * arg);
+void ?{}(processor & this, const char name[], cluster & cltr) with( this ) {
+static void start(processor * this);
+void ?{}(processor & this, const char * name, cluster & cltr) with( this ) {
         this.name = name;
         this.cltr = &cltr;
         terminated{ 0 };
-        destroyer = 0p;
         do_terminate = false;
         preemption_alarm = 0p;
+        preemption_alarm = NULL;
         pending_preemption = false;
         runner.proc = &this;
+        idle{};
+        __cfaabi_dbg_print_safe("Kernel : Starting core %p\n", &this);
+        this.stack = __create_pthread( &this.kernel_thread, __invoke_processor, (void *)&this );
+        __cfaabi_dbg_print_safe("Kernel : core %p started\n", &this);
+        idleLock{};
+        start( &this );
+}
 …
                 __atomic_store_n(&do_terminate, true, __ATOMIC_RELAXED);
                 __wake_proc( &this );
+                wake( &this );
                 P( terminated );
 …
+        }
+        pthread_join( kernel_thread, 0p );
+        free( this.stack );
+}
+void ?{}(cluster & this, const char name[], Duration preemption_rate) with( this ) {
+        pthread_join( kernel_thread, NULL );
+}
+void ?{}(cluster & this, const char * name, Duration preemption_rate) with( this ) {
         this.name = name;
         this.preemption_rate = preemption_rate;
 …
 // Kernel Scheduling logic
 //=============================================================================================
+static void runThread(processor * this, thread_desc * dst);
+static void finishRunning(processor * this);
+static void halt(processor * this);
 //Main of the processor contexts
 void main(processorCtx_t & runner) {
 …
                 __cfaabi_dbg_print_safe("Kernel : core %p started\n", this);
+                $thread * readyThread = 0p;
+                for( unsigned int spin_count = 0; ! __atomic_load_n(&this->do_terminate, __ATOMIC_SEQ_CST); spin_count++ ) {
+                        // Try to get the next thread
+                        readyThread = __next_thread( this->cltr );
+                        // If no ready thread
+                        if( readyThread == 0p ) {
+                                // Block until a thread is ready
+                                readyThread = __halt(this);
+                thread_desc * readyThread = NULL;
+                for( unsigned int spin_count = 0; ! __atomic_load_n(&this->do_terminate, __ATOMIC_SEQ_CST); spin_count++ )
+                {
+                        readyThread = nextThread( this->cltr );
+                        if(readyThread)
+                        {
+                                verify( ! kernelTLS.preemption_state.enabled );
+                                runThread(this, readyThread);
+                                verify( ! kernelTLS.preemption_state.enabled );
+                                //Some actions need to be taken from the kernel
+                                finishRunning(this);
+                                spin_count = 0;
+                        }
+                        // Check if we actually found a thread
+                        if( readyThread ) {
+                                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                                /* paranoid */ verifyf( readyThread->state == Ready || readyThread->preempted != __NO_PREEMPTION, "state : %d, preempted %d\n", readyThread->state, readyThread->preempted);
+                                /* paranoid */ verifyf( readyThread->next == 0p, "Expected null got %p", readyThread->next );
+                                // We found a thread run it
+                                __run_thread(this, readyThread);
+                                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                        else
+                        {
+                                // spin(this, &spin_count);
+                                halt(this);
+                        }
+                }
 …
         __cfaabi_dbg_print_safe("Kernel : core %p terminated\n", this);
-        // HACK : the coroutine context switch expects this_thread to be set
-        // and it make sense for it to be set in all other cases except here
-        // fake it
-        if( this == mainProcessor ) kernelTLS.this_thread = mainThread;
+}
 …
 // runThread runs a thread by context switching
 // from the processor coroutine to the target thread
+static void __run_thread(processor * this, $thread * thrd_dst) {
+        $coroutine * proc_cor = get_coroutine(this->runner);
+static void runThread(processor * this, thread_desc * thrd_dst) {
+        coroutine_desc * proc_cor = get_coroutine(this->runner);
+        // Reset the terminating actions here
+        this->finish.action_code = No_Action;
         // Update global state
         kernelTLS.this_thread = thrd_dst;
+        // set state of processor coroutine to inactive
+        verify(proc_cor->state == Active);
+        proc_cor->state = Blocked;
+        // Actually run the thread
+        RUNNING:  while(true) {
+                if(unlikely(thrd_dst->preempted)) {
+                        thrd_dst->preempted = __NO_PREEMPTION;
+                        verify(thrd_dst->state == Active  || thrd_dst->state == Rerun);
+                } else {
+                        verify(thrd_dst->state == Blocked || thrd_dst->state == Ready); // Ready means scheduled normally, blocked means rerun
+                        thrd_dst->state = Active;
+        // set state of processor coroutine to inactive and the thread to active
+        proc_cor->state = proc_cor->state == Halted ? Halted : Inactive;
+        thrd_dst->state = Active;
+        // set context switch to the thread that the processor is executing
+        verify( thrd_dst->context.SP );
+        CtxSwitch( &proc_cor->context, &thrd_dst->context );
+        // when CtxSwitch returns we are back in the processor coroutine
+        // set state of processor coroutine to active and the thread to inactive
+        thrd_dst->state = thrd_dst->state == Halted ? Halted : Inactive;
+        proc_cor->state = Active;
+}
+// KERNEL_ONLY
+static void returnToKernel() {
+        coroutine_desc * proc_cor = get_coroutine(kernelTLS.this_processor->runner);
+        thread_desc * thrd_src = kernelTLS.this_thread;
+        // set state of current coroutine to inactive
+        thrd_src->state = thrd_src->state == Halted ? Halted : Inactive;
+        proc_cor->state = Active;
+        int local_errno = *__volatile_errno();
+        #if defined( __i386 ) || defined( __x86_64 )
+                __x87_store;
+        #endif
+        // set new coroutine that the processor is executing
+        // and context switch to it
+        verify( proc_cor->context.SP );
+        CtxSwitch( &thrd_src->context, &proc_cor->context );
+        // set state of new coroutine to active
+        proc_cor->state = proc_cor->state == Halted ? Halted : Inactive;
+        thrd_src->state = Active;
+        #if defined( __i386 ) || defined( __x86_64 )
+                __x87_load;
+        #endif
+        *__volatile_errno() = local_errno;
+}
+// KERNEL_ONLY
+// Once a thread has finished running, some of
+// its final actions must be executed from the kernel
+static void finishRunning(processor * this) with( this->finish ) {
+        verify( ! kernelTLS.preemption_state.enabled );
+        choose( action_code ) {
+        case No_Action:
+                break;
+        case Release:
+                unlock( *lock );
+        case Schedule:
+                ScheduleThread( thrd );
+        case Release_Schedule:
+                unlock( *lock );
+                ScheduleThread( thrd );
+        case Release_Multi:
+                for(int i = 0; i < lock_count; i++) {
+                        unlock( *locks[i] );
+                }
+                __cfaabi_dbg_debug_do(
+                        thrd_dst->park_stale   = true;
+                        thrd_dst->unpark_stale = true;
+                )
+                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) < ((uintptr_t)__get_stack(thrd_dst->curr_cor)->base ) || thrd_dst->curr_cor == proc_cor, "ERROR : Destination $thread %p has been corrupted.\n StackPointer too small.\n", thrd_dst ); // add escape condition if we are setting up the processor
+                /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) > ((uintptr_t)__get_stack(thrd_dst->curr_cor)->limit) || thrd_dst->curr_cor == proc_cor, "ERROR : Destination $thread %p has been corrupted.\n StackPointer too large.\n", thrd_dst ); // add escape condition if we are setting up the processor
+                // set context switch to the thread that the processor is executing
+                verify( thrd_dst->context.SP );
+                __cfactx_switch( &proc_cor->context, &thrd_dst->context );
+                // when __cfactx_switch returns we are back in the processor coroutine
+                /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) > ((uintptr_t)__get_stack(thrd_dst->curr_cor)->limit), "ERROR : Destination $thread %p has been corrupted.\n StackPointer too large.\n", thrd_dst );
+                /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) < ((uintptr_t)__get_stack(thrd_dst->curr_cor)->base ), "ERROR : Destination $thread %p has been corrupted.\n StackPointer too small.\n", thrd_dst );
+                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                // We just finished running a thread, there are a few things that could have happened.
+                // 1 - Regular case : the thread has blocked and now one has scheduled it yet.
+                // 2 - Racy case    : the thread has blocked but someone has already tried to schedule it.
+                // 4 - Preempted
+                // In case 1, we may have won a race so we can't write to the state again.
+                // In case 2, we lost the race so we now own the thread.
+                if(unlikely(thrd_dst->preempted != __NO_PREEMPTION)) {
+                        // The thread was preempted, reschedule it and reset the flag
+                        __schedule_thread( thrd_dst );
+                        break RUNNING;
+        case Release_Multi_Schedule:
+                for(int i = 0; i < lock_count; i++) {
+                        unlock( *locks[i] );
+                }
+                // set state of processor coroutine to active and the thread to inactive
+                static_assert(sizeof(thrd_dst->state) == sizeof(int));
+                enum coroutine_state old_state = __atomic_exchange_n(&thrd_dst->state, Blocked, __ATOMIC_SEQ_CST);
+                __cfaabi_dbg_debug_do( thrd_dst->park_result = old_state; )
+                switch(old_state) {
+                        case Halted:
+                                // The thread has halted, it should never be scheduled/run again, leave it back to Halted and move on
+                                thrd_dst->state = Halted;
+                                // We may need to wake someone up here since
+                                unpark( this->destroyer __cfaabi_dbg_ctx2 );
+                                this->destroyer = 0p;
+                                break RUNNING;
+                        case Active:
+                                // This is case 1, the regular case, nothing more is needed
+                                break RUNNING;
+                        case Rerun:
+                                // This is case 2, the racy case, someone tried to run this thread before it finished blocking
+                                // In this case, just run it again.
+                                continue RUNNING;
+                        default:
+                                // This makes no sense, something is wrong abort
+                                abort("Finished running a thread that was Blocked/Start/Primed %d\n", old_state);
+                for(int i = 0; i < thrd_count; i++) {
+                        ScheduleThread( thrds[i] );
+                }
+        }
+        // Just before returning to the processor, set the processor coroutine to active
+        proc_cor->state = Active;
+        kernelTLS.this_thread = 0p;
+}
+// KERNEL_ONLY
+void returnToKernel() {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        $coroutine * proc_cor = get_coroutine(kernelTLS.this_processor->runner);
+        $thread * thrd_src = kernelTLS.this_thread;
+        // Run the thread on this processor
+        {
+                int local_errno = *__volatile_errno();
+                #if defined( __i386 ) || defined( __x86_64 )
+                        __x87_store;
+                #endif
+                verify( proc_cor->context.SP );
+                __cfactx_switch( &thrd_src->context, &proc_cor->context );
+                #if defined( __i386 ) || defined( __x86_64 )
+                        __x87_load;
+                #endif
+                *__volatile_errno() = local_errno;
+        }
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ verifyf( ((uintptr_t)thrd_src->context.SP) < ((uintptr_t)__get_stack(thrd_src->curr_cor)->base ), "ERROR : Returning $thread %p has been corrupted.\n StackPointer too small.\n", thrd_src );
+        /* paranoid */ verifyf( ((uintptr_t)thrd_src->context.SP) > ((uintptr_t)__get_stack(thrd_src->curr_cor)->limit), "ERROR : Returning $thread %p has been corrupted.\n StackPointer too large.\n", thrd_src );
+        case Callback:
+                callback();
+        default:
+                abort("KERNEL ERROR: Unexpected action to run after thread");
+        }
+}
 …
 // This is the entry point for processors (kernel threads)
 // It effectively constructs a coroutine by stealing the pthread stack
 static void * __invoke_processor(void * arg) {
+static void * CtxInvokeProcessor(void * arg) {
         processor * proc = (processor *) arg;
         kernelTLS.this_processor = proc;
         kernelTLS.this_thread    = 0p;
+        kernelTLS.this_thread    = NULL;
         kernelTLS.preemption_state.[enabled, disable_count] = [false, 1];
         // SKULLDUGGERY: We want to create a context for the processor coroutine
 …
         //Set global state
         kernelTLS.this_thread = 0p;
+        kernelTLS.this_thread    = NULL;
         //We now have a proper context from which to schedule threads
 …
         __cfaabi_dbg_print_safe("Kernel : core %p main ended (%p)\n", proc, &proc->runner);
+        return 0p;
+}
+static void Abort( int ret, const char func[] ) {
+        if ( ret ) {                                                                            // pthread routines return errno values
+                abort( "%s : internal error, error(%d) %s.", func, ret, strerror( ret ) );
+        } // if
+} // Abort
+void * __create_pthread( pthread_t * pthread, void * (*start)(void *), void * arg ) {
+        pthread_attr_t attr;
+        Abort( pthread_attr_init( &attr ), "pthread_attr_init" ); // initialize attribute
+        size_t stacksize;
+        // default stack size, normally defined by shell limit
+        Abort( pthread_attr_getstacksize( &attr, &stacksize ), "pthread_attr_getstacksize" );
+        assert( stacksize >= PTHREAD_STACK_MIN );
+        void * stack;
+        __cfaabi_dbg_debug_do(
+                stack = memalign( __page_size, stacksize + __page_size );
+                // pthread has no mechanism to create the guard page in user supplied stack.
+                if ( mprotect( stack, __page_size, PROT_NONE ) == -1 ) {
+                        abort( "mprotect : internal error, mprotect failure, error(%d) %s.", errno, strerror( errno ) );
+                } // if
+        );
+        __cfaabi_dbg_no_debug_do(
+                stack = malloc( stacksize );
+        );
+        Abort( pthread_attr_setstack( &attr, stack, stacksize ), "pthread_attr_setstack" );
+        Abort( pthread_create( pthread, &attr, start, arg ), "pthread_create" );
+        return stack;
+        return NULL;
+}
+static void start(processor * this) {
+        __cfaabi_dbg_print_safe("Kernel : Starting core %p\n", this);
+        pthread_create( &this->kernel_thread, NULL, CtxInvokeProcessor, (void*)this );
+        __cfaabi_dbg_print_safe("Kernel : core %p started\n", this);
+}
 // KERNEL_ONLY
+static void __kernel_first_resume( processor * this ) {
+        $thread * src = mainThread;
+        $coroutine * dst = get_coroutine(this->runner);
+        verify( ! kernelTLS.preemption_state.enabled );
+        kernelTLS.this_thread->curr_cor = dst;
+void kernel_first_resume( processor * this ) {
+        thread_desc * src = mainThread;
+        coroutine_desc * dst = get_coroutine(this->runner);
+        verify( ! kernelTLS.preemption_state.enabled );
         __stack_prepare( &dst->stack, 65000 );
         __cfactx_start(main, dst, this->runner, __cfactx_invoke_coroutine);
+        CtxStart(&this->runner, CtxInvokeCoroutine);
         verify( ! kernelTLS.preemption_state.enabled );
 …
         dst->starter = dst->starter ? dst->starter : &src->self_cor;
         // make sure the current state is still correct
         /* paranoid */ verify(src->state == Ready);
+        // set state of current coroutine to inactive
+        src->state = src->state == Halted ? Halted : Inactive;
         // context switch to specified coroutine
         verify( dst->context.SP );
+        __cfactx_switch( &src->context, &dst->context );
+        // when __cfactx_switch returns we are back in the src coroutine
+        mainThread->curr_cor = &mainThread->self_cor;
+        // make sure the current state has been update
+        /* paranoid */ verify(src->state == Active);
+        CtxSwitch( &src->context, &dst->context );
+        // when CtxSwitch returns we are back in the src coroutine
+        // set state of new coroutine to active
+        src->state = Active;
         verify( ! kernelTLS.preemption_state.enabled );
 …
 // KERNEL_ONLY
 static void __kernel_last_resume( processor * this ) {
         $coroutine * src = &mainThread->self_cor;
         $coroutine * dst = get_coroutine(this->runner);
+void kernel_last_resume( processor * this ) {
+        coroutine_desc * src = &mainThread->self_cor;
+        coroutine_desc * dst = get_coroutine(this->runner);
         verify( ! kernelTLS.preemption_state.enabled );
 …
         verify( dst->context.SP );
-        // SKULLDUGGERY in debug the processors check that the
-        // stack is still within the limit of the stack limits after running a thread.
-        // that check doesn't make sense if we context switch to the processor using the
-        // coroutine semantics. Since this is a special case, use the current context
-        // info to populate these fields.
-        __cfaabi_dbg_debug_do(
-                __stack_context_t ctx;
-                CtxGet( ctx );
-                mainThread->context.SP = ctx.SP;
-                mainThread->context.FP = ctx.FP;
+        )
         // context switch to the processor
         __cfactx_switch( &src->context, &dst->context );
+        CtxSwitch( &src->context, &dst->context );
+}
 //-----------------------------------------------------------------------------
 // Scheduler routines
 // KERNEL ONLY
+void __schedule_thread( $thread * thrd ) with( *thrd->curr_cluster ) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ #if defined( __CFA_WITH_VERIFY__ )
+        /* paranoid */ if( thrd->state == Blocked || thrd->state == Start ) assertf( thrd->preempted == __NO_PREEMPTION,
+                          "Error inactive thread marked as preempted, state %d, preemption %d\n", thrd->state, thrd->preempted );
+        /* paranoid */ if( thrd->preempted != __NO_PREEMPTION ) assertf(thrd->state == Active || thrd->state == Rerun,
+                          "Error preempted thread marked as not currently running, state %d, preemption %d\n", thrd->state, thrd->preempted );
+        /* paranoid */ #endif
+        /* paranoid */ verifyf( thrd->next == 0p, "Expected null got %p", thrd->next );
+        if (thrd->preempted == __NO_PREEMPTION) thrd->state = Ready;
+        lock  ( ready_queue_lock __cfaabi_dbg_ctx2 );
+        bool was_empty = !(ready_queue != 0);
+        append( ready_queue, thrd );
+void ScheduleThread( thread_desc * thrd ) {
+        verify( thrd );
+        verify( thrd->state != Halted );
+        verify( ! kernelTLS.preemption_state.enabled );
+        verifyf( thrd->next == NULL, "Expected null got %p", thrd->next );
+        with( *thrd->curr_cluster ) {
+                lock  ( ready_queue_lock __cfaabi_dbg_ctx2 );
+                bool was_empty = !(ready_queue != 0);
+                append( ready_queue, thrd );
+                unlock( ready_queue_lock );
+                if(was_empty) {
+                        lock      (proc_list_lock __cfaabi_dbg_ctx2);
+                        if(idles) {
+                                wake_fast(idles.head);
+                        }
+                        unlock    (proc_list_lock);
+                }
+                else if( struct processor * idle = idles.head ) {
+                        wake_fast(idle);
+                }
+        }
+        verify( ! kernelTLS.preemption_state.enabled );
+}
+// KERNEL ONLY
+thread_desc * nextThread(cluster * this) with( *this ) {
+        verify( ! kernelTLS.preemption_state.enabled );
+        lock( ready_queue_lock __cfaabi_dbg_ctx2 );
+        thread_desc * head = pop_head( ready_queue );
         unlock( ready_queue_lock );
+        __wake_one(thrd->curr_cluster, was_empty);
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        verify( ! kernelTLS.preemption_state.enabled );
+        return head;
+}
+void BlockInternal() {
+        disable_interrupts();
+        verify( ! kernelTLS.preemption_state.enabled );
+        returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+void BlockInternal( __spinlock_t * lock ) {
+        disable_interrupts();
+        with( *kernelTLS.this_processor ) {
+                finish.action_code = Release;
+                finish.lock        = lock;
+        }
+        verify( ! kernelTLS.preemption_state.enabled );
+        returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+void BlockInternal( thread_desc * thrd ) {
+        disable_interrupts();
+        with( * kernelTLS.this_processor ) {
+                finish.action_code = Schedule;
+                finish.thrd        = thrd;
+        }
+        verify( ! kernelTLS.preemption_state.enabled );
+        returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+void BlockInternal( __spinlock_t * lock, thread_desc * thrd ) {
+        assert(thrd);
+        disable_interrupts();
+        with( * kernelTLS.this_processor ) {
+                finish.action_code = Release_Schedule;
+                finish.lock        = lock;
+                finish.thrd        = thrd;
+        }
+        verify( ! kernelTLS.preemption_state.enabled );
+        returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+void BlockInternal(__spinlock_t * locks [], unsigned short count) {
+        disable_interrupts();
+        with( * kernelTLS.this_processor ) {
+                finish.action_code = Release_Multi;
+                finish.locks       = locks;
+                finish.lock_count  = count;
+        }
+        verify( ! kernelTLS.preemption_state.enabled );
+        returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+void BlockInternal(__spinlock_t * locks [], unsigned short lock_count, thread_desc * thrds [], unsigned short thrd_count) {
+        disable_interrupts();
+        with( *kernelTLS.this_processor ) {
+                finish.action_code = Release_Multi_Schedule;
+                finish.locks       = locks;
+                finish.lock_count  = lock_count;
+                finish.thrds       = thrds;
+                finish.thrd_count  = thrd_count;
+        }
+        verify( ! kernelTLS.preemption_state.enabled );
+        returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+void BlockInternal(__finish_callback_fptr_t callback) {
+        disable_interrupts();
+        with( *kernelTLS.this_processor ) {
+                finish.action_code = Callback;
+                finish.callback    = callback;
+        }
+        verify( ! kernelTLS.preemption_state.enabled );
+        returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
 // KERNEL ONLY
+static $thread * __next_thread(cluster * this) with( *this ) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        lock( ready_queue_lock __cfaabi_dbg_ctx2 );
+        $thread * head = pop_head( ready_queue );
+        unlock( ready_queue_lock );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        return head;
+}
+void unpark( $thread * thrd __cfaabi_dbg_ctx_param2 ) {
+        if( !thrd ) return;
+        disable_interrupts();
+        static_assert(sizeof(thrd->state) == sizeof(int));
+        // record activity
+        __cfaabi_dbg_record_thrd( *thrd, false, caller );
+        enum coroutine_state old_state = __atomic_exchange_n(&thrd->state, Rerun, __ATOMIC_SEQ_CST);
+        __cfaabi_dbg_debug_do( thrd->unpark_result = old_state; )
+        switch(old_state) {
+                case Active:
+                        // Wake won the race, the thread will reschedule/rerun itself
+                        break;
+                case Blocked:
+                        /* paranoid */ verify( ! thrd->preempted != __NO_PREEMPTION );
+                        // Wake lost the race,
+                        thrd->state = Blocked;
+                        __schedule_thread( thrd );
+                        break;
+                case Rerun:
+                        abort("More than one thread attempted to schedule thread %p\n", thrd);
+                        break;
+                case Halted:
+                case Start:
+                case Primed:
+                default:
+                        // This makes no sense, something is wrong abort
+                        abort();
+        }
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+void park( __cfaabi_dbg_ctx_param ) {
+        /* paranoid */ verify( kernelTLS.preemption_state.enabled );
+        disable_interrupts();
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ verify( kernelTLS.this_thread->preempted == __NO_PREEMPTION );
+        // record activity
+        __cfaabi_dbg_record_thrd( *kernelTLS.this_thread, true, caller );
+void LeaveThread(__spinlock_t * lock, thread_desc * thrd) {
+        verify( ! kernelTLS.preemption_state.enabled );
+        with( * kernelTLS.this_processor ) {
+                finish.action_code = thrd ? Release_Schedule : Release;
+                finish.lock        = lock;
+                finish.thrd        = thrd;
+        }
         returnToKernel();
-        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
-        enable_interrupts( __cfaabi_dbg_ctx );
-        /* paranoid */ verify( kernelTLS.preemption_state.enabled );
+}
-// KERNEL ONLY
-void __leave_thread() {
-        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
-        returnToKernel();
-        abort();
+}
-// KERNEL ONLY
-bool force_yield( __Preemption_Reason reason ) {
-        /* paranoid */ verify( kernelTLS.preemption_state.enabled );
-        disable_interrupts();
-        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
-        $thread * thrd = kernelTLS.this_thread;
-        /* paranoid */ verify(thrd->state == Active || thrd->state == Rerun);
-        // SKULLDUGGERY: It is possible that we are preempting this thread just before
-        // it was going to park itself. If that is the case and it is already using the
-        // intrusive fields then we can't use them to preempt the thread
-        // If that is the case, abandon the preemption.
-        bool preempted = false;
-        if(thrd->next == 0p) {
-                preempted = true;
-                thrd->preempted = reason;
-                returnToKernel();
+        }
-        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
-        enable_interrupts_noPoll();
-        /* paranoid */ verify( kernelTLS.preemption_state.enabled );
-        return preempted;
+}
 …
 //-----------------------------------------------------------------------------
 // Kernel boot procedures
 static void __kernel_startup(void) {
+static void kernel_startup(void) {
         verify( ! kernelTLS.preemption_state.enabled );
         __cfaabi_dbg_print_safe("Kernel : Starting\n");
 …
         // SKULLDUGGERY: the mainThread steals the process main thread
         // which will then be scheduled by the mainProcessor normally
         mainThread = ($thread *)&storage_mainThread;
+        mainThread = (thread_desc *)&storage_mainThread;
         current_stack_info_t info;
         info.storage = (__stack_t*)&storage_mainThreadCtx;
 …
         void ?{}(processorCtx_t & this, processor * proc) {
                 (this.__cor){ "Processor" };
                 this.__cor.starter = 0p;
+                this.__cor.starter = NULL;
                 this.proc = proc;
+        }
 …
                 terminated{ 0 };
                 do_terminate = false;
                 preemption_alarm = 0p;
+                preemption_alarm = NULL;
                 pending_preemption = false;
                 kernel_thread = pthread_self();
 …
         // Add the main thread to the ready queue
         // once resume is called on mainProcessor->runner the mainThread needs to be scheduled like any normal thread
         __schedule_thread(mainThread);
+        ScheduleThread(mainThread);
         // SKULLDUGGERY: Force a context switch to the main processor to set the main thread's context to the current UNIX
         // context. Hence, the main thread does not begin through __cfactx_invoke_thread, like all other threads. The trick here is that
+        // context. Hence, the main thread does not begin through CtxInvokeThread, like all other threads. The trick here is that
         // mainThread is on the ready queue when this call is made.
         __kernel_first_resume( kernelTLS.this_processor );
+        kernel_first_resume( kernelTLS.this_processor );
 …
+}
 static void __kernel_shutdown(void) {
+static void kernel_shutdown(void) {
         __cfaabi_dbg_print_safe("\n--------------------------------------------------\nKernel : Shutting down\n");
         /* paranoid */ verify( TL_GET( preemption_state.enabled ) );
+        verify( TL_GET( preemption_state.enabled ) );
         disable_interrupts();
         /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        verify( ! kernelTLS.preemption_state.enabled );
         // SKULLDUGGERY: Notify the mainProcessor it needs to terminates.
 …
         // which is currently here
         __atomic_store_n(&mainProcessor->do_terminate, true, __ATOMIC_RELEASE);
         __kernel_last_resume( kernelTLS.this_processor );
+        kernel_last_resume( kernelTLS.this_processor );
         mainThread->self_cor.state = Halted;
 …
         // Destroy the main processor and its context in reverse order of construction
         // These were manually constructed so we need manually destroy them
+        ^(*mainProcessor){};
+        ^(mainProcessor->runner){};
+        ^(mainProcessor){};
         // Final step, destroy the main thread since it is no longer needed
         // Since we provided a stack to this taxk it will not destroy anything
+        /* paranoid */ verify(mainThread->self_cor.stack.storage == (__stack_t*)(((uintptr_t)&storage_mainThreadCtx)| 0x1));
+        ^(*mainThread){};
+        ^(mainThread){};
         ^(__cfa_dbg_global_clusters.list){};
 …
 //=============================================================================================
 // Kernel Idle Sleep
+// Kernel Quiescing
 //=============================================================================================
+static $thread * __halt(processor * this) with( *this ) {
+        if( do_terminate ) return 0p;
+        // First, lock the cluster idle
+        lock( cltr->idle_lock __cfaabi_dbg_ctx2 );
+        // Check if we can find a thread
+        if( $thread * found = __next_thread( cltr ) ) {
+                unlock( cltr->idle_lock );
+                return found;
+        }
+        // Move this processor from the active list to the idle list
+        move_to_front(cltr->procs, cltr->idles, *this);
+        // Unlock the idle lock so we don't go to sleep with a lock
+        unlock    (cltr->idle_lock);
+        // We are ready to sleep
+static void halt(processor * this) with( *this ) {
+        // verify( ! __atomic_load_n(&do_terminate, __ATOMIC_SEQ_CST) );
+        with( *cltr ) {
+                lock      (proc_list_lock __cfaabi_dbg_ctx2);
+                remove    (procs, *this);
+                push_front(idles, *this);
+                unlock    (proc_list_lock);
+        }
         __cfaabi_dbg_print_safe("Kernel : Processor %p ready to sleep\n", this);
+        wait( idle );
+        // We have woken up
+        wait( idleLock );
         __cfaabi_dbg_print_safe("Kernel : Processor %p woke up and ready to run\n", this);
-        // Get ourself off the idle list
         with( *cltr ) {
+                lock  (idle_lock __cfaabi_dbg_ctx2);
+                move_to_front(idles, procs, *this);
+                unlock(idle_lock);
+        }
+        // Don't check the ready queue again, we may not be in a position to run a thread
+        return 0p;
+}
+// Wake a thread from the front if there are any
+static bool __wake_one(cluster * this, __attribute__((unused)) bool force) {
+        // if we don't want to force check if we know it's false
+        if( !this->idles.head && !force ) return false;
+        // First, lock the cluster idle
+        lock( this->idle_lock __cfaabi_dbg_ctx2 );
+        // Check if there is someone to wake up
+        if( !this->idles.head ) {
+                // Nope unlock and return false
+                unlock( this->idle_lock );
+                return false;
+        }
+        // Wake them up
+        post( this->idles.head->idle );
+        // Unlock and return true
+        unlock( this->idle_lock );
+        return true;
+}
+// Unconditionnaly wake a thread
+static bool __wake_proc(processor * this) {
+        return post( this->idle );
+                lock      (proc_list_lock __cfaabi_dbg_ctx2);
+                remove    (idles, *this);
+                push_front(procs, *this);
+                unlock    (proc_list_lock);
+        }
+}
 …
                 sigemptyset( &mask );
                 sigaddset( &mask, SIGALRM );            // block SIGALRM signals
+                sigaddset( &mask, SIGUSR1 );            // block SIGALRM signals
+                sigsuspend( &mask );                            // block the processor to prevent further damage during abort
+                _exit( EXIT_FAILURE );                          // if processor unblocks before it is killed, terminate it
+                sigsuspend( &mask );                    // block the processor to prevent further damage during abort
+                _exit( EXIT_FAILURE );                  // if processor unblocks before it is killed, terminate it
+        }
         else {
 …
 void kernel_abort_msg( void * kernel_data, char * abort_text, int abort_text_size ) {
         $thread * thrd = kernel_data;
+        thread_desc * thrd = kernel_data;
         if(thrd) {
                 int len = snprintf( abort_text, abort_text_size, "Error occurred while executing thread %.256s (%p)", thrd->self_cor.name, thrd );
                 __cfaabi_bits_write( STDERR_FILENO, abort_text, len );
+                __cfaabi_dbg_bits_write( abort_text, len );
                 if ( &thrd->self_cor != thrd->curr_cor ) {
                         len = snprintf( abort_text, abort_text_size, " in coroutine %.256s (%p).\n", thrd->curr_cor->name, thrd->curr_cor );
                         __cfaabi_bits_write( STDERR_FILENO, abort_text, len );
+                        __cfaabi_dbg_bits_write( abort_text, len );
+                }
                 else {
                         __cfaabi_bits_write( STDERR_FILENO, ".\n", 2 );
+                        __cfaabi_dbg_bits_write( ".\n", 2 );
+                }
+        }
         else {
                 int len = snprintf( abort_text, abort_text_size, "Error occurred outside of any thread.\n" );
                 __cfaabi_bits_write( STDERR_FILENO, abort_text, len );
+                __cfaabi_dbg_bits_write( abort_text, len );
+        }
+}
 …
 extern "C" {
         void __cfaabi_bits_acquire() {
+        void __cfaabi_dbg_bits_acquire() {
                 lock( kernel_debug_lock __cfaabi_dbg_ctx2 );
+        }
         void __cfaabi_bits_release() {
+        void __cfaabi_dbg_bits_release() {
                 unlock( kernel_debug_lock );
+        }
 …
                 // atomically release spin lock and block
+                unlock( lock );
+                park( __cfaabi_dbg_ctx );
+                BlockInternal( &lock );
+        }
         else {
 …
+}
 bool V(semaphore & this) with( this ) {
         $thread * thrd = 0p;
+void V(semaphore & this) with( this ) {
+        thread_desc * thrd = NULL;
         lock( lock __cfaabi_dbg_ctx2 );
         count += 1;
 …
         // make new owner
+        unpark( thrd __cfaabi_dbg_ctx2 );
+        return thrd != 0p;
+        WakeThread( thrd );
+}
 …
+}
 void doregister( cluster * cltr, $thread & thrd ) {
+void doregister( cluster * cltr, thread_desc & thrd ) {
         lock      (cltr->thread_list_lock __cfaabi_dbg_ctx2);
         cltr->nthreads += 1;
 …
+}
 void unregister( cluster * cltr, $thread & thrd ) {
+void unregister( cluster * cltr, thread_desc & thrd ) {
         lock  (cltr->thread_list_lock __cfaabi_dbg_ctx2);
         remove(cltr->threads, thrd );
 …
 void doregister( cluster * cltr, processor * proc ) {
         lock      (cltr->idle_lock __cfaabi_dbg_ctx2);
+        lock      (cltr->proc_list_lock __cfaabi_dbg_ctx2);
         cltr->nprocessors += 1;
         push_front(cltr->procs, *proc);
         unlock    (cltr->idle_lock);
+        unlock    (cltr->proc_list_lock);
+}
 void unregister( cluster * cltr, processor * proc ) {
         lock  (cltr->idle_lock __cfaabi_dbg_ctx2);
+        lock  (cltr->proc_list_lock __cfaabi_dbg_ctx2);
         remove(cltr->procs, *proc );
         cltr->nprocessors -= 1;
         unlock(cltr->idle_lock);
+        unlock(cltr->proc_list_lock);
+}
 …
 __cfaabi_dbg_debug_do(
         extern "C" {
                 void __cfaabi_dbg_record_lock(__spinlock_t & this, const char prev_name[]) {
+                void __cfaabi_dbg_record(__spinlock_t & this, const char * prev_name) {
                         this.prev_name = prev_name;
                         this.prev_thrd = kernelTLS.this_thread;
+                }
-                void __cfaabi_dbg_record_thrd($thread & this, bool park, const char prev_name[]) {
-                        if(park) {
-                                this.park_caller   = prev_name;
-                                this.park_stale    = false;
+                        }
-                        else {
-                                this.unpark_caller = prev_name;
-                                this.unpark_stale  = false;
+                        }
+                }
+        }
+)
 …
 //-----------------------------------------------------------------------------
 // Debug
 bool threading_enabled(void) __attribute__((const)) {
+bool threading_enabled(void) {
         return true;
+}

libcfa/src/concurrency/kernel.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jan 17 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 12:29:26 2020
 // Update Count     : 22
+// Last Modified On : Sat Jun 22 11:39:17 2019
+// Update Count     : 16
 //
 …
 #include "invoke.h"
 #include "time_t.hfa"
-#include "coroutine.hfa"
 extern "C" {
 …
         __spinlock_t lock;
         int count;
         __queue_t($thread) waiting;
+        __queue_t(thread_desc) waiting;
 };
 …
 void ^?{}(semaphore & this);
 void   P (semaphore & this);
 bool   V (semaphore & this);
+void   V (semaphore & this);
 …
 // Processor
 extern struct cluster * mainCluster;
+enum FinishOpCode { No_Action, Release, Schedule, Release_Schedule, Release_Multi, Release_Multi_Schedule, Callback };
+typedef void (*__finish_callback_fptr_t)(void);
+//TODO use union, many of these fields are mutually exclusive (i.e. MULTI vs NOMULTI)
+struct FinishAction {
+        FinishOpCode action_code;
+        /*
+        // Union of possible actions
+        union {
+                // Option 1 : locks and threads
+                struct {
+                        // 1 thread or N thread
+                        union {
+                                thread_desc * thrd;
+                                struct {
+                                        thread_desc ** thrds;
+                                        unsigned short thrd_count;
+                                };
+                        };
+                        // 1 lock or N lock
+                        union {
+                                __spinlock_t * lock;
+                                struct {
+                                        __spinlock_t ** locks;
+                                        unsigned short lock_count;
+                                };
+                        };
+                };
+                // Option 2 : action pointer
+                __finish_callback_fptr_t callback;
+        };
+        /*/
+        thread_desc * thrd;
+        thread_desc ** thrds;
+        unsigned short thrd_count;
+        __spinlock_t * lock;
+        __spinlock_t ** locks;
+        unsigned short lock_count;
+        __finish_callback_fptr_t callback;
+        //*/
+};
+static inline void ?{}(FinishAction & this) {
+        this.action_code = No_Action;
+        this.thrd = NULL;
+        this.lock = NULL;
+}
+static inline void ^?{}(FinishAction &) {}
 // Processor
 …
         // RunThread data
         // Action to do after a thread is ran
         $thread * destroyer;
+        struct FinishAction finish;
         // Preemption data
 …
         bool pending_preemption;
         // Idle lock (kernel semaphore)
         __bin_sem_t idle;
+        // Idle lock
+        __bin_sem_t idleLock;
         // Termination
 …
         volatile bool do_terminate;
         // Termination synchronisation (user semaphore)
+        // Termination synchronisation
         semaphore terminated;
-        // pthread Stack
-        void * stack;
         // Link lists fields
 …
 };
 void  ?{}(processor & this, const char name[], struct cluster & cltr);
+void  ?{}(processor & this, const char * name, struct cluster & cltr);
 void ^?{}(processor & this);
 static inline void  ?{}(processor & this)                    { this{ "Anonymous Processor", *mainCluster}; }
 static inline void  ?{}(processor & this, struct cluster & cltr)    { this{ "Anonymous Processor", cltr}; }
+static inline void  ?{}(processor & this, const char name[]) { this{name, *mainCluster }; }
+static inline [processor *&, processor *& ] __get( processor & this ) __attribute__((const)) { return this.node.[next, prev]; }
+static inline void  ?{}(processor & this, const char * name) { this{name, *mainCluster }; }
+static inline [processor *&, processor *& ] __get( processor & this ) {
+        return this.node.[next, prev];
+}
 //-----------------------------------------------------------------------------
 …
         // Ready queue for threads
         __queue_t($thread) ready_queue;
+        __queue_t(thread_desc) ready_queue;
         // Name of the cluster
 …
         // List of processors
         __spinlock_t idle_lock;
+        __spinlock_t proc_list_lock;
         __dllist_t(struct processor) procs;
         __dllist_t(struct processor) idles;
 …
         // List of threads
         __spinlock_t thread_list_lock;
         __dllist_t(struct $thread) threads;
+        __dllist_t(struct thread_desc) threads;
         unsigned int nthreads;
 …
 extern Duration default_preemption();
 void ?{} (cluster & this, const char name[], Duration preemption_rate);
+void ?{} (cluster & this, const char * name, Duration preemption_rate);
 void ^?{}(cluster & this);
 static inline void ?{} (cluster & this)                           { this{"Anonymous Cluster", default_preemption()}; }
 static inline void ?{} (cluster & this, Duration preemption_rate) { this{"Anonymous Cluster", preemption_rate}; }
+static inline void ?{} (cluster & this, const char name[])        { this{name, default_preemption()}; }
+static inline [cluster *&, cluster *& ] __get( cluster & this ) __attribute__((const)) { return this.node.[next, prev]; }
+static inline void ?{} (cluster & this, const char * name)        { this{name, default_preemption()}; }
+static inline [cluster *&, cluster *& ] __get( cluster & this ) {
+        return this.node.[next, prev];
+}
 static inline struct processor * active_processor() { return TL_GET( this_processor ); } // UNSAFE

libcfa/src/concurrency/kernel_private.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Feb 13 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Nov 30 19:25:02 2019
 // Update Count     : 8
+// Last Modified On : Thu Mar 29 14:06:40 2018
+// Update Count     : 3
 //
 …
+}
+void __schedule_thread( $thread * ) __attribute__((nonnull (1)));
+void ScheduleThread( thread_desc * );
+static inline void WakeThread( thread_desc * thrd ) {
+        if( !thrd ) return;
+        verify(thrd->state == Inactive);
+        disable_interrupts();
+        ScheduleThread( thrd );
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+thread_desc * nextThread(cluster * this);
 //Block current thread and release/wake-up the following resources
+void __leave_thread() __attribute__((noreturn));
+void BlockInternal(void);
+void BlockInternal(__spinlock_t * lock);
+void BlockInternal(thread_desc * thrd);
+void BlockInternal(__spinlock_t * lock, thread_desc * thrd);
+void BlockInternal(__spinlock_t * locks [], unsigned short count);
+void BlockInternal(__spinlock_t * locks [], unsigned short count, thread_desc * thrds [], unsigned short thrd_count);
+void BlockInternal(__finish_callback_fptr_t callback);
+void LeaveThread(__spinlock_t * lock, thread_desc * thrd);
 //-----------------------------------------------------------------------------
 …
 void main(processorCtx_t *);
+void * __create_pthread( pthread_t *, void * (*)(void *), void * );
+static inline void wake_fast(processor * this) {
+        __cfaabi_dbg_print_safe("Kernel : Waking up processor %p\n", this);
+        post( this->idleLock );
+}
+static inline void wake(processor * this) {
+        disable_interrupts();
+        wake_fast(this);
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
 struct event_kernel_t {
 …
 // Threads
 extern "C" {
+      void __cfactx_invoke_thread(void (*main)(void *), void * this);
+      forall(dtype T | is_thread(T))
+      void CtxInvokeThread(T * this);
+}
+extern void ThreadCtxSwitch(coroutine_desc * src, coroutine_desc * dst);
 __cfaabi_dbg_debug_do(
         extern void __cfaabi_dbg_thread_register  ( $thread * thrd );
         extern void __cfaabi_dbg_thread_unregister( $thread * thrd );
+        extern void __cfaabi_dbg_thread_register  ( thread_desc * thrd );
+        extern void __cfaabi_dbg_thread_unregister( thread_desc * thrd );
+)
 …
 #define KERNEL_STORAGE(T,X) static char storage_##X[sizeof(T)]
 static inline uint32_t __tls_rand() {
+static inline uint32_t tls_rand() {
         kernelTLS.rand_seed ^= kernelTLS.rand_seed << 6;
         kernelTLS.rand_seed ^= kernelTLS.rand_seed >> 21;
 …
 void unregister( struct cluster & cltr );
 void doregister( struct cluster * cltr, struct $thread & thrd );
 void unregister( struct cluster * cltr, struct $thread & thrd );
+void doregister( struct cluster * cltr, struct thread_desc & thrd );
+void unregister( struct cluster * cltr, struct thread_desc & thrd );
 void doregister( struct cluster * cltr, struct processor * proc );

libcfa/src/concurrency/monitor.cfa

-              r7030dab
+              r71d6bd8
 // file "LICENCE" distributed with Cforall.
 //
 // $monitor.c --
+// monitor_desc.c --
 //
 // Author           : Thierry Delisle
 // Created On       : Thd Feb 23 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec  4 07:55:14 2019
 // Update Count     : 10
+// Last Modified On : Fri Mar 30 14:30:26 2018
+// Update Count     : 9
 //
 …
 //-----------------------------------------------------------------------------
 // Forward declarations
 static inline void __set_owner ( $monitor * this, $thread * owner );
 static inline void __set_owner ( $monitor * storage [], __lock_size_t count, $thread * owner );
 static inline void set_mask  ( $monitor * storage [], __lock_size_t count, const __waitfor_mask_t & mask );
 static inline void reset_mask( $monitor * this );
 static inline $thread * next_thread( $monitor * this );
 static inline bool is_accepted( $monitor * this, const __monitor_group_t & monitors );
+static inline void set_owner ( monitor_desc * this, thread_desc * owner );
+static inline void set_owner ( monitor_desc * storage [], __lock_size_t count, thread_desc * owner );
+static inline void set_mask  ( monitor_desc * storage [], __lock_size_t count, const __waitfor_mask_t & mask );
+static inline void reset_mask( monitor_desc * this );
+static inline thread_desc * next_thread( monitor_desc * this );
+static inline bool is_accepted( monitor_desc * this, const __monitor_group_t & monitors );
 static inline void lock_all  ( __spinlock_t * locks [], __lock_size_t count );
 static inline void lock_all  ( $monitor * source [], __spinlock_t * /*out*/ locks [], __lock_size_t count );
+static inline void lock_all  ( monitor_desc * source [], __spinlock_t * /*out*/ locks [], __lock_size_t count );
 static inline void unlock_all( __spinlock_t * locks [], __lock_size_t count );
 static inline void unlock_all( $monitor * locks [], __lock_size_t count );
 static inline void save   ( $monitor * ctx [], __lock_size_t count, __spinlock_t * locks [], unsigned int /*out*/ recursions [], __waitfor_mask_t /*out*/ masks [] );
 static inline void restore( $monitor * ctx [], __lock_size_t count, __spinlock_t * locks [], unsigned int /*in */ recursions [], __waitfor_mask_t /*in */ masks [] );
 static inline void init     ( __lock_size_t count, $monitor * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] );
 static inline void init_push( __lock_size_t count, $monitor * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] );
 static inline $thread *        check_condition   ( __condition_criterion_t * );
+static inline void unlock_all( monitor_desc * locks [], __lock_size_t count );
+static inline void save   ( monitor_desc * ctx [], __lock_size_t count, __spinlock_t * locks [], unsigned int /*out*/ recursions [], __waitfor_mask_t /*out*/ masks [] );
+static inline void restore( monitor_desc * ctx [], __lock_size_t count, __spinlock_t * locks [], unsigned int /*in */ recursions [], __waitfor_mask_t /*in */ masks [] );
+static inline void init     ( __lock_size_t count, monitor_desc * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] );
+static inline void init_push( __lock_size_t count, monitor_desc * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] );
+static inline thread_desc *        check_condition   ( __condition_criterion_t * );
 static inline void                 brand_condition   ( condition & );
 static inline [$thread *, int] search_entry_queue( const __waitfor_mask_t &, $monitor * monitors [], __lock_size_t count );
+static inline [thread_desc *, int] search_entry_queue( const __waitfor_mask_t &, monitor_desc * monitors [], __lock_size_t count );
 forall(dtype T | sized( T ))
 static inline __lock_size_t insert_unique( T * array [], __lock_size_t & size, T * val );
 static inline __lock_size_t count_max    ( const __waitfor_mask_t & mask );
 static inline __lock_size_t aggregate    ( $monitor * storage [], const __waitfor_mask_t & mask );
+static inline __lock_size_t aggregate    ( monitor_desc * storage [], const __waitfor_mask_t & mask );
 //-----------------------------------------------------------------------------
 …
 #define monitor_ctx( mons, cnt )                                /* Define that create the necessary struct for internal/external scheduling operations */ \
         $monitor ** monitors = mons;                          /* Save the targeted monitors                                                          */ \
+        monitor_desc ** monitors = mons;                          /* Save the targeted monitors                                                          */ \
         __lock_size_t count = cnt;                                /* Save the count to a local variable                                                  */ \
         unsigned int recursions[ count ];                         /* Save the current recursion levels to restore them later                             */ \
 …
 //-----------------------------------------------------------------------------
 // Enter/Leave routines
+// Enter single monitor
+static void __enter( $monitor * this, const __monitor_group_t & group ) {
+        // Lock the monitor spinlock
+        lock( this->lock __cfaabi_dbg_ctx2 );
+        // Interrupts disable inside critical section
+        $thread * thrd = kernelTLS.this_thread;
+        __cfaabi_dbg_print_safe( "Kernel : %10p Entering mon %p (%p)\n", thrd, this, this->owner);
+        if( !this->owner ) {
+                // No one has the monitor, just take it
+                __set_owner( this, thrd );
+                __cfaabi_dbg_print_safe( "Kernel :  mon is free \n" );
+        }
+        else if( this->owner == thrd) {
+                // We already have the monitor, just note how many times we took it
+                this->recursion += 1;
+                __cfaabi_dbg_print_safe( "Kernel :  mon already owned \n" );
+        }
+        else if( is_accepted( this, group) ) {
+                // Some one was waiting for us, enter
+                __set_owner( this, thrd );
+                // Reset mask
+                reset_mask( this );
+                __cfaabi_dbg_print_safe( "Kernel :  mon accepts \n" );
+        }
+        else {
+                __cfaabi_dbg_print_safe( "Kernel :  blocking \n" );
+                // Some one else has the monitor, wait in line for it
+                /* paranoid */ verify( thrd->next == 0p );
+                append( this->entry_queue, thrd );
+                /* paranoid */ verify( thrd->next == 1p );
+                unlock( this->lock );
+                park( __cfaabi_dbg_ctx );
+extern "C" {
+        // Enter single monitor
+        static void __enter_monitor_desc( monitor_desc * this, const __monitor_group_t & group ) {
+                // Lock the monitor spinlock
+                lock( this->lock __cfaabi_dbg_ctx2 );
+                // Interrupts disable inside critical section
+                thread_desc * thrd = kernelTLS.this_thread;
+                __cfaabi_dbg_print_safe( "Kernel : %10p Entering mon %p (%p)\n", thrd, this, this->owner);
+                if( !this->owner ) {
+                        // No one has the monitor, just take it
+                        set_owner( this, thrd );
+                        __cfaabi_dbg_print_safe( "Kernel :  mon is free \n" );
+                }
+                else if( this->owner == thrd) {
+                        // We already have the monitor, just note how many times we took it
+                        this->recursion += 1;
+                        __cfaabi_dbg_print_safe( "Kernel :  mon already owned \n" );
+                }
+                else if( is_accepted( this, group) ) {
+                        // Some one was waiting for us, enter
+                        set_owner( this, thrd );
+                        // Reset mask
+                        reset_mask( this );
+                        __cfaabi_dbg_print_safe( "Kernel :  mon accepts \n" );
+                }
+                else {
+                        __cfaabi_dbg_print_safe( "Kernel :  blocking \n" );
+                        // Some one else has the monitor, wait in line for it
+                        append( this->entry_queue, thrd );
+                        BlockInternal( &this->lock );
+                        __cfaabi_dbg_print_safe( "Kernel : %10p Entered  mon %p\n", thrd, this);
+                        // BlockInternal will unlock spinlock, no need to unlock ourselves
+                        return;
+                }
                 __cfaabi_dbg_print_safe( "Kernel : %10p Entered  mon %p\n", thrd, this);
+                /* paranoid */ verifyf( kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+                return;
+        }
+        __cfaabi_dbg_print_safe( "Kernel : %10p Entered  mon %p\n", thrd, this);
+        /* paranoid */ verifyf( kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+        /* paranoid */ verify( this->lock.lock );
+        // Release the lock and leave
+        unlock( this->lock );
+        return;
+}
+static void __dtor_enter( $monitor * this, fptr_t func ) {
+        // Lock the monitor spinlock
+        lock( this->lock __cfaabi_dbg_ctx2 );
+        // Interrupts disable inside critical section
+        $thread * thrd = kernelTLS.this_thread;
+        __cfaabi_dbg_print_safe( "Kernel : %10p Entering dtor for mon %p (%p)\n", thrd, this, this->owner);
+        if( !this->owner ) {
+                __cfaabi_dbg_print_safe( "Kernel : Destroying free mon %p\n", this);
+                // No one has the monitor, just take it
+                __set_owner( this, thrd );
+                verifyf( kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+                // Release the lock and leave
                 unlock( this->lock );
                 return;
+        }
+        else if( this->owner == thrd) {
+                // We already have the monitor... but where about to destroy it so the nesting will fail
+                // Abort!
+                abort( "Attempt to destroy monitor %p by thread \"%.256s\" (%p) in nested mutex.", this, thrd->self_cor.name, thrd );
+        }
+        __lock_size_t count = 1;
+        $monitor ** monitors = &this;
+        __monitor_group_t group = { &this, 1, func };
+        if( is_accepted( this, group) ) {
+                __cfaabi_dbg_print_safe( "Kernel :  mon accepts dtor, block and signal it \n" );
+                // Wake the thread that is waiting for this
+                __condition_criterion_t * urgent = pop( this->signal_stack );
+                /* paranoid */ verify( urgent );
+                // Reset mask
+                reset_mask( this );
+                // Create the node specific to this wait operation
+                wait_ctx_primed( thrd, 0 )
+                // Some one else has the monitor, wait for him to finish and then run
+        static void __enter_monitor_dtor( monitor_desc * this, fptr_t func ) {
+                // Lock the monitor spinlock
+                lock( this->lock __cfaabi_dbg_ctx2 );
+                // Interrupts disable inside critical section
+                thread_desc * thrd = kernelTLS.this_thread;
+                __cfaabi_dbg_print_safe( "Kernel : %10p Entering dtor for mon %p (%p)\n", thrd, this, this->owner);
+                if( !this->owner ) {
+                        __cfaabi_dbg_print_safe( "Kernel : Destroying free mon %p\n", this);
+                        // No one has the monitor, just take it
+                        set_owner( this, thrd );
+                        unlock( this->lock );
+                        return;
+                }
+                else if( this->owner == thrd) {
+                        // We already have the monitor... but where about to destroy it so the nesting will fail
+                        // Abort!
+                        abort( "Attempt to destroy monitor %p by thread \"%.256s\" (%p) in nested mutex.", this, thrd->self_cor.name, thrd );
+                }
+                __lock_size_t count = 1;
+                monitor_desc ** monitors = &this;
+                __monitor_group_t group = { &this, 1, func };
+                if( is_accepted( this, group) ) {
+                        __cfaabi_dbg_print_safe( "Kernel :  mon accepts dtor, block and signal it \n" );
+                        // Wake the thread that is waiting for this
+                        __condition_criterion_t * urgent = pop( this->signal_stack );
+                        verify( urgent );
+                        // Reset mask
+                        reset_mask( this );
+                        // Create the node specific to this wait operation
+                        wait_ctx_primed( thrd, 0 )
+                        // Some one else has the monitor, wait for him to finish and then run
+                        BlockInternal( &this->lock, urgent->owner->waiting_thread );
+                        // Some one was waiting for us, enter
+                        set_owner( this, thrd );
+                }
+                else {
+                        __cfaabi_dbg_print_safe( "Kernel :  blocking \n" );
+                        wait_ctx( thrd, 0 )
+                        this->dtor_node = &waiter;
+                        // Some one else has the monitor, wait in line for it
+                        append( this->entry_queue, thrd );
+                        BlockInternal( &this->lock );
+                        // BlockInternal will unlock spinlock, no need to unlock ourselves
+                        return;
+                }
+                __cfaabi_dbg_print_safe( "Kernel : Destroying %p\n", this);
+        }
+        // Leave single monitor
+        void __leave_monitor_desc( monitor_desc * this ) {
+                // Lock the monitor spinlock
+                lock( this->lock __cfaabi_dbg_ctx2 );
+                __cfaabi_dbg_print_safe( "Kernel : %10p Leaving mon %p (%p)\n", kernelTLS.this_thread, this, this->owner);
+                verifyf( kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+                // Leaving a recursion level, decrement the counter
+                this->recursion -= 1;
+                // If we haven't left the last level of recursion
+                // it means we don't need to do anything
+                if( this->recursion != 0) {
+                        __cfaabi_dbg_print_safe( "Kernel :  recursion still %d\n", this->recursion);
+                        unlock( this->lock );
+                        return;
+                }
+                // Get the next thread, will be null on low contention monitor
+                thread_desc * new_owner = next_thread( this );
+                // We can now let other threads in safely
                 unlock( this->lock );
+                // Release the next thread
+                /* paranoid */ verifyf( urgent->owner->waiting_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+                unpark( urgent->owner->waiting_thread __cfaabi_dbg_ctx2 );
+                // Park current thread waiting
+                park( __cfaabi_dbg_ctx );
+                // Some one was waiting for us, enter
+                /* paranoid */ verifyf( kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+        }
+        else {
+                __cfaabi_dbg_print_safe( "Kernel :  blocking \n" );
+                wait_ctx( thrd, 0 )
+                this->dtor_node = &waiter;
+                // Some one else has the monitor, wait in line for it
+                /* paranoid */ verify( thrd->next == 0p );
+                append( this->entry_queue, thrd );
+                /* paranoid */ verify( thrd->next == 1p );
+                unlock( this->lock );
+                // Park current thread waiting
+                park( __cfaabi_dbg_ctx );
+                /* paranoid */ verifyf( kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+                return;
+        }
+        __cfaabi_dbg_print_safe( "Kernel : Destroying %p\n", this);
+}
+// Leave single monitor
+void __leave( $monitor * this ) {
+        // Lock the monitor spinlock
+        lock( this->lock __cfaabi_dbg_ctx2 );
+        __cfaabi_dbg_print_safe( "Kernel : %10p Leaving mon %p (%p)\n", kernelTLS.this_thread, this, this->owner);
+        /* paranoid */ verifyf( kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+        // Leaving a recursion level, decrement the counter
+        this->recursion -= 1;
+        // If we haven't left the last level of recursion
+        // it means we don't need to do anything
+        if( this->recursion != 0) {
+                __cfaabi_dbg_print_safe( "Kernel :  recursion still %d\n", this->recursion);
+                unlock( this->lock );
+                return;
+        }
+        // Get the next thread, will be null on low contention monitor
+        $thread * new_owner = next_thread( this );
+        // Check the new owner is consistent with who we wake-up
+        // new_owner might be null even if someone owns the monitor when the owner is still waiting for another monitor
+        /* paranoid */ verifyf( !new_owner || new_owner == this->owner, "Expected owner to be %p, got %p (m: %p)", new_owner, this->owner, this );
+        // We can now let other threads in safely
+        unlock( this->lock );
+        //We need to wake-up the thread
+        /* paranoid */ verifyf( !new_owner || new_owner == this->owner, "Expected owner to be %p, got %p (m: %p)", new_owner, this->owner, this );
+        unpark( new_owner __cfaabi_dbg_ctx2 );
+}
+// Leave single monitor for the last time
+void __dtor_leave( $monitor * this ) {
+        __cfaabi_dbg_debug_do(
+                if( TL_GET( this_thread ) != this->owner ) {
+                        abort( "Destroyed monitor %p has inconsistent owner, expected %p got %p.\n", this, TL_GET( this_thread ), this->owner);
+                }
+                if( this->recursion != 1 ) {
+                        abort( "Destroyed monitor %p has %d outstanding nested calls.\n", this, this->recursion - 1);
+                }
+        )
+}
+extern "C" {
+                //We need to wake-up the thread
+                WakeThread( new_owner );
+        }
+        // Leave single monitor for the last time
+        void __leave_dtor_monitor_desc( monitor_desc * this ) {
+                __cfaabi_dbg_debug_do(
+                        if( TL_GET( this_thread ) != this->owner ) {
+                                abort( "Destroyed monitor %p has inconsistent owner, expected %p got %p.\n", this, TL_GET( this_thread ), this->owner);
+                        }
+                        if( this->recursion != 1 ) {
+                                abort( "Destroyed monitor %p has %d outstanding nested calls.\n", this, this->recursion - 1);
+                        }
+                )
+        }
         // Leave the thread monitor
         // last routine called by a thread.
         // Should never return
+        void __cfactx_thrd_leave() {
+                $thread * thrd = TL_GET( this_thread );
+                $monitor * this = &thrd->self_mon;
+        void __leave_thread_monitor( thread_desc * thrd ) {
+                monitor_desc * this = &thrd->self_mon;
                 // Lock the monitor now
 …
                 disable_interrupts();
                 thrd->state = Halted;
                 /* paranoid */ verifyf( thrd == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", thrd, this->owner, this->recursion, this );
+                thrd->self_cor.state = Halted;
+                verifyf( thrd == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", thrd, this->owner, this->recursion, this );
                 // Leaving a recursion level, decrement the counter
 …
                 // Fetch the next thread, can be null
+                $thread * new_owner = next_thread( this );
+                // Release the monitor lock
+                unlock( this->lock );
+                // Unpark the next owner if needed
+                /* paranoid */ verifyf( !new_owner || new_owner == this->owner, "Expected owner to be %p, got %p (m: %p)", new_owner, this->owner, this );
+                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                /* paranoid */ verify( ! kernelTLS.this_processor->destroyer );
+                /* paranoid */ verify( thrd->state == Halted );
+                kernelTLS.this_processor->destroyer = new_owner;
+                // Leave the thread
+                __leave_thread();
+                thread_desc * new_owner = next_thread( this );
+                // Leave the thread, this will unlock the spinlock
+                // Use leave thread instead of BlockInternal which is
+                // specialized for this case and supports null new_owner
+                LeaveThread( &this->lock, new_owner );
                 // Control flow should never reach here!
 …
 static inline void enter( __monitor_group_t monitors ) {
         for( __lock_size_t i = 0; i < monitors.size; i++) {
                 __enter( monitors[i], monitors );
+                __enter_monitor_desc( monitors[i], monitors );
+        }
+}
 …
 // Leave multiple monitor
 // relies on the monitor array being sorted
 static inline void leave($monitor * monitors [], __lock_size_t count) {
+static inline void leave(monitor_desc * monitors [], __lock_size_t count) {
         for( __lock_size_t i = count - 1; i >= 0; i--) {
                 __leave( monitors[i] );
+                __leave_monitor_desc( monitors[i] );
+        }
+}
 …
 // Ctor for monitor guard
 // Sorts monitors before entering
 void ?{}( monitor_guard_t & this, $monitor * m [], __lock_size_t count, fptr_t func ) {
         $thread * thrd = TL_GET( this_thread );
+void ?{}( monitor_guard_t & this, monitor_desc * m [], __lock_size_t count, fptr_t func ) {
+        thread_desc * thrd = TL_GET( this_thread );
         // Store current array
 …
 // Ctor for monitor guard
 // Sorts monitors before entering
 void ?{}( monitor_dtor_guard_t & this, $monitor * m [], fptr_t func ) {
+void ?{}( monitor_dtor_guard_t & this, monitor_desc * m [], fptr_t func ) {
         // optimization
         $thread * thrd = TL_GET( this_thread );
+        thread_desc * thrd = TL_GET( this_thread );
         // Store current array
 …
         (thrd->monitors){m, 1, func};
         __dtor_enter( this.m, func );
+        __enter_monitor_dtor( this.m, func );
+}
 …
 void ^?{}( monitor_dtor_guard_t & this ) {
         // Leave the monitors in order
         __dtor_leave( this.m );
+        __leave_dtor_monitor_desc( this.m );
         // Restore thread context
 …
 //-----------------------------------------------------------------------------
 // Internal scheduling types
 void ?{}(__condition_node_t & this, $thread * waiting_thread, __lock_size_t count, uintptr_t user_info ) {
+void ?{}(__condition_node_t & this, thread_desc * waiting_thread, __lock_size_t count, uintptr_t user_info ) {
         this.waiting_thread = waiting_thread;
         this.count = count;
         this.next = 0p;
+        this.next = NULL;
         this.user_info = user_info;
+}
 …
 void ?{}(__condition_criterion_t & this ) with( this ) {
         ready  = false;
         target = 0p;
         owner  = 0p;
         next   = 0p;
+}
 void ?{}(__condition_criterion_t & this, $monitor * target, __condition_node_t & owner ) {
+        target = NULL;
+        owner  = NULL;
+        next   = NULL;
+}
+void ?{}(__condition_criterion_t & this, monitor_desc * target, __condition_node_t & owner ) {
         this.ready  = false;
         this.target = target;
         this.owner  = &owner;
         this.next   = 0p;
+        this.next   = NULL;
+}
 …
         // Check that everything is as expected
         assertf( this.monitors != 0p, "Waiting with no monitors (%p)", this.monitors );
+        assertf( this.monitors != NULL, "Waiting with no monitors (%p)", this.monitors );
         verifyf( this.monitor_count != 0, "Waiting with 0 monitors (%"PRIiFAST16")", this.monitor_count );
         verifyf( this.monitor_count < 32u, "Excessive monitor count (%"PRIiFAST16")", this.monitor_count );
 …
         // Append the current wait operation to the ones already queued on the condition
         // We don't need locks for that since conditions must always be waited on inside monitor mutual exclusion
-        /* paranoid */ verify( waiter.next == 0p );
         append( this.blocked, &waiter );
-        /* paranoid */ verify( waiter.next == 1p );
         // Lock all monitors (aggregates the locks as well)
 …
         // Find the next thread(s) to run
         __lock_size_t thread_count = 0;
         $thread * threads[ count ];
+        thread_desc * threads[ count ];
         __builtin_memset( threads, 0, sizeof( threads ) );
 …
         // Remove any duplicate threads
         for( __lock_size_t i = 0; i < count; i++) {
                 $thread * new_owner = next_thread( monitors[i] );
+                thread_desc * new_owner = next_thread( monitors[i] );
                 insert_unique( threads, thread_count, new_owner );
+        }
-        // Unlock the locks, we don't need them anymore
-        for(int i = 0; i < count; i++) {
-                unlock( *locks[i] );
+        }
-        // Wake the threads
-        for(int i = 0; i < thread_count; i++) {
-                unpark( threads[i] __cfaabi_dbg_ctx2 );
+        }
         // Everything is ready to go to sleep
         park( __cfaabi_dbg_ctx );
+        BlockInternal( locks, count, threads, thread_count );
         // We are back, restore the owners and recursions
 …
         //Some more checking in debug
         __cfaabi_dbg_debug_do(
                 $thread * this_thrd = TL_GET( this_thread );
+                thread_desc * this_thrd = TL_GET( this_thread );
                 if ( this.monitor_count != this_thrd->monitors.size ) {
                         abort( "Signal on condition %p made with different number of monitor(s), expected %zi got %zi", &this, this.monitor_count, this_thrd->monitors.size );
 …
         // Lock all monitors
         lock_all( this.monitors, 0p, count );
+        lock_all( this.monitors, NULL, count );
         //Pop the head of the waiting queue
 …
         //Check that everything is as expected
         verifyf( this.monitors != 0p, "Waiting with no monitors (%p)", this.monitors );
+        verifyf( this.monitors != NULL, "Waiting with no monitors (%p)", this.monitors );
         verifyf( this.monitor_count != 0, "Waiting with 0 monitors (%"PRIiFAST16")", this.monitor_count );
 …
         //Find the thread to run
         $thread * signallee = pop_head( this.blocked )->waiting_thread;
         __set_owner( monitors, count, signallee );
+        thread_desc * signallee = pop_head( this.blocked )->waiting_thread;
+        set_owner( monitors, count, signallee );
         __cfaabi_dbg_print_buffer_decl( "Kernel : signal_block condition %p (s: %p)\n", &this, signallee );
-        // unlock all the monitors
-        unlock_all( locks, count );
-        // unpark the thread we signalled
-        unpark( signallee __cfaabi_dbg_ctx2 );
         //Everything is ready to go to sleep
         park( __cfaabi_dbg_ctx );
+        BlockInternal( locks, count, &signallee, 1 );
 …
         // Create one!
         __lock_size_t max = count_max( mask );
         $monitor * mon_storage[max];
+        monitor_desc * mon_storage[max];
         __builtin_memset( mon_storage, 0, sizeof( mon_storage ) );
         __lock_size_t actual_count = aggregate( mon_storage, mask );
 …
+        {
                 // Check if the entry queue
                 $thread * next; int index;
+                thread_desc * next; int index;
                 [next, index] = search_entry_queue( mask, monitors, count );
 …
                                 verifyf( accepted.size == 1,  "ERROR: Accepted dtor has more than 1 mutex parameter." );
                                 $monitor * mon2dtor = accepted[0];
+                                monitor_desc * mon2dtor = accepted[0];
                                 verifyf( mon2dtor->dtor_node, "ERROR: Accepted monitor has no dtor_node." );
 …
                                 // Set the owners to be the next thread
+                                __set_owner( monitors, count, next );
+                                // unlock all the monitors
+                                unlock_all( locks, count );
+                                // unpark the thread we signalled
+                                unpark( next __cfaabi_dbg_ctx2 );
+                                //Everything is ready to go to sleep
+                                park( __cfaabi_dbg_ctx );
+                                set_owner( monitors, count, next );
+                                // Everything is ready to go to sleep
+                                BlockInternal( locks, count, &next, 1 );
                                 // We are back, restore the owners and recursions
 …
+        }
-        // unlock all the monitors
-        unlock_all( locks, count );
         //Everything is ready to go to sleep
         park( __cfaabi_dbg_ctx );
+        BlockInternal( locks, count );
 …
 // Utilities
 static inline void __set_owner( $monitor * this, $thread * owner ) {
         /* paranoid */ verify( this->lock.lock );
+static inline void set_owner( monitor_desc * this, thread_desc * owner ) {
+        // __cfaabi_dbg_print_safe( "Kernal :   Setting owner of %p to %p ( was %p)\n", this, owner, this->owner );
         //Pass the monitor appropriately
 …
+}
+static inline void __set_owner( $monitor * monitors [], __lock_size_t count, $thread * owner ) {
+        /* paranoid */ verify ( monitors[0]->lock.lock );
+        /* paranoid */ verifyf( monitors[0]->owner == kernelTLS.this_thread, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, monitors[0]->owner, monitors[0]->recursion, monitors[0] );
+        monitors[0]->owner        = owner;
+        monitors[0]->recursion    = 1;
+static inline void set_owner( monitor_desc * monitors [], __lock_size_t count, thread_desc * owner ) {
+        monitors[0]->owner     = owner;
+        monitors[0]->recursion = 1;
         for( __lock_size_t i = 1; i < count; i++ ) {
+                /* paranoid */ verify ( monitors[i]->lock.lock );
+                /* paranoid */ verifyf( monitors[i]->owner == kernelTLS.this_thread, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, monitors[i]->owner, monitors[i]->recursion, monitors[i] );
+                monitors[i]->owner        = owner;
+                monitors[i]->recursion    = 0;
+        }
+}
+static inline void set_mask( $monitor * storage [], __lock_size_t count, const __waitfor_mask_t & mask ) {
+                monitors[i]->owner     = owner;
+                monitors[i]->recursion = 0;
+        }
+}
+static inline void set_mask( monitor_desc * storage [], __lock_size_t count, const __waitfor_mask_t & mask ) {
         for( __lock_size_t i = 0; i < count; i++) {
                 storage[i]->mask = mask;
 …
+}
 static inline void reset_mask( $monitor * this ) {
         this->mask.accepted = 0p;
         this->mask.data = 0p;
+static inline void reset_mask( monitor_desc * this ) {
+        this->mask.accepted = NULL;
+        this->mask.data = NULL;
         this->mask.size = 0;
+}
 static inline $thread * next_thread( $monitor * this ) {
+static inline thread_desc * next_thread( monitor_desc * this ) {
         //Check the signaller stack
         __cfaabi_dbg_print_safe( "Kernel :  mon %p AS-stack top %p\n", this, this->signal_stack.top);
 …
                 //regardless of if we are ready to baton pass,
                 //we need to set the monitor as in use
+                /* paranoid */ verifyf( !this->owner || kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+                __set_owner( this,  urgent->owner->waiting_thread );
+                set_owner( this,  urgent->owner->waiting_thread );
                 return check_condition( urgent );
 …
         // No signaller thread
         // Get the next thread in the entry_queue
+        $thread * new_owner = pop_head( this->entry_queue );
+        /* paranoid */ verifyf( !this->owner || kernelTLS.this_thread == this->owner, "Expected owner to be %p, got %p (r: %i, m: %p)", kernelTLS.this_thread, this->owner, this->recursion, this );
+        /* paranoid */ verify( !new_owner || new_owner->next == 0p );
+        __set_owner( this, new_owner );
+        thread_desc * new_owner = pop_head( this->entry_queue );
+        set_owner( this, new_owner );
         return new_owner;
+}
 static inline bool is_accepted( $monitor * this, const __monitor_group_t & group ) {
+static inline bool is_accepted( monitor_desc * this, const __monitor_group_t & group ) {
         __acceptable_t * it = this->mask.data; // Optim
         __lock_size_t count = this->mask.size;
 …
+}
 static inline void init( __lock_size_t count, $monitor * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] ) {
+static inline void init( __lock_size_t count, monitor_desc * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] ) {
         for( __lock_size_t i = 0; i < count; i++) {
                 (criteria[i]){ monitors[i], waiter };
 …
+}
 static inline void init_push( __lock_size_t count, $monitor * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] ) {
+static inline void init_push( __lock_size_t count, monitor_desc * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] ) {
         for( __lock_size_t i = 0; i < count; i++) {
                 (criteria[i]){ monitors[i], waiter };
 …
+}
 static inline void lock_all( $monitor * source [], __spinlock_t * /*out*/ locks [], __lock_size_t count ) {
+static inline void lock_all( monitor_desc * source [], __spinlock_t * /*out*/ locks [], __lock_size_t count ) {
         for( __lock_size_t i = 0; i < count; i++ ) {
                 __spinlock_t * l = &source[i]->lock;
 …
+}
 static inline void unlock_all( $monitor * locks [], __lock_size_t count ) {
+static inline void unlock_all( monitor_desc * locks [], __lock_size_t count ) {
         for( __lock_size_t i = 0; i < count; i++ ) {
                 unlock( locks[i]->lock );
 …
 static inline void save(
         $monitor * ctx [],
+        monitor_desc * ctx [],
         __lock_size_t count,
         __attribute((unused)) __spinlock_t * locks [],
 …
 static inline void restore(
         $monitor * ctx [],
+        monitor_desc * ctx [],
         __lock_size_t count,
         __spinlock_t * locks [],
 …
 // 2 - Checks if all the monitors are ready to run
 //     if so return the thread to run
 static inline $thread * check_condition( __condition_criterion_t * target ) {
+static inline thread_desc * check_condition( __condition_criterion_t * target ) {
         __condition_node_t * node = target->owner;
         unsigned short count = node->count;
 …
+        }
         __cfaabi_dbg_print_safe( "Kernel :  Runing %i (%p)\n", ready2run, ready2run ? node->waiting_thread : 0p );
         return ready2run ? node->waiting_thread : 0p;
+        __cfaabi_dbg_print_safe( "Kernel :  Runing %i (%p)\n", ready2run, ready2run ? node->waiting_thread : NULL );
+        return ready2run ? node->waiting_thread : NULL;
+}
 static inline void brand_condition( condition & this ) {
         $thread * thrd = TL_GET( this_thread );
+        thread_desc * thrd = TL_GET( this_thread );
         if( !this.monitors ) {
                 // __cfaabi_dbg_print_safe( "Branding\n" );
                 assertf( thrd->monitors.data != 0p, "No current monitor to brand condition %p", thrd->monitors.data );
+                assertf( thrd->monitors.data != NULL, "No current monitor to brand condition %p", thrd->monitors.data );
                 this.monitor_count = thrd->monitors.size;
                 this.monitors = ($monitor **)malloc( this.monitor_count * sizeof( *this.monitors ) );
+                this.monitors = (monitor_desc **)malloc( this.monitor_count * sizeof( *this.monitors ) );
                 for( int i = 0; i < this.monitor_count; i++ ) {
                         this.monitors[i] = thrd->monitors[i];
 …
+}
 static inline [$thread *, int] search_entry_queue( const __waitfor_mask_t & mask, $monitor * monitors [], __lock_size_t count ) {
         __queue_t($thread) & entry_queue = monitors[0]->entry_queue;
+static inline [thread_desc *, int] search_entry_queue( const __waitfor_mask_t & mask, monitor_desc * monitors [], __lock_size_t count ) {
+        __queue_t(thread_desc) & entry_queue = monitors[0]->entry_queue;
         // For each thread in the entry-queue
         for(    $thread ** thrd_it = &entry_queue.head;
                 *thrd_it != 1p;
+        for(    thread_desc ** thrd_it = &entry_queue.head;
+                *thrd_it;
                 thrd_it = &(*thrd_it)->next
         ) {
 …
+}
 static inline __lock_size_t aggregate( $monitor * storage [], const __waitfor_mask_t & mask ) {
+static inline __lock_size_t aggregate( monitor_desc * storage [], const __waitfor_mask_t & mask ) {
         __lock_size_t size = 0;
         for( __lock_size_t i = 0; i < mask.size; i++ ) {

libcfa/src/concurrency/monitor.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Thd Feb 23 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec  4 07:55:32 2019
 // Update Count     : 11
+// Last Modified On : Sat Oct  7 18:06:45 2017
+// Update Count     : 10
 //
 …
 trait is_monitor(dtype T) {
         $monitor * get_monitor( T & );
+        monitor_desc * get_monitor( T & );
         void ^?{}( T & mutex );
 };
 static inline void ?{}($monitor & this) with( this ) {
+static inline void ?{}(monitor_desc & this) with( this ) {
         lock{};
         entry_queue{};
         signal_stack{};
         owner         = 0p;
+        owner         = NULL;
         recursion     = 0;
         mask.accepted = 0p;
         mask.data     = 0p;
+        mask.accepted = NULL;
+        mask.data     = NULL;
         mask.size     = 0;
         dtor_node     = 0p;
+        dtor_node     = NULL;
+}
-static inline void ^?{}($monitor & ) {}
 struct monitor_guard_t {
         $monitor **     m;
+        monitor_desc **         m;
         __lock_size_t           count;
         __monitor_group_t prev;
 };
 void ?{}( monitor_guard_t & this, $monitor ** m, __lock_size_t count, void (*func)() );
+void ?{}( monitor_guard_t & this, monitor_desc ** m, __lock_size_t count, void (*func)() );
 void ^?{}( monitor_guard_t & this );
 struct monitor_dtor_guard_t {
         $monitor *    m;
+        monitor_desc *    m;
         __monitor_group_t prev;
 };
 void ?{}( monitor_dtor_guard_t & this, $monitor ** m, void (*func)() );
+void ?{}( monitor_dtor_guard_t & this, monitor_desc ** m, void (*func)() );
 void ^?{}( monitor_dtor_guard_t & this );
 …
         // The monitor this criterion concerns
         $monitor * target;
+        monitor_desc * target;
         // The parent node to which this criterion belongs
 …
 struct __condition_node_t {
         // Thread that needs to be woken when all criteria are met
         $thread * waiting_thread;
+        thread_desc * waiting_thread;
         // Array of criteria (Criterions are contiguous in memory)
 …
+}
 void ?{}(__condition_node_t & this, $thread * waiting_thread, __lock_size_t count, uintptr_t user_info );
+void ?{}(__condition_node_t & this, thread_desc * waiting_thread, __lock_size_t count, uintptr_t user_info );
 void ?{}(__condition_criterion_t & this );
 void ?{}(__condition_criterion_t & this, $monitor * target, __condition_node_t * owner );
+void ?{}(__condition_criterion_t & this, monitor_desc * target, __condition_node_t * owner );
 struct condition {
 …
         // Array of monitor pointers (Monitors are NOT contiguous in memory)
         $monitor ** monitors;
+        monitor_desc ** monitors;
         // Number of monitors in the array
 …
 static inline void ?{}( condition & this ) {
         this.monitors = 0p;
+        this.monitors = NULL;
         this.monitor_count = 0;
+}
 …
               bool signal      ( condition & this );
               bool signal_block( condition & this );
 static inline bool is_empty    ( condition & this ) { return this.blocked.head == 1p; }
+static inline bool is_empty    ( condition & this ) { return !this.blocked.head; }
          uintptr_t front       ( condition & this );

libcfa/src/concurrency/mutex.cfa

-              r7030dab
+              r71d6bd8
 // Author           : Thierry Delisle
 // Created On       : Fri May 25 01:37:11 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec  4 09:16:39 2019
 // Update Count     : 1
+// Last Modified By : Thierry Delisle
+// Last Modified On : Fri May 25 01:37:51 2018
+// Update Count     : 0
 //
 …
         if( is_locked ) {
                 append( blocked_threads, kernelTLS.this_thread );
+                unlock( lock );
+                park( __cfaabi_dbg_ctx );
+                BlockInternal( &lock );
+        }
         else {
 …
         lock( this.lock __cfaabi_dbg_ctx2 );
         this.is_locked = (this.blocked_threads != 0);
         unpark(
                 pop_head( this.blocked_threads ) __cfaabi_dbg_ctx2
+        WakeThread(
+                pop_head( this.blocked_threads )
         );
         unlock( this.lock );
 …
         this.lock{};
         this.blocked_threads{};
         this.owner = 0p;
+        this.owner = NULL;
         this.recursion_count = 0;
+}
 …
 void lock(recursive_mutex_lock & this) with(this) {
         lock( lock __cfaabi_dbg_ctx2 );
         if( owner == 0p ) {
+        if( owner == NULL ) {
                 owner = kernelTLS.this_thread;
                 recursion_count = 1;
 …
         else {
                 append( blocked_threads, kernelTLS.this_thread );
+                unlock( lock );
+                park( __cfaabi_dbg_ctx );
+                BlockInternal( &lock );
+        }
+}
 …
         bool ret = false;
         lock( lock __cfaabi_dbg_ctx2 );
         if( owner == 0p ) {
+        if( owner == NULL ) {
                 owner = kernelTLS.this_thread;
                 recursion_count = 1;
 …
         recursion_count--;
         if( recursion_count == 0 ) {
                 $thread * thrd = pop_head( blocked_threads );
+                thread_desc * thrd = pop_head( blocked_threads );
                 owner = thrd;
                 recursion_count = (thrd ? 1 : 0);
                 unpark( thrd __cfaabi_dbg_ctx2 );
+                WakeThread( thrd );
+        }
         unlock( lock );
 …
 void notify_one(condition_variable & this) with(this) {
         lock( lock __cfaabi_dbg_ctx2 );
         unpark(
                 pop_head( this.blocked_threads ) __cfaabi_dbg_ctx2
+        WakeThread(
+                pop_head( this.blocked_threads )
         );
         unlock( lock );
 …
         lock( lock __cfaabi_dbg_ctx2 );
         while(this.blocked_threads) {
                 unpark(
                         pop_head( this.blocked_threads ) __cfaabi_dbg_ctx2
+                WakeThread(
+                        pop_head( this.blocked_threads )
                 );
+        }
 …
         lock( this.lock __cfaabi_dbg_ctx2 );
         append( this.blocked_threads, kernelTLS.this_thread );
+        unlock( this.lock );
+        park( __cfaabi_dbg_ctx );
+        BlockInternal( &this.lock );
+}
 …
         lock( this.lock __cfaabi_dbg_ctx2 );
         append( this.blocked_threads, kernelTLS.this_thread );
+        unlock(l);
+        unlock(this.lock);
+        park( __cfaabi_dbg_ctx );
+        void __unlock(void) {
+                unlock(l);
+                unlock(this.lock);
+        }
+        BlockInternal( __unlock );
         lock(l);
+}

libcfa/src/concurrency/mutex.hfa

-              r7030dab
+              r71d6bd8
 // Author           : Thierry Delisle
 // Created On       : Fri May 25 01:24:09 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec  4 09:16:53 2019
 // Update Count     : 1
+// Last Modified By : Thierry Delisle
+// Last Modified On : Fri May 25 01:24:12 2018
+// Update Count     : 0
 //
 …
         // List of blocked threads
         __queue_t(struct $thread) blocked_threads;
+        __queue_t(struct thread_desc) blocked_threads;
         // Locked flag
 …
         // List of blocked threads
         __queue_t(struct $thread) blocked_threads;
+        __queue_t(struct thread_desc) blocked_threads;
         // Current thread owning the lock
         struct $thread * owner;
+        struct thread_desc * owner;
         // Number of recursion level
 …
         // List of blocked threads
         __queue_t(struct $thread) blocked_threads;
+        __queue_t(struct thread_desc) blocked_threads;
 };
 …
         static inline void ?{}(lock_scope(L) & this) {
                 this.locks = 0p;
+                this.locks = NULL;
                 this.count = 0;
+        }

libcfa/src/concurrency/preemption.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Jun 5 14:20:42 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Dec  5 16:34:05 2019
 // Update Count     : 43
+// Last Modified On : Tue Jun  5 17:35:49 2018
+// Update Count     : 37
 //
 …
 #include <string.h>
 #include <unistd.h>
-#include <limits.h>                                                                             // PTHREAD_STACK_MIN
+}
 …
 // FwdDeclarations : timeout handlers
 static void preempt( processor   * this );
 static void timeout( $thread * this );
+static void timeout( thread_desc * this );
 // FwdDeclarations : Signal handlers
 …
 event_kernel_t * event_kernel;                        // kernel public handle to even kernel
 static pthread_t alarm_thread;                        // pthread handle to alarm thread
-static void * alarm_stack;                                                        // pthread stack for alarm thread
 static void ?{}(event_kernel_t & this) with( this ) {
 …
 // Get next expired node
 static inline alarm_node_t * get_expired( alarm_list_t * alarms, Time currtime ) {
         if( !alarms->head ) return 0p;                                          // If no alarms return null
         if( alarms->head->alarm >= currtime ) return 0p;        // If alarms head not expired return null
         return pop(alarms);                                                                     // Otherwise just pop head
+        if( !alarms->head ) return NULL;                          // If no alarms return null
+        if( alarms->head->alarm >= currtime ) return NULL;        // If alarms head not expired return null
+        return pop(alarms);                                       // Otherwise just pop head
+}
 // Tick one frame of the Discrete Event Simulation for alarms
 static void tick_preemption() {
         alarm_node_t * node = 0p;                                                       // Used in the while loop but cannot be declared in the while condition
         alarm_list_t * alarms = &event_kernel->alarms;          // Local copy for ease of reading
         Time currtime = __kernel_get_time();                            // Check current time once so everything "happens at once"
+        alarm_node_t * node = NULL;                     // Used in the while loop but cannot be declared in the while condition
+        alarm_list_t * alarms = &event_kernel->alarms;  // Local copy for ease of reading
+        Time currtime = __kernel_get_time();                    // Check current time once so we everything "happens at once"
         //Loop throught every thing expired
 …
         // Enable interrupts by decrementing the counter
         // If counter reaches 0, execute any pending __cfactx_switch
+        // If counter reaches 0, execute any pending CtxSwitch
         void enable_interrupts( __cfaabi_dbg_ctx_param ) {
                 processor   * proc = kernelTLS.this_processor; // Cache the processor now since interrupts can start happening after the atomic store
+                thread_desc * thrd = kernelTLS.this_thread;       // Cache the thread now since interrupts can start happening after the atomic store
                 with( kernelTLS.preemption_state ){
 …
                                 if( proc->pending_preemption ) {
                                         proc->pending_preemption = false;
                                         force_yield( __POLL_PREEMPTION );
+                                        BlockInternal( thrd );
+                                }
+                        }
 …
         // Disable interrupts by incrementint the counter
         // Don't execute any pending __cfactx_switch even if counter reaches 0
+        // Don't execute any pending CtxSwitch even if counter reaches 0
         void enable_interrupts_noPoll() {
                 unsigned short prev = kernelTLS.preemption_state.disable_count;
 …
         sigaddset( &mask, sig );
         if ( pthread_sigmask( SIG_UNBLOCK, &mask, 0p ) == -1 ) {
+        if ( pthread_sigmask( SIG_UNBLOCK, &mask, NULL ) == -1 ) {
             abort( "internal error, pthread_sigmask" );
+        }
 …
         sigaddset( &mask, sig );
         if ( pthread_sigmask( SIG_BLOCK, &mask, 0p ) == -1 ) {
+        if ( pthread_sigmask( SIG_BLOCK, &mask, NULL ) == -1 ) {
             abort( "internal error, pthread_sigmask" );
+        }
 …
 // reserved for future use
 static void timeout( $thread * this ) {
+static void timeout( thread_desc * this ) {
         //TODO : implement waking threads
+}
 // KERNEL ONLY
 // Check if a __cfactx_switch signal handler shoud defer
+// Check if a CtxSwitch signal handler shoud defer
 // If true  : preemption is safe
 // If false : preemption is unsafe and marked as pending
 …
         // Setup proper signal handlers
         __cfaabi_sigaction( SIGUSR1, sigHandler_ctxSwitch, SA_SIGINFO | SA_RESTART ); // __cfactx_switch handler
+        __cfaabi_sigaction( SIGUSR1, sigHandler_ctxSwitch, SA_SIGINFO | SA_RESTART );         // CtxSwitch handler
         signal_block( SIGALRM );
         alarm_stack = __create_pthread( &alarm_thread, alarm_loop, 0p );
+        pthread_create( &alarm_thread, NULL, alarm_loop, NULL );
+}
 …
         sigset_t mask;
         sigfillset( &mask );
         sigprocmask( SIG_BLOCK, &mask, 0p );
+        sigprocmask( SIG_BLOCK, &mask, NULL );
         // Notify the alarm thread of the shutdown
 …
         // Wait for the preemption thread to finish
+        pthread_join( alarm_thread, 0p );
+        free( alarm_stack );
+        pthread_join( alarm_thread, NULL );
         // Preemption is now fully stopped
 …
         static_assert( sizeof( sigset_t ) == sizeof( cxt->uc_sigmask ), "Expected cxt->uc_sigmask to be of sigset_t" );
         #endif
         if ( pthread_sigmask( SIG_SETMASK, (sigset_t *)&(cxt->uc_sigmask), 0p ) == -1 ) {
+        if ( pthread_sigmask( SIG_SETMASK, (sigset_t *)&(cxt->uc_sigmask), NULL ) == -1 ) {
                 abort( "internal error, sigprocmask" );
+        }
 …
         // Preemption can occur here
         force_yield( __ALARM_PREEMPTION ); // Do the actual __cfactx_switch
+        BlockInternal( kernelTLS.this_thread ); // Do the actual CtxSwitch
+}
 …
         sigset_t mask;
         sigfillset(&mask);
         if ( pthread_sigmask( SIG_BLOCK, &mask, 0p ) == -1 ) {
+        if ( pthread_sigmask( SIG_BLOCK, &mask, NULL ) == -1 ) {
             abort( "internal error, pthread_sigmask" );
+        }
 …
                                         {__cfaabi_dbg_print_buffer_decl( " KERNEL: Spurious wakeup %d.\n", err );}
                                         continue;
                                 case EINVAL :
+                        case EINVAL :
                                         abort( "Timeout was invalid." );
                                 default:
 …
 EXIT:
         __cfaabi_dbg_print_safe( "Kernel : Preemption thread stopping\n" );
         return 0p;
+        return NULL;
+}
 …
         sigset_t oldset;
         int ret;
         ret = pthread_sigmask(0, 0p, &oldset);
+        ret = pthread_sigmask(0, NULL, &oldset);
         if(ret != 0) { abort("ERROR sigprocmask returned %d", ret); }

libcfa/src/concurrency/thread.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jan 17 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec  4 09:17:49 2019
 // Update Count     : 9
+// Last Modified On : Fri Mar 30 17:19:52 2018
+// Update Count     : 8
 //
 …
 #include "invoke.h"
+extern "C" {
+        #include <fenv.h>
+        #include <stddef.h>
+}
+//extern volatile thread_local processor * this_processor;
 //-----------------------------------------------------------------------------
 // Thread ctors and dtors
 void ?{}($thread & this, const char * const name, cluster & cl, void * storage, size_t storageSize ) with( this ) {
         context{ 0p, 0p };
+void ?{}(thread_desc & this, const char * const name, cluster & cl, void * storage, size_t storageSize ) with( this ) {
+        context{ NULL, NULL };
         self_cor{ name, storage, storageSize };
         state = Start;
-        preempted = __NO_PREEMPTION;
         curr_cor = &self_cor;
         self_mon.owner = &this;
 …
         self_mon_p = &self_mon;
         curr_cluster = &cl;
         next = 0p;
+        next = NULL;
         node.next = 0p;
         node.prev = 0p;
+        node.next = NULL;
+        node.prev = NULL;
         doregister(curr_cluster, this);
 …
+}
 void ^?{}($thread& this) with( this ) {
+void ^?{}(thread_desc& this) with( this ) {
         unregister(curr_cluster, this);
         ^self_cor{};
+}
-//-----------------------------------------------------------------------------
-// Starting and stopping threads
-forall( dtype T | is_thread(T) )
-void __thrd_start( T & this, void (*main_p)(T &) ) {
-        $thread * this_thrd = get_thread(this);
-        disable_interrupts();
-        __cfactx_start(main_p, get_coroutine(this), this, __cfactx_invoke_thread);
-        this_thrd->context.[SP, FP] = this_thrd->self_cor.context.[SP, FP];
-        verify( this_thrd->context.SP );
-        __schedule_thread(this_thrd);
-        enable_interrupts( __cfaabi_dbg_ctx );
+}
-//-----------------------------------------------------------------------------
-// Support for threads that don't ues the thread keyword
 forall( dtype T | sized(T) | is_thread(T) | { void ?{}(T&); } )
 void ?{}( scoped(T)& this ) with( this ) {
         handle{};
         __thrd_start(handle, main);
+        __thrd_start(handle);
+}
 …
 void ?{}( scoped(T)& this, P params ) with( this ) {
         handle{ params };
         __thrd_start(handle, main);
+        __thrd_start(handle);
+}
 …
+}
+//-----------------------------------------------------------------------------
+// Starting and stopping threads
+forall( dtype T | is_thread(T) )
+void __thrd_start( T& this ) {
+        thread_desc * this_thrd = get_thread(this);
+        thread_desc * curr_thrd = TL_GET( this_thread );
+        disable_interrupts();
+        CtxStart(&this, CtxInvokeThread);
+        this_thrd->context.[SP, FP] = this_thrd->self_cor.context.[SP, FP];
+        verify( this_thrd->context.SP );
+        CtxSwitch( &curr_thrd->context, &this_thrd->context );
+        ScheduleThread(this_thrd);
+        enable_interrupts( __cfaabi_dbg_ctx );
+}
+extern "C" {
+        // KERNEL ONLY
+        void __finish_creation(thread_desc * this) {
+                // set new coroutine that the processor is executing
+                // and context switch to it
+                verify( kernelTLS.this_thread != this );
+                verify( kernelTLS.this_thread->context.SP );
+                CtxSwitch( &this->context, &kernelTLS.this_thread->context );
+        }
+}
+void yield( void ) {
+        // Safety note : This could cause some false positives due to preemption
+      verify( TL_GET( preemption_state.enabled ) );
+        BlockInternal( TL_GET( this_thread ) );
+        // Safety note : This could cause some false positives due to preemption
+      verify( TL_GET( preemption_state.enabled ) );
+}
+void yield( unsigned times ) {
+        for( unsigned i = 0; i < times; i++ ) {
+                yield();
+        }
+}
 // Local Variables: //
 // mode: c //

libcfa/src/concurrency/thread.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jan 17 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec  4 09:18:14 2019
 // Update Count     : 6
+// Last Modified On : Fri Jun 21 17:51:33 2019
+// Update Count     : 5
 //
 …
       void ^?{}(T& mutex this);
       void main(T& this);
       $thread* get_thread(T& this);
+      thread_desc* get_thread(T& this);
 };
+// define that satisfies the trait without using the thread keyword
+#define DECL_THREAD(X) $thread* get_thread(X& this) __attribute__((const)) { return &this.__thrd; } void main(X& this)
+// Inline getters for threads/coroutines/monitors
+forall( dtype T | is_thread(T) )
+static inline $coroutine* get_coroutine(T & this) __attribute__((const)) { return &get_thread(this)->self_cor; }
+#define DECL_THREAD(X) thread_desc* get_thread(X& this) { return &this.__thrd; } void main(X& this)
 forall( dtype T | is_thread(T) )
+static inline $monitor  * get_monitor  (T & this) __attribute__((const)) { return &get_thread(this)->self_mon; }
+static inline coroutine_desc* get_coroutine(T & this) {
+        return &get_thread(this)->self_cor;
+}
+static inline $coroutine* get_coroutine($thread * this) __attribute__((const)) { return &this->self_cor; }
+static inline $monitor  * get_monitor  ($thread * this) __attribute__((const)) { return &this->self_mon; }
+forall( dtype T | is_thread(T) )
+static inline monitor_desc* get_monitor(T & this) {
+        return &get_thread(this)->self_mon;
+}
+//-----------------------------------------------------------------------------
+// forward declarations needed for threads
+static inline coroutine_desc* get_coroutine(thread_desc * this) {
+        return &this->self_cor;
+}
+static inline monitor_desc* get_monitor(thread_desc * this) {
+        return &this->self_mon;
+}
 extern struct cluster * mainCluster;
 forall( dtype T | is_thread(T) )
 void __thrd_start( T & this, void (*)(T &) );
+void __thrd_start( T & this );
 //-----------------------------------------------------------------------------
 // Ctors and dtors
 void ?{}($thread & this, const char * const name, struct cluster & cl, void * storage, size_t storageSize );
 void ^?{}($thread & this);
+void ?{}(thread_desc & this, const char * const name, struct cluster & cl, void * storage, size_t storageSize );
+void ^?{}(thread_desc & this);
 static inline void ?{}($thread & this)                                                                  { this{ "Anonymous Thread", *mainCluster, 0p, 65000 }; }
 static inline void ?{}($thread & this, size_t stackSize )                                               { this{ "Anonymous Thread", *mainCluster, 0p, stackSize }; }
 static inline void ?{}($thread & this, void * storage, size_t storageSize )                             { this{ "Anonymous Thread", *mainCluster, storage, storageSize }; }
 static inline void ?{}($thread & this, struct cluster & cl )                                            { this{ "Anonymous Thread", cl, 0p, 65000 }; }
 static inline void ?{}($thread & this, struct cluster & cl, size_t stackSize )                          { this{ "Anonymous Thread", cl, 0p, stackSize }; }
 static inline void ?{}($thread & this, struct cluster & cl, void * storage, size_t storageSize )        { this{ "Anonymous Thread", cl, storage, storageSize }; }
 static inline void ?{}($thread & this, const char * const name)                                         { this{ name, *mainCluster, 0p, 65000 }; }
 static inline void ?{}($thread & this, const char * const name, struct cluster & cl )                   { this{ name, cl, 0p, 65000 }; }
 static inline void ?{}($thread & this, const char * const name, struct cluster & cl, size_t stackSize ) { this{ name, cl, 0p, stackSize }; }
+static inline void ?{}(thread_desc & this)                                                                  { this{ "Anonymous Thread", *mainCluster, NULL, 65000 }; }
+static inline void ?{}(thread_desc & this, size_t stackSize )                                               { this{ "Anonymous Thread", *mainCluster, NULL, stackSize }; }
+static inline void ?{}(thread_desc & this, void * storage, size_t storageSize )                             { this{ "Anonymous Thread", *mainCluster, storage, storageSize }; }
+static inline void ?{}(thread_desc & this, struct cluster & cl )                                            { this{ "Anonymous Thread", cl, NULL, 65000 }; }
+static inline void ?{}(thread_desc & this, struct cluster & cl, size_t stackSize )                          { this{ "Anonymous Thread", cl, NULL, stackSize }; }
+static inline void ?{}(thread_desc & this, struct cluster & cl, void * storage, size_t storageSize )        { this{ "Anonymous Thread", cl, storage, storageSize }; }
+static inline void ?{}(thread_desc & this, const char * const name)                                         { this{ name, *mainCluster, NULL, 65000 }; }
+static inline void ?{}(thread_desc & this, const char * const name, struct cluster & cl )                   { this{ name, cl, NULL, 65000 }; }
+static inline void ?{}(thread_desc & this, const char * const name, struct cluster & cl, size_t stackSize ) { this{ name, cl, NULL, stackSize }; }
 //-----------------------------------------------------------------------------
 …
 void ^?{}( scoped(T)& this );
+//-----------------------------------------------------------------------------
+// Thread getters
+static inline struct $thread * active_thread () { return TL_GET( this_thread ); }
+void yield();
+void yield( unsigned times );
+//-----------------------------------------------------------------------------
+// Scheduler API
+//----------
+// Park thread: block until corresponding call to unpark, won't block if unpark is already called
+void park( __cfaabi_dbg_ctx_param );
+//----------
+// Unpark a thread, if the thread is already blocked, schedule it
+//                  if the thread is not yet block, signal that it should rerun immediately
+void unpark( $thread * this __cfaabi_dbg_ctx_param2 );
+forall( dtype T | is_thread(T) )
+static inline void unpark( T & this __cfaabi_dbg_ctx_param2 ) { if(!&this) return; unpark( get_thread( this ) __cfaabi_dbg_ctx_fwd2 );}
+//----------
+// Yield: force thread to block and be rescheduled
+bool force_yield( enum __Preemption_Reason );
+static inline void yield() {
+        force_yield(__MANUAL_PREEMPTION);
+}
+// Yield: yield N times
+static inline void yield( unsigned times ) {
+        for( times ) {
+                yield();
+        }
+}
+static inline struct thread_desc * active_thread () { return TL_GET( this_thread ); }
 // Local Variables: //

libcfa/src/exception.c

-              r7030dab
+              r71d6bd8
 // Author           : Andrew Beach
 // Created On       : Mon Jun 26 15:13:00 2017
 // Last Modified By : Andrew Beach
 // Last Modified On : Fri Apr 03 11:57:00 2020
 // Update Count     : 14
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Thu Feb 22 18:17:34 2018
+// Update Count     : 11
 //
-// Normally we would get this from the CFA prelude.
 #include <stddef.h> // for size_t
 #include "exception.h"
+// Implementation of the secret header is hardware dependent.
+#if !( defined( __x86_64 ) || defined( __i386 ) )
+#error Exception Handling: No known architecture detected.
+#endif
+// Implementation of the secret header.
 #include <stdlib.h>
 …
 // FIX ME: temporary hack to keep ARM build working
 #ifndef _URC_FATAL_PHASE1_ERROR
 #define _URC_FATAL_PHASE1_ERROR 3
+#define _URC_FATAL_PHASE1_ERROR 2
 #endif // ! _URC_FATAL_PHASE1_ERROR
 #ifndef _URC_FATAL_PHASE2_ERROR
 …
 #include "lsda.h"
-/* The exception class for our exceptions. Because of the vendor component
- * its value would not be standard.
- * Vendor: UWPL
- * Language: CFA\0
- */
-const _Unwind_Exception_Class __cfaehm_exception_class = 0x4c50575500414643;
 // Base exception vtable is abstract, you should not have base exceptions.
 struct __cfaehm_base_exception_t_vtable
                 ___cfaehm_base_exception_t_vtable_instance = {
+struct __cfaabi_ehm__base_exception_t_vtable
+                ___cfaabi_ehm__base_exception_t_vtable_instance = {
         .parent = NULL,
         .size = 0,
 …
 // Temperary global exception context. Does not work with concurency.
 struct exception_context_t {
+        struct __cfaehm_try_resume_node * top_resume;
+        exception_t * current_exception;
+        int current_handler_index;
+} static shared_stack = {NULL, NULL, 0};
+    struct __cfaabi_ehm__try_resume_node * top_resume;
+    struct __cfaabi_ehm__try_resume_node * current_resume;
+    exception_t * current_exception;
+    int current_handler_index;
+} shared_stack = {NULL, NULL, 0, 0};
 // Get the current exception context.
 …
         return &shared_stack;
+}
+//#define SAVE_EXCEPTION_CONTEXT(to_name)
+//struct exception_context_t * to_name = this_exception_context();
+//exception * this_exception() {
+//    return this_exception_context()->current_exception;
+//}
+// This macro should be the only thing that needs to change across machines.  Used in the personality function, way down
+// in termination.
+// struct _Unwind_Context * -> _Unwind_Reason_Code(*)(exception_t *)
+#define MATCHER_FROM_CONTEXT(ptr_to_context) \
+        (*(_Unwind_Reason_Code(**)(exception_t *))(_Unwind_GetCFA(ptr_to_context) + 8))
 // RESUMPTION ================================================================
+void __cfaehm_throw_resume(exception_t * except) {
+        struct exception_context_t * context = this_exception_context();
+void __cfaabi_ehm__throw_resume(exception_t * except) {
         __cfaabi_dbg_print_safe("Throwing resumption exception\n");
+        struct __cfaehm_try_resume_node * original_head = context->top_resume;
+        struct __cfaehm_try_resume_node * current = context->top_resume;
+        struct __cfaabi_ehm__try_resume_node * original_head = shared_stack.current_resume;
+        struct __cfaabi_ehm__try_resume_node * current =
+                (original_head) ? original_head->next : shared_stack.top_resume;
         for ( ; current ; current = current->next) {
                 context->top_resume = current->next;
+                shared_stack.current_resume = current;
                 if (current->handler(except)) {
                         context->top_resume = original_head;
+                        shared_stack.current_resume = original_head;
                         return;
+                }
 …
         __cfaabi_dbg_print_safe("Unhandled exception\n");
         context->top_resume = original_head;
+        shared_stack.current_resume = original_head;
         // Fall back to termination:
         __cfaehm_throw_terminate(except);
+        __cfaabi_ehm__throw_terminate(except);
         // TODO: Default handler for resumption.
+}
+// Do we control where exceptions get thrown even with concurency?
+// If not these are not quite thread safe, the cleanup hook has to
+// be added after the node is built but before it is made the top node.
+void __cfaehm_try_resume_setup(struct __cfaehm_try_resume_node * node,
+// Do we control where exceptions get thrown even with concurency?  If not these are not quite thread safe, the cleanup
+// hook has to be added after the node is built but before it is made the top node.
+void __cfaabi_ehm__try_resume_setup(struct __cfaabi_ehm__try_resume_node * node,
                         _Bool (*handler)(exception_t * except)) {
+        struct exception_context_t * context = this_exception_context();
+        node->next = context->top_resume;
+        node->next = shared_stack.top_resume;
         node->handler = handler;
+        context->top_resume = node;
+}
+void __cfaehm_try_resume_cleanup(struct __cfaehm_try_resume_node * node) {
+        struct exception_context_t * context = this_exception_context();
+        context->top_resume = node->next;
+        shared_stack.top_resume = node;
+}
+void __cfaabi_ehm__try_resume_cleanup(struct __cfaabi_ehm__try_resume_node * node) {
+        shared_stack.top_resume = node->next;
+}
 …
 // May have to move to cfa for constructors and destructors (references).
+// How to clean up an exception in various situations.
+static void __cfaehm_exception_cleanup(
+                _Unwind_Reason_Code reason,
+                struct _Unwind_Exception * exception) {
+        switch (reason) {
+        case _URC_FOREIGN_EXCEPTION_CAUGHT:
+                // This one we could clean-up to allow cross-language exceptions.
+        case _URC_FATAL_PHASE1_ERROR:
+        case _URC_FATAL_PHASE2_ERROR:
+        default:
+                abort();
+        }
+}
+// We need a piece of storage to raise the exception, for now its a single
+// piece.
+static struct _Unwind_Exception this_exception_storage;
+struct __cfaehm_node {
+        struct __cfaehm_node * next;
+struct __cfaabi_ehm__node {
+        struct __cfaabi_ehm__node * next;
 };
 #define NODE_TO_EXCEPT(node) ((exception_t *)(1 + (node)))
 #define EXCEPT_TO_NODE(except) ((struct __cfaehm_node *)(except) - 1)
+#define EXCEPT_TO_NODE(except) ((struct __cfaabi_ehm__node *)(except) - 1)
 // Creates a copy of the indicated exception and sets current_exception to it.
 static void __cfaehm_allocate_exception( exception_t * except ) {
+static void __cfaabi_ehm__allocate_exception( exception_t * except ) {
         struct exception_context_t * context = this_exception_context();
         // Allocate memory for the exception.
         struct __cfaehm_node * store = malloc(
                 sizeof( struct __cfaehm_node ) + except->virtual_table->size );
+        struct __cfaabi_ehm__node * store = malloc(
+                sizeof( struct __cfaabi_ehm__node ) + except->virtual_table->size );
         if ( ! store ) {
 …
         // Copy the exception to storage.
         except->virtual_table->copy( context->current_exception, except );
-        // Set up the exception storage.
-        this_exception_storage.exception_class = __cfaehm_exception_class;
-        this_exception_storage.exception_cleanup = __cfaehm_exception_cleanup;
+}
 // Delete the provided exception, unsetting current_exception if relivant.
 static void __cfaehm_delete_exception( exception_t * except ) {
+static void __cfaabi_ehm__delete_exception( exception_t * except ) {
         struct exception_context_t * context = this_exception_context();
 …
         // Remove the exception from the list.
         struct __cfaehm_node * to_free = EXCEPT_TO_NODE(except);
         struct __cfaehm_node * node;
+        struct __cfaabi_ehm__node * to_free = EXCEPT_TO_NODE(except);
+        struct __cfaabi_ehm__node * node;
         if ( context->current_exception == except ) {
 …
                 node = EXCEPT_TO_NODE(context->current_exception);
                 // It may always be in the first or second position.
                 while ( to_free != node->next ) {
+                while( to_free != node->next ) {
                         node = node->next;
+                }
 …
 // If this isn't a rethrow (*except==0), delete the provided exception.
+void __cfaehm_cleanup_terminate( void * except ) {
+        if ( *(void**)except ) __cfaehm_delete_exception( *(exception_t **)except );
+}
+void __cfaabi_ehm__cleanup_terminate( void * except ) {
+        if ( *(void**)except ) __cfaabi_ehm__delete_exception( *(exception_t **)except );
+}
+// We need a piece of storage to raise the exception
+struct _Unwind_Exception this_exception_storage;
 // Function needed by force unwind
 …
                 int version,
                 _Unwind_Action actions,
                 _Unwind_Exception_Class exception_class,
+                _Unwind_Exception_Class exceptionClass,
                 struct _Unwind_Exception * unwind_exception,
                 struct _Unwind_Context * unwind_context,
                 void * stop_param) {
         if ( actions & _UA_END_OF_STACK  ) exit(1);
         if ( actions & _UA_CLEANUP_PHASE ) return _URC_NO_REASON;
+                struct _Unwind_Context * context,
+                void * some_param) {
+        if( actions & _UA_END_OF_STACK  ) exit(1);
+        if( actions & _UA_CLEANUP_PHASE ) return _URC_NO_REASON;
         return _URC_FATAL_PHASE2_ERROR;
 …
 // The exception that is being thrown must already be stored.
 static __attribute__((noreturn)) void __cfaehm_begin_unwind(void) {
+__attribute__((noreturn)) void __cfaabi_ehm__begin_unwind(void) {
         if ( ! this_exception_context()->current_exception ) {
                 printf("UNWIND ERROR missing exception in begin unwind\n");
 …
+        }
         // Call stdlibc to raise the exception
         _Unwind_Reason_Code ret = _Unwind_RaiseException( &this_exception_storage );
+        // If we reach here it means something happened. For resumption to work we need to find a way
+        // to return back to here. Most of them will probably boil down to setting a global flag and
+        // making the phase 1 either stop or fail. Causing an error on purpose may help avoiding
+        // unnecessary work but it might have some weird side effects. If we just pretend no handler
+        // was found that would work but may be expensive for no reason since we will always search
+        // the whole stack.
+        if ( ret == _URC_END_OF_STACK ) {
+                // No proper handler was found. This can be handled in many ways, C++ calls std::terminate.
+                // Here we force unwind the stack, basically raising a cancellation.
+        // If we reach here it means something happened.  For resumption to work we need to find a way to return back to
+        // here.  Most of them will probably boil down to setting a global flag and making the phase 1 either stop or
+        // fail.  Causing an error on purpose may help avoiding unnecessary work but it might have some weird side
+        // effects.  If we just pretend no handler was found that would work but may be expensive for no reason since we
+        // will always search the whole stack.
+        if( ret == _URC_END_OF_STACK ) {
+                // No proper handler was found.  This can be handled in several way.  C++ calls std::terminate Here we
+                // force unwind the stack, basically raising a cancellation.
                 printf("Uncaught exception %p\n", &this_exception_storage);
 …
+        }
         // We did not simply reach the end of the stack without finding a handler. This is an error.
+        // We did not simply reach the end of the stack without finding a handler.  Something wen't wrong
         printf("UNWIND ERROR %d after raise exception\n", ret);
         abort();
+}
 void __cfaehm_throw_terminate( exception_t * val ) {
+void __cfaabi_ehm__throw_terminate( exception_t * val ) {
         __cfaabi_dbg_print_safe("Throwing termination exception\n");
         __cfaehm_allocate_exception( val );
         __cfaehm_begin_unwind();
+}
 void __cfaehm_rethrow_terminate(void) {
+        __cfaabi_ehm__allocate_exception( val );
+        __cfaabi_ehm__begin_unwind();
+}
+void __cfaabi_ehm__rethrow_terminate(void) {
         __cfaabi_dbg_print_safe("Rethrowing termination exception\n");
+        __cfaehm_begin_unwind();
+}
+// This is our personality routine. For every stack frame annotated with
+// ".cfi_personality 0x3,__gcfa_personality_v0" this function will be called twice when unwinding.
+//  Once in the search phase and once in the cleanup phase.
+_Unwind_Reason_Code __gcfa_personality_v0(
+                int version,
+                _Unwind_Action actions,
+                unsigned long long exception_class,
+                struct _Unwind_Exception * unwind_exception,
+                struct _Unwind_Context * unwind_context)
+        __cfaabi_ehm__begin_unwind();
+}
+#if defined(PIC)
+#warning Exceptions not yet supported when using Position-Independent Code
+__attribute__((noinline))
+void __cfaabi_ehm__try_terminate(void (*try_block)(),
+                void (*catch_block)(int index, exception_t * except),
+                __attribute__((unused)) int (*match_block)(exception_t * except)) {
+        abort();
+}
+#else
+// This is our personality routine.  For every stack frame anotated with ".cfi_personality 0x3,__gcfa_personality_v0".
+// This function will be called twice when unwinding.  Once in the search phased and once in the cleanup phase.
+_Unwind_Reason_Code __gcfa_personality_v0 (
+                int version, _Unwind_Action actions, unsigned long long exceptionClass,
+                struct _Unwind_Exception* unwind_exception,
+                struct _Unwind_Context* context)
+{
         //__cfaabi_dbg_print_safe("CFA: 0x%lx\n", _Unwind_GetCFA(context));
+        __cfaabi_dbg_print_safe("Personality function (%d, %x, %llu, %p, %p):",
+                        version, actions, exception_class, unwind_exception, unwind_context);
+        __cfaabi_dbg_print_safe("Personality function (%d, %x, %llu, %p, %p):", version, actions, exceptionClass, unwind_exception, context);
         // If we've reached the end of the stack then there is nothing much we can do...
         if (actions & _UA_END_OF_STACK) return _URC_END_OF_STACK;
+        if( actions & _UA_END_OF_STACK ) return _URC_END_OF_STACK;
         if (actions & _UA_SEARCH_PHASE) {
 …
         // Get a pointer to the language specific data from which we will read what we need
         const unsigned char * lsd = _Unwind_GetLanguageSpecificData( unwind_context );
         if ( !lsd ) {   //Nothing to do, keep unwinding
+        const unsigned char * lsd = (const unsigned char*) _Unwind_GetLanguageSpecificData( context );
+        if( !lsd ) {    //Nothing to do, keep unwinding
                 printf(" no LSD");
                 goto UNWIND;
 …
         // Get the instuction pointer and a reading pointer into the exception table
         lsda_header_info lsd_info;
+        const unsigned char * cur_ptr = parse_lsda_header(unwind_context, lsd, &lsd_info);
+        _Unwind_Ptr instruction_ptr = _Unwind_GetIP(unwind_context);
+        struct exception_context_t * context = this_exception_context();
+        const unsigned char * cur_ptr = parse_lsda_header( context, lsd, &lsd_info);
+        _Unwind_Ptr instruction_ptr = _Unwind_GetIP( context );
         // Linearly search the table for stuff to do
         while ( cur_ptr < lsd_info.action_table ) {
+        while( cur_ptr < lsd_info.action_table ) {
                 _Unwind_Ptr callsite_start;
                 _Unwind_Ptr callsite_len;
 …
                 // Decode the common stuff we have in here
                 cur_ptr = read_encoded_value(0, lsd_info.call_site_encoding, cur_ptr, &callsite_start);
                 cur_ptr = read_encoded_value(0, lsd_info.call_site_encoding, cur_ptr, &callsite_len);
                 cur_ptr = read_encoded_value(0, lsd_info.call_site_encoding, cur_ptr, &callsite_landing_pad);
                 cur_ptr = read_uleb128(cur_ptr, &callsite_action);
+                cur_ptr = read_encoded_value (0, lsd_info.call_site_encoding, cur_ptr, &callsite_start);
+                cur_ptr = read_encoded_value (0, lsd_info.call_site_encoding, cur_ptr, &callsite_len);
+                cur_ptr = read_encoded_value (0, lsd_info.call_site_encoding, cur_ptr, &callsite_landing_pad);
+                cur_ptr = read_uleb128 (cur_ptr, &callsite_action);
                 // Have we reach the correct frame info yet?
                 if ( lsd_info.Start + callsite_start + callsite_len < instruction_ptr ) {
+                if( lsd_info.Start + callsite_start + callsite_len < instruction_ptr ) {
 #ifdef __CFA_DEBUG_PRINT__
                         void * ls = (void*)lsd_info.Start;
 …
                         void * ep = (void*)lsd_info.Start + callsite_start + callsite_len;
                         void * ip = (void*)instruction_ptr;
+                        __cfaabi_dbg_print_safe("\nfound %p - %p (%p, %p, %p), looking for %p\n",
+                                        bp, ep, ls, cs, cl, ip);
+                        __cfaabi_dbg_print_safe("\nfound %p - %p (%p, %p, %p), looking for %p\n", bp, ep, ls, cs, cl, ip);
 #endif // __CFA_DEBUG_PRINT__
                         continue;
+                }
                 // Have we gone too far?
                 if ( lsd_info.Start + callsite_start > instruction_ptr ) {
+                // Have we gone too far
+                if( lsd_info.Start + callsite_start > instruction_ptr ) {
                         printf(" gone too far");
                         break;
+                }
                 // Check for what we must do:
                 if ( 0 == callsite_landing_pad ) {
                         // Nothing to do, move along
                         __cfaabi_dbg_print_safe(" no landing pad");
                 } else if (actions & _UA_SEARCH_PHASE) {
+                        // In search phase, these means we found a potential handler we must check.
                         // We have arbitrarily decided that 0 means nothing to do and 1 means there is
                         // a potential handler. This doesn't seem to conflict the gcc default behavior.
                         if (callsite_action != 0) {
                                 // Now we want to run some code to see if the handler matches
                                 // This is the tricky part where we want to the power to run arbitrary code
                                 // However, generating a new exception table entry and try routine every time
                                 // is way more expansive than we might like
                                 // The information we have is :
                                 //  - The GR (Series of registers)
                                 //    GR1=GP Global Pointer of frame ref by context
                                 //  - The instruction pointer
                                 //  - The instruction pointer info (???)
                                 //  - The CFA (Canonical Frame Address)
+                                //  - The BSP (Probably the base stack pointer)
                                 // The current apprach uses one exception table entry per try block
                                 _uleb128_t imatcher;
                                 // Get the relative offset to the {...}?
                                 cur_ptr = read_uleb128(cur_ptr, &imatcher);
+#                               if defined( __x86_64 )
                                 _Unwind_Word match_pos = _Unwind_GetCFA(unwind_context) + 8;
+#                               elif defined( __i386 )
                                 _Unwind_Word match_pos = _Unwind_GetCFA(unwind_context) + 24;
+#                               endif
                                 int (*matcher)(exception_t *) = *(int(**)(exception_t *))match_pos;
                                 int index = matcher(context->current_exception);
                                 _Unwind_Reason_Code ret = (0 == index)
+                                        ? _URC_CONTINUE_UNWIND : _URC_HANDLER_FOUND;
                                 context->current_handler_index = index;
                                 // Based on the return value, check if we matched the exception
                                 if (ret == _URC_HANDLER_FOUND) {
                                         __cfaabi_dbg_print_safe(" handler found\n");
                                 } else {
                                         __cfaabi_dbg_print_safe(" no handler\n");
+                // Something to do?
+                if( callsite_landing_pad ) {
+                        // Which phase are we in
+                        if (actions & _UA_SEARCH_PHASE) {
+                                // Search phase, this means we probably found a potential handler and must check if it is a match
+                                // If we have arbitrarily decided that 0 means nothing to do and 1 means there is a potential handler
+                                // This doesn't seem to conflict the gcc default behavior
+                                if (callsite_action != 0) {
+                                        // Now we want to run some code to see if the handler matches
+                                        // This is the tricky part where we want to the power to run arbitrary code
+                                        // However, generating a new exception table entry and try routine every time
+                                        // is way more expansive than we might like
+                                        // The information we have is :
+                                        //  - The GR (Series of registers)
+                                        //    GR1=GP Global Pointer of frame ref by context
+                                        //  - The instruction pointer
+                                        //  - The instruction pointer info (???)
+                                        //  - The CFA (Canonical Frame Address)
+                                        //  - The BSP (Probably the base stack pointer)
+                                        // The current apprach uses one exception table entry per try block
+                                        _uleb128_t imatcher;
+                                        // Get the relative offset to the
+                                        cur_ptr = read_uleb128 (cur_ptr, &imatcher);
+                                        // Get a function pointer from the relative offset and call it
+                                        // _Unwind_Reason_Code (*matcher)() = (_Unwind_Reason_Code (*)())lsd_info.LPStart + imatcher;
+                                        _Unwind_Reason_Code (*matcher)(exception_t *) =
+                                                MATCHER_FROM_CONTEXT(context);
+                                        int index = matcher(shared_stack.current_exception);
+                                        _Unwind_Reason_Code ret = (0 == index)
+                                                ? _URC_CONTINUE_UNWIND : _URC_HANDLER_FOUND;
+                                        shared_stack.current_handler_index = index;
+                                        // Based on the return value, check if we matched the exception
+                                        if( ret == _URC_HANDLER_FOUND) {
+                                                __cfaabi_dbg_print_safe(" handler found\n");
+                                        } else {
+                                                __cfaabi_dbg_print_safe(" no handler\n");
+                                        }
+                                        return ret;
+                                }
+                                return ret;
+                                // This is only a cleanup handler, ignore it
+                                __cfaabi_dbg_print_safe(" no action");
+                        }
+                        // This is only a cleanup handler, ignore it
+                        __cfaabi_dbg_print_safe(" no action");
+                } else if (actions & _UA_CLEANUP_PHASE) {
+                        // In clean-up phase, no destructors here but this could be the handler.
+                        if ( (callsite_action != 0) && !(actions & _UA_HANDLER_FRAME) ){
+                                // If this is a potential exception handler
+                                // but not the one that matched the exception in the seach phase,
+                                // just ignore it
+                                goto UNWIND;
+                        else if (actions & _UA_CLEANUP_PHASE) {
+                                if( (callsite_action != 0) && !(actions & _UA_HANDLER_FRAME) ){
+                                        // If this is a potential exception handler
+                                        // but not the one that matched the exception in the seach phase,
+                                        // just ignore it
+                                        goto UNWIND;
+                                }
+                                // We need to run some clean-up or a handler
+                                // These statment do the right thing but I don't know any specifics at all
+                                _Unwind_SetGR( context, __builtin_eh_return_data_regno(0), (_Unwind_Ptr) unwind_exception );
+                                _Unwind_SetGR( context, __builtin_eh_return_data_regno(1), 0 );
+                                // I assume this sets the instruction pointer to the adress of the landing pad
+                                // It doesn't actually set it, it only state the value that needs to be set once we return _URC_INSTALL_CONTEXT
+                                _Unwind_SetIP( context, ((lsd_info.LPStart) + (callsite_landing_pad)) );
+                                __cfaabi_dbg_print_safe(" action\n");
+                                // Return have some action to run
+                                return _URC_INSTALL_CONTEXT;
+                        }
-                        // We need to run some clean-up or a handler
-                        // These statment do the right thing but I don't know any specifics at all
-                        _Unwind_SetGR( unwind_context, __builtin_eh_return_data_regno(0),
-                                (_Unwind_Ptr)unwind_exception );
-                        _Unwind_SetGR( unwind_context, __builtin_eh_return_data_regno(1), 0 );
-                        // I assume this sets the instruction pointer to the adress of the landing pad
-                        // It doesn't actually set it, it only state the value that needs to be set once we
-                        // return _URC_INSTALL_CONTEXT
-                        _Unwind_SetIP( unwind_context, ((lsd_info.LPStart) + (callsite_landing_pad)) );
-                        __cfaabi_dbg_print_safe(" action\n");
-                        // Return have some action to run
-                        return _URC_INSTALL_CONTEXT;
+                }
+                // Nothing to do, move along
+                __cfaabi_dbg_print_safe(" no landing pad");
+        }
         // No handling found
 …
+}
+#pragma GCC push_options
+#pragma GCC optimize("O0")
+// Try statements are hoisted out see comments for details. While this could probably be unique
+// and simply linked from libcfa but there is one problem left, see the exception table for details
+// Try statements are hoisted out see comments for details.  With this could probably be unique and simply linked from
+// libcfa but there is one problem left, see the exception table for details
 __attribute__((noinline))
 void __cfaehm_try_terminate(void (*try_block)(),
+void __cfaabi_ehm__try_terminate(void (*try_block)(),
                 void (*catch_block)(int index, exception_t * except),
                 __attribute__((unused)) int (*match_block)(exception_t * except)) {
 …
         //! printf("%p %p %p %p\n", &try_block, &catch_block, &match_block, &xy);
+        // Setup the personality routine and exception table.
+        // Unforturnately these clobber gcc cancellation support which means we can't get access to
+        // the attribute cleanup tables at the same time. We would have to inspect the assembly to
+        // create a new set ourselves.
+#ifdef __PIC__
+        asm volatile (".cfi_personality 0x9b,CFA.ref.__gcfa_personality_v0");
+        asm volatile (".cfi_lsda 0x1b, .LLSDACFA2");
+#else
+        // Setup statments: These 2 statments won't actually result in any code, they only setup global tables.
+        // However, they clobber gcc cancellation support from gcc.  We can replace the personality routine but
+        // replacing the exception table gcc generates is not really doable, it generates labels based on how the
+        // assembly works.
+        // Setup the personality routine
         asm volatile (".cfi_personality 0x3,__gcfa_personality_v0");
+        // Setup the exception table
         asm volatile (".cfi_lsda 0x3, .LLSDACFA2");
-#endif
         // Label which defines the start of the area for which the handler is setup.
 …
         asm volatile goto ("" : : : : CATCH );
         // Normal return for when there is no throw.
+        // Normal return
         return;
 …
         // Label which defines the end of the area for which the handler is setup.
         asm volatile (".TRYEND:");
         // Label which defines the start of the exception landing pad. Basically what is called when
         // the exception is caught. Note, if multiple handlers are given, the multiplexing should be
         // done by the generated code, not the exception runtime.
+        // Label which defines the start of the exception landing pad.  Basically what is called when the exception is
+        // caught.  Note, if multiple handlers are given, the multiplexing should be done by the generated code, not the
+        // exception runtime.
         asm volatile (".CATCH:");
         // Exception handler
+        // Note: Saving the exception context on the stack breaks termination exceptions.
+        catch_block( this_exception_context()->current_handler_index,
+                     this_exception_context()->current_exception );
+}
+// Exception table data we need to generate. While this is almost generic, the custom data refers
+// to {*}try_terminate, which is no way generic. Some more works need to be done if we want to
+// have a single call to the try routine.
+#ifdef __PIC__
+        catch_block( shared_stack.current_handler_index,
+                     shared_stack.current_exception );
+}
+// Exception table data we need to generate.  While this is almost generic, the custom data refers to foo_try_match try
+// match, which is no way generic.  Some more works need to be done if we want to have a single call to the try routine.
+#if defined( __i386 ) || defined( __x86_64 )
 asm (
         // HEADER
+        //HEADER
         ".LFECFA1:\n"
         "       .globl  __gcfa_personality_v0\n"
         "       .section        .gcc_except_table,\"a\",@progbits\n"
+        // TABLE HEADER (important field is the BODY length at the end)
+        ".LLSDACFA2:\n"
+        ".LLSDACFA2:\n"                                                 //TABLE header
         "       .byte   0xff\n"
         "       .byte   0xff\n"
         "       .byte   0x1\n"
+        "       .uleb128 .LLSDACSECFA2-.LLSDACSBCFA2\n"
+        // BODY (language specific data)
+        // This uses language specific data and can be modified arbitrarily
+        // We use handled area offset, handled area length,
+        // handler landing pad offset and 1 (action code, gcc seems to use 0).
+        ".LLSDACSBCFA2:\n"
+        "       .uleb128 .TRYSTART-__cfaehm_try_terminate\n"
+        "       .uleb128 .TRYEND-.TRYSTART\n"
+        "       .uleb128 .CATCH-__cfaehm_try_terminate\n"
+        "       .uleb128 1\n"
+        ".LLSDACSECFA2:\n"
+        // TABLE FOOTER
+        "       .text\n"
+        "       .size   __cfaehm_try_terminate, .-__cfaehm_try_terminate\n"
+        "       .uleb128 .LLSDACSECFA2-.LLSDACSBCFA2\n"         // BODY length
+        // Body uses language specific data and therefore could be modified arbitrarily
+        ".LLSDACSBCFA2:\n"                                              // BODY start
+        "       .uleb128 .TRYSTART-__cfaabi_ehm__try_terminate\n"               // Handled area start  (relative to start of function)
+        "       .uleb128 .TRYEND-.TRYSTART\n"                           // Handled area length
+        "       .uleb128 .CATCH-__cfaabi_ehm__try_terminate\n"                          // Hanlder landing pad adress  (relative to start of function)
+        "       .uleb128 1\n"                                           // Action code, gcc seems to use always 0
+        ".LLSDACSECFA2:\n"                                              // BODY end
+        "       .text\n"                                                        // TABLE footer
+        "       .size   __cfaabi_ehm__try_terminate, .-__cfaabi_ehm__try_terminate\n"
+        "       .ident  \"GCC: (Ubuntu 6.2.0-3ubuntu11~16.04) 6.2.0 20160901\"\n"
+//      "       .section        .note.GNU-stack,\"x\",@progbits\n"
 );
+// Somehow this piece of helps with the resolution of debug symbols.
+__attribute__((unused)) static const int dummy = 0;
+asm (
+        // Add a hidden symbol which points at the function.
+        "       .hidden CFA.ref.__gcfa_personality_v0\n"
+        "       .weak   CFA.ref.__gcfa_personality_v0\n"
+        // No clue what this does specifically
+        "       .section        .data.rel.local.CFA.ref.__gcfa_personality_v0,\"awG\",@progbits,CFA.ref.__gcfa_personality_v0,comdat\n"
+        "       .align 8\n"
+        "       .type CFA.ref.__gcfa_personality_v0, @object\n"
+        "       .size CFA.ref.__gcfa_personality_v0, 8\n"
+        "CFA.ref.__gcfa_personality_v0:\n"
+#if defined( __x86_64 )
+        "       .quad __gcfa_personality_v0\n"
+#else // then __i386
+        "   .long __gcfa_personality_v0\n"
+#endif
+);
+#else // __PIC__
+asm (
+        // HEADER
+        ".LFECFA1:\n"
+        "       .globl  __gcfa_personality_v0\n"
+        "       .section        .gcc_except_table,\"a\",@progbits\n"
+        // TABLE HEADER (important field is the BODY length at the end)
+        ".LLSDACFA2:\n"
+        "       .byte   0xff\n"
+        "       .byte   0xff\n"
+        "       .byte   0x1\n"
+        "       .uleb128 .LLSDACSECFA2-.LLSDACSBCFA2\n"
+        // BODY (language specific data)
+        ".LLSDACSBCFA2:\n"
+        //      Handled area start (relative to start of function)
+        "       .uleb128 .TRYSTART-__cfaehm_try_terminate\n"
+        //      Handled area length
+        "       .uleb128 .TRYEND-.TRYSTART\n"
+        //      Handler landing pad address (relative to start of function)
+        "       .uleb128 .CATCH-__cfaehm_try_terminate\n"
+        //      Action code, gcc seems to always use 0.
+        "       .uleb128 1\n"
+        // TABLE FOOTER
+        ".LLSDACSECFA2:\n"
+        "       .text\n"
+        "       .size   __cfaehm_try_terminate, .-__cfaehm_try_terminate\n"
+        "       .ident  \"GCC: (Ubuntu 6.2.0-3ubuntu11~16.04) 6.2.0 20160901\"\n"
+        "       .section        .note.GNU-stack,\"x\",@progbits\n"
+);
+#endif // __PIC__
+#pragma GCC pop_options
+#endif // __i386 || __x86_64
+#endif //PIC

libcfa/src/exception.h

-              r7030dab
+              r71d6bd8
 // Author           : Andrew Beach
 // Created On       : Mon Jun 26 15:11:00 2017
 // Last Modified By : Andrew Beach
 // Last Modified On : Fri Mar 27 10:16:00 2020
 // Update Count     : 9
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Thu Feb 22 18:11:15 2018
+// Update Count     : 8
 //
 …
 #endif
 struct __cfaehm_base_exception_t;
 typedef struct __cfaehm_base_exception_t exception_t;
 struct __cfaehm_base_exception_t_vtable {
         const struct __cfaehm_base_exception_t_vtable * parent;
+struct __cfaabi_ehm__base_exception_t;
+typedef struct __cfaabi_ehm__base_exception_t exception_t;
+struct __cfaabi_ehm__base_exception_t_vtable {
+        const struct __cfaabi_ehm__base_exception_t_vtable * parent;
         size_t size;
         void (*copy)(struct __cfaehm_base_exception_t *this,
                      struct __cfaehm_base_exception_t * other);
         void (*free)(struct __cfaehm_base_exception_t *this);
         const char * (*msg)(struct __cfaehm_base_exception_t *this);
+        void (*copy)(struct __cfaabi_ehm__base_exception_t *this,
+                     struct __cfaabi_ehm__base_exception_t * other);
+        void (*free)(struct __cfaabi_ehm__base_exception_t *this);
+        const char * (*msg)(struct __cfaabi_ehm__base_exception_t *this);
 };
 struct __cfaehm_base_exception_t {
         struct __cfaehm_base_exception_t_vtable const * virtual_table;
+struct __cfaabi_ehm__base_exception_t {
+        struct __cfaabi_ehm__base_exception_t_vtable const * virtual_table;
 };
 extern struct __cfaehm_base_exception_t_vtable
         ___cfaehm_base_exception_t_vtable_instance;
+extern struct __cfaabi_ehm__base_exception_t_vtable
+        ___cfaabi_ehm__base_exception_t_vtable_instance;
 // Used in throw statement translation.
 void __cfaehm_throw_terminate(exception_t * except) __attribute__((noreturn));
 void __cfaehm_rethrow_terminate() __attribute__((noreturn));
 void __cfaehm_throw_resume(exception_t * except);
+void __cfaabi_ehm__throw_terminate(exception_t * except) __attribute__((noreturn));
+void __cfaabi_ehm__rethrow_terminate() __attribute__((noreturn));
+void __cfaabi_ehm__throw_resume(exception_t * except);
 // Function catches termination exceptions.
 void __cfaehm_try_terminate(
+void __cfaabi_ehm__try_terminate(
     void (*try_block)(),
     void (*catch_block)(int index, exception_t * except),
 …
 // Clean-up the exception in catch blocks.
 void __cfaehm_cleanup_terminate(void * except);
+void __cfaabi_ehm__cleanup_terminate(void * except);
 // Data structure creates a list of resume handlers.
 struct __cfaehm_try_resume_node {
     struct __cfaehm_try_resume_node * next;
+struct __cfaabi_ehm__try_resume_node {
+    struct __cfaabi_ehm__try_resume_node * next;
     _Bool (*handler)(exception_t * except);
 };
 // These act as constructor and destructor for the resume node.
 void __cfaehm_try_resume_setup(
     struct __cfaehm_try_resume_node * node,
+void __cfaabi_ehm__try_resume_setup(
+    struct __cfaabi_ehm__try_resume_node * node,
     _Bool (*handler)(exception_t * except));
 void __cfaehm_try_resume_cleanup(
     struct __cfaehm_try_resume_node * node);
+void __cfaabi_ehm__try_resume_cleanup(
+    struct __cfaabi_ehm__try_resume_node * node);
 // Check for a standard way to call fake deconstructors.
 struct __cfaehm_cleanup_hook {};
+struct __cfaabi_ehm__cleanup_hook {};
 #ifdef __cforall

libcfa/src/executor.cfa

-              r7030dab
+              r71d6bd8
 #include <stdio.h>
 forall( dtype T )
 monitor Buffer {                                        // unbounded buffer
     __queue_t( T ) queue;                               // unbounded list of work requests
     condition delay;
 }; // Buffer
+forall( dtype T | is_node(T) ) {
+forall( otype T | is_node(T) | is_monitor(T) ) {
+    monitor Buffer {                                    // unbounded buffer
+        __queue_t( T ) queue;                           // unbounded list of work requests
+        condition delay;
+    }; // Buffer
     void insert( Buffer( T ) & mutex buf, T * elem ) with(buf) {
         append( queue, elem );                          // insert element into buffer
 …
     T * remove( Buffer( T ) & mutex buf ) with(buf) {
         if ( queue.head != 0 ) wait( delay );                   // no request to process ? => wait
 //      return pop_head( queue );
+        if ( ! queue ) wait( delay );                   // no request to process ? => wait
+        return pop_head( queue );
     } // remove
 } // distribution

libcfa/src/fstream.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:01:01 2020
 // Update Count     : 363
+// Last Modified On : Tue Sep 10 22:19:56 2019
+// Update Count     : 354
 //
 …
 void ?{}( ofstream & os, void * file ) {
+        os.$file = file;
+        os.$sepDefault = true;
+        os.$sepOnOff = false;
+        os.$nlOnOff = true;
+        os.$prt = false;
+        os.$sawNL = false;
+        $sepSetCur( os, sepGet( os ) );
+        os.file = file;
+        os.sepDefault = true;
+        os.sepOnOff = false;
+        os.nlOnOff = true;
+        os.prt = false;
+        os.sawNL = false;
         sepSet( os, " " );
+        sepSetCur( os, sepGet( os ) );
         sepSetTuple( os, ", " );
 } // ?{}
 // private
 bool $sepPrt( ofstream & os ) { $setNL( os, false ); return os.$sepOnOff; }
 void $sepReset( ofstream & os ) { os.$sepOnOff = os.$sepDefault; }
 void $sepReset( ofstream & os, bool reset ) { os.$sepDefault = reset; os.$sepOnOff = os.$sepDefault; }
 const char * $sepGetCur( ofstream & os ) { return os.$sepCur; }
 void $sepSetCur( ofstream & os, const char sepCur[] ) { os.$sepCur = sepCur; }
 bool $getNL( ofstream & os ) { return os.$sawNL; }
 void $setNL( ofstream & os, bool state ) { os.$sawNL = state; }
 bool $getANL( ofstream & os ) { return os.$nlOnOff; }
 bool $getPrt( ofstream & os ) { return os.$prt; }
 void $setPrt( ofstream & os, bool state ) { os.$prt = state; }
+bool sepPrt( ofstream & os ) { setNL( os, false ); return os.sepOnOff; }
+void sepReset( ofstream & os ) { os.sepOnOff = os.sepDefault; }
+void sepReset( ofstream & os, bool reset ) { os.sepDefault = reset; os.sepOnOff = os.sepDefault; }
+const char * sepGetCur( ofstream & os ) { return os.sepCur; }
+void sepSetCur( ofstream & os, const char * sepCur ) { os.sepCur = sepCur; }
+bool getNL( ofstream & os ) { return os.sawNL; }
+void setNL( ofstream & os, bool state ) { os.sawNL = state; }
+bool getANL( ofstream & os ) { return os.nlOnOff; }
+bool getPrt( ofstream & os ) { return os.prt; }
+void setPrt( ofstream & os, bool state ) { os.prt = state; }
 // public
 void ?{}( ofstream & os ) { os.$file = 0p; }
 void ?{}( ofstream & os, const char name[], const char mode[] ) {
+void ?{}( ofstream & os ) { os.file = 0; }
+void ?{}( ofstream & os, const char * name, const char * mode ) {
         open( os, name, mode );
 } // ?{}
 void ?{}( ofstream & os, const char name[] ) {
+void ?{}( ofstream & os, const char * name ) {
         open( os, name, "w" );
 } // ?{}
+void ^?{}( ofstream & os ) {
+        close( os );
+} // ^?{}
+void sepOn( ofstream & os ) { os.$sepOnOff = ! $getNL( os ); }
+void sepOff( ofstream & os ) { os.$sepOnOff = false; }
+void sepOn( ofstream & os ) { os.sepOnOff = ! getNL( os ); }
+void sepOff( ofstream & os ) { os.sepOnOff = false; }
 bool sepDisable( ofstream & os ) {
         bool temp = os.$sepDefault;
         os.$sepDefault = false;
         $sepReset( os );
+        bool temp = os.sepDefault;
+        os.sepDefault = false;
+        sepReset( os );
         return temp;
 } // sepDisable
 bool sepEnable( ofstream & os ) {
         bool temp = os.$sepDefault;
         os.$sepDefault = true;
         if ( os.$sepOnOff ) $sepReset( os );                            // start of line ?
+        bool temp = os.sepDefault;
+        os.sepDefault = true;
+        if ( os.sepOnOff ) sepReset( os );                                      // start of line ?
         return temp;
 } // sepEnable
 void nlOn( ofstream & os ) { os.$nlOnOff = true; }
 void nlOff( ofstream & os ) { os.$nlOnOff = false; }
 const char * sepGet( ofstream & os ) { return os.$separator; }
 void sepSet( ofstream & os, const char s[] ) {
+void nlOn( ofstream & os ) { os.nlOnOff = true; }
+void nlOff( ofstream & os ) { os.nlOnOff = false; }
+const char * sepGet( ofstream & os ) { return os.separator; }
+void sepSet( ofstream & os, const char * s ) {
         assert( s );
         strncpy( os.$separator, s, sepSize - 1 );
         os.$separator[sepSize - 1] = '\0';
+        strncpy( os.separator, s, sepSize - 1 );
+        os.separator[sepSize - 1] = '\0';
 } // sepSet
 const char * sepGetTuple( ofstream & os ) { return os.$tupleSeparator; }
 void sepSetTuple( ofstream & os, const char s[] ) {
+const char * sepGetTuple( ofstream & os ) { return os.tupleSeparator; }
+void sepSetTuple( ofstream & os, const char * s ) {
         assert( s );
         strncpy( os.$tupleSeparator, s, sepSize - 1 );
         os.$tupleSeparator[sepSize - 1] = '\0';
+        strncpy( os.tupleSeparator, s, sepSize - 1 );
+        os.tupleSeparator[sepSize - 1] = '\0';
 } // sepSet
 void ends( ofstream & os ) {
         if ( $getANL( os ) ) nl( os );
         else $setPrt( os, false );                                                      // turn off
+        if ( getANL( os ) ) nl( os );
+        else setPrt( os, false );                                                       // turn off
         if ( &os == &exit ) exit( EXIT_FAILURE );
         if ( &os == &abort ) abort();
 …
 int fail( ofstream & os ) {
         return os.$file == 0 || ferror( (FILE *)(os.$file) );
+        return os.file == 0 || ferror( (FILE *)(os.file) );
 } // fail
 int flush( ofstream & os ) {
         return fflush( (FILE *)(os.$file) );
+        return fflush( (FILE *)(os.file) );
 } // flush
 void open( ofstream & os, const char name[], const char mode[] ) {
+void open( ofstream & os, const char * name, const char * mode ) {
         FILE * file = fopen( name, mode );
         #ifdef __CFA_DEBUG__
         if ( file == 0p ) {
+        if ( file == 0 ) {
                 abort | IO_MSG "open output file \"" | name | "\"" | nl | strerror( errno );
         } // if
 …
 } // open
 void open( ofstream & os, const char name[] ) {
+void open( ofstream & os, const char * name ) {
         open( os, name, "w" );
 } // open
 void close( ofstream & os ) {
         if ( (FILE *)(os.$file) == stdout || (FILE *)(os.$file) == stderr ) return;
         if ( fclose( (FILE *)(os.$file) ) == EOF ) {
+        if ( (FILE *)(os.file) == stdout || (FILE *)(os.file) == stderr ) return;
+        if ( fclose( (FILE *)(os.file) ) == EOF ) {
                 abort | IO_MSG "close output" | nl | strerror( errno );
         } // if
 } // close
 ofstream & write( ofstream & os, const char data[], size_t size ) {
+ofstream & write( ofstream & os, const char * data, size_t size ) {
         if ( fail( os ) ) {
                 abort | IO_MSG "attempt write I/O on failed stream";
         } // if
         if ( fwrite( data, 1, size, (FILE *)(os.$file) ) != size ) {
+        if ( fwrite( data, 1, size, (FILE *)(os.file) ) != size ) {
                 abort | IO_MSG "write" | nl | strerror( errno );
         } // if
 …
         va_list args;
         va_start( args, format );
         int len = vfprintf( (FILE *)(os.$file), format, args );
+        int len = vfprintf( (FILE *)(os.file), format, args );
         if ( len == EOF ) {
                 if ( ferror( (FILE *)(os.$file) ) ) {
+                if ( ferror( (FILE *)(os.file) ) ) {
                         abort | IO_MSG "invalid write";
                 } // if
 …
         va_end( args );
         $setPrt( os, true );                                                            // called in output cascade
         $sepReset( os );                                                                        // reset separator
+        setPrt( os, true );                                                                     // called in output cascade
+        sepReset( os );                                                                         // reset separator
         return len;
 } // fmt
 …
 // private
 void ?{}( ifstream & is, void * file ) {
         is.$file = file;
         is.$nlOnOff = false;
+        is.file = file;
+        is.nlOnOff = false;
 } // ?{}
 // public
 void ?{}( ifstream & is ) { is.$file = 0p; }
 void ?{}( ifstream & is, const char name[], const char mode[] ) {
+void ?{}( ifstream & is ) {     is.file = 0; }
+void ?{}( ifstream & is, const char * name, const char * mode ) {
         open( is, name, mode );
 } // ?{}
 void ?{}( ifstream & is, const char name[] ) {
+void ?{}( ifstream & is, const char * name ) {
         open( is, name, "r" );
 } // ?{}
+void ^?{}( ifstream & is ) {
+        close( is );
+} // ^?{}
+void nlOn( ifstream & os ) { os.$nlOnOff = true; }
+void nlOff( ifstream & os ) { os.$nlOnOff = false; }
+bool getANL( ifstream & os ) { return os.$nlOnOff; }
+void nlOn( ifstream & os ) { os.nlOnOff = true; }
+void nlOff( ifstream & os ) { os.nlOnOff = false; }
+bool getANL( ifstream & os ) { return os.nlOnOff; }
 int fail( ifstream & is ) {
         return is.$file == 0p || ferror( (FILE *)(is.$file) );
+        return is.file == 0 || ferror( (FILE *)(is.file) );
 } // fail
 int eof( ifstream & is ) {
         return feof( (FILE *)(is.$file) );
+        return feof( (FILE *)(is.file) );
 } // eof
 void open( ifstream & is, const char name[], const char mode[] ) {
+void open( ifstream & is, const char * name, const char * mode ) {
         FILE * file = fopen( name, mode );
         #ifdef __CFA_DEBUG__
         if ( file == 0p ) {
+        if ( file == 0 ) {
                 abort | IO_MSG "open input file \"" | name | "\"" | nl | strerror( errno );
         } // if
         #endif // __CFA_DEBUG__
         is.$file = file;
 } // open
 void open( ifstream & is, const char name[] ) {
+        is.file = file;
+} // open
+void open( ifstream & is, const char * name ) {
         open( is, name, "r" );
 } // open
 void close( ifstream & is ) {
         if ( (FILE *)(is.$file) == stdin ) return;
         if ( fclose( (FILE *)(is.$file) ) == EOF ) {
+        if ( (FILE *)(is.file) == stdin ) return;
+        if ( fclose( (FILE *)(is.file) ) == EOF ) {
                 abort | IO_MSG "close input" | nl | strerror( errno );
         } // if
 …
         } // if
         if ( fread( data, size, 1, (FILE *)(is.$file) ) == 0 ) {
+        if ( fread( data, size, 1, (FILE *)(is.file) ) == 0 ) {
                 abort | IO_MSG "read" | nl | strerror( errno );
         } // if
 …
         } // if
         if ( ungetc( c, (FILE *)(is.$file) ) == EOF ) {
+        if ( ungetc( c, (FILE *)(is.file) ) == EOF ) {
                 abort | IO_MSG "ungetc" | nl | strerror( errno );
         } // if
 …
         va_start( args, format );
         int len = vfscanf( (FILE *)(is.$file), format, args );
+        int len = vfscanf( (FILE *)(is.file), format, args );
         if ( len == EOF ) {
                 if ( ferror( (FILE *)(is.$file) ) ) {
+                if ( ferror( (FILE *)(is.file) ) ) {
                         abort | IO_MSG "invalid read";
                 } // if

libcfa/src/fstream.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Feb 17 08:29:23 2020
 // Update Count     : 175
+// Last Modified On : Mon Jul 15 18:10:23 2019
+// Update Count     : 167
 //
 …
 enum { sepSize = 16 };
 struct ofstream {
         void * $file;
         bool $sepDefault;
         bool $sepOnOff;
         bool $nlOnOff;
         bool $prt;                                                                                      // print text
         bool $sawNL;
         const char * $sepCur;
         char $separator[sepSize];
         char $tupleSeparator[sepSize];
+        void * file;
+        bool sepDefault;
+        bool sepOnOff;
+        bool nlOnOff;
+        bool prt;                                                                                       // print text
+        bool sawNL;
+        const char * sepCur;
+        char separator[sepSize];
+        char tupleSeparator[sepSize];
 }; // ofstream
 // private
 bool $sepPrt( ofstream & );
 void $sepReset( ofstream & );
 void $sepReset( ofstream &, bool );
 const char * $sepGetCur( ofstream & );
 void $sepSetCur( ofstream &, const char [] );
 bool $getNL( ofstream & );
 void $setNL( ofstream &, bool );
 bool $getANL( ofstream & );
 bool $getPrt( ofstream & );
 void $setPrt( ofstream &, bool );
+bool sepPrt( ofstream & );
+void sepReset( ofstream & );
+void sepReset( ofstream &, bool );
+const char * sepGetCur( ofstream & );
+void sepSetCur( ofstream &, const char * );
+bool getNL( ofstream & );
+void setNL( ofstream &, bool );
+bool getANL( ofstream & );
+bool getPrt( ofstream & );
+void setPrt( ofstream &, bool );
 // public
 …
 const char * sepGet( ofstream & );
 void sepSet( ofstream &, const char [] );
+void sepSet( ofstream &, const char * );
 const char * sepGetTuple( ofstream & );
 void sepSetTuple( ofstream &, const char [] );
+void sepSetTuple( ofstream &, const char * );
 void ends( ofstream & os );
 int fail( ofstream & );
 int flush( ofstream & );
 void open( ofstream &, const char name[], const char mode[] );
 void open( ofstream &, const char name[] );
+void open( ofstream &, const char * name, const char * mode );
+void open( ofstream &, const char * name );
 void close( ofstream & );
 ofstream & write( ofstream &, const char data[], size_t size );
 int fmt( ofstream &, const char format[], ... ) __attribute__(( format(printf, 2, 3) ));
+ofstream & write( ofstream &, const char * data, size_t size );
+int fmt( ofstream &, const char format[], ... );
 void ?{}( ofstream & os );
+void ?{}( ofstream & os, const char name[], const char mode[] );
+void ?{}( ofstream & os, const char name[] );
+void ^?{}( ofstream & os );
+void ?{}( ofstream & os, const char * name, const char * mode );
+void ?{}( ofstream & os, const char * name );
 extern ofstream & sout, & stdout, & serr, & stderr;             // aliases
 …
 struct ifstream {
         void * $file;
         bool $nlOnOff;
+        void * file;
+        bool nlOnOff;
 }; // ifstream
 …
 int fail( ifstream & is );
 int eof( ifstream & is );
 void open( ifstream & is, const char name[], const char mode[] );
 void open( ifstream & is, const char name[] );
+void open( ifstream & is, const char * name, const char * mode );
+void open( ifstream & is, const char * name );
 void close( ifstream & is );
 ifstream & read( ifstream & is, char * data, size_t size );
 ifstream & ungetc( ifstream & is, char c );
 int fmt( ifstream &, const char format[], ... ) __attribute__(( format(scanf, 2, 3) ));
+int fmt( ifstream &, const char format[], ... );
 void ?{}( ifstream & is );
+void ?{}( ifstream & is, const char name[], const char mode[] );
+void ?{}( ifstream & is, const char name[] );
+void ^?{}( ifstream & is );
+void ?{}( ifstream & is, const char * name, const char * mode );
+void ?{}( ifstream & is, const char * name );
 extern ifstream & sin, & stdin;                                                 // aliases

libcfa/src/gmp.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Apr 19 08:43:43 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb  9 09:56:54 2020
 // Update Count     : 31
+// Last Modified On : Sat Jul 13 15:25:05 2019
+// Update Count     : 27
 //
 …
 static inline {
         // constructor, zero_t/one_t are unnecessary because of relationship with signed/unsigned int
+        // constructor
         void ?{}( Int & this ) { mpz_init( this.mpz ); }
         void ?{}( Int & this, Int init ) { mpz_init_set( this.mpz, init.mpz ); }
+        void ?{}( Int & this, zero_t ) { mpz_init_set_si( this.mpz, 0 ); }
+        void ?{}( Int & this, one_t ) { mpz_init_set_si( this.mpz, 1 ); }
         void ?{}( Int & this, signed long int init ) { mpz_init_set_si( this.mpz, init ); }
         void ?{}( Int & this, unsigned long int init ) { mpz_init_set_ui( this.mpz, init ); }
         void ?{}( Int & this, const char val[] ) { if ( mpz_init_set_str( this.mpz, val, 0 ) ) abort(); }
+        void ?{}( Int & this, const char * val ) { if ( mpz_init_set_str( this.mpz, val, 0 ) ) abort(); }
         void ^?{}( Int & this ) { mpz_clear( this.mpz ); }
 …
         Int ?`mp( signed long int init ) { return (Int){ init }; }
         Int ?`mp( unsigned long int init ) { return (Int){ init }; }
         Int ?`mp( const char init[] ) { return (Int){ init }; }
+        Int ?`mp( const char * init ) { return (Int){ init }; }
         // assignment
 …
         Int ?=?( Int & lhs, long int rhs ) { mpz_set_si( lhs.mpz, rhs ); return lhs; }
         Int ?=?( Int & lhs, unsigned long int rhs ) { mpz_set_ui( lhs.mpz, rhs ); return lhs; }
         Int ?=?( Int & lhs, const char rhs[] ) { if ( mpz_set_str( lhs.mpz, rhs, 0 ) ) { abort | "invalid string conversion"; } return lhs; }
+        Int ?=?( Int & lhs, const char * rhs ) { if ( mpz_set_str( lhs.mpz, rhs, 0 ) ) { abort | "invalid string conversion"; } return lhs; }
         char ?=?( char & lhs, Int rhs ) { char val = mpz_get_si( rhs.mpz ); lhs = val; return lhs; }
 …
         forall( dtype ostype | ostream( ostype ) ) {
                 ostype & ?|?( ostype & os, Int mp ) {
                         if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                        if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                         gmp_printf( "%Zd", mp.mpz );
                         sepOn( os );

libcfa/src/heap.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Dec 19 21:58:35 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Apr  1 15:59:53 2020
 // Update Count     : 692
+// Last Modified On : Fri Oct 18 07:42:09 2019
+// Update Count     : 556
 //
 …
 #include <stdio.h>                                                                              // snprintf, fileno
 #include <errno.h>                                                                              // errno
-#include <string.h>                                                                             // memset, memcpy
 extern "C" {
 #include <sys/mman.h>                                                                   // mmap, munmap
 …
 #include "bits/locks.hfa"                                                               // __spinlock_t
 #include "startup.hfa"                                                                  // STARTUP_PRIORITY_MEMORY
 //#include "stdlib.hfa"                                                                 // bsearchl
+#include "stdlib.hfa"                                                                   // bsearchl
 #include "malloc.h"
-#define MIN(x, y) (y > x ? x : y)
 static bool traceHeap = false;
+inline bool traceHeap() { return traceHeap; }
+inline bool traceHeap() {
+        return traceHeap;
+} // traceHeap
 bool traceHeapOn() {
 …
 } // traceHeapOff
+bool traceHeapTerm() { return false; }
+static bool prtFree = false;
+inline bool prtFree() {
+        return prtFree;
+} // prtFree
+bool prtFreeOn() {
+        bool temp = prtFree;
+        prtFree = true;
+static bool checkFree = false;
+inline bool checkFree() {
+        return checkFree;
+} // checkFree
+bool checkFreeOn() {
+        bool temp = checkFree;
+        checkFree = true;
         return temp;
 } // prtFreeOn
 bool prtFreeOff() {
         bool temp = prtFree;
         prtFree = false;
+} // checkFreeOn
+bool checkFreeOff() {
+        bool temp = checkFree;
+        checkFree = false;
         return temp;
+} // prtFreeOff
+} // checkFreeOff
+// static bool traceHeapTerm = false;
+// inline bool traceHeapTerm() {
+//      return traceHeapTerm;
+// } // traceHeapTerm
+// bool traceHeapTermOn() {
+//      bool temp = traceHeapTerm;
+//      traceHeapTerm = true;
+//      return temp;
+// } // traceHeapTermOn
+// bool traceHeapTermOff() {
+//      bool temp = traceHeapTerm;
+//      traceHeapTerm = false;
+//      return temp;
+// } // traceHeapTermOff
 enum {
+        // Define the default extension heap amount in units of bytes. When the uC++ supplied heap reaches the brk address,
+        // the brk address is extended by the extension amount.
+        __CFA_DEFAULT_MMAP_START__ = (512 * 1024 + 1),
         __CFA_DEFAULT_HEAP_EXPANSION__ = (1 * 1024 * 1024),
-        // Define the mmap crossover point during allocation. Allocations less than this amount are allocated from buckets;
-        // values greater than or equal to this value are mmap from the operating system.
-        __CFA_DEFAULT_MMAP_START__ = (512 * 1024 + 1),
 };
 …
 static unsigned int allocFree;                                                  // running total of allocations minus frees
 static void prtUnfreed() {
+static void checkUnfreed() {
         if ( allocFree != 0 ) {
                 // DO NOT USE STREAMS AS THEY MAY BE UNAVAILABLE AT THIS POINT.
                 char helpText[512];
                 int len = snprintf( helpText, sizeof(helpText), "CFA warning (UNIX pid:%ld) : program terminating with %u(0x%x) bytes of storage allocated but not freed.\n"
                                                         "Possible cause is unfreed storage allocated by the program or system/library routines called from the program.\n",
                                                         (long int)getpid(), allocFree, allocFree ); // always print the UNIX pid
                 __cfaabi_bits_write( STDERR_FILENO, helpText, len ); // print debug/nodebug
         } // if
 } // prtUnfreed
+                // char helpText[512];
+                // int len = snprintf( helpText, sizeof(helpText), "CFA warning (UNIX pid:%ld) : program terminating with %u(0x%x) bytes of storage allocated but not freed.\n"
+                //                                      "Possible cause is unfreed storage allocated by the program or system/library routines called from the program.\n",
+                //                                      (long int)getpid(), allocFree, allocFree ); // always print the UNIX pid
+                // __cfaabi_dbg_bits_write( helpText, len );
+        } // if
+} // checkUnfreed
 extern "C" {
 …
         void heapAppStop() {                                                            // called by __cfaabi_appready_startdown
                 fclose( stdin ); fclose( stdout );
                 prtUnfreed();
+                checkUnfreed();
         } // heapAppStop
 } // extern "C"
 #endif // __CFA_DEBUG__
 // statically allocated variables => zero filled.
 …
 static unsigned int maxBucketsUsed;                                             // maximum number of buckets in use
+// #comment TD : This defined is significantly different from the __ALIGN__ define from locks.hfa
+#define ALIGN 16
 #define SPINLOCK 0
 …
 // Recursive definitions: HeapManager needs size of bucket array and bucket area needs sizeof HeapManager storage.
 // Break recusion by hardcoding number of buckets and statically checking number is correct after bucket array defined.
 enum { NoBucketSizes = 91 };                                                    // number of buckets sizes
+enum { NoBucketSizes = 93 };                                                    // number of buckets sizes
 struct HeapManager {
 …
                                                         union {
 //                                                              FreeHeader * home;              // allocated block points back to home locations (must overlay alignment)
-                                                                // 2nd low-order bit => zero filled
                                                                 void * home;                    // allocated block points back to home locations (must overlay alignment)
                                                                 size_t blockSize;               // size for munmap (must overlay alignment)
 …
                                 struct FakeHeader {
                                         #if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
+                                        // 1st low-order bit => fake header & alignment
+                                        uint32_t alignment;
+                                        uint32_t alignment;                                     // low-order bits of home/blockSize used for tricks
                                         #endif // __ORDER_LITTLE_ENDIAN__
 …
                                 } fake; // FakeHeader
                         } kind; // Kind
-                        uint32_t dimension;                                                     // used by calloc-like to remember number of array elements
                 } header; // Header
                 char pad[libAlign() - sizeof( Header )];
+                char pad[ALIGN - sizeof( Header )];
                 char data[0];                                                                   // storage
         }; // Storage
         static_assert( libAlign() >= sizeof( Storage ), "libAlign() < sizeof( Storage )" );
+        static_assert( ALIGN >= sizeof( Storage ), "ALIGN < sizeof( Storage )" );
         struct FreeHeader {
 …
 #define __STATISTICS__
-// Bucket size must be multiple of 16.
 // Powers of 2 are common allocation sizes, so make powers of 2 generate the minimum required size.
 static const unsigned int bucketSizes[] @= {                    // different bucket sizes
+, 32, 48, 64 + sizeof(HeapManager.Storage), // 4
+, 112, 128 + sizeof(HeapManager.Storage), // 3
+, 192, 224, 256 + sizeof(HeapManager.Storage), // 4
+, 384, 448, 512 + sizeof(HeapManager.Storage), // 4
+, 768, 896, 1_024 + sizeof(HeapManager.Storage), // 4
+_536, 2_048 + sizeof(HeapManager.Storage), // 2
+_560, 3_072, 3_584, 4_096 + sizeof(HeapManager.Storage), // 4
+_144, 8_192 + sizeof(HeapManager.Storage), // 2
+_216, 10_240, 11_264, 12_288, 13_312, 14_336, 15_360, 16_384 + sizeof(HeapManager.Storage), // 8
+_432, 20_480, 22_528, 24_576, 26_624, 28_672, 30_720, 32_768 + sizeof(HeapManager.Storage), // 8
+_864, 40_960, 45_056, 49_152, 53_248, 57_344, 61_440, 65_536 + sizeof(HeapManager.Storage), // 8
+_728, 81_920, 90_112, 98_304, 106_496, 114_688, 122_880, 131_072 + sizeof(HeapManager.Storage), // 8
+_456, 163_840, 180_224, 196_608, 212_992, 229_376, 245_760, 262_144 + sizeof(HeapManager.Storage), // 8
+_912, 327_680, 360_448, 393_216, 425_984, 458_752, 491_520, 524_288 + sizeof(HeapManager.Storage), // 8
+_360, 786_432, 917_504, 1_048_576 + sizeof(HeapManager.Storage), // 4
+_179_648, 1_310_720, 1_441_792, 1_572_864, 1_703_936, 1_835_008, 1_966_080, 2_097_152 + sizeof(HeapManager.Storage), // 8
+_621_440, 3_145_728, 3_670_016, 4_194_304 + sizeof(HeapManager.Storage), // 4
+, 32, 48, 64,
++ sizeof(HeapManager.Storage), 96, 112, 128, 128 + sizeof(HeapManager.Storage), 160, 192, 224,
++ sizeof(HeapManager.Storage), 320, 384, 448, 512 + sizeof(HeapManager.Storage), 640, 768, 896,
+_024 + sizeof(HeapManager.Storage), 1_536, 2_048 + sizeof(HeapManager.Storage), 2_560, 3_072, 3_584, 4_096 + sizeof(HeapManager.Storage), 6_144,
+_192 + sizeof(HeapManager.Storage), 9_216, 10_240, 11_264, 12_288, 13_312, 14_336, 15_360,
+_384 + sizeof(HeapManager.Storage), 18_432, 20_480, 22_528, 24_576, 26_624, 28_672, 30_720,
+_768 + sizeof(HeapManager.Storage), 36_864, 40_960, 45_056, 49_152, 53_248, 57_344, 61_440,
+_536 + sizeof(HeapManager.Storage), 73_728, 81_920, 90_112, 98_304, 106_496, 114_688, 122_880,
+_072 + sizeof(HeapManager.Storage), 147_456, 163_840, 180_224, 196_608, 212_992, 229_376, 245_760,
+_144 + sizeof(HeapManager.Storage), 294_912, 327_680, 360_448, 393_216, 425_984, 458_752, 491_520,
+_288 + sizeof(HeapManager.Storage), 655_360, 786_432, 917_504, 1_048_576 + sizeof(HeapManager.Storage), 1_179_648, 1_310_720, 1_441_792,
+_572_864, 1_703_936, 1_835_008, 1_966_080, 2_097_152 + sizeof(HeapManager.Storage), 2_621_440, 3_145_728, 3_670_016,
+_194_304 + sizeof(HeapManager.Storage)
 };
 …
 static unsigned char lookup[LookupSizes];                               // O(1) lookup for small sizes
 #endif // FASTLOOKUP
 static int mmapFd = -1;                                                                 // fake or actual fd for anonymous file
 #ifdef __CFA_DEBUG__
 static bool heapBoot = 0;                                                               // detect recursion during boot
 …
 static HeapManager heapManager __attribute__(( aligned (128) )) @= {}; // size of cache line to prevent false sharing
+// #comment TD : The return type of this function should be commented
+static inline bool setMmapStart( size_t value ) {
+  if ( value < pageSize || bucketSizes[NoBucketSizes - 1] < value ) return true;
+        mmapStart = value;                                                                      // set global
+        // find the closest bucket size less than or equal to the mmapStart size
+        maxBucketsUsed = bsearchl( (unsigned int)mmapStart, bucketSizes, NoBucketSizes ); // binary search
+        assert( maxBucketsUsed < NoBucketSizes );                       // subscript failure ?
+        assert( mmapStart <= bucketSizes[maxBucketsUsed] ); // search failure ?
+        return false;
+} // setMmapStart
+static void ?{}( HeapManager & manager ) with ( manager ) {
+        pageSize = sysconf( _SC_PAGESIZE );
+        for ( unsigned int i = 0; i < NoBucketSizes; i += 1 ) { // initialize the free lists
+                freeLists[i].blockSize = bucketSizes[i];
+        } // for
+        #ifdef FASTLOOKUP
+        unsigned int idx = 0;
+        for ( unsigned int i = 0; i < LookupSizes; i += 1 ) {
+                if ( i > bucketSizes[idx] ) idx += 1;
+                lookup[i] = idx;
+        } // for
+        #endif // FASTLOOKUP
+        if ( setMmapStart( default_mmap_start() ) ) {
+                abort( "HeapManager : internal error, mmap start initialization failure." );
+        } // if
+        heapExpand = default_heap_expansion();
+        char * End = (char *)sbrk( 0 );
+        sbrk( (char *)libCeiling( (long unsigned int)End, libAlign() ) - End ); // move start of heap to multiple of alignment
+        heapBegin = heapEnd = sbrk( 0 );                                        // get new start point
+} // HeapManager
+static void ^?{}( HeapManager & ) {
+        #ifdef __STATISTICS__
+        // if ( traceHeapTerm() ) {
+        //      printStats();
+        //      if ( checkfree() ) checkFree( heapManager, true );
+        // } // if
+        #endif // __STATISTICS__
+} // ~HeapManager
+static void memory_startup( void ) __attribute__(( constructor( STARTUP_PRIORITY_MEMORY ) ));
+void memory_startup( void ) {
+        #ifdef __CFA_DEBUG__
+        if ( unlikely( heapBoot ) ) {                                           // check for recursion during system boot
+                // DO NOT USE STREAMS AS THEY MAY BE UNAVAILABLE AT THIS POINT.
+                abort( "boot() : internal error, recursively invoked during system boot." );
+        } // if
+        heapBoot = true;
+        #endif // __CFA_DEBUG__
+        //assert( heapManager.heapBegin != 0 );
+        //heapManager{};
+        if ( heapManager.heapBegin == 0 ) heapManager{};
+} // memory_startup
+static void memory_shutdown( void ) __attribute__(( destructor( STARTUP_PRIORITY_MEMORY ) ));
+void memory_shutdown( void ) {
+        ^heapManager{};
+} // memory_shutdown
 #ifdef __STATISTICS__
+// Heap statistics counters.
+static unsigned long long int mmap_storage;
+static unsigned long long int mmap_storage;                             // heap statistics counters
 static unsigned int mmap_calls;
 static unsigned long long int munmap_storage;
 …
 static unsigned long long int cmemalign_storage;
 static unsigned int cmemalign_calls;
-static unsigned long long int resize_storage;
-static unsigned int resize_calls;
 static unsigned long long int realloc_storage;
 static unsigned int realloc_calls;
+// Statistics file descriptor (changed by malloc_stats_fd).
+static int statfd = STDERR_FILENO;                                              // default stderr
+static int statfd;                                                                              // statistics file descriptor (changed by malloc_stats_fd)
 // Use "write" because streams may be shutdown when calls are made.
 static void printStats() {
         char helpText[512];
         __cfaabi_bits_print_buffer( STDERR_FILENO, helpText, sizeof(helpText),
+        __cfaabi_dbg_bits_print_buffer( helpText, sizeof(helpText),
                                                                         "\nHeap statistics:\n"
                                                                         "  malloc: calls %u / storage %llu\n"
 …
                                                                         "  memalign: calls %u / storage %llu\n"
                                                                         "  cmemalign: calls %u / storage %llu\n"
-                                                                        "  resize: calls %u / storage %llu\n"
                                                                         "  realloc: calls %u / storage %llu\n"
                                                                         "  free: calls %u / storage %llu\n"
 …
                                                                         memalign_calls, memalign_storage,
                                                                         cmemalign_calls, cmemalign_storage,
-                                                                        resize_calls, resize_storage,
                                                                         realloc_calls, realloc_storage,
                                                                         free_calls, free_storage,
 …
                                                 "<total type=\"memalign\" count=\"%u\" size=\"%llu\"/>\n"
                                                 "<total type=\"cmemalign\" count=\"%u\" size=\"%llu\"/>\n"
-                                                "<total type=\"resize\" count=\"%u\" size=\"%llu\"/>\n"
                                                 "<total type=\"realloc\" count=\"%u\" size=\"%llu\"/>\n"
                                                 "<total type=\"free\" count=\"%u\" size=\"%llu\"/>\n"
 …
                                                 memalign_calls, memalign_storage,
                                                 cmemalign_calls, cmemalign_storage,
-                                                resize_calls, resize_storage,
                                                 realloc_calls, realloc_storage,
                                                 free_calls, free_storage,
 …
                                                 sbrk_calls, sbrk_storage
                 );
+        __cfaabi_bits_write( fileno( stream ), helpText, len ); // ensures all bytes written or exit
+        return len;
+        return write( fileno( stream ), helpText, len );        // -1 => error
 } // printStatsXML
 #endif // __STATISTICS__
+// static inline void noMemory() {
+//      abort( "Heap memory exhausted at %zu bytes.\n"
+//                 "Possible cause is very large memory allocation and/or large amount of unfreed storage allocated by the program or system/library routines.",
+//                 ((char *)(sbrk( 0 )) - (char *)(heapManager.heapBegin)) );
+// } // noMemory
+// #comment TD : Is this the samething as Out-of-Memory?
+static inline void noMemory() {
+        abort( "Heap memory exhausted at %zu bytes.\n"
+                   "Possible cause is very large memory allocation and/or large amount of unfreed storage allocated by the program or system/library routines.",
+                   ((char *)(sbrk( 0 )) - (char *)(heapManager.heapBegin)) );
+} // noMemory
+static inline void checkAlign( size_t alignment ) {
+        if ( alignment < sizeof(void *) || ! libPow2( alignment ) ) {
+                abort( "Alignment %zu for memory allocation is less than sizeof(void *) and/or not a power of 2.", alignment );
+        } // if
+} // checkAlign
 …
+// thunk problem
+static inline void checkHeader( bool check, const char * name, void * addr ) {
+        if ( unlikely( check ) ) {                                                      // bad address ?
+                abort( "Attempt to %s storage %p with address outside the heap.\n"
+                           "Possible cause is duplicate free on same block or overwriting of memory.",
+                           name, addr );
+        } // if
+} // checkHeader
+// #comment TD : function should be commented and/or have a more evocative name
+//               this isn't either a check or a constructor which is what I would expect this function to be
+static inline void fakeHeader( HeapManager.Storage.Header *& header, size_t & size, size_t & alignment ) {
+        if ( unlikely( (header->kind.fake.alignment & 1) == 1 ) ) { // fake header ?
+                size_t offset = header->kind.fake.offset;
+                alignment = header->kind.fake.alignment & -2;   // remove flag from value
+                #ifdef __CFA_DEBUG__
+                checkAlign( alignment );                                                // check alignment
+                #endif // __CFA_DEBUG__
+                header = (HeapManager.Storage.Header *)((char *)header - offset);
+        } // if
+} // fakeHeader
+// #comment TD : Why is this a define
+#define headerAddr( addr ) ((HeapManager.Storage.Header *)( (char *)addr - sizeof(HeapManager.Storage) ))
+static inline bool headers( const char * name, void * addr, HeapManager.Storage.Header *& header, HeapManager.FreeHeader *& freeElem, size_t & size, size_t & alignment ) with ( heapManager ) {
+        header = headerAddr( addr );
+        if ( unlikely( heapEnd < addr ) ) {                                     // mmapped ?
+                fakeHeader( header, size, alignment );
+                size = header->kind.real.blockSize & -3;                // mmap size
+                return true;
+        } // if
+        #ifdef __CFA_DEBUG__
+        checkHeader( addr < heapBegin || header < (HeapManager.Storage.Header *)heapBegin, name, addr ); // bad low address ?
+        #endif // __CFA_DEBUG__
+        // #comment TD : This code looks weird...
+        //               It's called as the first statement of both branches of the last if, with the same parameters in all cases
+        // header may be safe to dereference
+        fakeHeader( header, size, alignment );
+        #ifdef __CFA_DEBUG__
+        checkHeader( header < (HeapManager.Storage.Header *)heapBegin || (HeapManager.Storage.Header *)heapEnd < header, name, addr ); // bad address ? (offset could be + or -)
+        #endif // __CFA_DEBUG__
+        freeElem = (HeapManager.FreeHeader *)((size_t)header->kind.real.home & -3);
+        #ifdef __CFA_DEBUG__
+        if ( freeElem < &freeLists[0] || &freeLists[NoBucketSizes] <= freeElem ) {
+                abort( "Attempt to %s storage %p with corrupted header.\n"
+                           "Possible cause is duplicate free on same block or overwriting of header information.",
+                           name, addr );
+        } // if
+        #endif // __CFA_DEBUG__
+        size = freeElem->blockSize;
+        return false;
+} // headers
+static inline void * extend( size_t size ) with ( heapManager ) {
+        lock( extlock __cfaabi_dbg_ctx2 );
+        ptrdiff_t rem = heapRemaining - size;
+        if ( rem < 0 ) {
+                // If the size requested is bigger than the current remaining storage, increase the size of the heap.
+                size_t increase = libCeiling( size > heapExpand ? size : heapExpand, libAlign() );
+                if ( sbrk( increase ) == (void *)-1 ) {
+                        unlock( extlock );
+                        errno = ENOMEM;
+                        return 0;
+                } // if
+                #ifdef __STATISTICS__
+                sbrk_calls += 1;
+                sbrk_storage += increase;
+                #endif // __STATISTICS__
+                #ifdef __CFA_DEBUG__
+                // Set new memory to garbage so subsequent uninitialized usages might fail.
+                memset( (char *)heapEnd + heapRemaining, '\377', increase );
+                #endif // __CFA_DEBUG__
+                rem = heapRemaining + increase - size;
+        } // if
+        HeapManager.Storage * block = (HeapManager.Storage *)heapEnd;
+        heapRemaining = rem;
+        heapEnd = (char *)heapEnd + size;
+        unlock( extlock );
+        return block;
+} // extend
 size_t Bsearchl( unsigned int key, const unsigned int * vals, size_t dim ) {
         size_t l = 0, m, h = dim;
 …
-static inline bool setMmapStart( size_t value ) {               // true => mmapped, false => sbrk
-  if ( value < pageSize || bucketSizes[NoBucketSizes - 1] < value ) return true;
-        mmapStart = value;                                                                      // set global
-        // find the closest bucket size less than or equal to the mmapStart size
-        maxBucketsUsed = Bsearchl( (unsigned int)mmapStart, bucketSizes, NoBucketSizes ); // binary search
-        assert( maxBucketsUsed < NoBucketSizes );                       // subscript failure ?
-        assert( mmapStart <= bucketSizes[maxBucketsUsed] ); // search failure ?
-        return false;
-} // setMmapStart
-// <-------+----------------------------------------------------> bsize (bucket size)
-// |header |addr
-//==================================================================================
-//                   align/offset |
-// <-----------------<------------+-----------------------------> bsize (bucket size)
-//                   |fake-header | addr
-#define headerAddr( addr ) ((HeapManager.Storage.Header *)( (char *)addr - sizeof(HeapManager.Storage) ))
-#define realHeader( header ) ((HeapManager.Storage.Header *)((char *)header - header->kind.fake.offset))
-// <-------<<--------------------- dsize ---------------------->> bsize (bucket size)
-// |header |addr
-//==================================================================================
-//                   align/offset |
-// <------------------------------<<---------- dsize --------->>> bsize (bucket size)
-//                   |fake-header |addr
-#define dataStorage( bsize, addr, header ) (bsize - ( (char *)addr - (char *)header ))
-static inline void checkAlign( size_t alignment ) {
-        if ( alignment < libAlign() || ! libPow2( alignment ) ) {
-                abort( "Alignment %zu for memory allocation is less than %d and/or not a power of 2.", alignment, libAlign() );
-        } // if
-} // checkAlign
-static inline void checkHeader( bool check, const char name[], void * addr ) {
-        if ( unlikely( check ) ) {                                                      // bad address ?
-                abort( "Attempt to %s storage %p with address outside the heap.\n"
-                           "Possible cause is duplicate free on same block or overwriting of memory.",
-                           name, addr );
-        } // if
-} // checkHeader
-static inline void fakeHeader( HeapManager.Storage.Header *& header, size_t & alignment ) {
-        if ( unlikely( (header->kind.fake.alignment & 1) == 1 ) ) { // fake header ?
-                alignment = header->kind.fake.alignment & -2;   // remove flag from value
-                #ifdef __CFA_DEBUG__
-                checkAlign( alignment );                                                // check alignment
-                #endif // __CFA_DEBUG__
-                header = realHeader( header );                                  // backup from fake to real header
-        } // if
-} // fakeHeader
-static inline bool headers( const char name[] __attribute__(( unused )), void * addr, HeapManager.Storage.Header *& header, HeapManager.FreeHeader *& freeElem, size_t & size, size_t & alignment ) with ( heapManager ) {
-        header = headerAddr( addr );
-        if ( unlikely( heapEnd < addr ) ) {                                     // mmapped ?
-                fakeHeader( header, alignment );
-                size = header->kind.real.blockSize & -3;                // mmap size
-                return true;
-        } // if
-        #ifdef __CFA_DEBUG__
-        checkHeader( addr < heapBegin || header < (HeapManager.Storage.Header *)heapBegin, name, addr ); // bad low address ?
-        #endif // __CFA_DEBUG__
-        // header may be safe to dereference
-        fakeHeader( header, alignment );
-        #ifdef __CFA_DEBUG__
-        checkHeader( header < (HeapManager.Storage.Header *)heapBegin || (HeapManager.Storage.Header *)heapEnd < header, name, addr ); // bad address ? (offset could be + or -)
-        #endif // __CFA_DEBUG__
-        freeElem = (HeapManager.FreeHeader *)((size_t)header->kind.real.home & -3);
-        #ifdef __CFA_DEBUG__
-        if ( freeElem < &freeLists[0] || &freeLists[NoBucketSizes] <= freeElem ) {
-                abort( "Attempt to %s storage %p with corrupted header.\n"
-                           "Possible cause is duplicate free on same block or overwriting of header information.",
-                           name, addr );
-        } // if
-        #endif // __CFA_DEBUG__
-        size = freeElem->blockSize;
-        return false;
-} // headers
-static inline void * extend( size_t size ) with ( heapManager ) {
-        lock( extlock __cfaabi_dbg_ctx2 );
-        ptrdiff_t rem = heapRemaining - size;
-        if ( rem < 0 ) {
-                // If the size requested is bigger than the current remaining storage, increase the size of the heap.
-                size_t increase = libCeiling( size > heapExpand ? size : heapExpand, libAlign() );
-                if ( sbrk( increase ) == (void *)-1 ) {
-                        unlock( extlock );
-                        errno = ENOMEM;
-                        return 0p;
-                } // if
-                #ifdef __STATISTICS__
-                sbrk_calls += 1;
-                sbrk_storage += increase;
-                #endif // __STATISTICS__
-                #ifdef __CFA_DEBUG__
-                // Set new memory to garbage so subsequent uninitialized usages might fail.
-                memset( (char *)heapEnd + heapRemaining, '\377', increase );
-                #endif // __CFA_DEBUG__
-                rem = heapRemaining + increase - size;
-        } // if
-        HeapManager.Storage * block = (HeapManager.Storage *)heapEnd;
-        heapRemaining = rem;
-        heapEnd = (char *)heapEnd + size;
-        unlock( extlock );
-        return block;
-} // extend
 static inline void * doMalloc( size_t size ) with ( heapManager ) {
         HeapManager.Storage * block;                                            // pointer to new block of storage
 …
         // along with the block and is a multiple of the alignment size.
   if ( unlikely( size > ~0ul - sizeof(HeapManager.Storage) ) ) return 0p;
+  if ( unlikely( size > ~0ul - sizeof(HeapManager.Storage) ) ) return 0;
         size_t tsize = size + sizeof(HeapManager.Storage);
         if ( likely( tsize < mmapStart ) ) {                            // small size => sbrk
 …
                 block = freeElem->freeList.pop();
                 #endif // SPINLOCK
                 if ( unlikely( block == 0p ) ) {                                // no free block ?
+                if ( unlikely( block == 0 ) ) {                                 // no free block ?
                         #if defined( SPINLOCK )
                         unlock( freeElem->lock );
 …
                         block = (HeapManager.Storage *)extend( tsize ); // mutual exclusion on call
   if ( unlikely( block == 0p ) ) return 0p;
                 #if defined( SPINLOCK )
+  if ( unlikely( block == 0 ) ) return 0;
+                        #if defined( SPINLOCK )
                 } else {
                         freeElem->freeList = block->header.kind.real.next;
                         unlock( freeElem->lock );
                 #endif // SPINLOCK
+                        #endif // SPINLOCK
                 } // if
                 block->header.kind.real.home = freeElem;                // pointer back to free list of apropriate size
         } else {                                                                                        // large size => mmap
   if ( unlikely( size > ~0ul - pageSize ) ) return 0p;
+  if ( unlikely( size > ~0ul - pageSize ) ) return 0;
                 tsize = libCeiling( tsize, pageSize );                  // must be multiple of page size
                 #ifdef __STATISTICS__
 …
         } // if
         void * addr = &(block->data);                                           // adjust off header to user bytes
+        void * area = &(block->data);                                           // adjust off header to user bytes
         #ifdef __CFA_DEBUG__
         assert( ((uintptr_t)addr & (libAlign() - 1)) == 0 ); // minimum alignment ?
+        assert( ((uintptr_t)area & (libAlign() - 1)) == 0 ); // minimum alignment ?
         __atomic_add_fetch( &allocFree, tsize, __ATOMIC_SEQ_CST );
         if ( traceHeap() ) {
                 enum { BufferSize = 64 };
                 char helpText[BufferSize];
                 int len = snprintf( helpText, BufferSize, "%p = Malloc( %zu ) (allocated %zu)\n", addr, size, tsize );
                 // int len = snprintf( helpText, BufferSize, "Malloc %p %zu\n", addr, size );
                 __cfaabi_bits_write( STDERR_FILENO, helpText, len ); // print debug/nodebug
+                int len = snprintf( helpText, BufferSize, "%p = Malloc( %zu ) (allocated %zu)\n", area, size, tsize );
+                // int len = snprintf( helpText, BufferSize, "Malloc %p %zu\n", area, size );
+                __cfaabi_dbg_bits_write( helpText, len );
         } // if
         #endif // __CFA_DEBUG__
         return addr;
+        return area;
 } // doMalloc
 …
 static inline void doFree( void * addr ) with ( heapManager ) {
         #ifdef __CFA_DEBUG__
         if ( unlikely( heapManager.heapBegin == 0p ) ) {
+        if ( unlikely( heapManager.heapBegin == 0 ) ) {
                 abort( "doFree( %p ) : internal error, called before heap is initialized.", addr );
         } // if
 …
                 char helpText[BufferSize];
                 int len = snprintf( helpText, sizeof(helpText), "Free( %p ) size:%zu\n", addr, size );
                 __cfaabi_bits_write( STDERR_FILENO, helpText, len ); // print debug/nodebug
+                __cfaabi_dbg_bits_write( helpText, len );
         } // if
         #endif // __CFA_DEBUG__
 …
 size_t prtFree( HeapManager & manager ) with ( manager ) {
+size_t checkFree( HeapManager & manager ) with ( manager ) {
         size_t total = 0;
         #ifdef __STATISTICS__
         __cfaabi_bits_acquire();
         __cfaabi_bits_print_nolock( STDERR_FILENO, "\nBin lists (bin size : free blocks on list)\n" );
+        __cfaabi_dbg_bits_acquire();
+        __cfaabi_dbg_bits_print_nolock( "\nBin lists (bin size : free blocks on list)\n" );
         #endif // __STATISTICS__
         for ( unsigned int i = 0; i < maxBucketsUsed; i += 1 ) {
 …
                 #if defined( SPINLOCK )
                 for ( HeapManager.Storage * p = freeLists[i].freeList; p != 0p; p = p->header.kind.real.next ) {
+                for ( HeapManager.Storage * p = freeLists[i].freeList; p != 0; p = p->header.kind.real.next ) {
                 #else
                 for ( HeapManager.Storage * p = freeLists[i].freeList.top(); p != 0p; p = p->header.kind.real.next.top ) {
+                for ( HeapManager.Storage * p = freeLists[i].freeList.top(); p != 0; p = p->header.kind.real.next.top ) {
                 #endif // SPINLOCK
                         total += size;
 …
                 #ifdef __STATISTICS__
                 __cfaabi_bits_print_nolock( STDERR_FILENO, "%7zu, %-7u  ", size, N );
                 if ( (i + 1) % 8 == 0 ) __cfaabi_bits_print_nolock( STDERR_FILENO, "\n" );
+                __cfaabi_dbg_bits_print_nolock( "%7zu, %-7u  ", size, N );
+                if ( (i + 1) % 8 == 0 ) __cfaabi_dbg_bits_print_nolock( "\n" );
                 #endif // __STATISTICS__
         } // for
         #ifdef __STATISTICS__
         __cfaabi_bits_print_nolock( STDERR_FILENO, "\ntotal free blocks:%zu\n", total );
         __cfaabi_bits_release();
+        __cfaabi_dbg_bits_print_nolock( "\ntotal free blocks:%zu\n", total );
+        __cfaabi_dbg_bits_release();
         #endif // __STATISTICS__
         return (char *)heapEnd - (char *)heapBegin - total;
+} // prtFree
+static void ?{}( HeapManager & manager ) with ( manager ) {
+        pageSize = sysconf( _SC_PAGESIZE );
+        for ( unsigned int i = 0; i < NoBucketSizes; i += 1 ) { // initialize the free lists
+                freeLists[i].blockSize = bucketSizes[i];
+        } // for
+        #ifdef FASTLOOKUP
+        unsigned int idx = 0;
+        for ( unsigned int i = 0; i < LookupSizes; i += 1 ) {
+                if ( i > bucketSizes[idx] ) idx += 1;
+                lookup[i] = idx;
+        } // for
+        #endif // FASTLOOKUP
+        if ( setMmapStart( default_mmap_start() ) ) {
+                abort( "HeapManager : internal error, mmap start initialization failure." );
+        } // if
+        heapExpand = default_heap_expansion();
+        char * end = (char *)sbrk( 0 );
+        sbrk( (char *)libCeiling( (long unsigned int)end, libAlign() ) - end ); // move start of heap to multiple of alignment
+        heapBegin = heapEnd = sbrk( 0 );                                        // get new start point
+} // HeapManager
+static void ^?{}( HeapManager & ) {
+        #ifdef __STATISTICS__
+        if ( traceHeapTerm() ) {
+                printStats();
+                // if ( prtfree() ) prtFree( heapManager, true );
+        } // if
+        #endif // __STATISTICS__
+} // ~HeapManager
+static void memory_startup( void ) __attribute__(( constructor( STARTUP_PRIORITY_MEMORY ) ));
+void memory_startup( void ) {
+        #ifdef __CFA_DEBUG__
+        if ( unlikely( heapBoot ) ) {                                           // check for recursion during system boot
+                // DO NOT USE STREAMS AS THEY MAY BE UNAVAILABLE AT THIS POINT.
+                abort( "boot() : internal error, recursively invoked during system boot." );
+        } // if
+        heapBoot = true;
+        #endif // __CFA_DEBUG__
+        //assert( heapManager.heapBegin != 0 );
+        //heapManager{};
+        if ( heapManager.heapBegin == 0p ) heapManager{};
+} // memory_startup
+static void memory_shutdown( void ) __attribute__(( destructor( STARTUP_PRIORITY_MEMORY ) ));
+void memory_shutdown( void ) {
+        ^heapManager{};
+} // memory_shutdown
+} // checkFree
 static inline void * mallocNoStats( size_t size ) {             // necessary for malloc statistics
         //assert( heapManager.heapBegin != 0 );
         if ( unlikely( heapManager.heapBegin == 0p ) ) heapManager{}; // called before memory_startup ?
         void * addr = doMalloc( size );
         if ( unlikely( addr == 0p ) ) errno = ENOMEM;           // POSIX
         return addr;
+        if ( unlikely( heapManager.heapBegin == 0 ) ) heapManager{}; // called before memory_startup ?
+        void * area = doMalloc( size );
+        if ( unlikely( area == 0 ) ) errno = ENOMEM;            // POSIX
+        return area;
 } // mallocNoStats
-static inline void * callocNoStats( size_t noOfElems, size_t elemSize ) {
-        size_t size = noOfElems * elemSize;
-        char * addr = (char *)mallocNoStats( size );
-  if ( unlikely( addr == 0p ) ) return 0p;
-        HeapManager.Storage.Header * header;
-        HeapManager.FreeHeader * freeElem;
-        size_t bsize, alignment;
-        bool mapped __attribute__(( unused )) = headers( "calloc", addr, header, freeElem, bsize, alignment );
-        #ifndef __CFA_DEBUG__
-        // Mapped storage is zero filled, but in debug mode mapped memory is scrubbed in doMalloc, so it has to be reset to zero.
-        if ( ! mapped )
-        #endif // __CFA_DEBUG__
-                // Zero entire data space even when > than size => realloc without a new allocation and zero fill works.
-                // <-------00000000000000000000000000000000000000000000000000000> bsize (bucket size)
-                // `-header`-addr                      `-size
-                memset( addr, '\0', bsize - sizeof(HeapManager.Storage) ); // set to zeros
-        assert( noOfElems <= UINT32_MAX );
-        header->dimension = noOfElems;                                          // store number of array elements
-        header->kind.real.blockSize |= 2;                                       // mark as zero filled
-        return addr;
-} // callocNoStats
 …
         // subtract libAlign() because it is already the minimum alignment
         // add sizeof(Storage) for fake header
+        char * addr = (char *)mallocNoStats( size + alignment - libAlign() + sizeof(HeapManager.Storage) );
+  if ( unlikely( addr == 0p ) ) return addr;
+        // #comment TD : this is the only place that calls doMalloc without calling mallocNoStats, why ?
+        char * area = (char *)doMalloc( size + alignment - libAlign() + sizeof(HeapManager.Storage) );
+  if ( unlikely( area == 0 ) ) return area;
         // address in the block of the "next" alignment address
         char * user = (char *)libCeiling( (uintptr_t)(addr + sizeof(HeapManager.Storage)), alignment );
+        char * user = (char *)libCeiling( (uintptr_t)(area + sizeof(HeapManager.Storage)), alignment );
         // address of header from malloc
         HeapManager.Storage.Header * realHeader = headerAddr( addr );
+        HeapManager.Storage.Header * realHeader = headerAddr( area );
         // address of fake header * before* the alignment location
         HeapManager.Storage.Header * fakeHeader = headerAddr( user );
 …
         return user;
 } // memalignNoStats
-static inline void * cmemalignNoStats( size_t alignment, size_t noOfElems, size_t elemSize ) {
-        size_t size = noOfElems * elemSize;
-        char * addr = (char *)memalignNoStats( alignment, size );
-  if ( unlikely( addr == 0p ) ) return 0p;
-        HeapManager.Storage.Header * header;
-        HeapManager.FreeHeader * freeElem;
-        size_t bsize;
-        bool mapped __attribute__(( unused )) = headers( "cmemalign", addr, header, freeElem, bsize, alignment );
-        #ifndef __CFA_DEBUG__
-        // Mapped storage is zero filled, but in debug mode mapped memory is scrubbed in doMalloc, so it has to be reset to zero.
-        if ( ! mapped )
-        #endif // __CFA_DEBUG__
-                memset( addr, '\0', dataStorage( bsize, addr, header ) ); // set to zeros
-        assert( noOfElems <= UINT32_MAX );
-        header->dimension = noOfElems;                                          // store initial array size
-        header->kind.real.blockSize |= 2;                                       // mark as zero filled
-        return addr;
-} // cmemalignNoStats
 …
 extern "C" {
+        // Allocates size bytes and returns a pointer to the allocated memory. The memory is not initialized. If size is 0,
+        // then malloc() returns either 0p, or a unique pointer value that can later be successfully passed to free().
+        // The malloc() function allocates size bytes and returns a pointer to the allocated memory. The memory is not
+        // initialized. If size is 0, then malloc() returns either NULL, or a unique pointer value that can later be
+        // successfully passed to free().
         void * malloc( size_t size ) {
                 #ifdef __STATISTICS__
 …
         } // malloc
         // Allocate memory for an array of nmemb elements of size bytes each and returns a pointer to the allocated
         // memory. The memory is set to zero. If nmemb or size is 0, then calloc() returns either 0p, or a unique pointer
         // value that can later be successfully passed to free().
+        // The calloc() function allocates memory for an array of nmemb elements of size bytes each and returns a pointer to
+        // the allocated memory. The memory is set to zero. If nmemb or size is 0, then calloc() returns either NULL, or a
+        // unique pointer value that can later be successfully passed to free().
         void * calloc( size_t noOfElems, size_t elemSize ) {
+                size_t size = noOfElems * elemSize;
                 #ifdef __STATISTICS__
                 __atomic_add_fetch( &calloc_calls, 1, __ATOMIC_SEQ_CST );
+                __atomic_add_fetch( &calloc_storage, noOfElems * elemSize, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                return callocNoStats( noOfElems, elemSize );
+        } // calloc
+        // Change the size of the memory block pointed to by ptr to size bytes. The contents are undefined.  If ptr is 0p,
+        // then the call is equivalent to malloc(size), for all values of size; if size is equal to zero, and ptr is not 0p,
+        // then the call is equivalent to free(ptr). Unless ptr is 0p, it must have been returned by an earlier call to
+        // malloc(), calloc() or realloc(). If the area pointed to was moved, a free(ptr) is done.
+        void * resize( void * oaddr, size_t size ) {
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &resize_calls, 1, __ATOMIC_SEQ_CST );
+                __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
+          if ( unlikely( size == 0 ) ) { free( oaddr ); return mallocNoStats( size ); } // special cases
+          if ( unlikely( oaddr == 0p ) ) return mallocNoStats( size );
+                __atomic_add_fetch( &calloc_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                char * area = (char *)mallocNoStats( size );
+          if ( unlikely( area == 0 ) ) return 0;
                 HeapManager.Storage.Header * header;
                 HeapManager.FreeHeader * freeElem;
+                size_t bsize, oalign = 0;
+                headers( "resize", oaddr, header, freeElem, bsize, oalign );
+                size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
+                // same size, DO NOT preserve STICKY PROPERTIES.
+                if ( oalign == 0 && size <= odsize && odsize <= size * 2 ) { // allow 50% wasted storage for smaller size
+                        header->kind.real.blockSize &= -2;                      // no alignment and turn off 0 fill
+                        return oaddr;
+                } // if
+                // change size, DO NOT preserve STICKY PROPERTIES.
+                void * naddr = mallocNoStats( size );                   // create new area
+                free( oaddr );
+                return naddr;
+        } // resize
+        // Same as resize but the contents shall be unchanged in the range from the start of the region up to the minimum of
+        // the old and new sizes.
+        void * realloc( void * oaddr, size_t size ) {
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &realloc_calls, 1, __ATOMIC_SEQ_CST );
+                __atomic_add_fetch( &realloc_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
+          if ( unlikely( size == 0 ) ) { free( oaddr ); return mallocNoStats( size ); } // special cases
+          if ( unlikely( oaddr == 0p ) ) return mallocNoStats( size );
+                size_t asize, alignment;
+                bool mapped __attribute__(( unused )) = headers( "calloc", area, header, freeElem, asize, alignment );
+                #ifndef __CFA_DEBUG__
+                // Mapped storage is zero filled, but in debug mode mapped memory is scrubbed in doMalloc, so it has to be reset to zero.
+                if ( ! mapped )
+                #endif // __CFA_DEBUG__
+                        memset( area, '\0', asize - sizeof(HeapManager.Storage) ); // set to zeros
+                header->kind.real.blockSize |= 2;                               // mark as zero filled
+                return area;
+        } // calloc
+        // #comment TD : Document this function
+        void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize ) {
+                size_t size = noOfElems * elemSize;
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &cmemalign_calls, 1, __ATOMIC_SEQ_CST );
+                __atomic_add_fetch( &cmemalign_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                char * area = (char *)memalignNoStats( alignment, size );
+          if ( unlikely( area == 0 ) ) return 0;
                 HeapManager.Storage.Header * header;
                 HeapManager.FreeHeader * freeElem;
+                size_t bsize, oalign = 0;
+                headers( "realloc", oaddr, header, freeElem, bsize, oalign );
+                size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
+          if ( size <= odsize && odsize <= size * 2 ) { // allow up to 50% wasted storage in smaller size
+                        // Do not know size of original allocation => cannot do 0 fill for any additional space because do not know
+                        // where to start filling, i.e., do not overwrite existing values in space.
+                        return oaddr;
+                size_t asize;
+                bool mapped __attribute__(( unused )) = headers( "cmemalign", area, header, freeElem, asize, alignment );
+                #ifndef __CFA_DEBUG__
+                // Mapped storage is zero filled, but in debug mode mapped memory is scrubbed in doMalloc, so it has to be reset to zero.
+                if ( ! mapped )
+                        #endif // __CFA_DEBUG__
+                        memset( area, '\0', asize - ( (char *)area - (char *)header ) ); // set to zeros
+                header->kind.real.blockSize |= 2;                               // mark as zero filled
+                return area;
+        } // cmemalign
+        // The realloc() function changes the size of the memory block pointed to by ptr to size bytes. The contents will be
+        // unchanged in the range from the start of the region up to the minimum of the old and new sizes. If the new size
+        // is larger than the old size, the added memory will not be initialized.  If ptr is NULL, then the call is
+        // equivalent to malloc(size), for all values of size; if size is equal to zero, and ptr is not NULL, then the call
+        // is equivalent to free(ptr). Unless ptr is NULL, it must have been returned by an earlier call to malloc(),
+        // calloc() or realloc(). If the area pointed to was moved, a free(ptr) is done.
+        void * realloc( void * addr, size_t size ) {
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &realloc_calls, 1, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+          if ( unlikely( addr == 0 ) ) return mallocNoStats( size ); // special cases
+          if ( unlikely( size == 0 ) ) { free( addr ); return 0; }
+                HeapManager.Storage.Header * header;
+                HeapManager.FreeHeader * freeElem;
+                size_t asize, alignment = 0;
+                headers( "realloc", addr, header, freeElem, asize, alignment );
+                size_t usize = asize - ( (char *)addr - (char *)header ); // compute the amount of user storage in the block
+                if ( usize >= size ) {                                                  // already sufficient storage
+                        // This case does not result in a new profiler entry because the previous one still exists and it must match with
+                        // the free for this memory.  Hence, this realloc does not appear in the profiler output.
+                        return addr;
                 } // if
+                // change size and copy old content to new storage
+                void * naddr;
+                if ( unlikely( oalign != 0 ) ) {                                // previous request memalign?
+                        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
+                                naddr = cmemalignNoStats( oalign, 1, size ); // create new aligned area
+                        } else {
+                                naddr = memalignNoStats( oalign, size ); // create new aligned area
+                        } // if
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &realloc_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                void * area;
+                if ( unlikely( alignment != 0 ) ) {                             // previous request memalign?
+                        area = memalign( alignment, size );                     // create new aligned area
                 } else {
+                        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
+                                naddr = callocNoStats( 1, size );               // create new area
+                        } else {
+                                naddr = mallocNoStats( size );                  // create new area
+                        } // if
+                        area = mallocNoStats( size );                           // create new area
                 } // if
+          if ( unlikely( naddr == 0p ) ) return 0p;
+                headers( "realloc", naddr, header, freeElem, bsize, oalign );
+                size_t ndsize = dataStorage( bsize, naddr, header ); // data storage avilable in bucket
+                // To preserve prior fill, the entire bucket must be copied versus the size.
+                memcpy( naddr, oaddr, MIN( odsize, ndsize ) );  // copy bytes
+                free( oaddr );
+                return naddr;
+          if ( unlikely( area == 0 ) ) return 0;
+                if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill (calloc/cmemalign) ?
+                        assert( (header->kind.real.blockSize & 1) == 0 );
+                        bool mapped __attribute__(( unused )) = headers( "realloc", area, header, freeElem, asize, alignment );
+                        #ifndef __CFA_DEBUG__
+                        // Mapped storage is zero filled, but in debug mode mapped memory is scrubbed in doMalloc, so it has to be reset to zero.
+                        if ( ! mapped )
+                        #endif // __CFA_DEBUG__
+                                memset( (char *)area + usize, '\0', asize - ( (char *)area - (char *)header ) - usize ); // zero-fill back part
+                        header->kind.real.blockSize |= 2;                       // mark new request as zero fill
+                } // if
+                memcpy( area, addr, usize );                                    // copy bytes
+                free( addr );
+                return area;
         } // realloc
         // Allocates size bytes and returns a pointer to the allocated memory. The memory address shall be a multiple of
         // alignment, which must be a power of two. (obsolete)
+        // The obsolete function memalign() allocates size bytes and returns a pointer to the allocated memory. The memory
+        // address will be a multiple of alignment, which must be a power of two.
         void * memalign( size_t alignment, size_t size ) {
                 #ifdef __STATISTICS__
 …
                 #endif // __STATISTICS__
+                return memalignNoStats( alignment, size );
+                void * area = memalignNoStats( alignment, size );
+                return area;
         } // memalign
+        // Same as calloc() with memory alignment.
+        void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize ) {
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &cmemalign_calls, 1, __ATOMIC_SEQ_CST );
+                __atomic_add_fetch( &cmemalign_storage, noOfElems * elemSize, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                return cmemalignNoStats( alignment, noOfElems, elemSize );
+        } // cmemalign
+        // Same as memalign(), but ISO/IEC 2011 C11 Section 7.22.2 states: the value of size shall be an integral multiple
+    // of alignment. This requirement is universally ignored.
+        // The function aligned_alloc() is the same as memalign(), except for the added restriction that size should be a
+        // multiple of alignment.
         void * aligned_alloc( size_t alignment, size_t size ) {
                 return memalign( alignment, size );
 …
         // Allocates size bytes and places the address of the allocated memory in *memptr. The address of the allocated
         // memory shall be a multiple of alignment, which must be a power of two and a multiple of sizeof(void *). If size
         // is 0, then posix_memalign() returns either 0p, or a unique pointer value that can later be successfully passed to
         // free(3).
+        // The function posix_memalign() allocates size bytes and places the address of the allocated memory in *memptr. The
+        // address of the allocated memory will be a multiple of alignment, which must be a power of two and a multiple of
+        // sizeof(void *). If size is 0, then posix_memalign() returns either NULL, or a unique pointer value that can later
+        // be successfully passed to free(3).
         int posix_memalign( void ** memptr, size_t alignment, size_t size ) {
           if ( alignment < sizeof(void *) || ! libPow2( alignment ) ) return EINVAL; // check alignment
                 * memptr = memalign( alignment, size );
           if ( unlikely( * memptr == 0p ) ) return ENOMEM;
+          if ( unlikely( * memptr == 0 ) ) return ENOMEM;
                 return 0;
         } // posix_memalign
         // Allocates size bytes and returns a pointer to the allocated memory. The memory address shall be a multiple of the
         // page size.  It is equivalent to memalign(sysconf(_SC_PAGESIZE),size).
+        // The obsolete function valloc() allocates size bytes and returns a pointer to the allocated memory. The memory
+        // address will be a multiple of the page size.  It is equivalent to memalign(sysconf(_SC_PAGESIZE),size).
         void * valloc( size_t size ) {
                 return memalign( pageSize, size );
 …
+        // Same as valloc but rounds size to multiple of page size.
+        void * pvalloc( size_t size ) {
+                return memalign( pageSize, libCeiling( size, pageSize ) );
+        } // pvalloc
+        // Frees the memory space pointed to by ptr, which must have been returned by a previous call to malloc(), calloc()
+        // or realloc().  Otherwise, or if free(ptr) has already been called before, undefined behavior occurs. If ptr is
+        // 0p, no operation is performed.
+        // The free() function frees the memory space pointed to by ptr, which must have been returned by a previous call to
+        // malloc(), calloc() or realloc().  Otherwise, or if free(ptr) has already been called before, undefined behavior
+        // occurs. If ptr is NULL, no operation is performed.
         void free( void * addr ) {
                 #ifdef __STATISTICS__
 …
                 #endif // __STATISTICS__
+          if ( unlikely( addr == 0p ) ) {                                       // special case
+                        // #ifdef __CFA_DEBUG__
+                        // if ( traceHeap() ) {
+                        //      #define nullmsg "Free( 0x0 ) size:0\n"
+                        //      // Do not debug print free( 0p ), as it can cause recursive entry from sprintf.
+                        //      __cfaabi_dbg_write( nullmsg, sizeof(nullmsg) - 1 );
+                        // } // if
+                        // #endif // __CFA_DEBUG__
+                // #comment TD : To decrease nesting I would but the special case in the
+                //               else instead, plus it reads more naturally to have the
+                //               short / normal case instead
+                if ( unlikely( addr == 0 ) ) {                                  // special case
+                        #ifdef __CFA_DEBUG__
+                        if ( traceHeap() ) {
+                                #define nullmsg "Free( 0x0 ) size:0\n"
+                                // Do not debug print free( 0 ), as it can cause recursive entry from sprintf.
+                                __cfaabi_dbg_bits_write( nullmsg, sizeof(nullmsg) - 1 );
+                        } // if
+                        #endif // __CFA_DEBUG__
                         return;
                 } // exit
 …
         } // free
+        // Returns the alignment of the allocation.
+        // The mallopt() function adjusts parameters that control the behavior of the memory-allocation functions (see
+        // malloc(3)). The param argument specifies the parameter to be modified, and value specifies the new value for that
+        // parameter.
+        int mallopt( int option, int value ) {
+                choose( option ) {
+                  case M_TOP_PAD:
+                        if ( setHeapExpand( value ) ) fallthru default;
+                  case M_MMAP_THRESHOLD:
+                        if ( setMmapStart( value ) ) fallthru default;
+                  default:
+                        // #comment TD : 1 for unsopported feels wrong
+                        return 1;                                                                       // success, or unsupported
+                } // switch
+                return 0;                                                                               // error
+        } // mallopt
+        // The malloc_trim() function attempts to release free memory at the top of the heap (by calling sbrk(2) with a
+        // suitable argument).
+        int malloc_trim( size_t ) {
+                return 0;                                                                               // => impossible to release memory
+        } // malloc_trim
+        // The malloc_usable_size() function returns the number of usable bytes in the block pointed to by ptr, a pointer to
+        // a block of memory allocated by malloc(3) or a related function.
+        size_t malloc_usable_size( void * addr ) {
+          if ( unlikely( addr == 0 ) ) return 0;                        // null allocation has 0 size
+                HeapManager.Storage.Header * header;
+                HeapManager.FreeHeader * freeElem;
+                size_t size, alignment;
+                headers( "malloc_usable_size", addr, header, freeElem, size, alignment );
+                size_t usize = size - ( (char *)addr - (char *)header ); // compute the amount of user storage in the block
+                return usize;
+        } // malloc_usable_size
+    // The malloc_alignment() function returns the alignment of the allocation.
         size_t malloc_alignment( void * addr ) {
           if ( unlikely( addr == 0p ) ) return libAlign();      // minimum alignment
+          if ( unlikely( addr == 0 ) ) return libAlign();       // minimum alignment
                 HeapManager.Storage.Header * header = headerAddr( addr );
                 if ( (header->kind.fake.alignment & 1) == 1 ) { // fake header ?
                         return header->kind.fake.alignment & -2;        // remove flag from value
                 } else {
                         return libAlign();                                                      // minimum alignment
+                        return libAlign ();                                                     // minimum alignment
                 } // if
         } // malloc_alignment
         // Returns true if the allocation is zero filled, i.e., initially allocated by calloc().
+    // The malloc_zero_fill() function returns true if the allocation is zero filled, i.e., initially allocated by calloc().
         bool malloc_zero_fill( void * addr ) {
           if ( unlikely( addr == 0p ) ) return false;           // null allocation is not zero fill
+          if ( unlikely( addr == 0 ) ) return false;            // null allocation is not zero fill
                 HeapManager.Storage.Header * header = headerAddr( addr );
                 if ( (header->kind.fake.alignment & 1) == 1 ) { // fake header ?
                         header = realHeader( header );                          // backup from fake to real header
+                        header = (HeapManager.Storage.Header *)((char *)header - header->kind.fake.offset);
                 } // if
                 return (header->kind.real.blockSize & 2) != 0;  // zero filled (calloc/cmemalign) ?
 …
+        // Returns number of elements if the allocation is for an array, i.e., by calloc().
+        size_t malloc_dimension( void * addr ) {
+          if ( unlikely( addr == 0p ) ) return false;           // null allocation is not zero fill
+                HeapManager.Storage.Header * header = headerAddr( addr );
+                if ( (header->kind.fake.alignment & 1) == 1 ) { // fake header ?
+                        header = realHeader( header );                          // backup from fake to real header
+                } // if
+                return header->dimension;                                               // array (calloc/cmemalign)
+        } // malloc_zero_fill
+        // Returns the number of usable bytes in the block pointed to by ptr, a pointer to a block of memory allocated by
+        // malloc or a related function.
+        size_t malloc_usable_size( void * addr ) {
+          if ( unlikely( addr == 0p ) ) return 0;                       // null allocation has 0 size
+                HeapManager.Storage.Header * header;
+                HeapManager.FreeHeader * freeElem;
+                size_t bsize, alignment;
+                headers( "malloc_usable_size", addr, header, freeElem, bsize, alignment );
+                return dataStorage( bsize, addr, header );      // data storage in bucket
+        } // malloc_usable_size
+        // Prints (on default standard error) statistics about memory allocated by malloc and related functions.
+    // The malloc_stats() function prints (on default standard error) statistics about memory allocated by malloc(3) and
+    // related functions.
         void malloc_stats( void ) {
                 #ifdef __STATISTICS__
                 printStats();
                 if ( prtFree() ) prtFree( heapManager );
+                if ( checkFree() ) checkFree( heapManager );
                 #endif // __STATISTICS__
         } // malloc_stats
         // Changes the file descripter where malloc_stats() writes statistics.
         int malloc_stats_fd( int fd __attribute__(( unused )) ) {
+        // The malloc_stats_fd() function changes the file descripter where malloc_stats() writes the statistics.
+        int malloc_stats_fd( int fd ) {
                 #ifdef __STATISTICS__
                 int temp = statfd;
 …
         } // malloc_stats_fd
+        // Adjusts parameters that control the behavior of the memory-allocation functions (see malloc). The param argument
+        // specifies the parameter to be modified, and value specifies the new value for that parameter.
+        int mallopt( int option, int value ) {
+                choose( option ) {
+                  case M_TOP_PAD:
+                        if ( setHeapExpand( value ) ) return 1;
+                  case M_MMAP_THRESHOLD:
+                        if ( setMmapStart( value ) ) return 1;
+                } // switch
+                return 0;                                                                               // error, unsupported
+        } // mallopt
+        // Attempt to release free memory at the top of the heap (by calling sbrk with a suitable argument).
+        int malloc_trim( size_t ) {
+                return 0;                                                                               // => impossible to release memory
+        } // malloc_trim
+        // Exports an XML string that describes the current state of the memory-allocation implementation in the caller.
+        // The string is printed on the file stream stream.  The exported string includes information about all arenas (see
+        // malloc).
+        // The malloc_info() function exports an XML string that describes the current state of the memory-allocation
+        // implementation in the caller.  The string is printed on the file stream stream.  The exported string includes
+        // information about all arenas (see malloc(3)).
         int malloc_info( int options, FILE * stream ) {
-                if ( options != 0 ) { errno = EINVAL; return -1; }
                 return printStatsXML( stream );
         } // malloc_info
         // Records the current state of all malloc internal bookkeeping variables (but not the actual contents of the heap
         // or the state of malloc_hook functions pointers).  The state is recorded in a system-dependent opaque data
         // structure dynamically allocated via malloc, and a pointer to that data structure is returned as the function
         // result.  (The caller must free this memory.)
+        // The malloc_get_state() function records the current state of all malloc(3) internal bookkeeping variables (but
+        // not the actual contents of the heap or the state of malloc_hook(3) functions pointers).  The state is recorded in
+        // a system-dependent opaque data structure dynamically allocated via malloc(3), and a pointer to that data
+        // structure is returned as the function result.  (It is the caller's responsibility to free(3) this memory.)
         void * malloc_get_state( void ) {
                 return 0p;                                                                              // unsupported
+                return 0;                                                                               // unsupported
         } // malloc_get_state
         // Restores the state of all malloc internal bookkeeping variables to the values recorded in the opaque data
         // structure pointed to by state.
+        // The malloc_set_state() function restores the state of all malloc(3) internal bookkeeping variables to the values
+        // recorded in the opaque data structure pointed to by state.
         int malloc_set_state( void * ptr ) {
                 return 0;                                                                               // unsupported
 …
-// Must have CFA linkage to overload with C linkage realloc.
-void * resize( void * oaddr, size_t nalign, size_t size ) {
-        #ifdef __STATISTICS__
-        __atomic_add_fetch( &resize_calls, 1, __ATOMIC_SEQ_CST );
-        __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
-        #endif // __STATISTICS__
-        // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
-  if ( unlikely( size == 0 ) ) { free( oaddr ); return memalignNoStats( nalign, size ); } // special cases
-  if ( unlikely( oaddr == 0p ) ) return memalignNoStats( nalign, size );
-        if ( unlikely( nalign == 0 ) ) nalign = libAlign();     // reset alignment to minimum
-        #ifdef __CFA_DEBUG__
-        else
-                checkAlign( nalign );                                                   // check alignment
-        #endif // __CFA_DEBUG__
-        HeapManager.Storage.Header * header;
-        HeapManager.FreeHeader * freeElem;
-        size_t bsize, oalign = 0;
-        headers( "resize", oaddr, header, freeElem, bsize, oalign );
-        size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
-        if ( oalign <= nalign && (uintptr_t)oaddr % nalign == 0 ) { // <= alignment and new alignment happens to match
-                if ( oalign >= libAlign() ) {                                   // fake header ?
-                        headerAddr( oaddr )->kind.fake.alignment = nalign | 1; // update alignment (could be the same)
-                } // if
-                if ( size <= odsize && odsize <= size * 2 ) {   // allow 50% wasted storage for smaller size
-                        header->kind.real.blockSize &= -2;                      // turn off 0 fill
-                        return oaddr;
-                } // if
-        } // if
-        // change size
-        void * naddr;
-        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
-                naddr = cmemalignNoStats( nalign, 1, size );    // create new aligned area
-        } else {
-                naddr = memalignNoStats( nalign, size );                // create new aligned area
-        } // if
-        free( oaddr );
-        return naddr;
-} // resize
-void * realloc( void * oaddr, size_t nalign, size_t size ) {
-        if ( unlikely( nalign == 0 ) ) nalign = libAlign();     // reset alignment to minimum
-        #ifdef __CFA_DEBUG__
-        else
-                checkAlign( nalign );                                                   // check alignment
-        #endif // __CFA_DEBUG__
-        HeapManager.Storage.Header * header;
-        HeapManager.FreeHeader * freeElem;
-        size_t bsize, oalign = 0;
-        headers( "realloc", oaddr, header, freeElem, bsize, oalign );
-        size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
-        if ( oalign <= nalign && (uintptr_t)oaddr % nalign == 0 ) { // <= alignment and new alignment happens to match
-                if ( oalign >= libAlign() ) {                                   // fake header ?
-                        headerAddr( oaddr )->kind.fake.alignment = nalign | 1; // update alignment (could be the same)
-                } // if
-                return realloc( oaddr, size );
-        } // if
-        // change size and copy old content to new storage
-        #ifdef __STATISTICS__
-        __atomic_add_fetch( &realloc_calls, 1, __ATOMIC_SEQ_CST );
-        __atomic_add_fetch( &realloc_storage, size, __ATOMIC_SEQ_CST );
-        #endif // __STATISTICS__
-        // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
-  if ( unlikely( size == 0 ) ) { free( oaddr ); return memalignNoStats( nalign, size ); } // special cases
-  if ( unlikely( oaddr == 0p ) ) return memalignNoStats( nalign, size );
-        void * naddr;
-        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
-                naddr = cmemalignNoStats( nalign, 1, size );    // create new aligned area
-        } else {
-                naddr = memalignNoStats( nalign, size );                // create new aligned area
-        } // if
-        headers( "realloc", naddr, header, freeElem, bsize, oalign );
-        size_t ndsize = dataStorage( bsize, naddr, header ); // data storage available in bucket
-        // To preserve prior fill, the entire bucket must be copied versus the size.
-        memcpy( naddr, oaddr, MIN( odsize, ndsize ) );          // copy bytes
-        free( oaddr );
-        return naddr;
-} // realloc
 // Local Variables: //
 // tab-width: 4 //

libcfa/src/interpose.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Mar 29 16:10:31 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Mar 13 17:35:37 2020
 // Update Count     : 178
+// Last Modified On : Sun Jul 14 22:57:16 2019
+// Update Count     : 116
 //
 …
 #include "bits/signal.hfa"                                                              // sigHandler_?
 #include "startup.hfa"                                                                  // STARTUP_PRIORITY_CORE
-#include <assert.h>
 //=============================================================================================
 …
 typedef void (* generic_fptr_t)(void);
 generic_fptr_t interpose_symbol( const char symbol[], const char version[] ) {
+generic_fptr_t interpose_symbol( const char * symbol, const char * version ) {
         const char * error;
 …
         void __cfaabi_interpose_startup(void)  __attribute__(( constructor( STARTUP_PRIORITY_CORE ) ));
         void __cfaabi_interpose_startup( void ) {
                 const char *version = 0p;
+                const char *version = NULL;
                 preload_libgcc();
 …
 #pragma GCC diagnostic pop
-                // As a precaution (and necessity), errors that result in termination are delivered on a separate stack because
-                // task stacks might be very small (4K) and the signal delivery corrupts memory to the point that a clean
-                // shutdown is impossible. Also, when a stack overflow encounters the non-accessible sentinel page (debug only)
-                // and generates a segment fault, the signal cannot be delivered on the sentinel page. Finally, calls to abort
-                // print a stack trace that uses substantial stack space.
-                #define MINSTKSZ SIGSTKSZ * 8
-                static char stack[MINSTKSZ] __attribute__(( aligned (16) ));
-                static stack_t ss;
-                ss.ss_sp = stack;
-                ss.ss_size = MINSTKSZ;
-                ss.ss_flags = 0;
-                if ( sigaltstack( &ss, 0p ) == -1 ) {
-                        abort( "__cfaabi_interpose_startup : internal error, sigaltstack error(%d) %s.", errno, strerror( errno ) );
-                } // if
                 // Failure handler
+                __cfaabi_sigaction( SIGSEGV, sigHandler_segv, SA_SIGINFO | SA_ONSTACK );
+                __cfaabi_sigaction( SIGBUS , sigHandler_segv, SA_SIGINFO | SA_ONSTACK );
+                __cfaabi_sigaction( SIGILL , sigHandler_ill , SA_SIGINFO | SA_ONSTACK );
+                __cfaabi_sigaction( SIGFPE , sigHandler_fpe , SA_SIGINFO | SA_ONSTACK );
+                __cfaabi_sigaction( SIGTERM, sigHandler_term, SA_SIGINFO | SA_ONSTACK | SA_RESETHAND ); // one shot handler, return to default
+                __cfaabi_sigaction( SIGINT , sigHandler_term, SA_SIGINFO | SA_ONSTACK | SA_RESETHAND );
+                __cfaabi_sigaction( SIGABRT, sigHandler_term, SA_SIGINFO | SA_ONSTACK | SA_RESETHAND );
+                __cfaabi_sigaction( SIGHUP , sigHandler_term, SA_SIGINFO | SA_ONSTACK | SA_RESETHAND ); // terminal hangup
+                __cfaabi_sigaction( SIGSEGV, sigHandler_segv , SA_SIGINFO );
+                __cfaabi_sigaction( SIGBUS , sigHandler_segv , SA_SIGINFO );
+                __cfaabi_sigaction( SIGILL , sigHandler_ill  , SA_SIGINFO );
+                __cfaabi_sigaction( SIGFPE , sigHandler_fpe  , SA_SIGINFO );
+                __cfaabi_sigaction( SIGABRT, sigHandler_abrt, SA_SIGINFO | SA_RESETHAND);
+                __cfaabi_sigaction( SIGTERM, sigHandler_term , SA_SIGINFO );
+                __cfaabi_sigaction( SIGINT , sigHandler_term , SA_SIGINFO );
+        }
+}
 …
 void exit( int status, const char fmt[], ... ) __attribute__(( format(printf, 2, 3), __nothrow__, __leaf__, __noreturn__ ));
 void abort( const char fmt[], ... ) __attribute__(( format(printf, 1, 2), __nothrow__, __leaf__, __noreturn__ ));
-void abort( bool signalAbort, const char fmt[], ... ) __attribute__(( format(printf, 2, 3), __nothrow__, __leaf__, __noreturn__ ));
-void __abort( bool signalAbort, const char fmt[], va_list args ) __attribute__(( __nothrow__, __leaf__, __noreturn__ ));
 extern "C" {
         void abort( void ) __attribute__(( __nothrow__, __leaf__, __noreturn__ )) {
                 abort( false, "%s", "" );
+                abort( NULL );
+        }
 …
                 va_list argp;
                 va_start( argp, fmt );
                 __abort( false, fmt, argp );
+                abort( fmt, argp );
                 va_end( argp );
+        }
 …
+}
+void * kernel_abort( void ) __attribute__(( __nothrow__, __leaf__, __weak__ )) { return 0p; }
+void kernel_abort_msg( void * data, char buffer[], int size ) __attribute__(( __nothrow__, __leaf__, __weak__ )) {}
+// See concurrency/kernel.cfa for strong definition used in multi-processor mode.
+void * kernel_abort    ( void ) __attribute__(( __nothrow__, __leaf__, __weak__ )) { return NULL; }
+void   kernel_abort_msg( void * data, char * buffer, int size ) __attribute__(( __nothrow__, __leaf__, __weak__ )) {}
 int kernel_abort_lastframe( void ) __attribute__(( __nothrow__, __leaf__, __weak__ )) { return 4; }
 enum { abort_text_size = 1024 };
 static char abort_text[ abort_text_size ];
+static void __cfaabi_backtrace( int start ) {
+        enum { Frames = 50, };                                                          // maximum number of stack frames
+        int last = kernel_abort_lastframe();                            // skip last N stack frames
+static int abort_lastframe;
+void exit( int status, const char fmt[], ... ) __attribute__(( format(printf, 2, 3), __nothrow__, __leaf__, __noreturn__ )) {
+    va_list args;
+    va_start( args, fmt );
+    vfprintf( stderr, fmt, args );
+    va_end( args );
+        __cabi_libc.exit( status );
+}
+void abort( const char fmt[], ... ) __attribute__(( format(printf, 1, 2), __nothrow__, __leaf__, __noreturn__ )) {
+        void * kernel_data = kernel_abort();                    // must be done here to lock down kernel
+        int len;
+        abort_lastframe = kernel_abort_lastframe();
+        len = snprintf( abort_text, abort_text_size, "Cforall Runtime error (UNIX pid:%ld) ", (long int)getpid() ); // use UNIX pid (versus getPid)
+        __cfaabi_dbg_bits_write( abort_text, len );
+        if ( fmt ) {
+                va_list args;
+                va_start( args, fmt );
+                len = vsnprintf( abort_text, abort_text_size, fmt, args );
+                va_end( args );
+                __cfaabi_dbg_bits_write( abort_text, len );
+                if ( fmt[strlen( fmt ) - 1] != '\n' ) {         // add optional newline if missing at the end of the format text
+                        __cfaabi_dbg_bits_write( "\n", 1 );
+                }
+        }
+        kernel_abort_msg( kernel_data, abort_text, abort_text_size );
+        __cabi_libc.abort();
+}
+static void __cfaabi_backtrace() {
+        enum {
+                Frames = 50,                                                                    // maximum number of stack frames
+                Start = 8,                                                                              // skip first N stack frames
+        };
         void * array[Frames];
         size_t size = backtrace( array, Frames );
+        char ** messages = backtrace_symbols( array, size ); // does not demangle names
+        *index( messages[0], '(' ) = '\0';                                      // find executable name
+        __cfaabi_bits_print_nolock( STDERR_FILENO, "Stack back trace for: %s\n", messages[0]);
+        for ( unsigned int i = start; i < size - last && messages != 0p; i += 1 ) {
+                char * name = 0p, * offset_begin = 0p, * offset_end = 0p;
+                for ( char * p = messages[i]; *p; p += 1 ) {    // find parantheses and +offset
+                        //__cfaabi_bits_print_nolock( "X %s\n", p);
+        char ** messages = backtrace_symbols( array, size );
+        // find executable name
+        *index( messages[0], '(' ) = '\0';
+        __cfaabi_dbg_bits_print_nolock( "Stack back trace for: %s\n", messages[0]);
+        for ( int i = Start; i < size - abort_lastframe && messages != NULL; i += 1 ) {
+                char * name = NULL, * offset_begin = NULL, * offset_end = NULL;
+                for ( char * p = messages[i]; *p; ++p ) {
+                        //__cfaabi_dbg_bits_print_nolock( "X %s\n", p);
+                        // find parantheses and +offset
                         if ( *p == '(' ) {
                                 name = p;
 …
+                }
                 // if line contains symbol, print it
                 int frameNo = i - start;
+                // if line contains symbol print it
+                int frameNo = i - Start;
                 if ( name && offset_begin && offset_end && name < offset_begin ) {
+                        *name++ = '\0';                                                         // delimit strings
+                        // delimit strings
+                        *name++ = '\0';
                         *offset_begin++ = '\0';
                         *offset_end++ = '\0';
                         __cfaabi_bits_print_nolock( STDERR_FILENO, "(%i) %s : %s + %s %s\n", frameNo, messages[i], name, offset_begin, offset_end);
+                        __cfaabi_dbg_bits_print_nolock( "(%i) %s : %s + %s %s\n", frameNo, messages[i], name, offset_begin, offset_end);
                 } else {                                                                                // otherwise, print the whole line
                         __cfaabi_bits_print_nolock( STDERR_FILENO, "(%i) %s\n", frameNo, messages[i] );
+                        __cfaabi_dbg_bits_print_nolock( "(%i) %s\n", frameNo, messages[i] );
+                }
+        }
 …
+}
-void exit( int status, const char fmt[], ... ) {
-        va_list args;
-        va_start( args, fmt );
-        vfprintf( stderr, fmt, args );
-        va_end( args );
-        __cabi_libc.exit( status );
+}
-// Cannot forward va_list.
-void __abort( bool signalAbort, const char fmt[], va_list args ) {
-        void * kernel_data = kernel_abort();                            // must be done here to lock down kernel
-        int len;
-        signal( SIGABRT, SIG_DFL );                                                     // prevent final "real" abort from recursing to handler
-        len = snprintf( abort_text, abort_text_size, "Cforall Runtime error (UNIX pid:%ld) ", (long int)getpid() ); // use UNIX pid (versus getPid)
-        __cfaabi_bits_write( STDERR_FILENO, abort_text, len );
-        assert( fmt );
-        len = vsnprintf( abort_text, abort_text_size, fmt, args );
-        __cfaabi_bits_write( STDERR_FILENO, abort_text, len );
-        if ( fmt[strlen( fmt ) - 1] != '\n' ) {                         // add optional newline if missing at the end of the format text
-                __cfaabi_bits_write( STDERR_FILENO, "\n", 1 );
-        } // if
-        kernel_abort_msg( kernel_data, abort_text, abort_text_size );
-        __cfaabi_backtrace( signalAbort ? 4 : 2 );
-        __cabi_libc.abort();                                                            // print stack trace in handler
+}
-void abort( const char fmt[], ... ) {
-        va_list args;
-        va_start( args, fmt );
-        __abort( false, fmt, args );
-    // CONTROL NEVER REACHES HERE!
-        va_end( args );
+}
-void abort( bool signalAbort, const char fmt[], ... ) {
-    va_list args;
-    va_start( args, fmt );
-    __abort( signalAbort, fmt, args );
-    // CONTROL NEVER REACHES HERE!
-    va_end( args );
+}
 void sigHandler_segv( __CFA_SIGPARMS__ ) {
+                if ( sfp->si_addr == 0p ) {
+                        abort( true, "Null pointer (0p) dereference.\n" );
+                } else {
+                        abort( true, "%s at memory location %p.\n"
+                                   "Possible cause is reading outside the address space or writing to a protected area within the address space with an invalid pointer or subscript.\n",
+                                   (sig == SIGSEGV ? "Segment fault" : "Bus error"), sfp->si_addr );
+                }
+        abort( "Addressing invalid memory at location %p\n"
+                        "Possible cause is reading outside the address space or writing to a protected area within the address space with an invalid pointer or subscript.\n",
+                        sfp->si_addr );
+}
 void sigHandler_ill( __CFA_SIGPARMS__ ) {
         abort( true, "Executing illegal instruction at location %p.\n"
+        abort( "Executing illegal instruction at location %p.\n"
                         "Possible cause is stack corruption.\n",
                         sfp->si_addr );
 …
           default: msg = "unknown";
         } // choose
+        abort( true, "Computation error %s at location %p.\n", msg, sfp->si_addr );
+        abort( "Computation error %s at location %p.\n", msg, sfp->si_addr );
+}
+void sigHandler_abrt( __CFA_SIGPARMS__ ) {
+        __cfaabi_backtrace();
+        // reset default signal handler
+        __cfaabi_sigdefault( SIGABRT );
+        raise( SIGABRT );
+}
 void sigHandler_term( __CFA_SIGPARMS__ ) {
         abort( true, "Application interrupted by signal: %s.\n", strsignal( sig ) );
+        abort( "Application stopped by %s signal.", sig == SIGINT ? "an interrupt (SIGINT)" : "a terminate (SIGTERM)" );
+}

libcfa/src/iostream.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Mar 11 14:35:35 2020
 // Update Count     : 860
+// Last Modified On : Sat Jul 13 08:07:59 2019
+// Update Count     : 821
 //
 …
 #include <stdio.h>
 #include <stdbool.h>                                                                    // true/false
-#include <stdint.h>                                                                             // UINT64_MAX
 //#include <string.h>                                                                   // strlen, strcmp
 extern size_t strlen (const char *__s) __attribute__ ((__nothrow__ , __leaf__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));
 …
 forall( dtype ostype | ostream( ostype ) ) {
         ostype & ?|?( ostype & os, zero_t ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%d", 0n );
                 return os;
 …
         ostype & ?|?( ostype & os, one_t ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%d", 1n );
                 return os;
 …
         ostype & ?|?( ostype & os, bool b ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%s", b ? "true" : "false" );
                 return os;
 …
         ostype & ?|?( ostype & os, char c ) {
                 fmt( os, "%c", c );
                 if ( c == '\n' ) $setNL( os, true );
+                if ( c == '\n' ) setNL( os, true );
                 return sepOff( os );
         } // ?|?
 …
         ostype & ?|?( ostype & os, signed char sc ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%hhd", sc );
                 return os;
 …
         ostype & ?|?( ostype & os, unsigned char usc ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%hhu", usc );
                 return os;
 …
         ostype & ?|?( ostype & os, short int si ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%hd", si );
                 return os;
 …
         ostype & ?|?( ostype & os, unsigned short int usi ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%hu", usi );
                 return os;
 …
         ostype & ?|?( ostype & os, int i ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%d", i );
                 return os;
 …
         ostype & ?|?( ostype & os, unsigned int ui ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%u", ui );
                 return os;
 …
         ostype & ?|?( ostype & os, long int li ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%ld", li );
                 return os;
 …
         ostype & ?|?( ostype & os, unsigned long int uli ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%lu", uli );
                 return os;
 …
         ostype & ?|?( ostype & os, long long int lli ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%lld", lli );
                 return os;
 …
         ostype & ?|?( ostype & os, unsigned long long int ulli ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%llu", ulli );
                 return os;
 …
                 (ostype &)(os | ulli); ends( os );
         } // ?|?
-#if defined( __SIZEOF_INT128__ )
-        //      UINT64_MAX 18_446_744_073_709_551_615_ULL
-        #define P10_UINT64 10_000_000_000_000_000_000_ULL       // 19 zeroes
-        static void base10_128( ostype & os, unsigned int128 val ) {
-                if ( val > UINT64_MAX ) {
-                        base10_128( os, val / P10_UINT64 );                     // recursive
-                        fmt( os, "%.19lu", (uint64_t)(val % P10_UINT64) );
-                } else {
-                        fmt( os, "%lu", (uint64_t)val );
-                } // if
-        } // base10_128
-        static void base10_128( ostype & os, int128 val ) {
-                if ( val < 0 ) {
-                        fmt( os, "-" );                                                         // leading negative sign
-                        val = -val;
-                } // if
-                base10_128( os, (unsigned int128)val );                 // print zero/positive value
-        } // base10_128
-        ostype & ?|?( ostype & os, int128 llli ) {
-                if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
-                base10_128( os, llli );
-                return os;
-        } // ?|?
-        void & ?|?( ostype & os, int128 llli ) {
-                (ostype &)(os | llli); ends( os );
-        } // ?|?
-        ostype & ?|?( ostype & os, unsigned int128 ullli ) {
-                if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
-                base10_128( os, ullli );
-                return os;
-        } // ?|?
-        void & ?|?( ostype & os, unsigned int128 ullli ) {
-                (ostype &)(os | ullli); ends( os );
-        } // ?|?
-#endif // __SIZEOF_INT128__
         #define PrintWithDP( os, format, val, ... ) \
 …
         ostype & ?|?( ostype & os, float f ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 PrintWithDP( os, "%g", f );
                 return os;
 …
         ostype & ?|?( ostype & os, double d ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 PrintWithDP( os, "%.*lg", d, DBL_DIG );
                 return os;
 …
         ostype & ?|?( ostype & os, long double ld ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 PrintWithDP( os, "%.*Lg", ld, LDBL_DIG );
                 return os;
 …
         ostype & ?|?( ostype & os, float _Complex fc ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
 //              os | crealf( fc ) | nonl;
                 PrintWithDP( os, "%g", crealf( fc ) );
 …
         ostype & ?|?( ostype & os, double _Complex dc ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
 //              os | creal( dc ) | nonl;
                 PrintWithDP( os, "%.*lg", creal( dc ), DBL_DIG );
 …
         ostype & ?|?( ostype & os, long double _Complex ldc ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
 //              os | creall( ldc ) || nonl;
                 PrintWithDP( os, "%.*Lg", creall( ldc ), LDBL_DIG );
 …
         } // ?|?
         ostype & ?|?( ostype & os, const char str[] ) {
+        ostype & ?|?( ostype & os, const char * str ) {
                 enum { Open = 1, Close, OpenClose };
                 static const unsigned char mask[256] @= {
 …
                 // first character IS NOT spacing or closing punctuation => add left separator
                 unsigned char ch = str[0];                                              // must make unsigned
                 if ( $sepPrt( os ) && mask[ ch ] != Close && mask[ ch ] != OpenClose ) {
                         fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) && mask[ ch ] != Close && mask[ ch ] != OpenClose ) {
+                        fmt( os, "%s", sepGetCur( os ) );
                 } // if
                 // if string starts line, must reset to determine open state because separator is off
                 $sepReset( os );                                                                // reset separator
+                sepReset( os );                                                                 // reset separator
                 // last character IS spacing or opening punctuation => turn off separator for next item
                 size_t len = strlen( str );
                 ch = str[len - 1];                                                              // must make unsigned
                 if ( $sepPrt( os ) && mask[ ch ] != Open && mask[ ch ] != OpenClose ) {
+                if ( sepPrt( os ) && mask[ ch ] != Open && mask[ ch ] != OpenClose ) {
                         sepOn( os );
                 } else {
                         sepOff( os );
                 } // if
                 if ( ch == '\n' ) $setNL( os, true );                   // check *AFTER* $sepPrt call above as it resets NL flag
+                if ( ch == '\n' ) setNL( os, true );                    // check *AFTER* sepPrt call above as it resets NL flag
                 return write( os, str, len );
         } // ?|?
+        void ?|?( ostype & os, const char str[] ) {
+        void ?|?( ostype & os, const char * str ) {
                 (ostype &)(os | str); ends( os );
         } // ?|?
 //      ostype & ?|?( ostype & os, const char16_t * str ) {
 //              if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+//              if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
 //              fmt( os, "%ls", str );
 //              return os;
 …
 // #if ! ( __ARM_ARCH_ISA_ARM == 1 && __ARM_32BIT_STATE == 1 ) // char32_t == wchar_t => ambiguous
 //      ostype & ?|?( ostype & os, const char32_t * str ) {
 //              if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+//              if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
 //              fmt( os, "%ls", str );
 //              return os;
 …
 //      ostype & ?|?( ostype & os, const wchar_t * str ) {
 //              if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+//              if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
 //              fmt( os, "%ls", str );
 //              return os;
 …
         ostype & ?|?( ostype & os, const void * p ) {
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 fmt( os, "%p", p );
                 return os;
 …
         void ?|?( ostype & os, ostype & (* manip)( ostype & ) ) {
                 (ostype &)(manip( os ));
                 if ( $getPrt( os ) ) ends( os );                                // something printed ?
                 $setPrt( os, false );                                                   // turn off
+                if ( getPrt( os ) ) ends( os );                                 // something printed ?
+                setPrt( os, false );                                                    // turn off
         } // ?|?
 …
         ostype & nl( ostype & os ) {
                 (ostype &)(os | '\n');
                 $setPrt( os, false );                                                   // turn off
                 $setNL( os, true );
+                setPrt( os, false );                                                    // turn off
+                setNL( os, true );
                 flush( os );
                 return sepOff( os );                                                    // prepare for next line
 …
         ostype & nonl( ostype & os ) {
                 $setPrt( os, false );                                                   // turn off
+                setPrt( os, false );                                                    // turn off
                 return os;
         } // nonl
 …
         ostype & ?|?( ostype & os, T arg, Params rest ) {
                 (ostype &)(os | arg);                                                   // print first argument
                 $sepSetCur( os, sepGetTuple( os ) );                    // switch to tuple separator
+                sepSetCur( os, sepGetTuple( os ) );                             // switch to tuple separator
                 (ostype &)(os | rest);                                                  // print remaining arguments
                 $sepSetCur( os, sepGet( os ) );                                 // switch to regular separator
+                sepSetCur( os, sepGet( os ) );                                  // switch to regular separator
                 return os;
         } // ?|?
 …
                 // (ostype &)(?|?( os, arg, rest )); ends( os );
                 (ostype &)(os | arg);                                                   // print first argument
                 $sepSetCur( os, sepGetTuple( os ) );                    // switch to tuple separator
+                sepSetCur( os, sepGetTuple( os ) );                             // switch to tuple separator
                 (ostype &)(os | rest);                                                  // print remaining arguments
                 $sepSetCur( os, sepGet( os ) );                                 // switch to regular separator
+                sepSetCur( os, sepGet( os ) );                                  // switch to regular separator
                 ends( os );
         } // ?|?
 …
 forall( dtype ostype | ostream( ostype ) ) { \
         ostype & ?|?( ostype & os, _Ostream_Manip(T) f ) { \
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) ); \
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) ); \
+\
                 if ( f.base == 'b' || f.base == 'B' ) {                 /* bespoke binary format */ \
 …
+\
                 if ( ! f.flags.pc ) {                                                   /* no precision */ \
+                        /* printf( "%s\n", &fmtstr[star] ); */ \
                         fmtstr[sizeof(IFMTNP)-2] = f.base;                      /* sizeof includes '\0' */ \
-                        /* printf( "%s %c %c\n", &fmtstr[star], f.base, CODE ); */ \
                         fmt( os, &fmtstr[star], f.wd, f.val ); \
                 } else {                                                                                /* precision */ \
                         fmtstr[sizeof(IFMTP)-2] = f.base;                       /* sizeof includes '\0' */ \
                         /* printf( "%s %c %c\n", &fmtstr[star], f.base, CODE ); */ \
+                        /* printf( "%s\n", &fmtstr[star] ); */ \
                         fmt( os, &fmtstr[star], f.wd, f.pc, f.val ); \
                 } /* if */ \
 …
 IntegralFMTImpl( signed long long int, 'd', "%    *ll ", "%    *.*ll " )
 IntegralFMTImpl( unsigned long long int, 'u', "%    *ll ", "%    *.*ll " )
-#if defined( __SIZEOF_INT128__ )
-// Default prefix for non-decimal prints is 0b, 0, 0x.
-#define IntegralFMTImpl128( T, SIGNED, CODE, IFMTNP, IFMTP ) \
-forall( dtype ostype | ostream( ostype ) ) \
-static void base10_128( ostype & os, _Ostream_Manip(T) f ) { \
-        if ( f.val > UINT64_MAX ) { \
-                unsigned long long int lsig = f.val % P10_UINT64; \
-                f.val /= P10_UINT64; /* msig */ \
-                base10_128( os, f ); /* recursion */ \
-                _Ostream_Manip(unsigned long long int) fmt @= { lsig, 0, 19, 'u', { .all : 0 } }; \
-                fmt.flags.nobsdp = true; \
-                /* printf( "fmt1 %c %lld %d\n", fmt.base, fmt.val, fmt.all ); */ \
-                sepOff( os ); \
-                (ostype &)(os | fmt); \
-        } else { \
-                /* printf( "fmt2 %c %lld %d\n", f.base, (unsigned long long int)f.val, f.all ); */ \
-                _Ostream_Manip(SIGNED long long int) fmt @= { (SIGNED long long int)f.val, f.wd, f.pc, f.base, { .all : f.all } }; \
-                (ostype &)(os | fmt); \
-        } /* if */ \
-} /* base10_128 */ \
-forall( dtype ostype | ostream( ostype ) ) { \
-        ostype & ?|?( ostype & os, _Ostream_Manip(T) f ) { \
-                if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) ); \
+\
-                if ( f.base == 'b' | f.base == 'B' | f.base == 'o' | f.base == 'x' | f.base == 'X' ) { \
-                        unsigned long long int msig = (unsigned long long int)(f.val >> 64); \
-                        unsigned long long int lsig = (unsigned long long int)(f.val); \
-                        _Ostream_Manip(SIGNED long long int) fmt @= { msig, f.wd, f.pc, f.base, { .all : f.all } }; \
-                        _Ostream_Manip(unsigned long long int) fmt2 @= { lsig, 0, 0, f.base, { .all : 0 } }; \
-                        if ( msig == 0 ) { \
-                                fmt.val = lsig; \
-                                (ostype &)(os | fmt); \
-                        } else { \
-                                fmt2.flags.pad0 = fmt2.flags.nobsdp = true;     \
-                                if ( f.base == 'b' | f.base == 'B' ) { \
-                                        if ( f.wd > 64 ) fmt.wd = f.wd - 64; \
-                                        if ( f.flags.pc && f.pc > 64 ) fmt.pc = f.pc - 64; \
-                                        fmt2.wd = 64; \
-                                        (ostype &)(os | fmt | "" | fmt2); \
-                                } else if ( f.base == 'o' ) { \
-                                        fmt.val = (unsigned long long int)fmt.val >> 2; \
-                                        if ( f.wd > 21 ) fmt.wd = f.wd - 21; \
-                                        if ( f.flags.pc && f.pc > 21 ) fmt.pc = f.pc - 21; \
-                                        fmt2.wd = 1; \
-                                        fmt2.val = ((msig & 0x3) << 1) + 1; \
-                                        (ostype &)(os | fmt | "" | fmt2); \
-                                        sepOff( os ); \
-                                        fmt2.wd = 21; \
-                                        fmt2.val = lsig & 0x7fffffffffffffff; \
-                                        (ostype &)(os | fmt2); \
-                                } else { \
-                                        if ( f.flags.left ) { \
-                                                if ( f.wd > 16 ) fmt2.wd = f.wd - 16; \
-                                                fmt.wd = 16; \
-                                        } else { \
-                                                if ( f.wd > 16 ) fmt.wd = f.wd - 16; \
-                                                if ( f.flags.pc && f.pc > 16 ) fmt.pc = f.pc - 16; \
-                                                fmt2.wd = 16; \
-                                        } /* if */ \
-                                        (ostype &)(os | fmt | "" | fmt2); \
-                                } /* if */ \
-                        } /* if */ \
-                } else { \
-                        if ( CODE == 'd' ) { \
-                                if ( f.val < 0 )  { fmt( os, "-" ); sepOff( os ); f.val = -f.val; f.flags.sign = false; } \
-                        } /* if */ \
-                        base10_128( os, f ); \
-                } /* if */ \
-                return os; \
-        } /* ?|? */ \
-        void ?|?( ostype & os, _Ostream_Manip(T) f ) { (ostype &)(os | f); ends( os ); } \
-} // distribution
-IntegralFMTImpl128( int128, signed, 'd', "%    *ll ", "%    *.*ll " )
-IntegralFMTImpl128( unsigned int128, unsigned, 'u', "%    *ll ", "%    *.*ll " )
-#endif // __SIZEOF_INT128__
 //*********************************** floating point ***********************************
 …
 forall( dtype ostype | ostream( ostype ) ) { \
         ostype & ?|?( ostype & os, _Ostream_Manip(T) f ) { \
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) ); \
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) ); \
                 char fmtstr[sizeof(DFMTP)];                                             /* sizeof includes '\0' */ \
                 if ( ! f.flags.pc ) memcpy( &fmtstr, DFMTNP, sizeof(DFMTNP) ); \
 …
                 return os; \
         } /* ?|? */ \
+\
         void ?|?( ostype & os, _Ostream_Manip(T) f ) { (ostype &)(os | f); ends( os ); } \
 } // distribution
 …
                 } // if
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 #define CFMTNP "% * "
 …
                 return os;
         } // ?|?
         void ?|?( ostype & os, _Ostream_Manip(char) f ) { (ostype &)(os | f); ends( os ); }
 } // distribution
 …
                 } // if
                 if ( $sepPrt( os ) ) fmt( os, "%s", $sepGetCur( os ) );
+                if ( sepPrt( os ) ) fmt( os, "%s", sepGetCur( os ) );
                 #define SFMTNP "% * "
 …
                 return os;
         } // ?|?
         void ?|?( ostype & os, _Ostream_Manip(const char *) f ) { (ostype &)(os | f); ends( os ); }
 } // distribution
 …
         } // ?|?
         // istype & ?|?( istype & is, const char fmt[] ) {
+        // istype & ?|?( istype & is, const char * fmt ) {
         //      fmt( is, fmt, "" );
         //      return is;

libcfa/src/iostream.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Feb 20 15:30:56 2020
 // Update Count     : 337
+// Last Modified On : Fri Jul 12 12:08:38 2019
+// Update Count     : 334
 //
 …
 trait ostream( dtype ostype ) {
         // private
         bool $sepPrt( ostype & );                                                       // get separator state (on/off)
         void $sepReset( ostype & );                                                     // set separator state to default state
         void $sepReset( ostype &, bool );                                       // set separator and default state
         const char * $sepGetCur( ostype & );                            // get current separator string
         void $sepSetCur( ostype &, const char [] );                     // set current separator string
         bool $getNL( ostype & );                                                        // check newline
         void $setNL( ostype &, bool );                                          // saw newline
         bool $getANL( ostype & );                                                       // get auto newline (on/off)
         bool $getPrt( ostype & );                                                       // get fmt called in output cascade
         void $setPrt( ostype &, bool );                                         // set fmt called in output cascade
+        bool sepPrt( ostype & );                                                        // get separator state (on/off)
+        void sepReset( ostype & );                                                      // set separator state to default state
+        void sepReset( ostype &, bool );                                        // set separator and default state
+        const char * sepGetCur( ostype & );                                     // get current separator string
+        void sepSetCur( ostype &, const char * );                       // set current separator string
+        bool getNL( ostype & );                                                         // check newline
+        void setNL( ostype &, bool );                                           // saw newline
+        bool getANL( ostype & );                                                        // get auto newline (on/off)
+        bool getPrt( ostype & );                                                        // get fmt called in output cascade
+        void setPrt( ostype &, bool );                                          // set fmt called in output cascade
         // public
         void sepOn( ostype & );                                                         // turn separator state on
 …
         const char * sepGet( ostype & );                                        // get separator string
         void sepSet( ostype &, const char [] );                         // set separator to string (15 character maximum)
+        void sepSet( ostype &, const char * );                          // set separator to string (15 character maximum)
         const char * sepGetTuple( ostype & );                           // get tuple separator string
         void sepSetTuple( ostype &, const char [] );            // set tuple separator to string (15 character maximum)
+        void sepSetTuple( ostype &, const char * );                     // set tuple separator to string (15 character maximum)
         void ends( ostype & os );                                                       // end of output statement
         int fail( ostype & );
         int flush( ostype & );
         void open( ostype & os, const char name[], const char mode[] );
+        void open( ostype & os, const char * name, const char * mode );
         void close( ostype & os );
         ostype & write( ostype &, const char [], size_t );
+        ostype & write( ostype &, const char *, size_t );
         int fmt( ostype &, const char format[], ... ) __attribute__(( format(printf, 2, 3) ));
 }; // ostream
 …
         ostype & ?|?( ostype &, unsigned long long int );
         void ?|?( ostype &, unsigned long long int );
-#if defined( __SIZEOF_INT128__ )
-        ostype & ?|?( ostype &, int128 );
-        void ?|?( ostype &, int128 );
-        ostype & ?|?( ostype &, unsigned int128 );
-        void ?|?( ostype &, unsigned int128 );
-#endif // __SIZEOF_INT128__
         ostype & ?|?( ostype &, float );
 …
         void ?|?( ostype &, long double _Complex );
         ostype & ?|?( ostype &, const char [] );
         void ?|?( ostype &, const char [] );
+        ostype & ?|?( ostype &, const char * );
+        void ?|?( ostype &, const char * );
         // ostype & ?|?( ostype &, const char16_t * );
 #if ! ( __ARM_ARCH_ISA_ARM == 1 && __ARM_32BIT_STATE == 1 ) // char32_t == wchar_t => ambiguous
 …
 IntegralFMTDecl( signed long long int, 'd' )
 IntegralFMTDecl( unsigned long long int, 'u' )
-#if defined( __SIZEOF_INT128__ )
-IntegralFMTDecl( int128, 'd' )
-IntegralFMTDecl( unsigned int128, 'u' )
-#endif
 //*********************************** floating point ***********************************
 …
 static inline {
         _Ostream_Manip(const char *) bin( const char s[] ) { return (_Ostream_Manip(const char *))@{ s, 1, 0, 'b', { .all : 0 } }; }
         _Ostream_Manip(const char *) oct( const char s[] ) { return (_Ostream_Manip(const char *))@{ s, 1, 0, 'o', { .all : 0 } }; }
         _Ostream_Manip(const char *) hex( const char s[] ) { return (_Ostream_Manip(const char *))@{ s, 1, 0, 'x', { .all : 0 } }; }
         _Ostream_Manip(const char *) wd( unsigned int w, const char s[] ) { return (_Ostream_Manip(const char *))@{ s, w, 0, 's', { .all : 0 } }; }
         _Ostream_Manip(const char *) wd( unsigned int w, unsigned char pc, const char s[] ) { return (_Ostream_Manip(const char *))@{ s, w, pc, 's', { .flags.pc : true } }; }
+        _Ostream_Manip(const char *) bin( const char * s ) { return (_Ostream_Manip(const char *))@{ s, 1, 0, 'b', { .all : 0 } }; }
+        _Ostream_Manip(const char *) oct( const char * s ) { return (_Ostream_Manip(const char *))@{ s, 1, 0, 'o', { .all : 0 } }; }
+        _Ostream_Manip(const char *) hex( const char * s ) { return (_Ostream_Manip(const char *))@{ s, 1, 0, 'x', { .all : 0 } }; }
+        _Ostream_Manip(const char *) wd( unsigned int w, const char * s ) { return (_Ostream_Manip(const char *))@{ s, w, 0, 's', { .all : 0 } }; }
+        _Ostream_Manip(const char *) wd( unsigned int w, unsigned char pc, const char * s ) { return (_Ostream_Manip(const char *))@{ s, w, pc, 's', { .flags.pc : true } }; }
         _Ostream_Manip(const char *) & wd( unsigned int w, _Ostream_Manip(const char *) & fmt ) { fmt.wd = w; return fmt; }
         _Ostream_Manip(const char *) & wd( unsigned int w, unsigned char pc, _Ostream_Manip(const char *) & fmt ) { fmt.wd = w; fmt.pc = pc; fmt.flags.pc = true; return fmt; }
 …
         int fail( istype & );
         int eof( istype & );
         void open( istype & is, const char name[] );
+        void open( istype & is, const char * name );
         void close( istype & is );
         istype & read( istype &, char *, size_t );
 …
         istype & ?|?( istype &, long double _Complex & );
 //      istype & ?|?( istype &, const char [] );
+//      istype & ?|?( istype &, const char * );
         istype & ?|?( istype &, char * );
 …
 static inline {
         _Istream_Cstr skip( unsigned int n ) { return (_Istream_Cstr){ 0p, 0p, n, { .all : 0 } }; }
         _Istream_Cstr skip( const char scanset[] ) { return (_Istream_Cstr){ 0p, scanset, -1, { .all : 0 } }; }
         _Istream_Cstr incl( const char scanset[], char * s ) { return (_Istream_Cstr){ s, scanset, -1, { .flags.inex : false } }; }
         _Istream_Cstr & incl( const char scanset[], _Istream_Cstr & fmt ) { fmt.scanset = scanset; fmt.flags.inex = false; return fmt; }
         _Istream_Cstr excl( const char scanset[], char * s ) { return (_Istream_Cstr){ s, scanset, -1, { .flags.inex : true } }; }
         _Istream_Cstr & excl( const char scanset[], _Istream_Cstr & fmt ) { fmt.scanset = scanset; fmt.flags.inex = true; return fmt; }
         _Istream_Cstr ignore( const char s[] ) { return (_Istream_Cstr)@{ s, 0p, -1, { .flags.ignore : true } }; }
+        _Istream_Cstr skip( const char * scanset ) { return (_Istream_Cstr){ 0p, scanset, -1, { .all : 0 } }; }
+        _Istream_Cstr incl( const char * scanset, char * s ) { return (_Istream_Cstr){ s, scanset, -1, { .flags.inex : false } }; }
+        _Istream_Cstr & incl( const char * scanset, _Istream_Cstr & fmt ) { fmt.scanset = scanset; fmt.flags.inex = false; return fmt; }
+        _Istream_Cstr excl( const char * scanset, char * s ) { return (_Istream_Cstr){ s, scanset, -1, { .flags.inex : true } }; }
+        _Istream_Cstr & excl( const char * scanset, _Istream_Cstr & fmt ) { fmt.scanset = scanset; fmt.flags.inex = true; return fmt; }
+        _Istream_Cstr ignore( const char * s ) { return (_Istream_Cstr)@{ s, 0p, -1, { .flags.ignore : true } }; }
         _Istream_Cstr & ignore( _Istream_Cstr & fmt ) { fmt.flags.ignore = true; return fmt; }
         _Istream_Cstr wdi( unsigned int w, char s[] ) { return (_Istream_Cstr)@{ s, 0p, w, { .all : 0 } }; }
+        _Istream_Cstr wdi( unsigned int w, char * s ) { return (_Istream_Cstr)@{ s, 0p, w, { .all : 0 } }; }
         _Istream_Cstr & wdi( unsigned int w, _Istream_Cstr & fmt ) { fmt.wd = w; return fmt; }
 } // distribution

libcfa/src/math.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Apr 18 23:37:04 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 10:27:11 2020
 // Update Count     : 117
+// Last Modified On : Fri Jul 13 11:02:15 2018
+// Update Count     : 116
 //
 …
 static inline long double fdim( long double x, long double y ) { return fdiml( x, y ); }
 static inline float nan( const char tag[] ) { return nanf( tag ); }
 // extern "C" { double nan( const char [] ); }
 static inline long double nan( const char tag[] ) { return nanl( tag ); }
+static inline float nan( const char * tag ) { return nanf( tag ); }
+// extern "C" { double nan( const char * ); }
+static inline long double nan( const char * tag ) { return nanl( tag ); }
 //---------------------- Exponential ----------------------

libcfa/src/rational.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Apr  6 17:54:28 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb  8 17:56:36 2020
 // Update Count     : 187
+// Last Modified On : Fri Jul 12 18:12:08 2019
+// Update Count     : 184
 //
 …
         } // rational
-        void ?{}( Rational(RationalImpl) & r, zero_t ) {
-                r{ (RationalImpl){0}, (RationalImpl){1} };
-        } // rational
-        void ?{}( Rational(RationalImpl) & r, one_t ) {
-                r{ (RationalImpl){1}, (RationalImpl){1} };
-        } // rational
         // getter for numerator/denominator

libcfa/src/startup.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jul 24 16:21:57 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 13:03:18 2020
 // Update Count     : 30
+// Last Modified On : Wed Jul 25 16:42:01 2018
+// Update Count     : 11
 //
-#include <time.h>                                                                               // tzset
 #include "startup.hfa"
+#include <unistd.h>
 extern "C" {
     void __cfaabi_appready_startup( void ) __attribute__(( constructor( STARTUP_PRIORITY_APPREADY ) ));
+    static void __cfaabi_appready_startup( void ) __attribute__(( constructor( STARTUP_PRIORITY_APPREADY ) ));
     void __cfaabi_appready_startup( void ) {
-                tzset();                                                                                // initialize time global variables
                 #ifdef __CFA_DEBUG__
                 extern void heapAppStart();
 …
     } // __cfaabi_appready_startup
     void __cfaabi_appready_shutdown( void ) __attribute__(( destructor( STARTUP_PRIORITY_APPREADY ) ));
+    static void __cfaabi_appready_shutdown( void ) __attribute__(( destructor( STARTUP_PRIORITY_APPREADY ) ));
     void __cfaabi_appready_shutdown( void ) {
                 #ifdef __CFA_DEBUG__
 …
 struct __spinlock_t;
 extern "C" {
         void __cfaabi_dbg_record_lock(struct __spinlock_t & this, const char prev_name[]) __attribute__(( weak )) {}
+        void __cfaabi_dbg_record(struct __spinlock_t & this, const char * prev_name) __attribute__(( weak )) {}
+}

libcfa/src/stdhdr/assert.h

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Jul  4 23:25:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 12:58:49 2020
 // Update Count     : 15
+// Last Modified On : Mon Jul 31 23:09:32 2017
+// Update Count     : 13
 //
 …
         #define assertf( expr, fmt, ... ) ((expr) ? ((void)0) : __assert_fail_f(__VSTRINGIFY__(expr), __FILE__, __LINE__, __PRETTY_FUNCTION__, fmt, ## __VA_ARGS__ ))
         void __assert_fail_f( const char assertion[], const char file[], unsigned int line, const char function[], const char fmt[], ... ) __attribute__((noreturn, format( printf, 5, 6) ));
+        void __assert_fail_f( const char *assertion, const char *file, unsigned int line, const char *function, const char *fmt, ... ) __attribute__((noreturn, format( printf, 5, 6) ));
 #endif

libcfa/src/stdhdr/bfdlink.h

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jul 18 07:26:04 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:05:08 2020
 // Update Count     : 6
+// Last Modified On : Sun Jul 22 13:49:30 2018
+// Update Count     : 4
 //
 // include file uses the CFA keyword "with".
 #if ! defined( with )                                                                   // nesting ?
 #define with ``with                                                                             // make keyword an identifier
+#define with `with`                                                                             // make keyword an identifier
 #define __CFA_BFDLINK_H__
 #endif

libcfa/src/stdhdr/hwloc.h

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jul 18 07:45:00 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:05:18 2020
 // Update Count     : 6
+// Last Modified On : Sun Jul 22 13:49:58 2018
+// Update Count     : 4
 //
 // include file uses the CFA keyword "thread".
 #if ! defined( thread )                                                                 // nesting ?
 #define thread ``thread                                                                 // make keyword an identifier
+#define thread `thread`                                                                 // make keyword an identifier
 #define __CFA_HWLOC_H__
 #endif

libcfa/src/stdhdr/krb5.h

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Jul 18 07:55:44 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:05:35 2020
 // Update Count     : 6
+// Last Modified On : Sun Jul 22 13:50:24 2018
+// Update Count     : 4
 //
 // include file uses the CFA keyword "enable".
 #if ! defined( enable )                                                                 // nesting ?
 #define enable ``enable                                                                 // make keyword an identifier
+#define enable `enable`                                                                 // make keyword an identifier
 #define __CFA_KRB5_H__
 #endif

libcfa/src/stdhdr/malloc.h

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Jul 20 15:58:16 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Mar  8 10:01:20 2020
 // Update Count     : 11
+// Last Modified On : Sat Aug 11 09:06:31 2018
+// Update Count     : 10
 //
 …
 extern "C" {
-void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize );
 size_t malloc_alignment( void * );
 bool malloc_zero_fill( void * );
-size_t malloc_dimension( void * );
 int malloc_stats_fd( int fd );
+void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize );
 } // extern "C"

libcfa/src/stdhdr/math.h

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Jul  4 23:25:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:05:27 2020
 // Update Count     : 15
+// Last Modified On : Thu Feb 22 18:16:07 2018
+// Update Count     : 13
 //
 extern "C" {
 #if ! defined( exception )                                                              // nesting ?
 #define exception ``exception                                                   // make keyword an identifier
+#define exception `exception`                                                   // make keyword an identifier
 #define __CFA_MATH_H__
 #endif

libcfa/src/stdhdr/sys/ucontext.h

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Feb  8 23:48:16 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:05:41 2020
 // Update Count     : 6
+// Last Modified On : Thu Feb  8 23:50:44 2018
+// Update Count     : 4
 //
 #if ! defined( ftype )                                                                  // nesting ?
 #define ftype ``ftype                                                                   // make keyword an identifier
+#define ftype `ftype`                                                                   // make keyword an identifier
 #define __CFA_UCONTEXT_H__
 #endif

libcfa/src/stdlib.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Jan 28 17:10:29 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Mar 31 13:26:46 2020
 // Update Count     : 495
+// Last Modified On : Tue Oct 22 08:57:52 2019
+// Update Count     : 478
 //
 …
 #define _XOPEN_SOURCE 600                                                               // posix_memalign, *rand48
 #include <string.h>                                                                             // memcpy, memset
+#include <malloc.h>                                                                             // malloc_usable_size
 //#include <math.h>                                                                             // fabsf, fabs, fabsl
 #include <complex.h>                                                                    // _Complex_I
 …
         T * alloc_set( T ptr[], size_t dim, char fill ) {       // realloc array with fill
                 size_t olen = malloc_usable_size( ptr );                // current allocation
                 void * nptr = (void *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
+                char * nptr = (char *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
                 size_t nlen = malloc_usable_size( nptr );               // new allocation
                 if ( nlen > olen ) {                                                    // larger ?
                         memset( (char *)nptr + olen, (int)fill, nlen - olen ); // initialize added storage
+                        memset( nptr + olen, (int)fill, nlen - olen ); // initialize added storage
                 } // if
                 return (T *)nptr;
         } // alloc_set
+        T * alloc_set( T ptr[], size_t dim, T fill ) { // realloc array with fill
+        T * alloc_align( T ptr[], size_t align ) {                      // aligned realloc array
+                char * nptr;
+                size_t alignment = malloc_alignment( ptr );
+                if ( align != alignment && (uintptr_t)ptr % align != 0 ) {
+                        size_t olen = malloc_usable_size( ptr );        // current allocation
+                        nptr = (char *)memalign( align, olen );
+                        size_t nlen = malloc_usable_size( nptr );       // new allocation
+                        size_t lnth = olen < nlen ? olen : nlen;        // min
+                        memcpy( nptr, ptr, lnth );                                      // initialize storage
+                        free( ptr );
+                } else {
+                        nptr = (char *)ptr;
+                } // if
+                return (T *)nptr;
+        } // alloc_align
+        T * alloc_align( T ptr[], size_t align, size_t dim ) { // aligned realloc array
+                char * nptr;
+                size_t alignment = malloc_alignment( ptr );
+                if ( align != alignment ) {
+                        size_t olen = malloc_usable_size( ptr );        // current allocation
+                        nptr = (char *)memalign( align, dim * sizeof(T) );
+                        size_t nlen = malloc_usable_size( nptr );       // new allocation
+                        size_t lnth = olen < nlen ? olen : nlen;        // min
+                        memcpy( nptr, ptr, lnth );                                      // initialize storage
+                        free( ptr );
+                } else {
+                        nptr = (char *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
+                } // if
+                return (T *)nptr;
+        } // alloc_align
+        T * alloc_align_set( T ptr[], size_t align, char fill ) { // aligned realloc with fill
                 size_t olen = malloc_usable_size( ptr );                // current allocation
                 void * nptr = (void *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
+                char * nptr = alloc_align( ptr, align );
                 size_t nlen = malloc_usable_size( nptr );               // new allocation
                 if ( nlen > olen ) {                                                    // larger ?
+                        for ( i; dim ) { memcpy( &ptr[i], &fill, sizeof(T) ); } // initialize with fill value
+                } // if
+                return (T *)nptr;
+        } // alloc_align_set
+        T * alloc_align_set( T ptr[], size_t align, char fill ) { // aligned realloc with fill
+                size_t olen = malloc_usable_size( ptr );                // current allocation
+                void * nptr = (void *)realloc( (void *)ptr, align, sizeof(T) ); // CFA realloc
+                // char * nptr = alloc_align( ptr, align );
+                size_t nlen = malloc_usable_size( nptr );               // new allocation
+                if ( nlen > olen ) {                                                    // larger ?
+                        memset( (char *)nptr + olen, (int)fill, nlen - olen ); // initialize added storage
+                } // if
+                return (T *)nptr;
+        } // alloc_align_set
+        T * alloc_align_set( T ptr[], size_t align, size_t dim, T fill ) { // aligned realloc with fill
+                size_t olen = malloc_usable_size( ptr );                // current allocation
+                void * nptr = (void *)realloc( (void *)ptr, align, sizeof(T) ); // CFA realloc
+                // char * nptr = alloc_align( ptr, align );
+                size_t nlen = malloc_usable_size( nptr );               // new allocation
+                if ( nlen > olen ) {                                                    // larger ?
+                        for ( i; dim ) { memcpy( &ptr[i], &fill, sizeof(T) ); } // initialize with fill value
+                        memset( nptr + olen, (int)fill, nlen - olen ); // initialize added storage
                 } // if
                 return (T *)nptr;
 …
 //---------------------------------------
 float _Complex strto( const char sptr[], char ** eptr ) {
+float _Complex strto( const char * sptr, char ** eptr ) {
         float re, im;
         char * eeptr;
 …
 } // strto
 double _Complex strto( const char sptr[], char ** eptr ) {
+double _Complex strto( const char * sptr, char ** eptr ) {
         double re, im;
         char * eeptr;
 …
 } // strto
 long double _Complex strto( const char sptr[], char ** eptr ) {
+long double _Complex strto( const char * sptr, char ** eptr ) {
         long double re, im;
         char * eeptr;

libcfa/src/stdlib.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Jan 28 17:12:35 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Apr  1 18:38:41 2020
 // Update Count     : 429
+// Last Modified On : Sun Oct 20 22:57:33 2019
+// Update Count     : 390
 //
 …
 #include <stdlib.h>                                                                             // *alloc, strto*, ato*
-// Reduce includes by explicitly defining these routines.
 extern "C" {
         void * memalign( size_t align, size_t size );           // malloc.h
-        size_t malloc_usable_size( void * ptr );                        // malloc.h
-        void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize ); // CFA heap
         void * memset( void * dest, int fill, size_t size ); // string.h
         void * memcpy( void * dest, const void * src, size_t size ); // string.h
         void * resize( void * oaddr, size_t size );                     // CFA heap
+    void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize ); // CFA heap
 } // extern "C"
-void * resize( void * oaddr, size_t nalign, size_t size ); // CFA heap
-void * realloc( void * oaddr, size_t nalign, size_t size ); // CFA heap
 //---------------------------------------
 …
 static inline forall( dtype T | sized(T) ) {
         // Cforall safe equivalents, i.e., implicit size specification
+        // C dynamic allocation
         T * malloc( void ) {
 …
         } // calloc
+        T * realloc( T * ptr, size_t size ) {                           // CFA realloc, eliminate return-type cast
+        T * realloc( T * ptr, size_t size ) {
+                if ( unlikely( ptr == 0 ) ) return malloc();
                 return (T *)(void *)realloc( (void *)ptr, size ); // C realloc
         } // realloc
 …
         } // memalign
-        T * cmemalign( size_t align, size_t dim  ) {
-                return (T *)cmemalign( align, dim, sizeof(T) ); // CFA cmemalign
-        } // cmemalign
         T * aligned_alloc( size_t align ) {
                 return (T *)aligned_alloc( align, sizeof(T) );  // C aligned_alloc
 …
                 return posix_memalign( (void **)ptr, align, sizeof(T) ); // C posix_memalign
         } // posix_memalign
+} // distribution
+static inline forall( dtype T | sized(T) ) {
+        // Cforall safe general allocation, fill, resize, array
+        // Cforall dynamic allocation
         T * alloc( void ) {
 …
         } // alloc
+        forall( dtype S | sized(S) )
+        T * alloc( S ptr[], size_t dim = 1 ) {                          // singleton/array resize
+                size_t len = malloc_usable_size( ptr );                 // current bucket size
+                if ( sizeof(T) * dim > len ) {                                  // not enough space ?
+                        T * temp = alloc( dim );                                        // new storage
+                        free( ptr );                                                            // free old storage
+                        return temp;
+                } else {
+                        return (T *)ptr;
+                } // if
+        } // alloc
+        T * alloc( T ptr[], size_t dim, bool copy = true ) {
+                if ( copy ) {                                                                   // realloc
+                        return (T *)(void *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
+                } else {
+                        struct __Unknown {};
+                        return alloc( (__Unknown *)ptr, dim );          // reuse, cheat making T/S different types
+                } // if
+        T * alloc( T ptr[], size_t dim ) {                                      // realloc
+                return realloc( ptr, dim * sizeof(T) );
         } // alloc
 …
 forall( dtype T | sized(T) ) {
         T * alloc_set( T ptr[], size_t dim, char fill );        // realloc array with fill
-        T * alloc_set( T ptr[], size_t dim, T fill );           // realloc array with fill
 } // distribution
 …
         T * alloc_align( size_t align, size_t dim ) {
                 return (T *)memalign( align, dim * sizeof(T) );
-        } // alloc_align
-        T * alloc_align( T ptr[], size_t align ) {                      // aligned realloc array
-                return (T *)(void *)realloc( (void *)ptr, align, sizeof(T) ); // CFA realloc
-        } // alloc_align
-        forall( dtype S | sized(S) )
-        T * alloc_align( S ptr[], size_t align ) {                      // aligned reuse array
-                return (T *)(void *)resize( (void *)ptr, align, sizeof(T) ); // CFA realloc
-        } // alloc_align
-        T * alloc_align( T ptr[], size_t align, size_t dim ) { // aligned realloc array
-                return (T *)(void *)realloc( (void *)ptr, align, dim * sizeof(T) ); // CFA realloc
         } // alloc_align
 …
 forall( dtype T | sized(T) ) {
         T * alloc_align_set( T ptr[], size_t align, char fill ); // aligned realloc with fill
         T * alloc_align_set( T ptr[], size_t align, T fill ); // aligned realloc with fill
+        T * alloc_align( T ptr[], size_t align );                       // realign
+        T * alloc_align( T ptr[], size_t align, size_t dim ); // aligned realloc array
         T * alloc_align_set( T ptr[], size_t align, size_t dim, char fill ); // aligned realloc array with fill
+        T * alloc_align_set( T ptr[], size_t align, size_t dim, T fill ); // aligned realloc array with fill
+} // distribution
+static inline forall( dtype T | sized(T) ) {
+        // Cforall safe initialization/copy, i.e., implicit size specification, non-array types
+} // distribution
+static inline forall( dtype T | sized(T) ) {
+        // data, non-array types
         T * memset( T * dest, char fill ) {
                 return (T *)memset( dest, fill, sizeof(T) );
 …
 static inline forall( dtype T | sized(T) ) {
         // Cforall safe initialization/copy, i.e., implicit size specification, array types
+        // data, array types
         T * amemset( T dest[], char fill, size_t dim ) {
                 return (T *)(void *)memset( dest, fill, dim * sizeof(T) ); // C memset
 …
 } // distribution
 // Cforall allocation/deallocation and constructor/destructor, non-array types
+// allocation/deallocation and constructor/destructor, non-array types
 forall( dtype T | sized(T), ttype Params | { void ?{}( T &, Params ); } ) T * new( Params p );
 forall( dtype T | sized(T) | { void ^?{}( T & ); } ) void delete( T * ptr );
 forall( dtype T, ttype Params | sized(T) | { void ^?{}( T & ); void delete( Params ); } ) void delete( T * ptr, Params rest );
 // Cforall allocation/deallocation and constructor/destructor, array types
+// allocation/deallocation and constructor/destructor, array types
 forall( dtype T | sized(T), ttype Params | { void ?{}( T &, Params ); } ) T * anew( size_t dim, Params p );
 forall( dtype T | sized(T) | { void ^?{}( T & ); } ) void adelete( size_t dim, T arr[] );
 …
 static inline {
         int strto( const char sptr[], char ** eptr, int base ) { return (int)strtol( sptr, eptr, base ); }
         unsigned int strto( const char sptr[], char ** eptr, int base ) { return (unsigned int)strtoul( sptr, eptr, base ); }
         long int strto( const char sptr[], char ** eptr, int base ) { return strtol( sptr, eptr, base ); }
         unsigned long int strto( const char sptr[], char ** eptr, int base ) { return strtoul( sptr, eptr, base ); }
         long long int strto( const char sptr[], char ** eptr, int base ) { return strtoll( sptr, eptr, base ); }
         unsigned long long int strto( const char sptr[], char ** eptr, int base ) { return strtoull( sptr, eptr, base ); }
         float strto( const char sptr[], char ** eptr ) { return strtof( sptr, eptr ); }
         double strto( const char sptr[], char ** eptr ) { return strtod( sptr, eptr ); }
         long double strto( const char sptr[], char ** eptr ) { return strtold( sptr, eptr ); }
 } // distribution
 float _Complex strto( const char sptr[], char ** eptr );
 double _Complex strto( const char sptr[], char ** eptr );
 long double _Complex strto( const char sptr[], char ** eptr );
+        int strto( const char * sptr, char ** eptr, int base ) { return (int)strtol( sptr, eptr, base ); }
+        unsigned int strto( const char * sptr, char ** eptr, int base ) { return (unsigned int)strtoul( sptr, eptr, base ); }
+        long int strto( const char * sptr, char ** eptr, int base ) { return strtol( sptr, eptr, base ); }
+        unsigned long int strto( const char * sptr, char ** eptr, int base ) { return strtoul( sptr, eptr, base ); }
+        long long int strto( const char * sptr, char ** eptr, int base ) { return strtoll( sptr, eptr, base ); }
+        unsigned long long int strto( const char * sptr, char ** eptr, int base ) { return strtoull( sptr, eptr, base ); }
+        float strto( const char * sptr, char ** eptr ) { return strtof( sptr, eptr ); }
+        double strto( const char * sptr, char ** eptr ) { return strtod( sptr, eptr ); }
+        long double strto( const char * sptr, char ** eptr ) { return strtold( sptr, eptr ); }
+} // distribution
+float _Complex strto( const char * sptr, char ** eptr );
+double _Complex strto( const char * sptr, char ** eptr );
+long double _Complex strto( const char * sptr, char ** eptr );
 static inline {
         int ato( const char sptr[] ) { return (int)strtol( sptr, 0p, 10 ); }
         unsigned int ato( const char sptr[] ) { return (unsigned int)strtoul( sptr, 0p, 10 ); }
         long int ato( const char sptr[] ) { return strtol( sptr, 0p, 10 ); }
         unsigned long int ato( const char sptr[] ) { return strtoul( sptr, 0p, 10 ); }
         long long int ato( const char sptr[] ) { return strtoll( sptr, 0p, 10 ); }
         unsigned long long int ato( const char sptr[] ) { return strtoull( sptr, 0p, 10 ); }
         float ato( const char sptr[] ) { return strtof( sptr, 0p ); }
         double ato( const char sptr[] ) { return strtod( sptr, 0p ); }
         long double ato( const char sptr[] ) { return strtold( sptr, 0p ); }
         float _Complex ato( const char sptr[] ) { return strto( sptr, 0p ); }
         double _Complex ato( const char sptr[] ) { return strto( sptr, 0p ); }
         long double _Complex ato( const char sptr[] ) { return strto( sptr, 0p ); }
+        int ato( const char * sptr ) { return (int)strtol( sptr, 0, 10 ); }
+        unsigned int ato( const char * sptr ) { return (unsigned int)strtoul( sptr, 0, 10 ); }
+        long int ato( const char * sptr ) { return strtol( sptr, 0, 10 ); }
+        unsigned long int ato( const char * sptr ) { return strtoul( sptr, 0, 10 ); }
+        long long int ato( const char * sptr ) { return strtoll( sptr, 0, 10 ); }
+        unsigned long long int ato( const char * sptr ) { return strtoull( sptr, 0, 10 ); }
+        float ato( const char * sptr ) { return strtof( sptr, 0 ); }
+        double ato( const char * sptr ) { return strtod( sptr, 0 ); }
+        long double ato( const char * sptr ) { return strtold( sptr, 0 ); }
+        float _Complex ato( const char * sptr ) { return strto( sptr, NULL ); }
+        double _Complex ato( const char * sptr ) { return strto( sptr, NULL ); }
+        long double _Complex ato( const char * sptr ) { return strto( sptr, NULL ); }
 } // distribution

libcfa/src/time.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Mar 27 13:33:14 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 08:24:18 2020
 // Update Count     : 70
+// Last Modified On : Sat Jul 13 08:41:55 2019
+// Update Count     : 65
 //
 …
 forall( dtype ostype | ostream( ostype ) ) {
         ostype & ?|?( ostype & os, Duration dur ) with( dur ) {
                 (ostype &)(os | tn / TIMEGRAN);                                 // print seconds
                 long int ns = (tn < 0 ? -tn : tn) % TIMEGRAN;   // compute nanoseconds
+                (ostype &)(os | tv / TIMEGRAN);                                 // print seconds
+                long int ns = (tv < 0 ? -tv : tv) % TIMEGRAN;   // compute nanoseconds
                 if ( ns != 0 ) {                                                                // some ?
                         char buf[16];
 …
 #ifdef __CFA_DEBUG__
 static void tabort( int year, int month, int day, int hour, int min, int sec, int64_t nsec ) {
+static void tabort( int year, int month, int day, int hour, int min, int sec, int nsec ) {
         abort | "Attempt to create Time( year=" | year | "(>=1970), month=" | month | "(1-12), day=" | day | "(1-31), hour=" | hour | "(0-23), min=" | min | "(0-59), sec=" | sec
                   | "(0-60), nsec=" | nsec | "(0-999_999_999), which is not in the range 00:00:00 UTC, January 1, 1970 to 03:14:07 UTC, January 19, 2038, where month and day have 1 origin.";
+                  | "(0-60), nsec=" | nsec | "(0-999_999_999), which exceeds range 00:00:00 UTC, January 1, 1970 to 03:14:07 UTC, January 19, 2038.";
 } // tabort
 #endif // __CFA_DEBUG__
 void ?{}( Time & time, int year, int month, int day, int hour, int min, int sec, int64_t nsec ) with( time ) {
+void ?{}( Time & time, int year, int month, int day, int hour, int min, int sec, int nsec ) with( time ) {
         tm tm;
         // Values can be in any range (+/-) but result must be in the epoch.
+        tm.tm_isdst = -1;                                                                       // let mktime determine if alternate timezone is in effect
         tm.tm_year = year - 1900;                                                       // mktime uses 1900 as its starting point
+        // Make month in range 1-12 to match with day.
+#ifdef __CFA_DEBUG__
+        if ( month < 1 || 12 < month ) {
+                tabort( year, month, day, hour, min, sec, nsec );
+        } // if
+#endif // __CFA_DEBUG__
         tm.tm_mon = month - 1;                                                          // mktime uses range 0-11
+#ifdef __CFA_DEBUG__
+        if ( day < 1 || 31 < day ) {
+                tabort( year, month, day, hour, min, sec, nsec );
+        } // if
+#endif // __CFA_DEBUG__
         tm.tm_mday = day;                                                                       // mktime uses range 1-31
         tm.tm_hour = hour;
         tm.tm_min = min;
         tm.tm_sec = sec;
-        tm.tm_isdst = -1;                                                                       // let mktime determine if alternate timezone is in effect
         time_t epochsec = mktime( &tm );
 #ifdef __CFA_DEBUG__
         if ( epochsec <= (time_t)-1 ) {                                         // MUST BE LESS THAN OR EQUAL!
+        if ( epochsec == (time_t)-1 ) {
                 tabort( year, month, day, hour, min, sec, nsec );
         } // if
 #endif // __CFA_DEBUG__
         tn = (int64_t)(epochsec) * TIMEGRAN + nsec;                     // convert to nanoseconds
+        tv = (int64_t)(epochsec) * TIMEGRAN + nsec;                     // convert to nanoseconds
 #ifdef __CFA_DEBUG__
         if ( tn > 2147483647LL * TIMEGRAN ) {                           // between 00:00:00 UTC, January 1, 1970 and 03:14:07 UTC, January 19, 2038.
+        if ( tv > 2147483647LL * TIMEGRAN ) {                           // between 00:00:00 UTC, January 1, 1970 and 03:14:07 UTC, January 19, 2038.
                 tabort( year, month, day, hour, min, sec, nsec );
         } // if
 …
 char * yy_mm_dd( Time time, char * buf ) with( time ) {
         time_t s = tn / TIMEGRAN;
+        time_t s = tv / TIMEGRAN;
         tm tm;
         gmtime_r( &s, &tm );                                                            // tm_mon <= 11, tm_mday <= 31
 …
 char * mm_dd_yy( Time time, char * buf ) with( time ) {
         time_t s = tn / TIMEGRAN;
+        time_t s = tv / TIMEGRAN;
         tm tm;
         gmtime_r( &s, &tm );                                                            // tm_mon <= 11, tm_mday <= 31
 …
 char * dd_mm_yy( Time time, char * buf ) with( time ) {
         time_t s = tn / TIMEGRAN;
+        time_t s = tv / TIMEGRAN;
         tm tm;
         gmtime_r( &s, &tm );                                                            // tm_mon <= 11, tm_mday <= 31
 …
 } // dd_mm_yy
 size_t strftime( char buf[], size_t size, const char fmt[], Time time ) with( time ) {
         time_t s = tn / TIMEGRAN;
+size_t strftime( char * buf, size_t size, const char * fmt, Time time ) with( time ) {
+        time_t s = tv / TIMEGRAN;
         tm tm;
         gmtime_r( &s, &tm );
 …
         ostype & ?|?( ostype & os, Time time ) with( time ) {
                 char buf[32];                                                                   // at least 26
                 time_t s = tn / TIMEGRAN;
+                time_t s = tv / TIMEGRAN;
                 ctime_r( &s, (char *)&buf );                                    // 26 characters: "Wed Jun 30 21:49:08 1993\n"
                 buf[24] = '\0';                                                                 // remove trailing '\n'
                 long int ns = (tn < 0 ? -tn : tn) % TIMEGRAN;   // compute nanoseconds
+                long int ns = (tv < 0 ? -tv : tv) % TIMEGRAN;   // compute nanoseconds
                 if ( ns == 0 ) {                                                                // none ?
                         (ostype &)(os | buf);                                           // print date/time/year

libcfa/src/time.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Mar 14 23:18:57 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 08:24:32 2020
 // Update Count     : 654
+// Last Modified On : Sat Sep 22 12:25:34 2018
+// Update Count     : 643
 //
 …
         Duration ?=?( Duration & dur, __attribute__((unused)) zero_t ) { return dur{ 0 }; }
         Duration +?( Duration rhs ) with( rhs ) { return (Duration)@{ +tn }; }
         Duration ?+?( Duration & lhs, Duration rhs ) { return (Duration)@{ lhs.tn + rhs.tn }; }
+        Duration +?( Duration rhs ) with( rhs ) {       return (Duration)@{ +tv }; }
+        Duration ?+?( Duration & lhs, Duration rhs ) { return (Duration)@{ lhs.tv + rhs.tv }; }
         Duration ?+=?( Duration & lhs, Duration rhs ) { lhs = lhs + rhs; return lhs; }
         Duration -?( Duration rhs ) with( rhs ) { return (Duration)@{ -tn }; }
         Duration ?-?( Duration & lhs, Duration rhs ) { return (Duration)@{ lhs.tn - rhs.tn }; }
+        Duration -?( Duration rhs ) with( rhs ) { return (Duration)@{ -tv }; }
+        Duration ?-?( Duration & lhs, Duration rhs ) { return (Duration)@{ lhs.tv - rhs.tv }; }
         Duration ?-=?( Duration & lhs, Duration rhs ) { lhs = lhs - rhs; return lhs; }
         Duration ?*?( Duration lhs, int64_t rhs ) { return (Duration)@{ lhs.tn * rhs }; }
         Duration ?*?( int64_t lhs, Duration rhs ) { return (Duration)@{ lhs * rhs.tn }; }
+        Duration ?*?( Duration lhs, int64_t rhs ) { return (Duration)@{ lhs.tv * rhs }; }
+        Duration ?*?( int64_t lhs, Duration rhs ) { return (Duration)@{ lhs * rhs.tv }; }
         Duration ?*=?( Duration & lhs, int64_t rhs ) { lhs = lhs * rhs; return lhs; }
         int64_t ?/?( Duration lhs, Duration rhs ) { return lhs.tn / rhs.tn; }
         Duration ?/?( Duration lhs, int64_t rhs ) { return (Duration)@{ lhs.tn / rhs }; }
+        int64_t ?/?( Duration lhs, Duration rhs ) { return lhs.tv / rhs.tv; }
+        Duration ?/?( Duration lhs, int64_t rhs ) { return (Duration)@{ lhs.tv / rhs }; }
         Duration ?/=?( Duration & lhs, int64_t rhs ) { lhs = lhs / rhs; return lhs; }
         double div( Duration lhs, Duration rhs ) { return (double)lhs.tn / (double)rhs.tn; }
         Duration ?%?( Duration lhs, Duration rhs ) { return (Duration)@{ lhs.tn % rhs.tn }; }
+        double div( Duration lhs, Duration rhs ) { return (double)lhs.tv / (double)rhs.tv; }
+        Duration ?%?( Duration lhs, Duration rhs ) { return (Duration)@{ lhs.tv % rhs.tv }; }
         Duration ?%=?( Duration & lhs, Duration rhs ) { lhs = lhs % rhs; return lhs; }
         bool ?==?( Duration lhs, Duration rhs ) { return lhs.tn == rhs.tn; }
         bool ?!=?( Duration lhs, Duration rhs ) { return lhs.tn != rhs.tn; }
         bool ?<? ( Duration lhs, Duration rhs ) { return lhs.tn <  rhs.tn; }
         bool ?<=?( Duration lhs, Duration rhs ) { return lhs.tn <= rhs.tn; }
         bool ?>? ( Duration lhs, Duration rhs ) { return lhs.tn >  rhs.tn; }
         bool ?>=?( Duration lhs, Duration rhs ) { return lhs.tn >= rhs.tn; }
         bool ?==?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tn == 0; }
         bool ?!=?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tn != 0; }
         bool ?<? ( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tn <  0; }
         bool ?<=?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tn <= 0; }
         bool ?>? ( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tn >  0; }
         bool ?>=?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tn >= 0; }
         Duration abs( Duration rhs ) { return rhs.tn >= 0 ? rhs : -rhs; }
+        bool ?==?( Duration lhs, Duration rhs ) { return lhs.tv == rhs.tv; }
+        bool ?!=?( Duration lhs, Duration rhs ) { return lhs.tv != rhs.tv; }
+        bool ?<? ( Duration lhs, Duration rhs ) { return lhs.tv <  rhs.tv; }
+        bool ?<=?( Duration lhs, Duration rhs ) { return lhs.tv <= rhs.tv; }
+        bool ?>? ( Duration lhs, Duration rhs ) { return lhs.tv >  rhs.tv; }
+        bool ?>=?( Duration lhs, Duration rhs ) { return lhs.tv >= rhs.tv; }
+        bool ?==?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tv == 0; }
+        bool ?!=?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tv != 0; }
+        bool ?<? ( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tv <  0; }
+        bool ?<=?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tv <= 0; }
+        bool ?>? ( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tv >  0; }
+        bool ?>=?( Duration lhs, __attribute__((unused)) zero_t ) { return lhs.tv >= 0; }
+        Duration abs( Duration rhs ) { return rhs.tv >= 0 ? rhs : -rhs; }
         Duration ?`ns( int64_t nsec ) { return (Duration)@{ nsec }; }
 …
         Duration ?`w( double weeks ) { return (Duration)@{ weeks * (7LL * 24LL * 60LL * 60LL * TIMEGRAN) }; }
         int64_t ?`ns( Duration dur ) { return dur.tn; }
         int64_t ?`us( Duration dur ) { return dur.tn / (TIMEGRAN / 1_000_000LL); }
         int64_t ?`ms( Duration dur ) { return dur.tn / (TIMEGRAN / 1_000LL); }
         int64_t ?`s( Duration dur ) { return dur.tn / TIMEGRAN; }
         int64_t ?`m( Duration dur ) { return dur.tn / (60LL * TIMEGRAN); }
         int64_t ?`h( Duration dur ) { return dur.tn / (60LL * 60LL * TIMEGRAN); }
         int64_t ?`d( Duration dur ) { return dur.tn / (24LL * 60LL * 60LL * TIMEGRAN); }
         int64_t ?`w( Duration dur ) { return dur.tn / (7LL * 24LL * 60LL * 60LL * TIMEGRAN); }
         Duration max( Duration lhs, Duration rhs ) { return  (lhs.tn < rhs.tn) ? rhs : lhs;}
         Duration min( Duration lhs, Duration rhs ) { return !(rhs.tn < lhs.tn) ? lhs : rhs;}
+        int64_t ?`ns( Duration dur ) { return dur.tv; }
+        int64_t ?`us( Duration dur ) { return dur.tv / (TIMEGRAN / 1_000_000LL); }
+        int64_t ?`ms( Duration dur ) { return dur.tv / (TIMEGRAN / 1_000LL); }
+        int64_t ?`s( Duration dur ) { return dur.tv / TIMEGRAN; }
+        int64_t ?`m( Duration dur ) { return dur.tv / (60LL * TIMEGRAN); }
+        int64_t ?`h( Duration dur ) { return dur.tv / (60LL * 60LL * TIMEGRAN); }
+        int64_t ?`d( Duration dur ) { return dur.tv / (24LL * 60LL * 60LL * TIMEGRAN); }
+        int64_t ?`w( Duration dur ) { return dur.tv / (7LL * 24LL * 60LL * 60LL * TIMEGRAN); }
+        Duration max( Duration lhs, Duration rhs ) { return  (lhs.tv < rhs.tv) ? rhs : lhs;}
+        Duration min( Duration lhs, Duration rhs ) { return !(rhs.tv < lhs.tv) ? lhs : rhs;}
 } // distribution
 …
 //######################### Time #########################
 void ?{}( Time & time, int year, int month = 1, int day = 1, int hour = 0, int min = 0, int sec = 0, int64_t nsec = 0 );
+void ?{}( Time & time, int year, int month = 0, int day = 0, int hour = 0, int min = 0, int sec = 0, int nsec = 0 );
 static inline {
         Time ?=?( Time & time, __attribute__((unused)) zero_t ) { return time{ 0 }; }
         void ?{}( Time & time, timeval t ) with( time ) { tn = (int64_t)t.tv_sec * TIMEGRAN + t.tv_usec * 1000; }
+        void ?{}( Time & time, timeval t ) with( time ) { tv = (int64_t)t.tv_sec * TIMEGRAN + t.tv_usec * 1000; }
         Time ?=?( Time & time, timeval t ) with( time ) {
                 tn = (int64_t)t.tv_sec * TIMEGRAN + t.tv_usec * (TIMEGRAN / 1_000_000LL);
+                tv = (int64_t)t.tv_sec * TIMEGRAN + t.tv_usec * (TIMEGRAN / 1_000_000LL);
                 return time;
         } // ?=?
         void ?{}( Time & time, timespec t ) with( time ) { tn = (int64_t)t.tv_sec * TIMEGRAN + t.tv_nsec; }
+        void ?{}( Time & time, timespec t ) with( time ) { tv = (int64_t)t.tv_sec * TIMEGRAN + t.tv_nsec; }
         Time ?=?( Time & time, timespec t ) with( time ) {
                 tn = (int64_t)t.tv_sec * TIMEGRAN + t.tv_nsec;
+                tv = (int64_t)t.tv_sec * TIMEGRAN + t.tv_nsec;
                 return time;
         } // ?=?
         Time ?+?( Time & lhs, Duration rhs ) { return (Time)@{ lhs.tn + rhs.tn }; }
+        Time ?+?( Time & lhs, Duration rhs ) { return (Time)@{ lhs.tv + rhs.tv }; }
         Time ?+?( Duration lhs, Time rhs ) { return rhs + lhs; }
         Time ?+=?( Time & lhs, Duration rhs ) { lhs = lhs + rhs; return lhs; }
         Duration ?-?( Time lhs, Time rhs ) { return (Duration)@{ lhs.tn - rhs.tn }; }
         Time ?-?( Time lhs, Duration rhs ) { return (Time)@{ lhs.tn - rhs.tn }; }
+        Duration ?-?( Time lhs, Time rhs ) { return (Duration)@{ lhs.tv - rhs.tv }; }
+        Time ?-?( Time lhs, Duration rhs ) { return (Time)@{ lhs.tv - rhs.tv }; }
         Time ?-=?( Time & lhs, Duration rhs ) { lhs = lhs - rhs; return lhs; }
+        bool ?==?( Time lhs, Time rhs ) { return lhs.tn == rhs.tn; }
+        bool ?!=?( Time lhs, Time rhs ) { return lhs.tn != rhs.tn; }
+        bool ?<?( Time lhs, Time rhs ) { return lhs.tn < rhs.tn; }
+        bool ?<=?( Time lhs, Time rhs ) { return lhs.tn <= rhs.tn; }
+        bool ?>?( Time lhs, Time rhs ) { return lhs.tn > rhs.tn; }
+        bool ?>=?( Time lhs, Time rhs ) { return lhs.tn >= rhs.tn; }
+        int64_t ?`ns( Time t ) { return t.tn; }
+        bool ?==?( Time lhs, Time rhs ) { return lhs.tv == rhs.tv; }
+        bool ?!=?( Time lhs, Time rhs ) { return lhs.tv != rhs.tv; }
+        bool ?<?( Time lhs, Time rhs ) { return lhs.tv < rhs.tv; }
+        bool ?<=?( Time lhs, Time rhs ) { return lhs.tv <= rhs.tv; }
+        bool ?>?( Time lhs, Time rhs ) { return lhs.tv > rhs.tv; }
+        bool ?>=?( Time lhs, Time rhs ) { return lhs.tv >= rhs.tv; }
 } // distribution
 …
 } // dmy
 size_t strftime( char buf[], size_t size, const char fmt[], Time time );
+size_t strftime( char * buf, size_t size, const char * fmt, Time time );
 //------------------------- timeval (cont) -------------------------
 static inline void ?{}( timeval & t, Time time ) with( t, time ) {
         tv_sec = tn / TIMEGRAN;                                                         // seconds
         tv_usec = tn % TIMEGRAN / (TIMEGRAN / 1_000_000LL);     // microseconds
+        tv_sec = tv / TIMEGRAN;                                                         // seconds
+        tv_usec = tv % TIMEGRAN / (TIMEGRAN / 1_000_000LL);     // microseconds
 } // ?{}
 …
 static inline void ?{}( timespec & t, Time time ) with( t, time ) {
         tv_sec = tn / TIMEGRAN;                                                         // seconds
         tv_nsec = tn % TIMEGRAN;                                                        // nanoseconds
+        tv_sec = tv / TIMEGRAN;                                                         // seconds
+        tv_nsec = tv % TIMEGRAN;                                                        // nanoseconds
 } // ?{}

libcfa/src/time_t.hfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Apr 10 14:42:03 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Jan  5 08:22:46 2020
 // Update Count     : 7
+// Last Modified On : Fri Apr 13 07:51:47 2018
+// Update Count     : 6
 //
 …
 struct Duration {                                                                               // private
         int64_t tn;                                                                                     // nanoseconds
+        int64_t tv;                                                                                     // nanoseconds
 }; // Duration
 static inline void ?{}( Duration & dur ) with( dur ) { tn = 0; }
 static inline void ?{}( Duration & dur, __attribute__((unused)) zero_t ) with( dur ) { tn = 0; }
+static inline void ?{}( Duration & dur ) with( dur ) { tv = 0; }
+static inline void ?{}( Duration & dur, __attribute__((unused)) zero_t ) with( dur ) { tv = 0; }
 …
 struct Time {                                                                                   // private
         uint64_t tn;                                                                            // nanoseconds since UNIX epoch
+        uint64_t tv;                                                                            // nanoseconds since UNIX epoch
 }; // Time
 static inline void ?{}( Time & time ) with( time ) { tn = 0; }
 static inline void ?{}( Time & time, __attribute__((unused)) zero_t ) with( time ) { tn = 0; }
+static inline void ?{}( Time & time ) with( time ) { tv = 0; }
+static inline void ?{}( Time & time, __attribute__((unused)) zero_t ) with( time ) { tv = 0; }
 // Local Variables: //

longrun_tests/Makefile.in

-              r7030dab
+              r71d6bd8
 LTCFACOMPILE = $(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) \
         $(LIBTOOLFLAGS) --mode=compile $(CFACC) $(DEFS) \
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(AM_CFLAGS) $(CFAFLAGS) $(CFLAGS)
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) \
+        $(AM_CFLAGS) $(CFLAGS)
 AM_V_CFA = $(am__v_CFA_@AM_V@)
 …
 am__v_CFA_0 = @echo "  CFA     " $@;
 am__v_CFA_1 =
+AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
+am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
+am__v_JAVAC_0 = @echo "  JAVAC   " $@;
+am__v_JAVAC_1 =
+AM_V_GOC = $(am__v_GOC_@AM_V@)
+am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
+am__v_GOC_0 = @echo "  GOC     " $@;
+am__v_GOC_1 =
 UPPCC = u++
 UPPCOMPILE = $(UPPCC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_UPPFLAGS) $(UPPFLAGS) $(AM_CXXFLAGS) $(CXXFLAGS) $(AM_CFLAGS) $(CFLAGS)
 …
 am__v_UPP_0 = @echo "  UPP     " $@;
 am__v_UPP_1 =
-AM_V_GOC = $(am__v_GOC_@AM_V@)
-am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
-am__v_GOC_0 = @echo "  GOC     " $@;
-am__v_GOC_1 =
-AM_V_PY = $(am__v_PY_@AM_V@)
-am__v_PY_ = $(am__v_PY_@AM_DEFAULT_V@)
-am__v_PY_0 = @echo "  PYTHON  " $@;
-am__v_PY_1 =
-AM_V_RUST = $(am__v_RUST_@AM_V@)
-am__v_RUST_ = $(am__v_RUST_@AM_DEFAULT_V@)
-am__v_RUST_0 = @echo "  RUST    " $@;
-am__v_RUST_1 =
-AM_V_NODEJS = $(am__v_NODEJS_@AM_V@)
-am__v_NODEJS_ = $(am__v_NODEJS_@AM_DEFAULT_V@)
-am__v_NODEJS_0 = @echo "  NODEJS  " $@;
-am__v_NODEJS_1 =
-AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
-am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
-am__v_JAVAC_0 = @echo "  JAVAC   " $@;
-am__v_JAVAC_1 =
 repeats = 10
 max_time = 600

src/AST/Convert.cpp

-              r7030dab
+              r71d6bd8
 // Created On       : Thu May 09 15::37::05 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 21:39:32 2019
 // Update Count     : 33
+// Last Modified On : Thu Jul 25 22:21:46 2019
+// Update Count     : 13
 //
 …
                 auto decl = new StructDecl(
                         node->name,
                         (AggregateDecl::Aggregate)node->kind,
+                        node->kind,
                         get<Attribute>().acceptL( node->attributes ),
                         LinkageSpec::Spec( node->linkage.val )
 …
+        }
-        const ast::Stmt * visit(const ast::SuspendStmt * node ) override final {
-                if ( inCache( node ) ) return nullptr;
-                auto stmt = new SuspendStmt();
-                stmt->then   = get<CompoundStmt>().accept1( node->then   );
-                switch(node->type) {
-                        case ast::SuspendStmt::None     : stmt->type = SuspendStmt::None     ; break;
-                        case ast::SuspendStmt::Coroutine: stmt->type = SuspendStmt::Coroutine; break;
-                        case ast::SuspendStmt::Generator: stmt->type = SuspendStmt::Generator; break;
+                }
-                return stmtPostamble( stmt, node );
+        }
         const ast::Stmt * visit( const ast::WaitForStmt * node ) override final {
                 if ( inCache( node ) ) return nullptr;
 …
         const ast::Expr * visit( const ast::KeywordCastExpr * node ) override final {
+                AggregateDecl::Aggregate castTarget = (AggregateDecl::Aggregate)node->target;
+                assert( AggregateDecl::Generator <= castTarget && castTarget <= AggregateDecl::Thread );
+                KeywordCastExpr::Target castTarget = KeywordCastExpr::NUMBER_OF_TARGETS;
+                switch (node->target) {
+                        case ast::KeywordCastExpr::Coroutine:
+                                castTarget = KeywordCastExpr::Coroutine;
+                                break;
+                        case ast::KeywordCastExpr::Thread:
+                                castTarget = KeywordCastExpr::Thread;
+                                break;
+                        case ast::KeywordCastExpr::Monitor:
+                                castTarget = KeywordCastExpr::Monitor;
+                                break;
+                        default:
+                                break;
+                }
+                assert ( castTarget < KeywordCastExpr::NUMBER_OF_TARGETS );
                 auto expr = visitBaseExpr( node,
                         new KeywordCastExpr(
 …
                                 cv( node ),
                                 node->name,
                                 node->kind == ast::TypeDecl::Ftype,
+                                node->kind == ast::TypeVar::Ftype,
                                 get<Attribute>().acceptL( node->attributes )
                         };
 …
                         old->location,
                         old->name,
                         (ast::AggregateDecl::Aggregate)old->kind,
+                        old->kind,
                         GET_ACCEPT_V(attributes, Attribute),
                         { old->linkage.val }
 …
                         { old->storageClasses.val },
                         GET_ACCEPT_1(base, Type),
                         (ast::TypeDecl::Kind)(unsigned)old->kind,
+                        (ast::TypeVar::Kind)(unsigned)old->kind,
                         old->sized,
                         GET_ACCEPT_1(init, Type)
 …
+        }
-        virtual void visit( const SuspendStmt * old ) override final {
-                if ( inCache( old ) ) return;
-                ast::SuspendStmt::Type type;
-                switch (old->type) {
-                        case SuspendStmt::Coroutine: type = ast::SuspendStmt::Coroutine; break;
-                        case SuspendStmt::Generator: type = ast::SuspendStmt::Generator; break;
-                        case SuspendStmt::None     : type = ast::SuspendStmt::None     ; break;
-                        default: abort();
+                }
-                this->node = new ast::SuspendStmt(
-                        old->location,
-                        GET_ACCEPT_1(then  , CompoundStmt),
-                        type,
-                        GET_LABELS_V(old->labels)
-                );
-                cache.emplace( old, this->node );
+        }
         virtual void visit( const WaitForStmt * old ) override final {
                 if ( inCache( old ) ) return;
 …
+        }
+        virtual void visit( const KeywordCastExpr * old ) override final {
+                ast::AggregateDecl::Aggregate castTarget = (ast::AggregateDecl::Aggregate)old->target;
+                assert( ast::AggregateDecl::Generator <= castTarget && castTarget <= ast::AggregateDecl::Thread );
+        virtual void visit( const KeywordCastExpr * old) override final {
+                ast::KeywordCastExpr::Target castTarget = ast::KeywordCastExpr::NUMBER_OF_TARGETS;
+                switch (old->target) {
+                        case KeywordCastExpr::Coroutine:
+                                castTarget = ast::KeywordCastExpr::Coroutine;
+                                break;
+                        case KeywordCastExpr::Thread:
+                                castTarget = ast::KeywordCastExpr::Thread;
+                                break;
+                        case KeywordCastExpr::Monitor:
+                                castTarget = ast::KeywordCastExpr::Monitor;
+                                break;
+                        default:
+                                break;
+                }
+                assert ( castTarget < ast::KeywordCastExpr::NUMBER_OF_TARGETS );
                 this->node = visitBaseExpr( old,
                         new ast::KeywordCastExpr(
 …
                         ty = new ast::TypeInstType{
                                 old->name,
                                 old->isFtype ? ast::TypeDecl::Ftype : ast::TypeDecl::Dtype,
+                                old->isFtype ? ast::TypeVar::Ftype : ast::TypeVar::Dtype,
                                 cv( old ),
                                 GET_ACCEPT_V( attributes, Attribute )

src/AST/Decl.cpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Thu May 9 10:00:00 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 16:23:15 2019
 // Update Count     : 20
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Thu May 9 10:00:00 2019
+// Update Count     : 1
 //
 …
 #include <cassert>             // for assert, strict_dynamic_cast
 #include <iostream>
+#include <string>
 #include <unordered_map>
 …
 #include "Node.hpp"            // for readonly
 #include "Type.hpp"            // for readonly
+#include "Parser/ParseNode.h"  // for DeclarationNode
 namespace ast {
 …
 // --- TypeDecl
+const char * TypeDecl::typeString() const {
+        static const char * kindNames[] = { "sized data type", "sized object type", "sized function type", "sized tuple type" };
+        static_assert( sizeof(kindNames)/sizeof(kindNames[0]) == TypeDecl::NUMBER_OF_KINDS, "typeString: kindNames is out of sync." );
+        assertf( kind < TypeDecl::NUMBER_OF_KINDS, "TypeDecl kind is out of bounds." );
+        return sized ? kindNames[ kind ] : &kindNames[ kind ][ sizeof("sized") ]; // sizeof includes '\0'
+std::string TypeDecl::typeString() const {
+        static const std::string kindNames[] = { "object type", "function type", "tuple type" };
+        assertf( sizeof(kindNames)/sizeof(kindNames[0]) == DeclarationNode::NoTypeClass-1,
+                "typeString: kindNames is out of sync." );
+        assertf( kind < sizeof(kindNames)/sizeof(kindNames[0]), "TypeDecl's kind is out of bounds." );
+        return (sized ? "sized " : "") + kindNames[ kind ];
+}
 const char * TypeDecl::genTypeString() const {
         static const char * kindNames[] = { "dtype", "otype", "ftype", "ttype" };
         static_assert( sizeof(kindNames)/sizeof(kindNames[0]) == TypeDecl::NUMBER_OF_KINDS, "genTypeString: kindNames is out of sync." );
         assertf( kind < TypeDecl::NUMBER_OF_KINDS, "TypeDecl kind is out of bounds." );
+std::string TypeDecl::genTypeString() const {
+        static const std::string kindNames[] = { "dtype", "ftype", "ttype" };
+        assertf( sizeof(kindNames)/sizeof(kindNames[0]) == DeclarationNode::NoTypeClass-1, "genTypeString: kindNames is out of sync." );
+        assertf( kind < sizeof(kindNames)/sizeof(kindNames[0]), "TypeDecl's kind is out of bounds." );
         return kindNames[ kind ];
+}
 …
 std::ostream & operator<< ( std::ostream & out, const TypeDecl::Data & data ) {
         return out << data.kind << ", " << data.isComplete;
+}
-// --- AggregateDecl
-// These must harmonize with the corresponding AggregateDecl::Aggregate enumerations.
-static const char * aggregateNames[] = { "struct", "union", "enum", "exception", "trait", "generator", "coroutine", "monitor", "thread", "NoAggregateName" };
-const char * AggregateDecl::aggrString( AggregateDecl::Aggregate aggr ) {
-        return aggregateNames[aggr];
+}

src/AST/Decl.hpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Thu May 9 10:00:00 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 17:38:33 2019
 // Update Count     : 29
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Thu May 9 10:00:00 2019
+// Update Count     : 1
 //
 …
 #include <unordered_map>
 #include <vector>
-#include <algorithm>
 #include "FunctionSpec.hpp"
 …
 #include "ParseNode.hpp"
 #include "StorageClasses.hpp"
+#include "TypeVar.hpp"
 #include "Visitor.hpp"
+#include "Common/utility.h"
+#include "Common/SemanticError.h"                                               // error_str
+#include "Parser/ParseNode.h"  // for DeclarationNode::Aggregate
 // Must be included in *all* AST classes; should be #undef'd at the end of the file
 …
         std::vector< ptr<Expr> > withExprs;
         FunctionDecl( const CodeLocation & loc, const std::string & name, FunctionType * type,
+        FunctionDecl( const CodeLocation & loc, const std::string &name, FunctionType * type,
                 CompoundStmt * stmts, Storage::Classes storage = {}, Linkage::Spec linkage = Linkage::C,
                 std::vector<ptr<Attribute>>&& attrs = {}, Function::Specs fs = {})
 …
         bool has_body() const { return stmts; }
         const DeclWithType * accept( Visitor & v ) const override { return v.visit( this ); }
+        const DeclWithType * accept( Visitor &v ) const override { return v.visit( this ); }
 private:
         FunctionDecl * clone() const override { return new FunctionDecl( *this ); }
 …
         std::vector<ptr<DeclWithType>> assertions;
         NamedTypeDecl(
+        NamedTypeDecl(
                 const CodeLocation & loc, const std::string & name, Storage::Classes storage,
                 const Type * b, Linkage::Spec spec = Linkage::Cforall )
 …
         /// Produces a name for the kind of alias
         virtual const char * typeString() const = 0;
+        virtual std::string typeString() const = 0;
 private:
 …
 /// Cforall type variable: `dtype T`
 class TypeDecl final : public NamedTypeDecl {
+  public:
+        enum Kind { Dtype, Otype, Ftype, Ttype, NUMBER_OF_KINDS };
+        Kind kind;
+public:
+        TypeVar::Kind kind;
         bool sized;
         ptr<Type> init;
 …
         /// Data extracted from a type decl
         struct Data {
                 Kind kind;
+                TypeVar::Kind kind;
                 bool isComplete;
                 Data() : kind( NUMBER_OF_KINDS ), isComplete( false ) {}
+                Data() : kind( (TypeVar::Kind)-1 ), isComplete( false ) {}
                 Data( const TypeDecl * d ) : kind( d->kind ), isComplete( d->sized ) {}
                 Data( Kind k, bool c ) : kind( k ), isComplete( c ) {}
+                Data( TypeVar::Kind k, bool c ) : kind( k ), isComplete( c ) {}
                 Data( const Data & d1, const Data & d2 )
+                        : kind( d1.kind ), isComplete( d1.isComplete || d2.isComplete ) {}
+                bool operator==( const Data & o ) const { return kind == o.kind && isComplete == o.isComplete; }
+                bool operator!=( const Data & o ) const { return !(*this == o); }
+                : kind( d1.kind ), isComplete( d1.isComplete || d2.isComplete ) {}
+                bool operator== ( const Data & o ) const {
+                        return kind == o.kind && isComplete == o.isComplete;
+                }
+                bool operator!= ( const Data & o ) const { return !(*this == o); }
         };
         TypeDecl(
                 const CodeLocation & loc, const std::string & name, Storage::Classes storage,
+        TypeDecl(
+                const CodeLocation & loc, const std::string & name, Storage::Classes storage,
                 const Type * b, TypeVar::Kind k, bool s, const Type * i = nullptr )
         : NamedTypeDecl( loc, name, storage, b ), kind( k ), sized( k == TypeVar::Ttype || s ),
           init( i ) {}
         const char * typeString() const override;
+        std::string typeString() const override;
         /// Produces a name for generated code
         const char * genTypeString() const;
+        std::string genTypeString() const;
         /// convenience accessor to match Type::isComplete()
 …
         const Decl * accept( Visitor & v ) const override { return v.visit( this ); }
   private:
+private:
         TypeDecl * clone() const override { return new TypeDecl{ *this }; }
         MUTATE_FRIEND
 …
         : NamedTypeDecl( loc, name, storage, b, spec ) {}
         const char * typeString() const override { return "typedef"; }
+        std::string typeString() const override { return "typedef"; }
         const Decl * accept( Visitor & v ) const override { return v.visit( this ); }
 …
 class AggregateDecl : public Decl {
 public:
-        enum Aggregate { Struct, Union, Enum, Exception, Trait, Generator, Coroutine, Monitor, Thread, NoAggregate };
-        static const char * aggrString( Aggregate aggr );
         std::vector<ptr<Decl>> members;
         std::vector<ptr<TypeDecl>> params;
 …
         /// Produces a name for the kind of aggregate
         virtual const char * typeString() const = 0;
+        virtual std::string typeString() const = 0;
 private:
 …
 class StructDecl final : public AggregateDecl {
 public:
         Aggregate kind;
+        DeclarationNode::Aggregate kind;
         StructDecl( const CodeLocation& loc, const std::string& name,
                 Aggregate kind = Struct,
+                DeclarationNode::Aggregate kind = DeclarationNode::Struct,
                 std::vector<ptr<Attribute>>&& attrs = {}, Linkage::Spec linkage = Linkage::Cforall )
         : AggregateDecl( loc, name, std::move(attrs), linkage ), kind( kind ) {}
+        bool is_coroutine() { return kind == Coroutine; }
+        bool is_generator() { return kind == Generator; }
+        bool is_monitor  () { return kind == Monitor  ; }
+        bool is_thread   () { return kind == Thread   ; }
+        const Decl * accept( Visitor & v ) const override { return v.visit( this ); }
+        const char * typeString() const override { return aggrString( kind ); }
+        bool is_coroutine() { return kind == DeclarationNode::Coroutine; }
+        bool is_monitor() { return kind == DeclarationNode::Monitor; }
+        bool is_thread() { return kind == DeclarationNode::Thread; }
+        const Decl * accept( Visitor & v ) const override { return v.visit( this ); }
+        std::string typeString() const override { return "struct"; }
 private:
 …
         const Decl * accept( Visitor& v ) const override { return v.visit( this ); }
         const char * typeString() const override { return aggrString( Union ); }
+        std::string typeString() const override { return "union"; }
 private:
 …
         const Decl * accept( Visitor & v ) const override { return v.visit( this ); }
         const char * typeString() const override { return aggrString( Enum ); }
+        std::string typeString() const override { return "enum"; }
 private:
 …
         const Decl * accept( Visitor & v ) const override { return v.visit( this ); }
         const char * typeString() const override { return "trait"; }
+        std::string typeString() const override { return "trait"; }
 private:
 …
         ptr<AsmStmt> stmt;
         AsmDecl( const CodeLocation & loc, AsmStmt * stmt )
+        AsmDecl( const CodeLocation & loc, AsmStmt *stmt )
         : Decl( loc, "", {}, {} ), stmt(stmt) {}
         const AsmDecl * accept( Visitor & v ) const override { return v.visit( this ); }
 private:
         AsmDecl * clone() const override { return new AsmDecl( *this ); }
+        const AsmDecl * accept( Visitor &v ) const override { return v.visit( this ); }
+private:
+        AsmDecl *clone() const override { return new AsmDecl( *this ); }
         MUTATE_FRIEND
 };
 …
         : Decl( loc, "", {}, {} ), cond( condition ), msg( msg ) {}
         const StaticAssertDecl * accept( Visitor & v ) const override { return v.visit( this ); }
+        const StaticAssertDecl * accept( Visitor &v ) const override { return v.visit( this ); }
 private:
         StaticAssertDecl * clone() const override { return new StaticAssertDecl( *this ); }

src/AST/Expr.cpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Wed May 15 17:00:00 2019
 // Last Modified By : Peter A. Buhr
 // Created On       : Thr Jun 13 13:38:00 2019
 // Update Count     : 6
+// Last Modified By : Andrew Beach
+// Created On       : Fri Oct  4 15:34:00 2019
+// Update Count     : 4
 //
 …
 // --- KeywordCastExpr
+const char * KeywordCastExpr::targetString() const {
+        return AggregateDecl::aggrString( target );
+const std::string & KeywordCastExpr::targetString() const {
+        static const std::string targetStrs[] = {
+                "coroutine", "thread", "monitor"
+        };
+        static_assert(
+                (sizeof(targetStrs) / sizeof(targetStrs[0])) == ((unsigned long)NUMBER_OF_TARGETS),
+                "Each KeywordCastExpr::Target should have a corresponding string representation"
+        );
+        return targetStrs[(unsigned long)target];
+}

src/AST/Expr.hpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Fri May 10 10:30:00 2019
 // Last Modified By : Peter A. Buhr
 // Created On       : Fri May 10 10:30:00 2019
 // Update Count     : 7
+// Last Modified By : Andrew Beach
+// Created On       : Thr Sep 26 12:51:00 2019
+// Update Count     : 2
 //
 …
 #include "Fwd.hpp"        // for UniqueId
 #include "Label.hpp"
-#include "Decl.hpp"
 #include "ParseNode.hpp"
 #include "Visitor.hpp"
 …
 public:
         ptr<Expr> arg;
         ast::AggregateDecl::Aggregate target;
         KeywordCastExpr( const CodeLocation & loc, const Expr * a, ast::AggregateDecl::Aggregate t )
+        enum Target { Coroutine, Thread, Monitor, NUMBER_OF_TARGETS } target;
+        KeywordCastExpr( const CodeLocation & loc, const Expr * a, Target t )
         : Expr( loc ), arg( a ), target( t ) {}
         /// Get a name for the target type
         const char * targetString() const;
+        const std::string& targetString() const;
         const Expr * accept( Visitor & v ) const override { return v.visit( this ); }

src/AST/Fwd.hpp

r7030dab	r71d6bd8
53	53	class CatchStmt;
54	54	class FinallyStmt;
55		~~class SuspendStmt;~~
56	55	class WaitForStmt;
57	56	class WithStmt;

src/AST/Pass.hpp

r7030dab	r71d6bd8
114	114	const ast::Stmt * visit( const ast::CatchStmt * ) override final;
115	115	const ast::Stmt * visit( const ast::FinallyStmt * ) override final;
116		~~const ast::Stmt * visit( const ast::SuspendStmt * ) override final;~~
117	116	const ast::Stmt * visit( const ast::WaitForStmt * ) override final;
118	117	const ast::Decl * visit( const ast::WithStmt * ) override final;

src/AST/Pass.impl.hpp

-              r7030dab
+              r71d6bd8
 //--------------------------------------------------------------------------
-// FinallyStmt
-template< typename pass_t >
-const ast::Stmt * ast::Pass< pass_t >::visit( const ast::SuspendStmt * node ) {
-        VISIT_START( node );
-        VISIT(
-                maybe_accept( node, &SuspendStmt::then   );
+        )
-        VISIT_END( Stmt, node );
+}
-//--------------------------------------------------------------------------
 // WaitForStmt
 template< typename pass_t >

src/AST/Print.cpp

-              r7030dab
+              r71d6bd8
                 safe_print( node->body );
                 --indent;
-                return node;
+        }
-        virtual const ast::Stmt * visit( const ast::SuspendStmt * node ) override final {
-                os << "Suspend Statement";
-                switch (node->type) {
-                        case ast::SuspendStmt::None     : os << " with implicit target"; break;
-                        case ast::SuspendStmt::Generator: os << " for generator"; break;
-                        case ast::SuspendStmt::Coroutine: os << " for coroutine"; break;
+                }
-                os << endl;
-                ++indent;
-                if(node->then) {
-                        os << indent << " with post statement :" << endl;
-                        safe_print( node->then );
+                }
-                ++indent;
                 return node;
 …
                 preprint( node );
                 os << "instance of type " << node->name
                    << " (" << (node->kind == ast::TypeDecl::Ftype ? "" : "not ") << "function type)";
+                   << " (" << (node->kind == ast::TypeVar::Ftype ? "" : "not ") << "function type)";
                 print( node->params );

src/AST/Stmt.hpp

-              r7030dab
+              r71d6bd8
 };
-/// Suspend statement
-class SuspendStmt final : public Stmt {
-public:
-        ptr<CompoundStmt> then;
-        enum Type { None, Coroutine, Generator } type = None;
-        SuspendStmt( const CodeLocation & loc, const CompoundStmt * then, Type type, std::vector<Label> && labels = {} )
-        : Stmt(loc, std::move(labels)), then(then), type(type) {}
-        const Stmt * accept( Visitor & v ) const override { return v.visit( this ); }
-private:
-        SuspendStmt * clone() const override { return new SuspendStmt{ *this }; }
-        MUTATE_FRIEND
-};
 /// Wait for concurrency statement `when (...) waitfor (... , ...) ... timeout(...) ... else ...`
 class WaitForStmt final : public Stmt {

src/AST/Type.cpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Mon May 13 15:00:00 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Dec 15 16:56:28 2019
 // Update Count     : 4
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Mon May 13 15:00:00 2019
+// Update Count     : 1
 //
 …
 // --- BasicType
+// GENERATED START, DO NOT EDIT
+// GENERATED BY BasicTypes-gen.cc
+const char * BasicType::typeNames[] = {
+const char *BasicType::typeNames[] = {
         "_Bool",
         "char",
 …
         "_Float128x _Complex",
 };
+// GENERATED END
+static_assert(
+        sizeof(BasicType::typeNames)/sizeof(BasicType::typeNames[0]) == BasicType::NUMBER_OF_BASIC_TYPES,
+        "Each basic type name should have a corresponding kind enum value"
+);
 // --- ParameterizedType

src/AST/Type.hpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Thu May 9 10:00:00 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 21:56:46 2019
 // Update Count     : 5
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Thu May 9 10:00:00 2019
+// Update Count     : 1
 //
 …
 #include "Fwd.hpp"
 #include "Node.hpp"          // for Node, ptr, ptr_base
+#include "TypeVar.hpp"
 #include "Visitor.hpp"
 …
 public:
         readonly<TypeDecl> base;
         TypeDecl::Kind kind;
+        TypeVar::Kind kind;
         TypeInstType(
 …
                 std::vector<ptr<Attribute>> && as = {} )
         : ReferenceToType( n, q, std::move(as) ), base( b ), kind( b->kind ) {}
+        TypeInstType( const std::string& n, TypeDecl::Kind k, CV::Qualifiers q = {},
+        TypeInstType(
+                const std::string& n, TypeVar::Kind k, CV::Qualifiers q = {},
                 std::vector<ptr<Attribute>> && as = {} )
         : ReferenceToType( n, q, std::move(as) ), base(), kind( k ) {}

src/AST/TypeEnvironment.cpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Wed May 29 11:00:00 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 21:49:13 2019
 // Update Count     : 4
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Wed May 29 11:00:00 2019
+// Update Count     : 1
 //
 …
                 return true;
         } else if ( auto typeInst = dynamic_cast< const TypeInstType * >( type ) ) {
                 return typeInst->kind == TypeDecl::Ftype;
+                return typeInst->kind == TypeVar::Ftype;
         } else return false;
+}
 …
         bool tyVarCompatible( const TypeDecl::Data & data, const Type * type ) {
                 switch ( data.kind ) {
                   case TypeDecl::Dtype:
+                  case TypeVar::Dtype:
                         // to bind to an object type variable, the type must not be a function type.
                         // if the type variable is specified to be a complete type then the incoming
 …
                         // xxx - should this also check that type is not a tuple type and that it's not a ttype?
                         return ! isFtype( type ) && ( ! data.isComplete || type->isComplete() );
                   case TypeDecl::Ftype:
+                  case TypeVar::Ftype:
                         return isFtype( type );
                   case TypeDecl::Ttype:
+                  case TypeVar::Ttype:
                         // ttype unifies with any tuple type
                         return dynamic_cast< const TupleType * >( type ) || Tuples::isTtype( type );

src/AST/TypeEnvironment.hpp

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Wed May 29 11:00:00 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 21:55:54 2019
 // Update Count     : 3
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Wed May 29 11:00:00 2019
+// Update Count     : 1
 //
 …
 #include "Type.hpp"
 #include "TypeSubstitution.hpp"
+#include "TypeVar.hpp"
 #include "Common/Indenter.h"
 #include "ResolvExpr/WidenMode.h"
 …
         /// Singleton class constructor from substitution
         EqvClass( const std::string & v, const Type * b )
         : vars{ v }, bound( b ), allowWidening( false ), data( TypeDecl::Dtype, false ) {}
+        : vars{ v }, bound( b ), allowWidening( false ), data( TypeVar::Dtype, false ) {}
         /// Single-var constructor (strips qualifiers from bound type)

src/AST/Visitor.hpp

r7030dab	r71d6bd8
47	47	virtual const ast::Stmt * visit( const ast::CatchStmt * ) = 0;
48	48	virtual const ast::Stmt * visit( const ast::FinallyStmt * ) = 0;
49		~~virtual const ast::Stmt * visit( const ast::SuspendStmt * ) = 0;~~
50	49	virtual const ast::Stmt * visit( const ast::WaitForStmt * ) = 0;
51	50	virtual const ast::Decl * visit( const ast::WithStmt * ) = 0;

src/AST/module.mk

-              r7030dab
+              r71d6bd8
 ## Author           : Thierry Delisle
 ## Created On       : Thu May 09 16:05:36 2019
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Sat Dec 14 07:29:10 2019
 ## Update Count     : 3
+## Last Modified By :
+## Last Modified On :
+## Update Count     :
 ###############################################################################
 …
         AST/TypeSubstitution.cpp
 SRC += $(SRC_AST)
 SRCDEMANGLE += $(SRC_AST)

src/BasicTypes-gen.cc

-              r7030dab
+              r71d6bd8
         #define TypeH TOP_SRCDIR "src/SynTree/Type.h"
         resetInput( file, TypeH, buffer, code, str );
         if ( (start = str.find( STARTMK )) == string::npos ) Abort( "start", TypeH );
+        #define Type TOP_SRCDIR "src/SynTree/Type.h"
+        resetInput( file, Type, buffer, code, str );
+        if ( (start = str.find( STARTMK )) == string::npos ) Abort( "start", Type );
         start += sizeof( STARTMK );                                                     // includes newline
         code << str.substr( 0, start );
 …
         code << "\t";                                                                           // indentation for end marker
         if ( (start = str.find( ENDMK, start + 1 )) == string::npos ) Abort( "end", TypeH );
+        if ( (start = str.find( ENDMK, start + 1 )) == string::npos ) Abort( "end", Type );
         code << str.substr( start );
         output( file, TypeH, code );
+        output( file, Type, code );
         // cout << code.str();
-        #define TypeC TOP_SRCDIR "src/SynTree/Type.cc"
-        resetInput( file, TypeC, buffer, code, str );
-        if ( (start = str.find( STARTMK )) == string::npos ) Abort( "start", TypeC );
-        start += sizeof( STARTMK );                                                     // includes newline
-        code << str.substr( 0, start );
-        code << BYMK << endl;
-        code << "const char * BasicType::typeNames[] = {" << endl;
-        for ( int r = 0; r < NUMBER_OF_BASIC_TYPES; r += 1 ) {
-                code << "\t\"" << graph[r].type << "\"," << endl;
-        } // for
-        code << "};" << endl;
-        if ( (start = str.find( ENDMK, start + 1 )) == string::npos ) Abort( "end", TypeC );
-        code << str.substr( start );
-        output( file, TypeC, code );
-        // cout << code.str();
         // TEMPORARY DURING CHANGE OVER
         #define TypeH_AST TOP_SRCDIR "src/AST/Type.hpp"
         resetInput( file, TypeH_AST, buffer, code, str );
         if ( (start = str.find( STARTMK )) == string::npos ) Abort( "start", TypeH_AST );
+        #define TypeAST TOP_SRCDIR "src/AST/Type.hpp"
+        resetInput( file, TypeAST, buffer, code, str );
+        if ( (start = str.find( STARTMK )) == string::npos ) Abort( "start", TypeAST );
         start += sizeof( STARTMK );                                                     // includes newline
         code << str.substr( 0, start );
 …
         code << "\t";                                                                           // indentation for end marker
         if ( (start = str.find( ENDMK, start + 1 )) == string::npos ) Abort( "end", TypeH_AST );
+        if ( (start = str.find( ENDMK, start + 1 )) == string::npos ) Abort( "end", TypeAST );
         code << str.substr( start );
+        output( file, TypeH_AST, code );
+        // cout << code.str();
+        #define TypeC_AST TOP_SRCDIR "src/AST/Type.cpp"
+        resetInput( file, TypeC_AST, buffer, code, str );
+        if ( (start = str.find( STARTMK )) == string::npos ) Abort( "start", TypeC_AST );
+        start += sizeof( STARTMK );                                                     // includes newline
+        code << str.substr( 0, start );
+        code << BYMK << endl;
+        code << "const char * BasicType::typeNames[] = {" << endl;
+        for ( int r = 0; r < NUMBER_OF_BASIC_TYPES; r += 1 ) {
+                code << "\t\"" << graph[r].type << "\"," << endl;
+        } // for
+        code << "};" << endl;
+        if ( (start = str.find( ENDMK, start + 1 )) == string::npos ) Abort( "end", TypeC_AST );
+        code << str.substr( start );
+        output( file, TypeC_AST, code );
+        output( file, TypeAST, code );
         // cout << code.str();

src/CodeGen/CodeGenerator.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 08:32:48 2020
 // Update Count     : 532
+// Last Modified On : Sat Oct 19 19:30:38 2019
+// Update Count     : 506
 //
 #include "CodeGenerator.h"
 …
 #include "InitTweak/InitTweak.h"     // for getPointerBase
 #include "OperatorTable.h"           // for OperatorInfo, operatorLookup
 #include "SynTree/LinkageSpec.h"     // for Spec, Intrinsic
+#include "Parser/LinkageSpec.h"      // for Spec, Intrinsic
 #include "SynTree/Attribute.h"       // for Attribute
 #include "SynTree/BaseSyntaxNode.h"  // for BaseSyntaxNode
 …
         int CodeGenerator::tabsize = 4;
         // The kinds of statements that would ideally be followed by whitespace.
+        // the kinds of statements that would ideally be followed by whitespace
         bool wantSpacing( Statement * stmt) {
                 return dynamic_cast< IfStmt * >( stmt ) || dynamic_cast< CompoundStmt * >( stmt ) ||
 …
+        }
+        // Using updateLocation at the beginning of a node and endl within a node should become the method of formating.
+        /* Using updateLocation at the beginning of a node and endl
+         * within a node should become the method of formating.
+         */
         void CodeGenerator::updateLocation( CodeLocation const & to ) {
                 // skip if linemarks shouldn't appear or if codelocation is unset
 …
                 } else {
                         output << "\n# " << to.first_line << " \"" << to.filename
                                    << "\"\n" << indent;
+                               << "\"\n" << indent;
                         currentLocation = to;
+                }
 …
         void CodeGenerator::genAttributes( list< Attribute * > & attributes ) {
                 if ( attributes.empty() ) return;
+          if ( attributes.empty() ) return;
                 output << "__attribute__ ((";
                 for ( list< Attribute * >::iterator attr( attributes.begin() );; ) {
 …
                                 output << ")";
                         } // if
                         if ( ++attr == attributes.end() ) break;
+                  if ( ++attr == attributes.end() ) break;
                         output << ",";                                                          // separator
                 } // for
 …
                 previsit( (BaseSyntaxNode *)node );
                 GuardAction( [this, node](){
                                 if ( options.printExprTypes && node->result ) {
                                         output << " /* " << genType( node->result, "", options ) << " */ ";
+                                }
                         } );
+                        if ( options.printExprTypes && node->result ) {
+                                output << " /* " << genType( node->result, "", options ) << " */ ";
+                        }
+                } );
+        }
 …
                 extension( applicationExpr );
                 if ( VariableExpr * varExpr = dynamic_cast< VariableExpr* >( applicationExpr->get_function() ) ) {
                         const OperatorInfo * opInfo;
                         if ( varExpr->get_var()->get_linkage() == LinkageSpec::Intrinsic && ( opInfo = operatorLookup( varExpr->get_var()->get_name() ) ) ) {
+                        OperatorInfo opInfo;
+                        if ( varExpr->get_var()->get_linkage() == LinkageSpec::Intrinsic && operatorLookup( varExpr->get_var()->get_name(), opInfo ) ) {
                                 std::list< Expression* >::iterator arg = applicationExpr->get_args().begin();
                                 switch ( opInfo->type ) {
+                                switch ( opInfo.type ) {
                                   case OT_INDEX:
                                         assert( applicationExpr->get_args().size() == 2 );
 …
                                                 output << "(";
                                                 (*arg++)->accept( *visitor );
                                                 output << ") /* " << opInfo->inputName << " */";
+                                                output << ") /* " << opInfo.inputName << " */";
                                         } else if ( applicationExpr->get_args().size() == 2 ) {
                                                 // intrinsic two parameter constructors are essentially bitwise assignment
                                                 output << "(";
                                                 (*arg++)->accept( *visitor );
                                                 output << opInfo->symbol;
+                                                output << opInfo.symbol;
                                                 (*arg)->accept( *visitor );
                                                 output << ") /* " << opInfo->inputName << " */";
+                                                output << ") /* " << opInfo.inputName << " */";
                                         } else {
                                                 // no constructors with 0 or more than 2 parameters
 …
                                         assert( applicationExpr->get_args().size() == 1 );
                                         output << "(";
                                         output << opInfo->symbol;
+                                        output << opInfo.symbol;
                                         (*arg)->accept( *visitor );
                                         output << ")";
 …
                                         assert( applicationExpr->get_args().size() == 1 );
                                         (*arg)->accept( *visitor );
                                         output << opInfo->symbol;
+                                        output << opInfo.symbol;
                                         break;
 …
                                         output << "(";
                                         (*arg++)->accept( *visitor );
                                         output << opInfo->symbol;
+                                        output << opInfo.symbol;
                                         (*arg)->accept( *visitor );
                                         output << ")";
 …
                 extension( untypedExpr );
                 if ( NameExpr * nameExpr = dynamic_cast< NameExpr* >( untypedExpr->function ) ) {
                         const OperatorInfo * opInfo = operatorLookup( nameExpr->name );
                         if ( opInfo ) {
+                        OperatorInfo opInfo;
+                        if ( operatorLookup( nameExpr->name, opInfo ) ) {
                                 std::list< Expression* >::iterator arg = untypedExpr->args.begin();
                                 switch ( opInfo->type ) {
+                                switch ( opInfo.type ) {
                                   case OT_INDEX:
                                         assert( untypedExpr->args.size() == 2 );
 …
                                                 output << "(";
                                                 (*arg++)->accept( *visitor );
                                                 output << ") /* " << opInfo->inputName << " */";
+                                                output << ") /* " << opInfo.inputName << " */";
                                         } else if ( untypedExpr->get_args().size() == 2 ) {
                                                 // intrinsic two parameter constructors are essentially bitwise assignment
                                                 output << "(";
                                                 (*arg++)->accept( *visitor );
                                                 output << opInfo->symbol;
+                                                output << opInfo.symbol;
                                                 (*arg)->accept( *visitor );
                                                 output << ") /* " << opInfo->inputName << " */";
+                                                output << ") /* " << opInfo.inputName << " */";
                                         } else {
                                                 // no constructors with 0 or more than 2 parameters
 …
                                                 output << "(";
                                                 (*arg++)->accept( *visitor );
                                                 output << opInfo->symbol << "{ ";
+                                                output << opInfo.symbol << "{ ";
                                                 genCommaList( arg, untypedExpr->args.end() );
                                                 output << "}) /* " << opInfo->inputName << " */";
+                                                output << "}) /* " << opInfo.inputName << " */";
                                         } // if
                                         break;
 …
                                         assert( untypedExpr->args.size() == 1 );
                                         output << "(";
                                         output << opInfo->symbol;
+                                        output << opInfo.symbol;
                                         (*arg)->accept( *visitor );
                                         output << ")";
 …
                                         assert( untypedExpr->args.size() == 1 );
                                         (*arg)->accept( *visitor );
                                         output << opInfo->symbol;
+                                        output << opInfo.symbol;
                                         break;
 …
                                         output << "(";
                                         (*arg++)->accept( *visitor );
                                         output << opInfo->symbol;
+                                        output << opInfo.symbol;
                                         (*arg)->accept( *visitor );
                                         output << ")";
 …
         void CodeGenerator::postvisit( NameExpr * nameExpr ) {
                 extension( nameExpr );
                 const OperatorInfo * opInfo = operatorLookup( nameExpr->name );
                 if ( opInfo ) {
                         if ( opInfo->type == OT_CONSTANT ) {
                                 output << opInfo->symbol;
+                OperatorInfo opInfo;
+                if ( operatorLookup( nameExpr->name, opInfo ) ) {
+                        if ( opInfo.type == OT_CONSTANT ) {
+                                output << opInfo.symbol;
                         } else {
                                 output << opInfo->outputName;
+                                output << opInfo.outputName;
+                        }
                 } else {
 …
         void CodeGenerator::postvisit( VariableExpr * variableExpr ) {
                 extension( variableExpr );
                 const OperatorInfo * opInfo;
                 if ( variableExpr->get_var()->get_linkage() == LinkageSpec::Intrinsic && (opInfo = operatorLookup( variableExpr->get_var()->get_name() )) && opInfo->type == OT_CONSTANT ) {
                         output << opInfo->symbol;
+                OperatorInfo opInfo;
+                if ( variableExpr->get_var()->get_linkage() == LinkageSpec::Intrinsic && operatorLookup( variableExpr->get_var()->get_name(), opInfo ) && opInfo.type == OT_CONSTANT ) {
+                        output << opInfo.symbol;
                 } else {
                         output << mangleName( variableExpr->get_var() );
 …
                   case BranchStmt::FallThroughDefault:
                         assertf( ! options.genC, "fallthru should not reach code generation." );
                         output << "fallthru";
+                  output << "fallthru";
                         break;
                 } // switch
 …
                 output << ((throwStmt->get_kind() == ThrowStmt::Terminate) ?
                                    "throw" : "throwResume");
+                           "throw" : "throwResume");
                 if (throwStmt->get_expr()) {
                         output << " ";
 …
                 output << ((stmt->get_kind() == CatchStmt::Terminate) ?
                                    "catch" : "catchResume");
+                "catch" : "catchResume");
                 output << "( ";
                 stmt->decl->accept( *visitor );
 …
         std::string genName( DeclarationWithType * decl ) {
                 const OperatorInfo * opInfo = operatorLookup( decl->get_name() );
                 if ( opInfo ) {
                         return opInfo->outputName;
+                CodeGen::OperatorInfo opInfo;
+                if ( operatorLookup( decl->get_name(), opInfo ) ) {
+                        return opInfo.outputName;
                 } else {
                         return decl->get_name();

src/CodeGen/CodeGenerator.h

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 03:58:31 2020
 // Update Count     : 62
+// Last Modified By : Andrew Beach
+// Last Modified On : Tue Apr 30 12:01:00 2019
+// Update Count     : 57
 //
 …
 namespace CodeGen {
         struct CodeGenerator : public WithShortCircuiting, public WithGuards, public WithVisitorRef<CodeGenerator> {
                 static int tabsize;
+          static int tabsize;
                 CodeGenerator( std::ostream &os, bool pretty = false, bool genC = false, bool lineMarks = false, bool printExprTypes = false );
 …
                 void postvisit( AsmStmt * );
                 void postvisit( DirectiveStmt * );
                 void postvisit( AsmDecl * );                                    // special: statement in declaration context
+                void postvisit( AsmDecl * );                            // special: statement in declaration context
                 void postvisit( IfStmt * );
                 void postvisit( SwitchStmt * );
 …
                 LabelPrinter printLabels;
                 Options options;
           public:
+        public:
                 LineEnder endl;
           private:
+        private:
                 CodeLocation currentLocation;
 …
         template< class Iterator >
         void CodeGenerator::genCommaList( Iterator begin, Iterator end ) {
                 if ( begin == end ) return;
+          if ( begin == end ) return;
                 for ( ;; ) {
                         (*begin++)->accept( *visitor );
                         if ( begin == end ) break;
+                  if ( begin == end ) break;
                         output << ", ";                                                         // separator
                 } // for

src/CodeGen/FixMain.h

-              r7030dab
+              r71d6bd8
 // Created On       : Thr Jan 12 14:11:09 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 03:24:32 2020
 // Update Count     : 5
+// Last Modified On : Fri Jul 21 22:16:59 2017
+// Update Count     : 1
 //
 …
 #include <memory>
 #include "SynTree/LinkageSpec.h"
+#include "Parser/LinkageSpec.h"
 class FunctionDecl;
 …
                 static std::unique_ptr<FunctionDecl> main_signature;
         };
 } // namespace CodeGen
+};

src/CodeGen/FixNames.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:39:14 2019
 // Update Count     : 21
+// Last Modified By : Andrew Beach
+// Last Modified On : Wed Jun 28 15:26:00 2017
+// Update Count     : 20
 //
 …
 #include "Common/SemanticError.h"  // for SemanticError
 #include "FixMain.h"               // for FixMain
+#include "Parser/LinkageSpec.h"    // for Cforall, isMangled
 #include "SymTab/Mangler.h"        // for Mangler
-#include "SynTree/LinkageSpec.h"   // for Cforall, isMangled
 #include "SynTree/Constant.h"      // for Constant
 #include "SynTree/Declaration.h"   // for FunctionDecl, ObjectDecl, Declarat...

src/CodeGen/GenType.h

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 04:11:40 2020
 // Update Count     : 5
+// Last Modified By : Andrew Beach
+// Last Modified On : Tue Apr 30 11:47:00 2019
+// Update Count     : 3
 //
 …
         std::string genType( Type *type, const std::string &baseString, const Options &options );
         std::string genType( Type *type, const std::string &baseString, bool pretty = false, bool genC = false, bool lineMarks = false );
         std::string genPrettyType( Type * type, const std::string & baseString );
+  std::string genPrettyType( Type * type, const std::string & baseString );
 } // namespace CodeGen

src/CodeGen/Generate.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 03:01:51 2020
 // Update Count     : 9
+// Last Modified By : Andrew Beach
+// Last Modified On : Fri Aug 18 15:39:00 2017
+// Update Count     : 7
 //
 #include "Generate.h"
 …
 #include "GenType.h"                 // for genPrettyType
 #include "Common/PassVisitor.h"      // for PassVisitor
 #include "SynTree/LinkageSpec.h"     // for isBuiltin, isGeneratable
+#include "Parser/LinkageSpec.h"      // for isBuiltin, isGeneratable
 #include "SynTree/BaseSyntaxNode.h"  // for BaseSyntaxNode
 #include "SynTree/Declaration.h"     // for Declaration
 …
         void generate( BaseSyntaxNode * node, std::ostream & os ) {
                 if ( Type * type = dynamic_cast< Type * >( node ) ) {
                         os << genPrettyType( type, "" );
+                        os << CodeGen::genPrettyType( type, "" );
                 } else {
                         PassVisitor<CodeGenerator> cgv( os, true, false, false, false );

src/CodeGen/OperatorTable.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb 18 15:55:01 2020
 // Update Count     : 55
+// Last Modified On : Sat Jul 15 17:12:22 2017
+// Update Count     : 15
 //
 …
 #include <map>        // for map, _Rb_tree_const_iterator, map<>::const_iterator
 #include <utility>    // for pair
-using namespace std;
 #include "OperatorTable.h"
 …
 namespace CodeGen {
         const OperatorInfo CodeGen::tableValues[] = {
                 // inputName symbol   outputName                     friendlyName                  type
                 {       "?[?]",   "",     "_operator_index",             "Index",                      OT_INDEX          },
                 {       "?{}",    "=",    "_constructor",                "Constructor",                OT_CTOR           },
                 {       "^?{}",   "",     "_destructor",                 "Destructor",                 OT_DTOR           },
                 {       "?()",    "",     "_operator_call",              "Call Operator",              OT_CALL           },
                 {       "?++",    "++",   "_operator_postincr",          "Postfix Increment",          OT_POSTFIXASSIGN  },
                 {       "?--",    "--",   "_operator_postdecr",          "Postfix Decrement",          OT_POSTFIXASSIGN  },
                 {       "*?",     "*",    "_operator_deref",             "Dereference",                OT_PREFIX         },
                 {       "+?",     "+",    "_operator_unaryplus",         "Plus",                       OT_PREFIX         },
                 {       "-?",     "-",    "_operator_unaryminus",        "Minus",                      OT_PREFIX         },
                 {       "~?",     "~",    "_operator_bitnot",            "Bitwise Not",                OT_PREFIX         },
                 {       "!?",     "!",    "_operator_lognot",            "Logical Not",                OT_PREFIX         },
                 {       "++?",    "++",   "_operator_preincr",           "Prefix Increment",           OT_PREFIXASSIGN   },
                 {       "--?",    "--",   "_operator_predecr",           "Prefix Decrement",           OT_PREFIXASSIGN   },
                 {       "?\\?",   "\\",   "_operator_exponential",       "Exponentiation",             OT_INFIX          },
                 {       "?*?",    "*",    "_operator_multiply",          "Multiplication",             OT_INFIX          },
                 {       "?/?",    "/",    "_operator_divide",            "Division",                   OT_INFIX          },
                 {       "?%?",    "%",    "_operator_modulus",           "Modulo",                     OT_INFIX          },
                 {       "?+?",    "+",    "_operator_add",               "Addition",                   OT_INFIX          },
                 {       "?-?",    "-",    "_operator_subtract",          "Substraction",               OT_INFIX          },
                 {       "?<<?",   "<<",   "_operator_shiftleft",         "Shift Left",                 OT_INFIX          },
                 {       "?>>?",   ">>",   "_operator_shiftright",        "Shift Right",                OT_INFIX          },
                 {       "?<?",    "<",    "_operator_less",              "Less-than",                  OT_INFIX          },
                 {       "?>?",    ">",    "_operator_greater",           "Greater-than",               OT_INFIX          },
                 {       "?<=?",   "<=",   "_operator_lessequal",         "Less-than-or-Equal",         OT_INFIX          },
                 {       "?>=?",   ">=",   "_operator_greaterequal",      "Greater-than-or-Equal",      OT_INFIX          },
                 {       "?==?",   "==",   "_operator_equal",             "Equality",                   OT_INFIX          },
                 {       "?!=?",   "!=",   "_operator_notequal",          "Not-Equal",                  OT_INFIX          },
                 {       "?&?",    "&",    "_operator_bitand",            "Bitwise And",                OT_INFIX          },
                 {       "?^?",    "^",    "_operator_bitxor",            "Bitwise Xor",                OT_INFIX          },
                 {       "?|?",    "|",    "_operator_bitor",             "Bitwise Or",                 OT_INFIX          },
                 {       "?=?",    "=",    "_operator_assign",            "Assignment",                 OT_INFIXASSIGN    },
                 {       "?\\=?",  "\\=",  "_operator_expassign",         "Exponentiation Assignment",  OT_INFIXASSIGN    },
                 {       "?*=?",   "*=",   "_operator_multassign",        "Multiplication Assignment",  OT_INFIXASSIGN    },
                 {       "?/=?",   "/=",   "_operator_divassign",         "Division Assignment",        OT_INFIXASSIGN    },
                 {       "?%=?",   "%=",   "_operator_modassign",         "Modulo Assignment",          OT_INFIXASSIGN    },
                 {       "?+=?",   "+=",   "_operator_addassign",         "Addition Assignment",        OT_INFIXASSIGN    },
                 {       "?-=?",   "-=",   "_operator_subassign",         "Substrction Assignment",     OT_INFIXASSIGN    },
                 {       "?<<=?",  "<<=",  "_operator_shiftleftassign",   "Shift Left Assignment",      OT_INFIXASSIGN    },
                 {       "?>>=?",  ">>=",  "_operator_shiftrightassign",  "Shift Right Assignment",     OT_INFIXASSIGN    },
                 {       "?&=?",   "&=",   "_operator_bitandassign",      "Bitwise And Assignment",     OT_INFIXASSIGN    },
                 {       "?^=?",   "^=",   "_operator_bitxorassign",      "Bitwise Xor Assignment",     OT_INFIXASSIGN    },
                 {       "?|=?",   "|=",   "_operator_bitorassign",       "Bitwise Or Assignment",      OT_INFIXASSIGN    },
         }; // tableValues
+        namespace {
+                const OperatorInfo tableValues[] = {
+                        {       "?[?]",         "",             "_operator_index",                              OT_INDEX                        },
+                        {       "?{}",          "=",    "_constructor",                                 OT_CTOR                         },
+                        {       "^?{}",         "",             "_destructor",                                  OT_DTOR                         },
+                        {       "?()",          "",             "_operator_call",                               OT_CALL                         },
+                        {       "?++",          "++",   "_operator_postincr",                   OT_POSTFIXASSIGN        },
+                        {       "?--",          "--",   "_operator_postdecr",                   OT_POSTFIXASSIGN        },
+                        {       "*?",           "*",    "_operator_deref",                              OT_PREFIX                       },
+                        {       "+?",           "+",    "_operator_unaryplus",                  OT_PREFIX                       },
+                        {       "-?",           "-",    "_operator_unaryminus",                 OT_PREFIX                       },
+                        {       "~?",           "~",    "_operator_bitnot",                             OT_PREFIX                       },
+                        {       "!?",           "!",    "_operator_lognot",                             OT_PREFIX                       },
+                        {       "++?",          "++",   "_operator_preincr",                    OT_PREFIXASSIGN         },
+                        {       "--?",          "--",   "_operator_predecr",                    OT_PREFIXASSIGN         },
+                        {       "?\\?",         "\\",   "_operator_exponential",                OT_INFIX                        },
+                        {       "?*?",          "*",    "_operator_multiply",                   OT_INFIX                        },
+                        {       "?/?",          "/",    "_operator_divide",                             OT_INFIX                        },
+                        {       "?%?",          "%",    "_operator_modulus",                    OT_INFIX                        },
+                        {       "?+?",          "+",    "_operator_add",                                OT_INFIX                        },
+                        {       "?-?",          "-",    "_operator_subtract",                   OT_INFIX                        },
+                        {       "?<<?",         "<<",   "_operator_shiftleft",                  OT_INFIX                        },
+                        {       "?>>?",         ">>",   "_operator_shiftright",                 OT_INFIX                        },
+                        {       "?<?",          "<",    "_operator_less",                               OT_INFIX                        },
+                        {       "?>?",          ">",    "_operator_greater",                    OT_INFIX                        },
+                        {       "?<=?",         "<=",   "_operator_lessequal",                  OT_INFIX                        },
+                        {       "?>=?",         ">=",   "_operator_greaterequal",               OT_INFIX                        },
+                        {       "?==?",         "==",   "_operator_equal",                              OT_INFIX                        },
+                        {       "?!=?",         "!=",   "_operator_notequal",                   OT_INFIX                        },
+                        {       "?&?",          "&",    "_operator_bitand",                             OT_INFIX                        },
+                        {       "?^?",          "^",    "_operator_bitxor",                             OT_INFIX                        },
+                        {       "?|?",          "|",    "_operator_bitor",                              OT_INFIX                        },
+                        {       "?=?",          "=",    "_operator_assign",                             OT_INFIXASSIGN          },
+                        {       "?\\=?",        "\\=",  "_operator_expassign",                  OT_INFIXASSIGN          },
+                        {       "?*=?",         "*=",   "_operator_multassign",                 OT_INFIXASSIGN          },
+                        {       "?/=?",         "/=",   "_operator_divassign",                  OT_INFIXASSIGN          },
+                        {       "?%=?",         "%=",   "_operator_modassign",                  OT_INFIXASSIGN          },
+                        {       "?+=?",         "+=",   "_operator_addassign",                  OT_INFIXASSIGN          },
+                        {       "?-=?",         "-=",   "_operator_subassign",                  OT_INFIXASSIGN          },
+                        {       "?<<=?",        "<<=",  "_operator_shiftleftassign",    OT_INFIXASSIGN          },
+                        {       "?>>=?",        ">>=",  "_operator_shiftrightassign",   OT_INFIXASSIGN          },
+                        {       "?&=?",         "&=",   "_operator_bitandassign",               OT_INFIXASSIGN          },
+                        {       "?^=?",         "^=",   "_operator_bitxorassign",               OT_INFIXASSIGN          },
+                        {       "?|=?",         "|=",   "_operator_bitorassign",                OT_INFIXASSIGN          },
+                };
         std::map< std::string, OperatorInfo > CodeGen::table;
+                const int numOps = sizeof( tableValues ) / sizeof( OperatorInfo );
+        CodeGen::CodeGen() {
+                enum { numOps = sizeof( tableValues ) / sizeof( OperatorInfo ) };
+                for ( int i = 0; i < numOps; i += 1 ) {
+                        table[ tableValues[i].inputName ] = tableValues[i];
+                } // for
+                std::map< std::string, OperatorInfo > table;
+                void initialize() {
+                        for ( int i = 0; i < numOps; ++i ) {
+                                table[ tableValues[i].inputName ] = tableValues[i];
+                        } // for
+                }
+        } // namespace
+        bool operatorLookup( const std::string & funcName, OperatorInfo & info ) {
+                static bool init = false;
+                if ( ! init ) {
+                        initialize();
+                } // if
+                std::map< std::string, OperatorInfo >::const_iterator i = table.find( funcName );
+                if ( i == table.end() ) {
+                        if ( isPrefix( funcName, "?`" ) ) {
+                                // handle literal suffixes, which are user-defined postfix operators
+                                info.inputName = funcName;
+                                info.symbol = funcName.substr(2);
+                                info.outputName = toString( "__operator_literal_", info.symbol );
+                                info.type = OT_POSTFIX;
+                                return true;
+                        }
+                        return false;
+                } else {
+                        info = i->second;
+                        return true;
+                } // if
+        }
+        const OperatorInfo * operatorLookup( const string & funcName ) {
+                if ( funcName.find_first_of( "?^*+-!", 0, 1 ) == string::npos ) return nullptr; // prefilter
+                const OperatorInfo * ret = &CodeGen::table.find( funcName )->second; // must be in the table
+                assert( ret );
+                return ret;
+        bool isOperator( const std::string & funcName ) {
+                OperatorInfo info;
+                return operatorLookup( funcName, info );
+        }
+        bool isOperator( const string & funcName ) {
+                return operatorLookup( funcName ) != nullptr;
+        }
+        string operatorFriendlyName( const string & funcName ) {
+                const OperatorInfo * info = operatorLookup( funcName );
+                if ( info ) return info->friendlyName;
+                return "";
+        }
+        bool isConstructor( const string & funcName ) {
+                const OperatorInfo * info = operatorLookup( funcName );
+                if ( info ) return info->type == OT_CTOR;
+        /// determines if a given function name is one of the operator types between [begin, end)
+        template<typename Iterator>
+        bool isOperatorType( const std::string & funcName, Iterator begin, Iterator end ) {
+                OperatorInfo info;
+                if ( operatorLookup( funcName, info ) ) {
+                        return std::find( begin, end, info.type ) != end;
+                }
                 return false;
+        }
+        bool isDestructor( const string & funcName ) {
+                const OperatorInfo * info = operatorLookup( funcName );
+                if ( info ) return info->type == OT_DTOR;
+                return false;
+        bool isConstructor( const std::string & funcName ) {
+                static OperatorType types[] = { OT_CTOR };
+                return isOperatorType( funcName, std::begin(types), std::end(types) );
+        }
+        bool isCtorDtor( const string & funcName ) {
+                const OperatorInfo * info = operatorLookup( funcName );
+                if ( info ) return info->type <= OT_CONSTRUCTOR;
+                return false;
+        bool isDestructor( const std::string & funcName ) {
+                static OperatorType types[] = { OT_DTOR };
+                return isOperatorType( funcName, std::begin(types), std::end(types) );
+        }
+        bool isAssignment( const string & funcName ) {
+                const OperatorInfo * info = operatorLookup( funcName );
+                if ( info ) return info->type > OT_CONSTRUCTOR && info->type <= OT_ASSIGNMENT;
+                return false;
+        bool isAssignment( const std::string & funcName ) {
+                static OperatorType types[] = { OT_PREFIXASSIGN, OT_POSTFIXASSIGN, OT_INFIXASSIGN };
+                return isOperatorType( funcName, std::begin(types), std::end(types) );
+        }
+        bool isCtorDtorAssign( const string & funcName ) {
+                const OperatorInfo * info = operatorLookup( funcName );
+                if ( info ) return info->type <= OT_ASSIGNMENT;
+                return false;
+        bool isCtorDtor( const std::string & funcName ) {
+                static OperatorType types[] = { OT_CTOR, OT_DTOR };
+                return isOperatorType( funcName, std::begin(types), std::end(types) );
+        }
+        CodeGen codegen;                                                                        // initialize singleton package
+        bool isCtorDtorAssign( const std::string & funcName ) {
+                static OperatorType types[] = { OT_CTOR, OT_DTOR, OT_PREFIXASSIGN, OT_POSTFIXASSIGN, OT_INFIXASSIGN };
+                return isOperatorType( funcName, std::begin(types), std::end(types) );
+        }
 } // namespace CodeGen
 // Local Variables: //
 // tab-width: 4 //
+// mode: c++ //
+// compile-command: "make install" //
 // End: //

src/CodeGen/OperatorTable.h

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 08:13:34 2020
 // Update Count     : 26
+// Last Modified On : Fri Jul 21 22:17:11 2017
+// Update Count     : 6
 //
 …
 #include <string>
-#include <map>
 namespace CodeGen {
         enum OperatorType {
+                OT_INDEX,
                 OT_CTOR,
                 OT_DTOR,
+                OT_CONSTRUCTOR = OT_DTOR,
+                OT_CALL,
+                OT_PREFIX,
+                OT_POSTFIX,
+                OT_INFIX,
                 OT_PREFIXASSIGN,
                 OT_POSTFIXASSIGN,
                 OT_INFIXASSIGN,
-                OT_ASSIGNMENT = OT_INFIXASSIGN,
-                OT_CALL,
-                OT_PREFIX,
-                OT_INFIX,
-                OT_POSTFIX,
-                OT_INDEX,
                 OT_LABELADDRESS,
                 OT_CONSTANT
 …
                 std::string symbol;
                 std::string outputName;
-                std::string friendlyName;
                 OperatorType type;
         };
-        class CodeGen {
-                friend const OperatorInfo * operatorLookup( const std::string & funcName );
-                static const OperatorInfo tableValues[];
-                static std::map< std::string, OperatorInfo > table;
-          public:
-                CodeGen();
-        }; // CodeGen
         bool isOperator( const std::string & funcName );
+        const OperatorInfo * operatorLookup( const std::string & funcName );
+        std::string operatorFriendlyName( const std::string & funcName );
+        bool operatorLookup( const std::string & funcName, OperatorInfo & info );
         bool isConstructor( const std::string & );

src/CodeGen/Options.h

-              r7030dab
+              r71d6bd8
 // Author           : Andrew Beach
 // Created On       : Tue Apr 30 11:36:00 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb 15 18:37:06 2020
 // Update Count     : 3
+// Last Modified By : Andrew Beach
+// Last Modified On : Thr May  2 10:45:00 2019
+// Update Count     : 2
 //
 #pragma once
+struct Options {
+        // External Options: Same thoughout a pass.
+        bool pretty;
+        bool genC;
+        bool lineMarks;
+        bool printExprTypes;
+namespace CodeGen {
+        struct Options {
+                // External Options: Same thoughout a pass.
+                bool pretty;
+                bool genC;
+                bool lineMarks;
+                bool printExprTypes;
         // Internal Options: Changed on some recurisive calls.
         bool anonymousUnused = false;
+                // Internal Options: Changed on some recurisive calls.
+                bool anonymousUnused = false;
         Options(bool pretty, bool genC, bool lineMarks, bool printExprTypes) :
                 pretty(pretty), genC(genC), lineMarks(lineMarks), printExprTypes(printExprTypes)
+                Options(bool pretty, bool genC, bool lineMarks, bool printExprTypes) :
+                        pretty(pretty), genC(genC), lineMarks(lineMarks), printExprTypes(printExprTypes)
                 {}
+};
+        };
+} // namespace CodeGen
 // Local Variables: //

src/CodeGen/module.mk

-              r7030dab
+              r71d6bd8
 ## Created On       : Mon Jun  1 17:49:17 2015
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Sat Dec 14 07:29:42 2019
 ## Update Count     : 4
+## Last Modified On : Tue Jun  2 11:17:02 2015
+## Update Count     : 3
 ###############################################################################
 …
         CodeGen/OperatorTable.cc
 SRC += $(SRC_CODEGEN) CodeGen/Generate.cc CodeGen/FixNames.cc
 SRCDEMANGLE += $(SRC_CODEGEN)

src/CodeTools/DeclStats.cc

-              r7030dab
+              r71d6bd8
 // Author           : Aaron Moss
 // Created On       : Wed Jan 31 16:40:00 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:39:33 2019
 // Update Count     : 2
+// Last Modified By : Aaron Moss
+// Last Modified On : Wed Jan 31 16:40:00 2016
+// Update Count     : 1
 //
 …
 #include "Common/VectorMap.h"      // for VectorMap
 #include "GenPoly/GenPoly.h"       // for hasPolyBase
 #include "SynTree/LinkageSpec.h"   // for ::NoOfSpecs, Spec
+#include "Parser/LinkageSpec.h"    // for ::NoOfSpecs, Spec
 #include "SynTree/Declaration.h"   // for FunctionDecl, TypeDecl, Declaration
 #include "SynTree/Expression.h"    // for UntypedExpr, Expression

src/CodeTools/ResolvProtoDump.cc

-              r7030dab
+              r71d6bd8
 // Author           : Aaron Moss
 // Created On       : Tue Sep 11 09:04:00 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb 15 13:50:11 2020
 // Update Count     : 3
+// Last Modified By : Aaron Moss
+// Last Modified On : Tue Sep 11 09:04:00 2018
+// Update Count     : 1
 //
 …
                         // replace operator names
                         const CodeGen::OperatorInfo * opInfo = CodeGen::operatorLookup( name );
                         if ( opInfo ) {
+                        CodeGen::OperatorInfo info;
+                        if ( CodeGen::operatorLookup( name, info ) ) {
                                 ss << new_prefix(pre, "");
                                 op_name( opInfo->outputName, ss );
+                                op_name( info.outputName, ss );
                                 return;
+                        }

src/Common/Debug.h

-              r7030dab
+              r71d6bd8
 // Author           : Rob Schluntz
 // Created On       : Fri Sep 1 11:09:14 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:39:42 2019
 // Update Count     : 3
+// Last Modified By : Rob Schluntz
+// Last Modified On : Fri Sep 1 11:09:36 2017
+// Update Count     : 2
 //
 …
 #include "CodeGen/Generate.h"
 #include "SynTree/LinkageSpec.h"
+#include "Parser/LinkageSpec.h"
 #include "SynTree/Declaration.h"

src/Common/PassVisitor.h

-              r7030dab
+              r71d6bd8
         virtual void visit( FinallyStmt * finallyStmt ) override final;
         virtual void visit( const FinallyStmt * finallyStmt ) override final;
-        virtual void visit( SuspendStmt * suspendStmt ) override final;
-        virtual void visit( const SuspendStmt * suspendStmt ) override final;
         virtual void visit( WaitForStmt * waitforStmt ) override final;
         virtual void visit( const WaitForStmt * waitforStmt ) override final;
 …
         virtual Statement * mutate( CatchStmt * catchStmt ) override final;
         virtual Statement * mutate( FinallyStmt * finallyStmt ) override final;
-        virtual Statement * mutate( SuspendStmt * suspendStmt ) override final;
         virtual Statement * mutate( WaitForStmt * waitforStmt ) override final;
         virtual Declaration * mutate( WithStmt * withStmt ) override final;

src/Common/PassVisitor.impl.h

-              r7030dab
+              r71d6bd8
 //--------------------------------------------------------------------------
-// SuspendStmt
-template< typename pass_type >
-void PassVisitor< pass_type >::visit( SuspendStmt * node ) {
-        VISIT_START( node );
-        maybeAccept_impl( node->then  , *this );
-        VISIT_END( node );
+}
-template< typename pass_type >
-void PassVisitor< pass_type >::visit( const SuspendStmt * node ) {
-        VISIT_START( node );
-        maybeAccept_impl( node->then  , *this );
-        VISIT_END( node );
+}
-template< typename pass_type >
-Statement * PassVisitor< pass_type >::mutate( SuspendStmt * node ) {
-        MUTATE_START( node );
-        maybeMutate_impl( node->then  , *this );
-        MUTATE_END( Statement, node );
+}
-//--------------------------------------------------------------------------
 // WaitForStmt
 template< typename pass_type >
 …
         VISIT_START( node );
         indexerAddUnion( node->name );
+        indexerAddStruct( node->name );
+        {
 …
         VISIT_START( node );
         indexerAddUnion( node->name );
+        indexerAddStruct( node->name );
+        {
 …
         MUTATE_START( node );
         indexerAddUnion( node->name );
+        indexerAddStruct( node->name );
+        {

src/Common/SemanticError.cc

-              r7030dab
+              r71d6bd8
 // Helpers
 namespace ErrorHelpers {
-        Colors colors = Colors::Auto;
-        static inline bool with_colors() {
-                return colors == Colors::Auto ? isatty( STDERR_FILENO ) : bool(colors);
+        }
         const std::string & error_str() {
                 static std::string str = with_colors() ? "\e[31merror:\e[39m " : "error: ";
+                static std::string str = isatty( STDERR_FILENO ) ? "\e[31merror:\e[39m " : "error: ";
                 return str;
+        }
         const std::string & warning_str() {
                 static std::string str = with_colors() ? "\e[95mwarning:\e[39m " : "warning: ";
+                static std::string str = isatty( STDERR_FILENO ) ? "\e[95mwarning:\e[39m " : "warning: ";
                 return str;
+        }
         const std::string & bold_ttycode() {
                 static std::string str = with_colors() ? "\e[1m" : "";
+                static std::string str = isatty( STDERR_FILENO ) ? "\e[1m" : "";
                 return str;
+        }
         const std::string & reset_font_ttycode() {
                 static std::string str = with_colors() ? "\e[0m" : "";
+                static std::string str = isatty( STDERR_FILENO ) ? "\e[0m" : "";
                 return str;
+        }

src/Common/SemanticError.h

-              r7030dab
+              r71d6bd8
 struct WarningData {
         const char * const name;
+        const char * const message;
         const Severity default_severity;
-        const char * const message;
 };
 constexpr WarningData WarningFormats[] = {
+        {"self-assign"            , Severity::Warn    , "self assignment of expression: %s"                          },
+        {"reference-conversion"   , Severity::Warn    , "rvalue to reference conversion of rvalue: %s"               },
+        {"qualifiers-zero_t-one_t", Severity::Warn    , "questionable use of type qualifier %s with %s"              },
+        {"aggregate-forward-decl" , Severity::Warn    , "forward declaration of nested aggregate: %s"                },
+        {"superfluous-decl"       , Severity::Warn    , "declaration does not allocate storage: %s"                  },
+        {"gcc-attributes"         , Severity::Warn    , "invalid attribute: %s"                                      },
+        {"c++-like-copy"          , Severity::Warn    , "Constructor from reference is not a valid copy constructor" },
+        {"self-assign"            , "self assignment of expression: %s"            , Severity::Warn},
+        {"reference-conversion"   , "rvalue to reference conversion of rvalue: %s" , Severity::Warn},
+        {"qualifiers-zero_t-one_t", "questionable use of type qualifier %s with %s", Severity::Warn},
+        {"aggregate-forward-decl" , "forward declaration of nested aggregate: %s"  , Severity::Warn},
+        {"superfluous-decl"       , "declaration does not allocate storage: %s"    , Severity::Warn},
+        {"gcc-attributes"         , "invalid attribute: %s"                        , Severity::Warn},
 };
 …
         SuperfluousDecl,
         GccAttributes,
-        CppCopy,
         NUMBER_OF_WARNINGS, // This MUST be the last warning
 };
 …
 // Helpers
 namespace ErrorHelpers {
-        enum class Colors {
-                Never = false,
-                Always = true,
-                Auto,
-        };
-        extern Colors colors;
         const std::string & error_str();
         const std::string & warning_str();

src/Common/Stats/Time.h

-              r7030dab
+              r71d6bd8
 // Author           : Thierry Delisle
 // Created On       : Fri Mar 01 15:14:11 2019
 // Last Modified By : Andrew Beach
+// Last Modified By :
 // Last Modified On :
 // Update Count     :
 …
                                 f();
+                        }
-                        template<typename ret_t = void, typename func_t, typename... arg_t>
-                        inline ret_t TimeCall(
-                                        const char *, func_t func, arg_t&&... arg) {
-                                return func(std::forward<arg_t>(arg)...);
+                        }
 #               else
                         void StartGlobal();
 …
                                 func();
+                        }
-                        template<typename ret_t = void, typename func_t, typename... arg_t>
-                        inline ret_t TimeCall(
-                                        const char * name, func_t func, arg_t&&... arg) {
-                                BlockGuard guard(name);
-                                return func(std::forward<arg_t>(arg)...);
+                        }
 #               endif
+        }

src/Common/utility.h

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb 11 13:00:36 2020
 // Update Count     : 50
+// Last Modified On : Wed Jul 24 14:28:19 2019
+// Update Count     : 41
 //
 …
 #include <utility>
 #include <vector>
-#include <cstring>                                                                              // memcmp
 #include "Common/Indenter.h"
 …
+}
 // determines if pref is a prefix of str
 static inline bool isPrefix( const std::string & str, const std::string & pref, unsigned int start = 0 ) {
+/// determines if `pref` is a prefix of `str`
+static inline bool isPrefix( const std::string & str, const std::string & pref ) {
         if ( pref.size() > str.size() ) return false;
     return 0 == memcmp( str.c_str() + start, pref.c_str(), pref.size() );
         // return prefix == full.substr(0, prefix.size()); // for future, requires c++17
+        auto its = std::mismatch( pref.begin(), pref.end(), str.begin() );
+        return its.first == pref.end();
+}

src/Concurrency/Keywords.cc

-              r7030dab
+              r71d6bd8
 // Last Modified By :
 // Last Modified On :
 // Update Count     : 10
+// Update Count     : 5
 //
 #include "Concurrency/Keywords.h"
+#include <cassert>                        // for assert
+#include <string>                         // for string, operator==
+#include "Common/PassVisitor.h"           // for PassVisitor
+#include "Common/SemanticError.h"         // for SemanticError
+#include "Common/utility.h"               // for deleteAll, map_range
+#include "CodeGen/OperatorTable.h"        // for isConstructor
+#include "ControlStruct/LabelGenerator.h" // for LebelGenerator
+#include "InitTweak/InitTweak.h"          // for getPointerBase
+#include "SynTree/LinkageSpec.h"          // for Cforall
+#include "SynTree/Constant.h"             // for Constant
+#include "SynTree/Declaration.h"          // for StructDecl, FunctionDecl, ObjectDecl
+#include "SynTree/Expression.h"           // for VariableExpr, ConstantExpr, Untype...
+#include "SynTree/Initializer.h"          // for SingleInit, ListInit, Initializer ...
+#include "SynTree/Label.h"                // for Label
+#include "SynTree/Statement.h"            // for CompoundStmt, DeclStmt, ExprStmt
+#include "SynTree/Type.h"                 // for StructInstType, Type, PointerType
+#include "SynTree/Visitor.h"              // for Visitor, acceptAll
+#include <cassert>                 // for assert
+#include <string>                  // for string, operator==
+#include "Common/PassVisitor.h"    // for PassVisitor
+#include "Common/SemanticError.h"  // for SemanticError
+#include "Common/utility.h"        // for deleteAll, map_range
+#include "CodeGen/OperatorTable.h" // for isConstructor
+#include "InitTweak/InitTweak.h"   // for getPointerBase
+#include "Parser/LinkageSpec.h"    // for Cforall
+#include "SynTree/Constant.h"      // for Constant
+#include "SynTree/Declaration.h"   // for StructDecl, FunctionDecl, ObjectDecl
+#include "SynTree/Expression.h"    // for VariableExpr, ConstantExpr, Untype...
+#include "SynTree/Initializer.h"   // for SingleInit, ListInit, Initializer ...
+#include "SynTree/Label.h"         // for Label
+#include "SynTree/Statement.h"     // for CompoundStmt, DeclStmt, ExprStmt
+#include "SynTree/Type.h"          // for StructInstType, Type, PointerType
+#include "SynTree/Visitor.h"       // for Visitor, acceptAll
 class Attribute;
 …
           public:
                 ConcurrentSueKeyword( std::string&& type_name, std::string&& field_name, std::string&& getter_name, std::string&& context_error, bool needs_main, AggregateDecl::Aggregate cast_target ) :
+                ConcurrentSueKeyword( std::string&& type_name, std::string&& field_name, std::string&& getter_name, std::string&& context_error, bool needs_main, KeywordCastExpr::Target cast_target ) :
                   type_name( type_name ), field_name( field_name ), getter_name( getter_name ), context_error( context_error ), needs_main( needs_main ), cast_target( cast_target ) {}
 …
                 Declaration * postmutate( StructDecl * decl );
-                DeclarationWithType * postmutate( FunctionDecl * decl );
                 void handle( StructDecl * );
 …
                 const std::string context_error;
                 bool needs_main;
+                AggregateDecl::Aggregate cast_target;
+                StructDecl   * type_decl = nullptr;
+                FunctionDecl * dtor_decl = nullptr;
+                KeywordCastExpr::Target cast_target;
+                StructDecl* type_decl = nullptr;
         };
 …
         //      int data;                                  int data;
         //      a_struct_t more_data;                      a_struct_t more_data;
         //                                =>             $thread __thrd_d;
+        //                                =>             thread_desc __thrd_d;
         // };                                        };
         //                                           static inline $thread * get_thread( MyThread * this ) { return &this->__thrd_d; }
+        //                                           static inline thread_desc * get_thread( MyThread * this ) { return &this->__thrd_d; }
         //
         class ThreadKeyword final : public ConcurrentSueKeyword {
 …
                 ThreadKeyword() : ConcurrentSueKeyword(
                         "$thread",
+                        "thread_desc",
                         "__thrd",
                         "get_thread",
                         "thread keyword requires threads to be in scope, add #include <thread.hfa>\n",
+                        "thread keyword requires threads to be in scope, add #include <thread.hfa>",
                         true,
                         AggregateDecl::Thread
+                        KeywordCastExpr::Thread
+                )
                 {}
 …
         //      int data;                                  int data;
         //      a_struct_t more_data;                      a_struct_t more_data;
         //                                =>             $coroutine __cor_d;
+        //                                =>             coroutine_desc __cor_d;
         // };                                        };
         //                                           static inline $coroutine * get_coroutine( MyCoroutine * this ) { return &this->__cor_d; }
+        //                                           static inline coroutine_desc * get_coroutine( MyCoroutine * this ) { return &this->__cor_d; }
         //
         class CoroutineKeyword final : public ConcurrentSueKeyword {
 …
                 CoroutineKeyword() : ConcurrentSueKeyword(
                         "$coroutine",
+                        "coroutine_desc",
                         "__cor",
                         "get_coroutine",
                         "coroutine keyword requires coroutines to be in scope, add #include <coroutine.hfa>\n",
+                        "coroutine keyword requires coroutines to be in scope, add #include <coroutine.hfa>",
                         true,
                         AggregateDecl::Coroutine
+                        KeywordCastExpr::Coroutine
+                )
                 {}
 …
+                }
         };
         //-----------------------------------------------------------------------------
 …
         //      int data;                                  int data;
         //      a_struct_t more_data;                      a_struct_t more_data;
         //                                =>             $monitor __mon_d;
+        //                                =>             monitor_desc __mon_d;
         // };                                        };
         //                                           static inline $monitor * get_coroutine( MyMonitor * this ) { return &this->__cor_d; }
+        //                                           static inline monitor_desc * get_coroutine( MyMonitor * this ) { return &this->__cor_d; }
         //
         class MonitorKeyword final : public ConcurrentSueKeyword {
 …
                 MonitorKeyword() : ConcurrentSueKeyword(
                         "$monitor",
+                        "monitor_desc",
                         "__mon",
                         "get_monitor",
                         "monitor keyword requires monitors to be in scope, add #include <monitor.hfa>\n",
+                        "monitor keyword requires monitors to be in scope, add #include <monitor.hfa>",
                         false,
                         AggregateDecl::Monitor
+                        KeywordCastExpr::Monitor
+                )
                 {}
 …
         //-----------------------------------------------------------------------------
-        //Handles generator type declarations :
-        // generator MyGenerator {                   struct MyGenerator {
-        //      int data;                                  int data;
-        //      a_struct_t more_data;                      a_struct_t more_data;
-        //                                =>             int __gen_next;
-        // };                                        };
-        //
-        class GeneratorKeyword final : public ConcurrentSueKeyword {
-          public:
-                GeneratorKeyword() : ConcurrentSueKeyword(
-                        "$generator",
-                        "__generator_state",
-                        "get_generator",
-                        "Unable to find builtin type $generator\n",
-                        true,
-                        AggregateDecl::Generator
+                )
-                {}
-                virtual ~GeneratorKeyword() {}
-                virtual bool is_target( StructDecl * decl ) override final { return decl->is_generator(); }
-                static void implement( std::list< Declaration * > & translationUnit ) {
-                        PassVisitor< GeneratorKeyword > impl;
-                        mutateAll( translationUnit, impl );
+                }
-        };
-        //-----------------------------------------------------------------------------
-        class SuspendKeyword final : public WithStmtsToAdd, public WithGuards {
-        public:
-                SuspendKeyword() = default;
-                virtual ~SuspendKeyword() = default;
-                void  premutate( FunctionDecl * );
-                DeclarationWithType * postmutate( FunctionDecl * );
-                Statement * postmutate( SuspendStmt * );
-                static void implement( std::list< Declaration * > & translationUnit ) {
-                        PassVisitor< SuspendKeyword > impl;
-                        mutateAll( translationUnit, impl );
+                }
-        private:
-                DeclarationWithType * is_main( FunctionDecl * );
-                bool is_real_suspend( FunctionDecl * );
-                Statement * make_generator_suspend( SuspendStmt * );
-                Statement * make_coroutine_suspend( SuspendStmt * );
-                struct LabelPair {
-                        Label obj;
-                        int   idx;
-                };
-                LabelPair make_label() {
-                        labels.push_back( gen.newLabel("generator") );
-                        return { labels.back(), int(labels.size()) };
+                }
-                DeclarationWithType * in_generator = nullptr;
-                FunctionDecl * decl_suspend = nullptr;
-                std::vector<Label> labels;
-                ControlStruct::LabelGenerator & gen = *ControlStruct::LabelGenerator::getGenerator();
-        };
-        //-----------------------------------------------------------------------------
         //Handles mutex routines definitions :
         // void foo( A * mutex a, B * mutex b,  int i ) {                  void foo( A * a, B * b,  int i ) {
         //                                                                       $monitor * __monitors[] = { get_monitor(a), get_monitor(b) };
+        //                                                                       monitor_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
         //                                                                       monitor_guard_t __guard = { __monitors, 2 };
         //    /*Some code*/                                       =>           /*Some code*/
 …
         //Handles mutex routines definitions :
         // void foo( A * mutex a, B * mutex b,  int i ) {                  void foo( A * a, B * b,  int i ) {
         //                                                                       $monitor * __monitors[] = { get_monitor(a), get_monitor(b) };
+        //                                                                       monitor_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
         //                                                                       monitor_guard_t __guard = { __monitors, 2 };
         //    /*Some code*/                                       =>           /*Some code*/
 …
                 CoroutineKeyword        ::implement( translationUnit );
                 MonitorKeyword  ::implement( translationUnit );
-                GeneratorKeyword  ::implement( translationUnit );
-                SuspendKeyword    ::implement( translationUnit );
+        }
 …
+        }
-        DeclarationWithType * ConcurrentSueKeyword::postmutate( FunctionDecl * decl ) {
-                if( !type_decl ) return decl;
-                if( !CodeGen::isDestructor( decl->name ) ) return decl;
-                auto params = decl->type->parameters;
-                if( params.size() != 1 ) return decl;
-                auto type = dynamic_cast<ReferenceType*>( params.front()->get_type() );
-                if( !type ) return decl;
-                auto stype = dynamic_cast<StructInstType*>( type->base );
-                if( !stype ) return decl;
-                if( stype->baseStruct != type_decl ) return decl;
-                if( !dtor_decl ) dtor_decl = decl;
-                return decl;
+        }
         Expression * ConcurrentSueKeyword::postmutate( KeywordCastExpr * cast ) {
                 if ( cast_target == cast->target ) {
                         // convert (thread &)t to ($thread &)*get_thread(t), etc.
+                        // convert (thread &)t to (thread_desc &)*get_thread(t), etc.
                         if( !type_decl ) SemanticError( cast, context_error );
+                        if( !dtor_decl ) SemanticError( cast, context_error );
+                        assert( cast->result == nullptr );
+                        cast->set_result( new ReferenceType( noQualifiers, new StructInstType( noQualifiers, type_decl ) ) );
+                        cast->concrete_target.field  = field_name;
+                        cast->concrete_target.getter = getter_name;
+                        Expression * arg = cast->arg;
+                        cast->arg = nullptr;
+                        delete cast;
+                        return new CastExpr(
+                                UntypedExpr::createDeref(
+                                        new UntypedExpr( new NameExpr( getter_name ), { arg } )
+                                ),
+                                new ReferenceType(
+                                        noQualifiers,
+                                        new StructInstType( noQualifiers, type_decl ) )
+                                );
+                }
                 return cast;
 …
                 if( !type_decl ) SemanticError( decl, context_error );
-                if( !dtor_decl ) SemanticError( decl, context_error );
                 FunctionDecl * func = forwardDeclare( decl );
 …
                         get_type,
                         nullptr,
                         { new Attribute("const") },
+                        noAttributes,
                         Type::Inline
                 );
 …
                 declsToAddAfter.push_back( get_decl );
+        }
+        //=============================================================================================
+        // Suspend keyword implementation
+        //=============================================================================================
+        DeclarationWithType * SuspendKeyword::is_main( FunctionDecl * func) {
+                if(func->name != "main") return nullptr;
+                if(func->type->parameters.size() != 1) return nullptr;
+                auto param = func->type->parameters.front();
+                auto type  = dynamic_cast<ReferenceType * >(param->get_type());
+                if(!type) return nullptr;
+                auto obj   = dynamic_cast<StructInstType *>(type->base);
+                if(!obj) return nullptr;
+                if(!obj->baseStruct->is_generator()) return nullptr;
+                return param;
+        }
+        bool SuspendKeyword::is_real_suspend( FunctionDecl * func ) {
+                if(isMangled(func->linkage)) return false; // the real suspend isn't mangled
+                if(func->name != "__cfactx_suspend") return false; // the real suspend has a specific name
+                if(func->type->parameters.size() != 0) return false; // Too many parameters
+                if(func->type->returnVals.size() != 0) return false; // Too many return values
+                return true;
+        }
+        void SuspendKeyword::premutate( FunctionDecl * func ) {
+                GuardValue(in_generator);
+                in_generator = nullptr;
+                // Is this the real suspend?
+                if(is_real_suspend(func)) {
+                        decl_suspend = decl_suspend ? decl_suspend : func;
+                        return;
+                }
+                // Is this the main of a generator?
+                auto param = is_main( func );
+                if(!param) return;
+                if(func->type->returnVals.size() != 0) SemanticError(func->location, "Generator main must return void");
+                in_generator = param;
+                GuardValue(labels);
+                labels.clear();
+        }
+        DeclarationWithType * SuspendKeyword::postmutate( FunctionDecl * func ) {
+                if( !func->statements ) return func; // Not the actual definition, don't do anything
+                if( !in_generator     ) return func; // Not in a generator, don't do anything
+                if( labels.empty()    ) return func; // Generator has no states, nothing to do, could throw a warning
+                // This is a generator main, we need to add the following code to the top
+                // static void * __generator_labels[] = {&&s0, &&s1, ...};
+                // goto * __generator_labels[gen.__generator_state];
+                const auto & loc = func->location;
+                const auto first_label = gen.newLabel("generator");
+                // for each label add to declaration
+                std::list<Initializer*> inits = { new SingleInit( new LabelAddressExpr( first_label ) ) };
+                for(const auto & label : labels) {
+                        inits.push_back(
+                                new SingleInit(
+                                        new LabelAddressExpr( label )
+                                )
+                        );
+                }
+                auto init = new ListInit(std::move(inits), noDesignators, true);
+                labels.clear();
+                // create decl
+                auto decl = new ObjectDecl(
+                        "__generator_labels",
+                        Type::StorageClasses( Type::Static ),
+                        LinkageSpec::AutoGen,
+                        nullptr,
+                        new ArrayType(
+                                Type::Qualifiers(),
+                                new PointerType(
+                                        Type::Qualifiers(),
+                                        new VoidType( Type::Qualifiers() )
+                                ),
+                                nullptr,
+                                false, false
+                        ),
+                        init
+                );
+                // create the goto
+                assert(in_generator);
+                auto go_decl = new ObjectDecl(
+                        "__generator_label",
+                        noStorageClasses,
+                        LinkageSpec::AutoGen,
+                        nullptr,
+                        new PointerType(
+                                Type::Qualifiers(),
+                                new VoidType( Type::Qualifiers() )
+                        ),
+                        new SingleInit(
+                                new UntypedExpr(
+                                        new NameExpr("?[?]"),
+                                        {
+                                                new NameExpr("__generator_labels"),
+                                                new UntypedMemberExpr(
+                                                        new NameExpr("__generator_state"),
+                                                        new VariableExpr( in_generator )
+                                                )
+                                        }
+                                )
+                        )
+                );
+                go_decl->location = loc;
+                auto go = new BranchStmt(
+                        new VariableExpr( go_decl ),
+                        BranchStmt::Goto
+                );
+                go->location = loc;
+                go->computedTarget->location = loc;
+                auto noop = new NullStmt({ first_label });
+                noop->location = loc;
+                // wrap everything in a nice compound
+                auto body = new CompoundStmt({
+                        new DeclStmt( decl ),
+                        new DeclStmt( go_decl ),
+                        go,
+                        noop,
+                        func->statements
+                });
+                body->location   = loc;
+                func->statements = body;
+                return func;
+        }
+        Statement * SuspendKeyword::postmutate( SuspendStmt * stmt ) {
+                SuspendStmt::Type type = stmt->type;
+                if(type == SuspendStmt::None) {
+                        // This suspend has a implicit target, find it
+                        type = in_generator ? SuspendStmt::Generator : SuspendStmt::Coroutine;
+                }
+                // Check that the target makes sense
+                if(!in_generator && type == SuspendStmt::Generator) SemanticError( stmt->location, "'suspend generator' must be used inside main of generator type.");
+                // Act appropriately
+                switch(type) {
+                        case SuspendStmt::Generator: return make_generator_suspend(stmt);
+                        case SuspendStmt::Coroutine: return make_coroutine_suspend(stmt);
+                        default: abort();
+                }
+        }
+        Statement * SuspendKeyword::make_generator_suspend( SuspendStmt * stmt ) {
+                assert(in_generator);
+                // Target code is :
+                //   gen.__generator_state = X;
+                //   { THEN }
+                //   return;
+                //   __gen_X:;
+                // Save the location and delete the old statement, we only need the location from this point on
+                auto loc = stmt->location;
+                // Build the label and get its index
+                auto label = make_label();
+                // Create the context saving statement
+                auto save = new ExprStmt( new UntypedExpr(
+                        new NameExpr( "?=?" ),
+                        {
+                                new UntypedMemberExpr(
+                                        new NameExpr("__generator_state"),
+                                        new VariableExpr( in_generator )
+                                ),
+                                new ConstantExpr(
+                                        Constant::from_int( label.idx )
+                                )
+                        }
+                ));
+                assert(save->expr);
+                save->location = loc;
+                stmtsToAddBefore.push_back( save );
+                // if we have a then add it here
+                auto then = stmt->then;
+                stmt->then = nullptr;
+                delete stmt;
+                if(then) stmtsToAddBefore.push_back( then );
+                // Create the return statement
+                auto ret = new ReturnStmt( nullptr );
+                ret->location = loc;
+                stmtsToAddBefore.push_back( ret );
+                // Create the null statement with the created label
+                auto noop = new NullStmt({ label.obj });
+                noop->location = loc;
+                // Return the null statement to take the place of the previous statement
+                return noop;
+        }
+        Statement * SuspendKeyword::make_coroutine_suspend( SuspendStmt * stmt ) {
+                if(stmt->then) SemanticError( stmt->location, "Compound statement following coroutines is not implemented.");
+                // Save the location and delete the old statement, we only need the location from this point on
+                auto loc = stmt->location;
+                delete stmt;
+                // Create the call expression
+                if(!decl_suspend) SemanticError( loc, "suspend keyword applied to coroutines requires coroutines to be in scope, add #include <coroutine.hfa>\n");
+                auto expr = new UntypedExpr( VariableExpr::functionPointer( decl_suspend ) );
+                expr->location = loc;
+                // Change this statement into a regular expr
+                assert(expr);
+                auto nstmt = new ExprStmt( expr );
+                nstmt->location = loc;
+                return nstmt;
+        }
+                // get_decl->fixUniqueId();
+        }
         //=============================================================================================
 …
         void MutexKeyword::postvisit(StructDecl* decl) {
                 if( decl->name == "$monitor" && decl->body ) {
+                if( decl->name == "monitor_desc" && decl->body ) {
                         assert( !monitor_decl );
                         monitor_decl = decl;
 …
                 );
                 //$monitor * __monitors[] = { get_monitor(a), get_monitor(b) };
+                //monitor_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
                 body->push_front( new DeclStmt( monitors) );
+        }
 …
                 );
                 //$monitor * __monitors[] = { get_monitor(a), get_monitor(b) };
+                //monitor_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
                 body->push_front( new DeclStmt( monitors) );
+        }
 …
         //=============================================================================================
         void ThreadStarter::previsit( StructDecl * decl ) {
                 if( decl->name == "$thread" && decl->body ) {
+                if( decl->name == "thread_desc" && decl->body ) {
                         assert( !thread_decl );
                         thread_decl = decl;
 …
                                 new UntypedExpr(
                                         new NameExpr( "__thrd_start" ),
                                         { new VariableExpr( param ), new NameExpr("main") }
+                                        { new VariableExpr( param ) }
+                                )
+                        )

src/Concurrency/Waitfor.cc

-              r7030dab
+              r71d6bd8
 // Last Modified By :
 // Last Modified On :
 // Update Count     : 12
+// Update Count     : 7
 //
 …
 #include "Common/PassVisitor.h"    // for PassVisitor
 #include "Common/SemanticError.h"  // for SemanticError
-#include "Common/UniqueName.h"     // for UniqueName
 #include "Common/utility.h"        // for deleteAll, map_range
 #include "CodeGen/OperatorTable.h" // for isConstructor
 #include "InitTweak/InitTweak.h"   // for getPointerBase
+#include "Parser/LinkageSpec.h"    // for Cforall
 #include "ResolvExpr/Resolver.h"   // for findVoidExpression
-#include "SynTree/LinkageSpec.h"   // for Cforall
 #include "SynTree/Constant.h"      // for Constant
 #include "SynTree/Declaration.h"   // for StructDecl, FunctionDecl, ObjectDecl
 …
 void foo() {
         while( true ) {
                 when( a < 1 ) waitfor( f : a ) { bar(); }
+                when( a < 1 ) waitfor( f, a ) { bar(); }
                 or timeout( swagl() );
                 or waitfor( g : a ) { baz(); }
                 or waitfor( ^?{} : a ) { break; }
+                or waitfor( g, a ) { baz(); }
+                or waitfor( ^?{}, a ) { break; }
                 or waitfor( ^?{} ) { break; }
+        }
 …
                         decl_mask = decl;
+                }
                 else if( decl->name == "$monitor" ) {
+                else if( decl->name == "monitor_desc" ) {
                         assert( !decl_monitor );
                         decl_monitor = decl;

src/ControlStruct/ExceptTranslate.cc

-              r7030dab
+              r71d6bd8
 // Author           : Andrew Beach
 // Created On       : Wed Jun 14 16:49:00 2017
 // Last Modified By : Andrew Beach
 // Last Modified On : Fri Mar 27 11:58:00 2020
 // Update Count     : 13
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Wed Feb 13 18:15:29 2019
+// Update Count     : 11
 //
 …
 #include "Common/SemanticError.h"     // for SemanticError
 #include "Common/utility.h"           // for CodeLocation
 #include "SynTree/LinkageSpec.h"      // for Cforall
+#include "Parser/LinkageSpec.h"       // for Cforall
 #include "SynTree/Attribute.h"        // for Attribute
 #include "SynTree/Constant.h"         // for Constant
 …
                         ThrowStmt *throwStmt ) {
                 // __throw_terminate( `throwStmt->get_name()` ); }
                 return create_given_throw( "__cfaehm_throw_terminate", throwStmt );
+                return create_given_throw( "__cfaabi_ehm__throw_terminate", throwStmt );
+        }
 …
                         ) ) );
                 result->push_back( new ExprStmt(
                         new UntypedExpr( new NameExpr( "__cfaehm_rethrow_terminate" ) )
+                        new UntypedExpr( new NameExpr( "__cfaabi_ehm__rethrow_terminate" ) )
                         ) );
                 delete throwStmt;
 …
                         ThrowStmt *throwStmt ) {
                 // __throw_resume( `throwStmt->get_name` );
                 return create_given_throw( "__cfaehm_throw_resume", throwStmt );
+                return create_given_throw( "__cfaabi_ehm__throw_resume", throwStmt );
+        }
 …
                         local_except->get_attributes().push_back( new Attribute(
                                 "cleanup",
                                 { new NameExpr( "__cfaehm_cleanup_terminate" ) }
+                                { new NameExpr( "__cfaabi_ehm__cleanup_terminate" ) }
                                 ) );
 …
                         FunctionDecl * terminate_catch,
                         FunctionDecl * terminate_match ) {
                 // { __cfaehm_try_terminate(`try`, `catch`, `match`); }
+                // { __cfaabi_ehm__try_terminate(`try`, `catch`, `match`); }
                 UntypedExpr * caller = new UntypedExpr( new NameExpr(
                         "__cfaehm_try_terminate" ) );
+                        "__cfaabi_ehm__try_terminate" ) );
                 std::list<Expression *>& args = caller->get_args();
                 args.push_back( nameOf( try_wrapper ) );
 …
                 // struct __try_resume_node __resume_node
                 //      __attribute__((cleanup( __cfaehm_try_resume_cleanup )));
+                //      __attribute__((cleanup( __cfaabi_ehm__try_resume_cleanup )));
                 // ** unwinding of the stack here could cause problems **
                 // ** however I don't think that can happen currently **
                 // __cfaehm_try_resume_setup( &__resume_node, resume_handler );
+                // __cfaabi_ehm__try_resume_setup( &__resume_node, resume_handler );
                 std::list< Attribute * > attributes;
 …
                         std::list< Expression * > attr_params;
                         attr_params.push_back( new NameExpr(
                                 "__cfaehm_try_resume_cleanup" ) );
+                                "__cfaabi_ehm__try_resume_cleanup" ) );
                         attributes.push_back( new Attribute( "cleanup", attr_params ) );
+                }
 …
                 UntypedExpr *setup = new UntypedExpr( new NameExpr(
                         "__cfaehm_try_resume_setup" ) );
+                        "__cfaabi_ehm__try_resume_setup" ) );
                 setup->get_args().push_back( new AddressExpr( nameOf( obj ) ) );
                 setup->get_args().push_back( nameOf( resume_handler ) );
 …
         ObjectDecl * ExceptionMutatorCore::create_finally_hook(
                         FunctionDecl * finally_wrapper ) {
                 // struct __cfaehm_cleanup_hook __finally_hook
+                // struct __cfaabi_ehm__cleanup_hook __finally_hook
                 //      __attribute__((cleanup( finally_wrapper )));
 …
                         // Skip children?
                         return;
                 } else if ( structDecl->get_name() == "__cfaehm_base_exception_t" ) {
+                } else if ( structDecl->get_name() == "__cfaabi_ehm__base_exception_t" ) {
                         assert( nullptr == except_decl );
                         except_decl = structDecl;
                         init_func_types();
                 } else if ( structDecl->get_name() == "__cfaehm_try_resume_node" ) {
+                } else if ( structDecl->get_name() == "__cfaabi_ehm__try_resume_node" ) {
                         assert( nullptr == node_decl );
                         node_decl = structDecl;
                 } else if ( structDecl->get_name() == "__cfaehm_cleanup_hook" ) {
+                } else if ( structDecl->get_name() == "__cfaabi_ehm__cleanup_hook" ) {
                         assert( nullptr == hook_decl );
                         hook_decl = structDecl;

src/ControlStruct/LabelFixer.cc

-              r7030dab
+              r71d6bd8
 // Author           : Rodolfo G. Esteves
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Tue Jan 21 10:32:00 2020
 // Update Count     : 160
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Mon Mar 11 22:26:02 2019
+// Update Count     : 159
 //
 …
 #include "ControlStruct/LabelGenerator.h"  // for LabelGenerator
 #include "LabelFixer.h"
 #include "MLEMutator.h"                    // for MultiLevelExitMutator
+#include "MLEMutator.h"                    // for MLEMutator
 #include "SynTree/Declaration.h"           // for FunctionDecl
 #include "SynTree/Expression.h"            // for NameExpr, Expression, Unty...
 …
         void LabelFixer::postvisit( FunctionDecl * functionDecl ) {
+                PassVisitor<MultiLevelExitMutator> mlem( resolveJumps(), generator );
+                // We start in the body so we can stop when we hit another FunctionDecl.
+                maybeMutate( functionDecl->statements, mlem );
+                PassVisitor<MLEMutator> mlemut( resolveJumps(), generator );
+                functionDecl->acceptMutator( mlemut );
+        }
 …
         // Sets the definition of the labelTable entry to be the provided statement for every label in
         // the list parameter. Happens for every kind of statement.
+        // sets the definition of the labelTable entry to be the provided statement for every label in the list
+        // parameter. Happens for every kind of statement
         Label LabelFixer::setLabelsDef( std::list< Label > & llabel, Statement * definition ) {
                 assert( definition != 0 );
                 assert( llabel.size() > 0 );
+                Entry * e = new Entry( definition );
                 for ( std::list< Label >::iterator i = llabel.begin(); i != llabel.end(); i++ ) {
 …
                         l.set_statement( definition ); // attach statement to the label to be used later
                         if ( labelTable.find( l ) == labelTable.end() ) {
+                                // All labels on this statement need to use the same entry,
+                                // so this should only be created once.
+                                // all labels on this statement need to use the same entry, so this should only be created once
                                 // undefined and unused until now, add an entry
                                 labelTable[ l ] = new Entry( definition );
+                                labelTable[ l ] =  e;
                         } else if ( labelTable[ l ]->defined() ) {
                                 // defined twice, error
+                                SemanticError( l.get_statement()->location,
+                                        "Duplicate definition of label: " + l.get_name() );
+                        } else {
+                                SemanticError( l.get_statement()->location, "Duplicate definition of label: " + l.get_name() );
+                        }       else {
                                 // used previously, but undefined until now -> link with this entry
-                                // Question: Is changing objects important?
                                 delete labelTable[ l ];
                                 labelTable[ l ] = new Entry( definition );
+                                labelTable[ l ] = e;
                         } // if
                 } // for
+                // Produce one of the labels attached to this statement to be temporarily used as the
+                // canonical label.
+                // produce one of the labels attached to this statement to be temporarily used as the canonical label
                 return labelTable[ llabel.front() ]->get_label();
+        }

src/ControlStruct/MLEMutator.cc

-              r7030dab
+              r71d6bd8
 // Author           : Rodolfo G. Esteves
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Wed Jan 22 11:50:00 2020
 // Update Count     : 223
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Tue Oct 22 17:22:44 2019
+// Update Count     : 220
 //
 …
 namespace ControlStruct {
         MultiLevelExitMutator::~MultiLevelExitMutator() {
+        MLEMutator::~MLEMutator() {
                 delete targetTable;
                 targetTable = 0;
+        }
         namespace {
+                bool isLoop( const MultiLevelExitMutator::Entry & e ) {
+                        return dynamic_cast< WhileStmt * >( e.get_controlStructure() )
+                                || dynamic_cast< ForStmt * >( e.get_controlStructure() );
+                }
+                bool isSwitch( const MultiLevelExitMutator::Entry & e ) {
+                        return dynamic_cast< SwitchStmt *>( e.get_controlStructure() );
+                }
+                bool isBreakTarget( const MultiLevelExitMutator::Entry & e ) {
+                        return isLoop( e ) || isSwitch( e )
+                                || dynamic_cast< CompoundStmt *>( e.get_controlStructure() );
+                }
+                bool isContinueTarget( const MultiLevelExitMutator::Entry & e ) {
+                        return isLoop( e );
+                }
+                bool isFallthroughTarget( const MultiLevelExitMutator::Entry & e ) {
+                        return dynamic_cast< CaseStmt *>( e.get_controlStructure() );
+                }
+                bool isFallthroughDefaultTarget( const MultiLevelExitMutator::Entry & e ) {
+                        return isSwitch( e );
+                }
+                bool isLoop( const MLEMutator::Entry & e ) { return dynamic_cast< WhileStmt * >( e.get_controlStructure() ) || dynamic_cast< ForStmt * >( e.get_controlStructure() ); }
+                bool isSwitch( const MLEMutator::Entry & e ) { return dynamic_cast< SwitchStmt *>( e.get_controlStructure() ); }
+                bool isBreakTarget( const MLEMutator::Entry & e ) { return isLoop( e ) || isSwitch( e ) || dynamic_cast< CompoundStmt *>( e.get_controlStructure() ); }
+                bool isContinueTarget( const MLEMutator::Entry & e ) { return isLoop( e ); }
+                bool isFallthroughTarget( const MLEMutator::Entry & e ) { return dynamic_cast< CaseStmt *>( e.get_controlStructure() );; }
+                bool isFallthroughDefaultTarget( const MLEMutator::Entry & e ) { return isSwitch( e ); }
         } // namespace
-        void MultiLevelExitMutator::premutate( FunctionDecl * ) {
-                visit_children = false;
+        }
         // break labels have to come after the statement they break out of, so mutate a statement, then if they inform us
         // through the breakLabel field tha they need a place to jump to on a break statement, add the break label to the
         // body of statements
         void MultiLevelExitMutator::fixBlock( std::list< Statement * > &kids, bool caseClause ) {
+        void MLEMutator::fixBlock( std::list< Statement * > &kids, bool caseClause ) {
                 SemanticErrorException errors;
 …
+        }
         void MultiLevelExitMutator::premutate( CompoundStmt *cmpndStmt ) {
+        void MLEMutator::premutate( CompoundStmt *cmpndStmt ) {
                 visit_children = false;
                 bool labeledBlock = !(cmpndStmt->labels.empty());
 …
+                        }
+                }
+                assertf( false, "Could not find label '%s' on statement %s",
+                        originalTarget.get_name().c_str(), toString( stmt ).c_str() );
+        }
+        Statement *MultiLevelExitMutator::postmutate( BranchStmt *branchStmt )
+                        throw ( SemanticErrorException ) {
+                assertf( false, "Could not find label '%s' on statement %s", originalTarget.get_name().c_str(), toString( stmt ).c_str() );
+        }
+        Statement *MLEMutator::postmutate( BranchStmt *branchStmt ) throw ( SemanticErrorException ) {
                 std::string originalTarget = branchStmt->originalTarget;
 …
+        }
+        Statement *MultiLevelExitMutator::mutateLoop( Statement *bodyLoop, Entry &e ) {
+        Statement *MLEMutator::mutateLoop( Statement *bodyLoop, Entry &e ) {
+                // ensure loop body is a block
+                CompoundStmt *newBody;
+                if ( ! (newBody = dynamic_cast<CompoundStmt *>( bodyLoop )) ) {
+                        newBody = new CompoundStmt();
+                        newBody->get_kids().push_back( bodyLoop );
+                } // if
                 // only generate these when needed
-                if( !e.isContUsed() && !e.isBreakUsed() ) return bodyLoop;
-                // ensure loop body is a block
-                CompoundStmt * newBody = new CompoundStmt();
-                newBody->get_kids().push_back( bodyLoop );
                 if ( e.isContUsed() ) {
 …
         template< typename LoopClass >
         void MultiLevelExitMutator::prehandleLoopStmt( LoopClass * loopStmt ) {
+        void MLEMutator::prehandleLoopStmt( LoopClass * loopStmt ) {
                 // remember this as the most recent enclosing loop, then mutate the body of the loop -- this will determine
                 // whether brkLabel and contLabel are used with branch statements and will recursively do the same to nested
 …
         template< typename LoopClass >
         Statement * MultiLevelExitMutator::posthandleLoopStmt( LoopClass * loopStmt ) {
+        Statement * MLEMutator::posthandleLoopStmt( LoopClass * loopStmt ) {
                 assert( ! enclosingControlStructures.empty() );
                 Entry &e = enclosingControlStructures.back();
 …
+        }
         void MultiLevelExitMutator::premutate( WhileStmt * whileStmt ) {
+        void MLEMutator::premutate( WhileStmt * whileStmt ) {
                 return prehandleLoopStmt( whileStmt );
+        }
         void MultiLevelExitMutator::premutate( ForStmt * forStmt ) {
+        void MLEMutator::premutate( ForStmt * forStmt ) {
                 return prehandleLoopStmt( forStmt );
+        }
         Statement * MultiLevelExitMutator::postmutate( WhileStmt * whileStmt ) {
+        Statement * MLEMutator::postmutate( WhileStmt * whileStmt ) {
                 return posthandleLoopStmt( whileStmt );
+        }
         Statement * MultiLevelExitMutator::postmutate( ForStmt * forStmt ) {
+        Statement * MLEMutator::postmutate( ForStmt * forStmt ) {
                 return posthandleLoopStmt( forStmt );
+        }
         void MultiLevelExitMutator::premutate( IfStmt * ifStmt ) {
+        void MLEMutator::premutate( IfStmt * ifStmt ) {
                 // generate a label for breaking out of a labeled if
                 bool labeledBlock = !(ifStmt->get_labels().empty());
 …
+        }
         Statement * MultiLevelExitMutator::postmutate( IfStmt * ifStmt ) {
+        Statement * MLEMutator::postmutate( IfStmt * ifStmt ) {
                 bool labeledBlock = !(ifStmt->get_labels().empty());
                 if ( labeledBlock ) {
 …
+        }
         void MultiLevelExitMutator::premutate( TryStmt * tryStmt ) {
+        void MLEMutator::premutate( TryStmt * tryStmt ) {
                 // generate a label for breaking out of a labeled if
                 bool labeledBlock = !(tryStmt->get_labels().empty());
 …
+        }
         Statement * MultiLevelExitMutator::postmutate( TryStmt * tryStmt ) {
+        Statement * MLEMutator::postmutate( TryStmt * tryStmt ) {
                 bool labeledBlock = !(tryStmt->get_labels().empty());
                 if ( labeledBlock ) {
 …
+        }
+        void MultiLevelExitMutator::premutate( FinallyStmt * ) {
+                GuardAction([this, old = std::move(enclosingControlStructures)]() {
+                        enclosingControlStructures = std::move(old);
+                });
+                enclosingControlStructures = std::list<Entry>();
+                GuardValue( inFinally );
+                inFinally = true;
+        }
+        void MultiLevelExitMutator::premutate( ReturnStmt *returnStmt ) {
+                if ( inFinally ) {
+                        SemanticError( returnStmt->location, "'return' may not appear in a finally clause" );
+                }
+        }
+        void MultiLevelExitMutator::premutate( CaseStmt *caseStmt ) {
+        void MLEMutator::premutate( CaseStmt *caseStmt ) {
                 visit_children = false;
 …
+        }
         void MultiLevelExitMutator::premutate( SwitchStmt *switchStmt ) {
+        void MLEMutator::premutate( SwitchStmt *switchStmt ) {
                 // generate a label for breaking out of a labeled switch
                 Label brkLabel = generator->newLabel("switchBreak", switchStmt);
 …
+        }
         Statement * MultiLevelExitMutator::postmutate( SwitchStmt * switchStmt ) {
+        Statement * MLEMutator::postmutate( SwitchStmt * switchStmt ) {
                 Entry &e = enclosingControlStructures.back();
                 assert ( e == switchStmt );

src/ControlStruct/MLEMutator.h

-              r7030dab
+              r71d6bd8
 // Author           : Rodolfo G. Esteves
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Wed Jan 22 11:50:00 2020
 // Update Count     : 48
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Tue Oct 22 17:22:47 2019
+// Update Count     : 45
 //
 …
         class LabelGenerator;
+        class MultiLevelExitMutator : public WithVisitorRef<MultiLevelExitMutator>,
+                        public WithShortCircuiting, public WithGuards {
+        class MLEMutator : public WithVisitorRef<MLEMutator>, public WithShortCircuiting, public WithGuards {
           public:
                 class Entry;
+                MultiLevelExitMutator( std::map<Label, Statement *> *t, LabelGenerator *gen = 0 ) :
+                        targetTable( t ), breakLabel(std::string("")), generator( gen ) {}
+                ~MultiLevelExitMutator();
+                void premutate( FunctionDecl * );
+                MLEMutator( std::map<Label, Statement *> *t, LabelGenerator *gen = 0 ) : targetTable( t ), breakLabel(std::string("")), generator( gen ) {}
+                ~MLEMutator();
                 void premutate( CompoundStmt *cmpndStmt );
 …
                 void premutate( SwitchStmt *switchStmt );
                 Statement * postmutate( SwitchStmt *switchStmt );
-                void premutate( ReturnStmt *returnStmt );
                 void premutate( TryStmt *tryStmt );
                 Statement * postmutate( TryStmt *tryStmt );
-                void premutate( FinallyStmt *finallyStmt );
                 Statement *mutateLoop( Statement *bodyLoop, Entry &e );
 …
                 Label breakLabel;
                 LabelGenerator *generator;
-                bool inFinally = false;
                 template< typename LoopClass >

src/ControlStruct/Mutate.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 03:22:07 2020
 // Update Count     : 10
+// Last Modified On : Thu Aug  4 11:39:08 2016
+// Update Count     : 9
 //
 …
                 mutateAll( translationUnit, formut );
+        }
 } // namespace ControlStruct
+} // namespace CodeGen
 // Local Variables: //

src/GenPoly/Box.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:40:34 2019
 // Update Count     : 347
+// Last Modified On : Wed Jun 21 15:49:59 2017
+// Update Count     : 346
 //
 …
 #include "InitTweak/InitTweak.h"         // for getFunctionName, isAssignment
 #include "Lvalue.h"                      // for generalizedLvalue
+#include "Parser/LinkageSpec.h"          // for C, Spec, Cforall, Intrinsic
 #include "ResolvExpr/TypeEnvironment.h"  // for EqvClass
 #include "ResolvExpr/typeops.h"          // for typesCompatible
 …
 #include "SymTab/Indexer.h"              // for Indexer
 #include "SymTab/Mangler.h"              // for Mangler
-#include "SynTree/LinkageSpec.h"         // for C, Spec, Cforall, Intrinsic
 #include "SynTree/Attribute.h"           // for Attribute
 #include "SynTree/Constant.h"            // for Constant

src/GenPoly/Lvalue.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:14:38 2019
 // Update Count     : 7
+// Last Modified On : Fri Mar 17 09:11:18 2017
+// Update Count     : 5
 //
 …
 #include <string>                        // for string
-#include "Common/UniqueName.h"
 #include "Common/PassVisitor.h"
 #include "GenPoly.h"                     // for isPolyType
 …
 #include "InitTweak/InitTweak.h"
+#include "Parser/LinkageSpec.h"          // for Spec, isBuiltin, Intrinsic
 #include "ResolvExpr/TypeEnvironment.h"  // for AssertionSet, OpenVarSet
 #include "ResolvExpr/Unify.h"            // for unify
 #include "ResolvExpr/typeops.h"
 #include "SymTab/Indexer.h"              // for Indexer
-#include "SynTree/LinkageSpec.h"         // for Spec, isBuiltin, Intrinsic
 #include "SynTree/Declaration.h"         // for Declaration, FunctionDecl
 #include "SynTree/Expression.h"          // for Expression, ConditionalExpr
 …
+                }
                 struct ReferenceConversions final : public WithStmtsToAdd, public WithGuards {
+                struct ReferenceConversions final : public WithStmtsToAdd {
                         Expression * postmutate( CastExpr * castExpr );
                         Expression * postmutate( AddressExpr * addrExpr );
 …
                 struct FixIntrinsicResult final : public WithGuards {
-                        enum {
-                                NoSkip,
-                                Skip,
-                                SkipInProgress
-                        } skip = NoSkip;
-                        void premutate( AsmExpr * ) { GuardValue( skip ); skip = Skip; }
-                        void premutate( ApplicationExpr * ) { GuardValue( skip ); skip = (skip == Skip) ? SkipInProgress : NoSkip; }
                         Expression * postmutate( ApplicationExpr * appExpr );
                         void premutate( FunctionDecl * funcDecl );
 …
                 Expression * FixIntrinsicResult::postmutate( ApplicationExpr * appExpr ) {
                         if ( skip != SkipInProgress && isIntrinsicReference( appExpr ) ) {
+                        if ( isIntrinsicReference( appExpr ) ) {
                                 // eliminate reference types from intrinsic applications - now they return lvalues
                                 ReferenceType * result = strict_dynamic_cast< ReferenceType * >( appExpr->result );

src/GenPoly/Specialize.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:40:49 2019
 // Update Count     : 32
+// Last Modified On : Thu Mar 16 07:53:59 2017
+// Update Count     : 31
 //
 …
 #include "GenPoly.h"                     // for getFunctionType
 #include "InitTweak/InitTweak.h"         // for isIntrinsicCallExpr
+#include "Parser/LinkageSpec.h"          // for C
 #include "ResolvExpr/FindOpenVars.h"     // for findOpenVars
 #include "ResolvExpr/TypeEnvironment.h"  // for OpenVarSet, AssertionSet
 #include "Specialize.h"
-#include "SynTree/LinkageSpec.h"         // for C
 #include "SynTree/Attribute.h"           // for Attribute
 #include "SynTree/Declaration.h"         // for FunctionDecl, DeclarationWit...

src/InitTweak/FixGlobalInit.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 04 15:14:56 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:41:10 2019
 // Update Count     : 19
+// Last Modified On : Thu Mar 16 07:53:11 2017
+// Update Count     : 18
 //
 …
 #include "Common/UniqueName.h"     // for UniqueName
 #include "InitTweak.h"             // for isIntrinsicSingleArgCallStmt
 #include "SynTree/LinkageSpec.h"   // for C
+#include "Parser/LinkageSpec.h"    // for C
 #include "SynTree/Attribute.h"     // for Attribute
 #include "SynTree/Constant.h"      // for Constant

src/InitTweak/FixInit.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Jan 13 16:29:30 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 04:17:07 2020
 // Update Count     : 82
+// Last Modified On : Wed Feb 13 18:15:56 2019
+// Update Count     : 76
 //
 #include "FixInit.h"
 …
 #include "GenPoly/GenPoly.h"           // for getFunctionType
 #include "InitTweak.h"                 // for getFunctionName, getCallArg
+#include "Parser/LinkageSpec.h"        // for C, Spec, Cforall, isBuiltin
 #include "ResolvExpr/Resolver.h"       // for findVoidExpression
 #include "ResolvExpr/typeops.h"        // for typesCompatible
 …
 #include "SymTab/Indexer.h"            // for Indexer
 #include "SymTab/Mangler.h"            // for Mangler
-#include "SynTree/LinkageSpec.h"       // for C, Spec, Cforall, isBuiltin
 #include "SynTree/Attribute.h"         // for Attribute
 #include "SynTree/Constant.h"          // for Constant
 …
+                }
                 // to prevent warnings ('_unq0' may be used uninitialized in this function),
+                // to prevent warnings (‘_unq0’ may be used uninitialized in this function),
                 // insert an appropriate zero initializer for UniqueExpr temporaries.
                 Initializer * makeInit( Type * t ) {

src/InitTweak/FixInit.h

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Jan 13 16:29:30 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 07:54:50 2020
 // Update Count     : 8
+// Last Modified On : Sat Jul 22 09:31:06 2017
+// Update Count     : 6
 //
 …
 namespace InitTweak {
+        /// replace constructor initializers with expression statements and unwrap basic C-style initializers
+  /// replace constructor initializers with expression statements
+  /// and unwrap basic C-style initializers
         void fix( std::list< Declaration * > & translationUnit, bool inLibrary );
 } // namespace

src/InitTweak/GenInit.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:15:10 2019
 // Update Count     : 184
+// Last Modified On : Fri Mar 17 09:12:36 2017
+// Update Count     : 183
 //
 #include "GenInit.h"
 …
 #include "GenPoly/ScopedSet.h"         // for ScopedSet, ScopedSet<>::const_iter...
 #include "InitTweak.h"                 // for isConstExpr, InitExpander, checkIn...
+#include "Parser/LinkageSpec.h"        // for isOverridable, C
 #include "ResolvExpr/Resolver.h"
 #include "SymTab/Autogen.h"            // for genImplicitCall
 #include "SymTab/Mangler.h"            // for Mangler
-#include "SynTree/LinkageSpec.h"       // for isOverridable, C
 #include "SynTree/Declaration.h"       // for ObjectDecl, DeclarationWithType
 #include "SynTree/Expression.h"        // for VariableExpr, UntypedExpr, Address...

src/InitTweak/InitTweak.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Fri May 13 11:26:36 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:15:52 2019
 // Update Count     : 8
+// Last Modified On : Thu Jul 25 22:21:48 2019
+// Update Count     : 7
 //
 …
 #include "GenPoly/GenPoly.h"       // for getFunctionType
 #include "InitTweak.h"
+#include "Parser/LinkageSpec.h"    // for Spec, isBuiltin, Intrinsic
 #include "ResolvExpr/typeops.h"    // for typesCompatibleIgnoreQualifiers
 #include "SymTab/Autogen.h"
 #include "SymTab/Indexer.h"        // for Indexer
-#include "SynTree/LinkageSpec.h"   // for Spec, isBuiltin, Intrinsic
 #include "SynTree/Attribute.h"     // for Attribute
 #include "SynTree/Constant.h"      // for Constant

src/MakeLibCfa.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 10:33:33 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Feb 16 03:49:49 2020
 // Update Count     : 45
+// Last Modified On : Sun Feb 17 21:08:09 2019
+// Update Count     : 41
 //
 …
 #include "Common/SemanticError.h"   // for SemanticError
 #include "Common/UniqueName.h"      // for UniqueName
 #include "SynTree/LinkageSpec.h"    // for Spec, Intrinsic, C
+#include "Parser/LinkageSpec.h"     // for Spec, Intrinsic, C
 #include "SynTree/Declaration.h"    // for FunctionDecl, ObjectDecl, Declara...
 #include "SynTree/Expression.h"     // for NameExpr, UntypedExpr, VariableExpr
 …
                         FunctionDecl *funcDecl = origFuncDecl->clone();
                         const CodeGen::OperatorInfo * opInfo;
                         opInfo = CodeGen::operatorLookup( funcDecl->get_name() );
                         assert( opInfo );
+                        CodeGen::OperatorInfo opInfo;
+                        bool lookResult = CodeGen::operatorLookup( funcDecl->get_name(), opInfo );
+                        assert( lookResult );
                         assert( ! funcDecl->get_statements() );
                         // build a recursive call - this is okay, as the call will actually be codegen'd using operator syntax
 …
                         Statement * stmt = nullptr;
                         switch ( opInfo->type ) {
+                        switch ( opInfo.type ) {
                           case CodeGen::OT_INDEX:
                           case CodeGen::OT_CALL:

src/Makefile.in

-              r7030dab
+              r71d6bd8
         SymTab/Indexer.$(OBJEXT) SymTab/Mangler.$(OBJEXT) \
         SymTab/ManglerCommon.$(OBJEXT) SymTab/Validate.$(OBJEXT)
+am__objects_7 = SynTree/AddressExpr.$(OBJEXT) \
+        SynTree/AggregateDecl.$(OBJEXT) \
+        SynTree/ApplicationExpr.$(OBJEXT) SynTree/ArrayType.$(OBJEXT) \
+        SynTree/AttrType.$(OBJEXT) SynTree/Attribute.$(OBJEXT) \
+        SynTree/BasicType.$(OBJEXT) SynTree/CommaExpr.$(OBJEXT) \
+        SynTree/CompoundStmt.$(OBJEXT) SynTree/Constant.$(OBJEXT) \
+        SynTree/DeclReplacer.$(OBJEXT) SynTree/DeclStmt.$(OBJEXT) \
+am__objects_7 = SynTree/Type.$(OBJEXT) SynTree/VoidType.$(OBJEXT) \
+        SynTree/BasicType.$(OBJEXT) SynTree/PointerType.$(OBJEXT) \
+        SynTree/ArrayType.$(OBJEXT) SynTree/ReferenceType.$(OBJEXT) \
+        SynTree/FunctionType.$(OBJEXT) \
+        SynTree/ReferenceToType.$(OBJEXT) SynTree/TupleType.$(OBJEXT) \
+        SynTree/TypeofType.$(OBJEXT) SynTree/AttrType.$(OBJEXT) \
+        SynTree/VarArgsType.$(OBJEXT) SynTree/ZeroOneType.$(OBJEXT) \
+        SynTree/Constant.$(OBJEXT) SynTree/Expression.$(OBJEXT) \
+        SynTree/TupleExpr.$(OBJEXT) SynTree/CommaExpr.$(OBJEXT) \
+        SynTree/TypeExpr.$(OBJEXT) SynTree/ApplicationExpr.$(OBJEXT) \
+        SynTree/AddressExpr.$(OBJEXT) SynTree/Statement.$(OBJEXT) \
+        SynTree/CompoundStmt.$(OBJEXT) SynTree/DeclStmt.$(OBJEXT) \
         SynTree/Declaration.$(OBJEXT) \
         SynTree/DeclarationWithType.$(OBJEXT) \
+        SynTree/Expression.$(OBJEXT) SynTree/FunctionDecl.$(OBJEXT) \
+        SynTree/FunctionType.$(OBJEXT) SynTree/Initializer.$(OBJEXT) \
+        SynTree/LinkageSpec.$(OBJEXT) SynTree/NamedTypeDecl.$(OBJEXT) \
+        SynTree/ObjectDecl.$(OBJEXT) SynTree/PointerType.$(OBJEXT) \
+        SynTree/ReferenceToType.$(OBJEXT) \
+        SynTree/ReferenceType.$(OBJEXT) SynTree/Statement.$(OBJEXT) \
+        SynTree/TupleExpr.$(OBJEXT) SynTree/TupleType.$(OBJEXT) \
+        SynTree/Type.$(OBJEXT) SynTree/TypeDecl.$(OBJEXT) \
+        SynTree/TypeExpr.$(OBJEXT) SynTree/TypeSubstitution.$(OBJEXT) \
+        SynTree/TypeofType.$(OBJEXT) SynTree/VarArgsType.$(OBJEXT) \
+        SynTree/VoidType.$(OBJEXT) SynTree/ZeroOneType.$(OBJEXT)
+        SynTree/ObjectDecl.$(OBJEXT) SynTree/FunctionDecl.$(OBJEXT) \
+        SynTree/AggregateDecl.$(OBJEXT) \
+        SynTree/NamedTypeDecl.$(OBJEXT) SynTree/TypeDecl.$(OBJEXT) \
+        SynTree/Initializer.$(OBJEXT) \
+        SynTree/TypeSubstitution.$(OBJEXT) SynTree/Attribute.$(OBJEXT) \
+        SynTree/DeclReplacer.$(OBJEXT)
 am__objects_8 = CompilationState.$(OBJEXT) $(am__objects_1) \
         $(am__objects_2) Concurrency/Keywords.$(OBJEXT) \
         $(am__objects_3) $(am__objects_4) GenPoly/GenPoly.$(OBJEXT) \
         GenPoly/Lvalue.$(OBJEXT) InitTweak/GenInit.$(OBJEXT) \
         InitTweak/InitTweak.$(OBJEXT) $(am__objects_5) \
         $(am__objects_6) SymTab/Demangle.$(OBJEXT) $(am__objects_7) \
         Tuples/TupleAssignment.$(OBJEXT) \
+        InitTweak/InitTweak.$(OBJEXT) Parser/LinkageSpec.$(OBJEXT) \
+        $(am__objects_5) $(am__objects_6) SymTab/Demangle.$(OBJEXT) \
+        $(am__objects_7) Tuples/TupleAssignment.$(OBJEXT) \
         Tuples/TupleExpansion.$(OBJEXT) Tuples/Explode.$(OBJEXT) \
         Tuples/Tuples.$(OBJEXT) Validate/HandleAttributes.$(OBJEXT) \
 …
         InitTweak/GenInit.$(OBJEXT) InitTweak/FixInit.$(OBJEXT) \
         InitTweak/FixGlobalInit.$(OBJEXT) \
         InitTweak/InitTweak.$(OBJEXT) Parser/DeclarationNode.$(OBJEXT) \
         Parser/ExpressionNode.$(OBJEXT) \
         Parser/InitializerNode.$(OBJEXT) Parser/ParseNode.$(OBJEXT) \
         Parser/StatementNode.$(OBJEXT) Parser/TypeData.$(OBJEXT) \
         Parser/TypedefTable.$(OBJEXT) Parser/lex.$(OBJEXT) \
         Parser/parser.$(OBJEXT) Parser/parserutility.$(OBJEXT) \
+        InitTweak/InitTweak.$(OBJEXT) Parser/parser.$(OBJEXT) \
+        Parser/lex.$(OBJEXT) Parser/TypedefTable.$(OBJEXT) \
+        Parser/ParseNode.$(OBJEXT) Parser/DeclarationNode.$(OBJEXT) \
+        Parser/ExpressionNode.$(OBJEXT) Parser/StatementNode.$(OBJEXT) \
+        Parser/InitializerNode.$(OBJEXT) Parser/TypeData.$(OBJEXT) \
+        Parser/LinkageSpec.$(OBJEXT) Parser/parserutility.$(OBJEXT) \
         $(am__objects_5) ResolvExpr/AlternativePrinter.$(OBJEXT) \
         $(am__objects_6) $(am__objects_7) \
 …
         InitTweak/GenInit.cc InitTweak/FixInit.cc \
         InitTweak/FixGlobalInit.cc InitTweak/InitTweak.cc \
         Parser/DeclarationNode.cc Parser/ExpressionNode.cc \
         Parser/InitializerNode.cc Parser/ParseNode.cc \
         Parser/StatementNode.cc Parser/TypeData.cc \
         Parser/TypedefTable.cc Parser/lex.ll Parser/parser.yy \
         Parser/parserutility.cc $(SRC_RESOLVEXPR) \
         ResolvExpr/AlternativePrinter.cc $(SRC_SYMTAB) $(SRC_SYNTREE) \
         Tuples/TupleAssignment.cc Tuples/TupleExpansion.cc \
         Tuples/Explode.cc Tuples/Tuples.cc \
+        Parser/parser.yy Parser/lex.ll Parser/TypedefTable.cc \
+        Parser/ParseNode.cc Parser/DeclarationNode.cc \
+        Parser/ExpressionNode.cc Parser/StatementNode.cc \
+        Parser/InitializerNode.cc Parser/TypeData.cc \
+        Parser/LinkageSpec.cc Parser/parserutility.cc \
+        $(SRC_RESOLVEXPR) ResolvExpr/AlternativePrinter.cc \
+        $(SRC_SYMTAB) $(SRC_SYNTREE) Tuples/TupleAssignment.cc \
+        Tuples/TupleExpansion.cc Tuples/Explode.cc Tuples/Tuples.cc \
         Validate/HandleAttributes.cc Validate/FindSpecialDecls.cc \
         Virtual/ExpandCasts.cc
 …
         Concurrency/Keywords.cc $(SRC_COMMON) $(SRC_CONTROLSTRUCT) \
         GenPoly/GenPoly.cc GenPoly/Lvalue.cc InitTweak/GenInit.cc \
+        InitTweak/InitTweak.cc $(SRC_RESOLVEXPR) $(SRC_SYMTAB) \
+        SymTab/Demangle.cc $(SRC_SYNTREE) Tuples/TupleAssignment.cc \
+        Tuples/TupleExpansion.cc Tuples/Explode.cc Tuples/Tuples.cc \
+        InitTweak/InitTweak.cc Parser/LinkageSpec.cc $(SRC_RESOLVEXPR) \
+        $(SRC_SYMTAB) SymTab/Demangle.cc $(SRC_SYNTREE) \
+        Tuples/TupleAssignment.cc Tuples/TupleExpansion.cc \
+        Tuples/Explode.cc Tuples/Tuples.cc \
         Validate/HandleAttributes.cc Validate/FindSpecialDecls.cc
 MAINTAINERCLEANFILES = ${libdir}/${notdir ${cfa_cpplib_PROGRAMS}}
 …
 SRC_SYNTREE = \
+      SynTree/Type.cc \
+      SynTree/VoidType.cc \
+      SynTree/BasicType.cc \
+      SynTree/PointerType.cc \
+      SynTree/ArrayType.cc \
+      SynTree/ReferenceType.cc \
+      SynTree/FunctionType.cc \
+      SynTree/ReferenceToType.cc \
+      SynTree/TupleType.cc \
+      SynTree/TypeofType.cc \
+      SynTree/AttrType.cc \
+      SynTree/VarArgsType.cc \
+      SynTree/ZeroOneType.cc \
+      SynTree/Constant.cc \
+      SynTree/Expression.cc \
+      SynTree/TupleExpr.cc \
+      SynTree/CommaExpr.cc \
+      SynTree/TypeExpr.cc \
+      SynTree/ApplicationExpr.cc \
       SynTree/AddressExpr.cc \
+      SynTree/AggregateDecl.cc \
+      SynTree/ApplicationExpr.cc \
+      SynTree/ArrayType.cc \
+      SynTree/AttrType.cc \
+      SynTree/Attribute.cc \
+      SynTree/BasicType.cc \
+      SynTree/CommaExpr.cc \
+      SynTree/Statement.cc \
       SynTree/CompoundStmt.cc \
-      SynTree/Constant.cc \
-      SynTree/DeclReplacer.cc \
       SynTree/DeclStmt.cc \
       SynTree/Declaration.cc \
       SynTree/DeclarationWithType.cc \
       SynTree/Expression.cc \
+      SynTree/ObjectDecl.cc \
       SynTree/FunctionDecl.cc \
+      SynTree/FunctionType.cc \
+      SynTree/AggregateDecl.cc \
+      SynTree/NamedTypeDecl.cc \
+      SynTree/TypeDecl.cc \
       SynTree/Initializer.cc \
-      SynTree/LinkageSpec.cc \
-      SynTree/NamedTypeDecl.cc \
-      SynTree/ObjectDecl.cc \
-      SynTree/PointerType.cc \
-      SynTree/ReferenceToType.cc \
-      SynTree/ReferenceType.cc \
-      SynTree/Statement.cc \
-      SynTree/TupleExpr.cc \
-      SynTree/TupleType.cc \
-      SynTree/Type.cc \
-      SynTree/TypeDecl.cc \
-      SynTree/TypeExpr.cc \
       SynTree/TypeSubstitution.cc \
+      SynTree/TypeofType.cc \
+      SynTree/VarArgsType.cc \
+      SynTree/VoidType.cc \
+      SynTree/ZeroOneType.cc
+      SynTree/Attribute.cc \
+      SynTree/DeclReplacer.cc
 …
 InitTweak/InitTweak.$(OBJEXT): InitTweak/$(am__dirstamp) \
         InitTweak/$(DEPDIR)/$(am__dirstamp)
+Parser/$(am__dirstamp):
+        @$(MKDIR_P) Parser
+        @: > Parser/$(am__dirstamp)
+Parser/$(DEPDIR)/$(am__dirstamp):
+        @$(MKDIR_P) Parser/$(DEPDIR)
+        @: > Parser/$(DEPDIR)/$(am__dirstamp)
+Parser/LinkageSpec.$(OBJEXT): Parser/$(am__dirstamp) \
+        Parser/$(DEPDIR)/$(am__dirstamp)
 ResolvExpr/$(am__dirstamp):
         @$(MKDIR_P) ResolvExpr
 …
         @$(MKDIR_P) SynTree/$(DEPDIR)
         @: > SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/Type.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/VoidType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/BasicType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/PointerType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ArrayType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ReferenceType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/FunctionType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ReferenceToType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/TupleType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/TypeofType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/AttrType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/VarArgsType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ZeroOneType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/Constant.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/Expression.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/TupleExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/CommaExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/TypeExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ApplicationExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
 SynTree/AddressExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
         SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/Statement.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/CompoundStmt.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/DeclStmt.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/Declaration.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/DeclarationWithType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ObjectDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/FunctionDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
 SynTree/AggregateDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
         SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ApplicationExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/ArrayType.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/AttrType.$(OBJEXT): SynTree/$(am__dirstamp) \
+SynTree/NamedTypeDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/TypeDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/Initializer.$(OBJEXT): SynTree/$(am__dirstamp) \
+        SynTree/$(DEPDIR)/$(am__dirstamp)
+SynTree/TypeSubstitution.$(OBJEXT): SynTree/$(am__dirstamp) \
         SynTree/$(DEPDIR)/$(am__dirstamp)
 SynTree/Attribute.$(OBJEXT): SynTree/$(am__dirstamp) \
         SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/BasicType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/CommaExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/CompoundStmt.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/Constant.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
 SynTree/DeclReplacer.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/DeclStmt.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/Declaration.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/DeclarationWithType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/Expression.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/FunctionDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/FunctionType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/Initializer.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/LinkageSpec.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/NamedTypeDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/ObjectDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/PointerType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/ReferenceToType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/ReferenceType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/Statement.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/TupleExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/TupleType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/Type.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/TypeDecl.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/TypeExpr.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/TypeSubstitution.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/TypeofType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/VarArgsType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/VoidType.$(OBJEXT): SynTree/$(am__dirstamp) \
-        SynTree/$(DEPDIR)/$(am__dirstamp)
-SynTree/ZeroOneType.$(OBJEXT): SynTree/$(am__dirstamp) \
         SynTree/$(DEPDIR)/$(am__dirstamp)
 Tuples/$(am__dirstamp):
 …
 InitTweak/FixGlobalInit.$(OBJEXT): InitTweak/$(am__dirstamp) \
         InitTweak/$(DEPDIR)/$(am__dirstamp)
+Parser/$(am__dirstamp):
+        @$(MKDIR_P) Parser
+        @: > Parser/$(am__dirstamp)
+Parser/$(DEPDIR)/$(am__dirstamp):
+        @$(MKDIR_P) Parser/$(DEPDIR)
+        @: > Parser/$(DEPDIR)/$(am__dirstamp)
+Parser/parser.hh: Parser/parser.cc
+        @if test ! -f $@; then rm -f Parser/parser.cc; else :; fi
+        @if test ! -f $@; then $(MAKE) $(AM_MAKEFLAGS) Parser/parser.cc; else :; fi
+Parser/parser.$(OBJEXT): Parser/$(am__dirstamp) \
+        Parser/$(DEPDIR)/$(am__dirstamp)
+Parser/lex.$(OBJEXT): Parser/$(am__dirstamp) \
+        Parser/$(DEPDIR)/$(am__dirstamp)
+Parser/TypedefTable.$(OBJEXT): Parser/$(am__dirstamp) \
+        Parser/$(DEPDIR)/$(am__dirstamp)
+Parser/ParseNode.$(OBJEXT): Parser/$(am__dirstamp) \
+        Parser/$(DEPDIR)/$(am__dirstamp)
 Parser/DeclarationNode.$(OBJEXT): Parser/$(am__dirstamp) \
         Parser/$(DEPDIR)/$(am__dirstamp)
 Parser/ExpressionNode.$(OBJEXT): Parser/$(am__dirstamp) \
         Parser/$(DEPDIR)/$(am__dirstamp)
+Parser/StatementNode.$(OBJEXT): Parser/$(am__dirstamp) \
+        Parser/$(DEPDIR)/$(am__dirstamp)
 Parser/InitializerNode.$(OBJEXT): Parser/$(am__dirstamp) \
         Parser/$(DEPDIR)/$(am__dirstamp)
-Parser/ParseNode.$(OBJEXT): Parser/$(am__dirstamp) \
-        Parser/$(DEPDIR)/$(am__dirstamp)
-Parser/StatementNode.$(OBJEXT): Parser/$(am__dirstamp) \
-        Parser/$(DEPDIR)/$(am__dirstamp)
 Parser/TypeData.$(OBJEXT): Parser/$(am__dirstamp) \
-        Parser/$(DEPDIR)/$(am__dirstamp)
-Parser/TypedefTable.$(OBJEXT): Parser/$(am__dirstamp) \
-        Parser/$(DEPDIR)/$(am__dirstamp)
-Parser/lex.$(OBJEXT): Parser/$(am__dirstamp) \
-        Parser/$(DEPDIR)/$(am__dirstamp)
-Parser/parser.hh: Parser/parser.cc
-        @if test ! -f $@; then rm -f Parser/parser.cc; else :; fi
-        @if test ! -f $@; then $(MAKE) $(AM_MAKEFLAGS) Parser/parser.cc; else :; fi
-Parser/parser.$(OBJEXT): Parser/$(am__dirstamp) \
         Parser/$(DEPDIR)/$(am__dirstamp)
 Parser/parserutility.$(OBJEXT): Parser/$(am__dirstamp) \
 …
 @AMDEP_TRUE@@am__include@ @am__quote@Parser/$(DEPDIR)/ExpressionNode.Po@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@Parser/$(DEPDIR)/InitializerNode.Po@am__quote@
+@AMDEP_TRUE@@am__include@ @am__quote@Parser/$(DEPDIR)/LinkageSpec.Po@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@Parser/$(DEPDIR)/ParseNode.Po@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@Parser/$(DEPDIR)/StatementNode.Po@am__quote@
 …
 @AMDEP_TRUE@@am__include@ @am__quote@SynTree/$(DEPDIR)/FunctionType.Po@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@SynTree/$(DEPDIR)/Initializer.Po@am__quote@
-@AMDEP_TRUE@@am__include@ @am__quote@SynTree/$(DEPDIR)/LinkageSpec.Po@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@SynTree/$(DEPDIR)/NamedTypeDecl.Po@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@SynTree/$(DEPDIR)/ObjectDecl.Po@am__quote@

src/Parser/DeclarationNode.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 12:34:05 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Dec 16 15:32:22 2019
 // Update Count     : 1133
+// Last Modified On : Thu Jul 25 22:17:10 2019
+// Update Count     : 1116
 //
 …
 #include "Common/UniqueName.h"     // for UniqueName
 #include "Common/utility.h"        // for maybeClone, maybeBuild, CodeLocation
+#include "Parser/LinkageSpec.h"    // for Spec, linkageName, Cforall
 #include "Parser/ParseNode.h"      // for DeclarationNode, ExpressionNode
-#include "SynTree/LinkageSpec.h"   // for Spec, linkageName, Cforall
 #include "SynTree/Attribute.h"     // for Attribute
 #include "SynTree/Declaration.h"   // for TypeDecl, ObjectDecl, Declaration
 …
 const char * DeclarationNode::signednessNames[] = { "signed", "unsigned", "NoSignednessNames" };
 const char * DeclarationNode::lengthNames[] = { "short", "long", "long long", "NoLengthNames" };
+const char * DeclarationNode::aggregateNames[] = { "struct", "union", "trait", "coroutine", "monitor", "thread", "NoAggregateNames" };
+const char * DeclarationNode::typeClassNames[] = { "otype", "dtype", "ftype", "NoTypeClassNames" };
 const char * DeclarationNode::builtinTypeNames[] = { "__builtin_va_list", "__auto_type", "zero_t", "one_t", "NoBuiltinTypeNames" };
 …
 //      variable.name = nullptr;
         variable.tyClass = TypeDecl::NUMBER_OF_KINDS;
+        variable.tyClass = NoTypeClass;
         variable.assertions = nullptr;
         variable.initializer = nullptr;
 …
         if ( linkage != LinkageSpec::Cforall ) {
                 os << LinkageSpec::name( linkage ) << " ";
+                os << LinkageSpec::linkageName( linkage ) << " ";
         } // if
 …
+}
 DeclarationNode * DeclarationNode::newAggregate( AggregateDecl::Aggregate kind, const string * name, ExpressionNode * actuals, DeclarationNode * fields, bool body ) {
+DeclarationNode * DeclarationNode::newAggregate( Aggregate kind, const string * name, ExpressionNode * actuals, DeclarationNode * fields, bool body ) {
         DeclarationNode * newnode = new DeclarationNode;
         newnode->type = new TypeData( TypeData::Aggregate );
 …
 } // DeclarationNode::newFromTypeGen
 DeclarationNode * DeclarationNode::newTypeParam( TypeDecl::Kind tc, const string * name ) {
+DeclarationNode * DeclarationNode::newTypeParam( TypeClass tc, const string * name ) {
         DeclarationNode * newnode = new DeclarationNode;
         newnode->type = nullptr;
 …
         newnode->type = new TypeData( TypeData::Aggregate );
         newnode->type->aggregate.name = name;
         newnode->type->aggregate.kind = AggregateDecl::Trait;
+        newnode->type->aggregate.kind = Trait;
         newnode->type->aggregate.params = params;
         newnode->type->aggregate.fields = asserts;
 …
         newnode->type = new TypeData( TypeData::AggregateInst );
         newnode->type->aggInst.aggregate = new TypeData( TypeData::Aggregate );
         newnode->type->aggInst.aggregate->aggregate.kind = AggregateDecl::Trait;
+        newnode->type->aggInst.aggregate->aggregate.kind = Trait;
         newnode->type->aggInst.aggregate->aggregate.name = name;
         newnode->type->aggInst.params = params;
 …
 DeclarationNode * DeclarationNode::addAssertions( DeclarationNode * assertions ) {
         if ( variable.tyClass != TypeDecl::NUMBER_OF_KINDS ) {
+        if ( variable.tyClass != NoTypeClass ) {
                 if ( variable.assertions ) {
                         variable.assertions->appendList( assertions );
 …
 DeclarationNode * DeclarationNode::addTypeInitializer( DeclarationNode * init ) {
         assertf( variable.tyClass != TypeDecl::NUMBER_OF_KINDS, "Called addTypeInitializer on something that isn't a type variable." );
+        assertf( variable.tyClass != NoTypeClass, "Called addTypeInitializer on something that isn't a type variable." );
         variable.initializer = init;
         return this;
 …
         } // if
         if ( variable.tyClass != TypeDecl::NUMBER_OF_KINDS ) {
+        if ( variable.tyClass != NoTypeClass ) {
                 // otype is internally converted to dtype + otype parameters
                 static const TypeDecl::Kind kindMap[] = { TypeDecl::Dtype, TypeDecl::Dtype, TypeDecl::Ftype, TypeDecl::Ttype };
                 static_assert( sizeof(kindMap)/sizeof(kindMap[0]) == TypeDecl::NUMBER_OF_KINDS, "DeclarationNode::build: kindMap is out of sync." );
+                assertf( sizeof(kindMap)/sizeof(kindMap[0]) == NoTypeClass, "DeclarationNode::build: kindMap is out of sync." );
                 assertf( variable.tyClass < sizeof(kindMap)/sizeof(kindMap[0]), "Variable's tyClass is out of bounds." );
                 TypeDecl * ret = new TypeDecl( *name, Type::StorageClasses(), nullptr, kindMap[ variable.tyClass ], variable.tyClass == TypeDecl::Otype, variable.initializer ? variable.initializer->buildType() : nullptr );
+                TypeDecl * ret = new TypeDecl( *name, Type::StorageClasses(), nullptr, kindMap[ variable.tyClass ], variable.tyClass == Otype, variable.initializer ? variable.initializer->buildType() : nullptr );
                 buildList( variable.assertions, ret->get_assertions() );
                 return ret;

src/Parser/ExpressionNode.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 13:17:07 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 18 21:14:58 2019
 // Update Count     : 981
+// Last Modified On : Sun Aug  4 20:57:55 2019
+// Update Count     : 978
 //
 …
         static const BasicType::Kind kind[2][12] = {
                 { BasicType::Float, BasicType::Double, BasicType::LongDouble, BasicType::uuFloat80, BasicType::uuFloat128, BasicType::uFloat16, BasicType::uFloat32, BasicType::uFloat32x, BasicType::uFloat64, BasicType::uFloat64x, BasicType::uFloat128, BasicType::uFloat128x },
                 { BasicType::FloatComplex, BasicType::DoubleComplex, BasicType::LongDoubleComplex, BasicType::NUMBER_OF_BASIC_TYPES, BasicType::NUMBER_OF_BASIC_TYPES, BasicType::uFloat16Complex, BasicType::uFloat32Complex, BasicType::uFloat32xComplex, BasicType::uFloat64Complex, BasicType::uFloat64xComplex, BasicType::uFloat128Complex, BasicType::uFloat128xComplex },
+                { BasicType::FloatComplex, BasicType::DoubleComplex, BasicType::LongDoubleComplex, (BasicType::Kind)-1, (BasicType::Kind)-1, BasicType::uFloat16Complex, BasicType::uFloat32Complex, BasicType::uFloat32xComplex, BasicType::uFloat64Complex, BasicType::uFloat64xComplex, BasicType::uFloat128Complex, BasicType::uFloat128xComplex },
         };
 …
 Expression * build_field_name_FLOATING_DECIMALconstant( const string & str ) {
         if ( str[str.size() - 1] != '.' ) SemanticError( yylloc, "invalid tuple index " + str );
+        if ( str[str.size()-1] != '.' ) SemanticError( yylloc, "invalid tuple index " + str );
         Expression * ret = build_constantInteger( *new string( str.substr( 0, str.size()-1 ) ) );
         delete &str;
 …
 } // build_cast
 Expression * build_keyword_cast( AggregateDecl::Aggregate target, ExpressionNode * expr_node ) {
+Expression * build_keyword_cast( KeywordCastExpr::Target target, ExpressionNode * expr_node ) {
         return new KeywordCastExpr( maybeMoveBuild< Expression >(expr_node), target );
+}

src/Parser/ParseNode.h

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 13:28:16 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 17:56:02 2020
 // Update Count     : 891
+// Last Modified On : Thu Jul 25 22:17:10 2019
+// Update Count     : 876
 //
 …
 #include "Common/UniqueName.h"     // for UniqueName
 #include "Common/utility.h"        // for maybeClone, maybeBuild
+#include "SynTree/LinkageSpec.h"   // for Spec
+#include "SynTree/Declaration.h"   // for Aggregate
+#include "Parser/LinkageSpec.h"    // for Spec
 #include "SynTree/Expression.h"    // for Expression, ConstantExpr (ptr only)
 #include "SynTree/Label.h"         // for Label
 …
 Expression * build_cast( DeclarationNode * decl_node, ExpressionNode * expr_node );
 Expression * build_keyword_cast( AggregateDecl::Aggregate target, ExpressionNode * expr_node );
+Expression * build_keyword_cast( KeywordCastExpr::Target target, ExpressionNode * expr_node );
 Expression * build_virtual_cast( DeclarationNode * decl_node, ExpressionNode * expr_node );
 Expression * build_fieldSel( ExpressionNode * expr_node, Expression * member );
 …
         enum Length { Short, Long, LongLong, NoLength };
         static const char * lengthNames[];
+        enum Aggregate { Struct, Union, Exception, Trait, Coroutine, Monitor, Thread, NoAggregate };
+        static const char * aggregateNames[];
+        enum TypeClass { Otype, Dtype, Ftype, Ttype, NoTypeClass };
+        static const char * typeClassNames[];
         enum BuiltinType { Valist, AutoType, Zero, One, NoBuiltinType };
         static const char * builtinTypeNames[];
 …
         static DeclarationNode * newQualifiedType( DeclarationNode *, DeclarationNode * );
         static DeclarationNode * newFunction( const std::string * name, DeclarationNode * ret, DeclarationNode * param, StatementNode * body );
         static DeclarationNode * newAggregate( AggregateDecl::Aggregate kind, const std::string * name, ExpressionNode * actuals, DeclarationNode * fields, bool body );
+        static DeclarationNode * newAggregate( Aggregate kind, const std::string * name, ExpressionNode * actuals, DeclarationNode * fields, bool body );
         static DeclarationNode * newEnum( const std::string * name, DeclarationNode * constants, bool body );
         static DeclarationNode * newEnumConstant( const std::string * name, ExpressionNode * constant );
         static DeclarationNode * newName( const std::string * );
         static DeclarationNode * newFromTypeGen( const std::string *, ExpressionNode * params );
         static DeclarationNode * newTypeParam( TypeDecl::Kind, const std::string * );
+        static DeclarationNode * newTypeParam( TypeClass, const std::string * );
         static DeclarationNode * newTrait( const std::string * name, DeclarationNode * params, DeclarationNode * asserts );
         static DeclarationNode * newTraitUse( const std::string * name, ExpressionNode * params );
 …
         struct Variable_t {
 //              const std::string * name;
                 TypeDecl::Kind tyClass;
+                DeclarationNode::TypeClass tyClass;
                 DeclarationNode * assertions;
                 DeclarationNode * initializer;
 …
 Statement * build_asm( bool voltile, Expression * instruction, ExpressionNode * output = nullptr, ExpressionNode * input = nullptr, ExpressionNode * clobber = nullptr, LabelNode * gotolabels = nullptr );
 Statement * build_directive( std::string * directive );
-SuspendStmt * build_suspend( StatementNode *, SuspendStmt::Type = SuspendStmt::None);
 WaitForStmt * build_waitfor( ExpressionNode * target, StatementNode * stmt, ExpressionNode * when );
 WaitForStmt * build_waitfor( ExpressionNode * target, StatementNode * stmt, ExpressionNode * when, WaitForStmt * existing );
 …
                                 * out++ = result;
                         } else {
                                 SemanticError( cur->location, "type specifier declaration in forall clause is currently unimplemented." );
+                                assertf(false, "buildList unknown type");
                         } // if
                 } catch( SemanticErrorException & e ) {

src/Parser/ParserTypes.h

-              r7030dab
+              r71d6bd8
 // Created On       : Sat Sep 22 08:58:10 2001
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb 15 11:04:40 2020
 // Update Count     : 351
+// Last Modified On : Sat Jul 22 09:33:28 2017
+// Update Count     : 350
 //
 …
 // current location in the input
 extern int yylineno;
 extern char * yyfilename;
+extern char *yyfilename;
 struct Location {
     char * file;
+    char *file;
     int line;
 }; // Location
 struct Token {
     std::string * str;                                                                  // must be pointer as used in union
+    std::string *str;                                                                   // must be pointer as used in union
     Location loc;

src/Parser/StatementNode.cc

-              r7030dab
+              r71d6bd8
 } // build_finally
-SuspendStmt * build_suspend( StatementNode * then, SuspendStmt::Type type ) {
-        auto node = new SuspendStmt();
-        node->type = type;
-        std::list< Statement * > stmts;
-        buildMoveList< Statement, StatementNode >( then, stmts );
-        if(!stmts.empty()) {
-                assert( stmts.size() == 1 );
-                node->then = dynamic_cast< CompoundStmt * >( stmts.front() );
+        }
-        return node;
+}
 WaitForStmt * build_waitfor( ExpressionNode * targetExpr, StatementNode * stmt, ExpressionNode * when ) {
         auto node = new WaitForStmt();

src/Parser/TypeData.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 15:12:51 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Dec 16 07:56:46 2019
 // Update Count     : 662
+// Last Modified On : Wed Feb 13 18:16:23 2019
+// Update Count     : 649
 //
 …
           case Aggregate:
                 // aggregate = new Aggregate_t;
                 aggregate.kind = AggregateDecl::NoAggregate;
+                aggregate.kind = DeclarationNode::NoAggregate;
                 aggregate.name = nullptr;
                 aggregate.params = nullptr;
 …
                 break;
           case Aggregate:
                 os << AggregateDecl::aggrString( aggregate.kind ) << ' ' << *aggregate.name << endl;
+                os << DeclarationNode::aggregateNames[ aggregate.kind ] << ' ' << *aggregate.name << endl;
                 if ( aggregate.params ) {
                         os << string( indent + 2, ' ' ) << "with type parameters" << endl;
 …
         for ( typename ForallList::iterator i = outputList.begin(); i != outputList.end(); ++i, n = (DeclarationNode*)n->get_next() ) {
                 TypeDecl * td = static_cast<TypeDecl *>(*i);
                 if ( n->variable.tyClass == TypeDecl::Otype ) {
+                if ( n->variable.tyClass == DeclarationNode::Otype ) {
                         // add assertion parameters to `type' tyvars in reverse order
                         // add dtor:  void ^?{}(T *)
 …
         switch ( td->kind ) {
           case TypeData::Unknown:
                 // fill in implicit int
                 return new BasicType( buildQualifiers( td ), BasicType::SignedInt );
+                        // fill in implicit int
+                        return new BasicType( buildQualifiers( td ), BasicType::SignedInt );
           case TypeData::Basic:
                 return buildBasicType( td );
+                        return buildBasicType( td );
           case TypeData::Pointer:
                 return buildPointer( td );
+                        return buildPointer( td );
           case TypeData::Array:
                 return buildArray( td );
+                        return buildArray( td );
           case TypeData::Reference:
                 return buildReference( td );
+                        return buildReference( td );
           case TypeData::Function:
                 return buildFunction( td );
+                        return buildFunction( td );
           case TypeData::AggregateInst:
                 return buildAggInst( td );
+                        return buildAggInst( td );
           case TypeData::EnumConstant:
                 // the name gets filled in later -- by SymTab::Validate
                 return new EnumInstType( buildQualifiers( td ), "" );
+                        // the name gets filled in later -- by SymTab::Validate
+                        return new EnumInstType( buildQualifiers( td ), "" );
           case TypeData::SymbolicInst:
                 return buildSymbolicInst( td );
+                        return buildSymbolicInst( td );
           case TypeData::Tuple:
                 return buildTuple( td );
+                        return buildTuple( td );
           case TypeData::Typeof:
           case TypeData::Basetypeof:
                 return buildTypeof( td );
+                        return buildTypeof( td );
           case TypeData::Builtin:
+                switch ( td->builtintype ) {
+                  case DeclarationNode::Zero:
+                        return new ZeroType( noQualifiers );
+                  case DeclarationNode::One:
+                        return new OneType( noQualifiers );
+                  default:
+                        return new VarArgsType( buildQualifiers( td ) );
+                } // switch
+                        if (td->builtintype == DeclarationNode::Zero) {
+                                return new ZeroType( noQualifiers );
+                        }
+                        else if (td->builtintype == DeclarationNode::One) {
+                                return new OneType( noQualifiers );
+                        }
+                        else {
+                                return new VarArgsType( buildQualifiers( td ) );
+                        }
           case TypeData::GlobalScope:
                 return new GlobalScopeType();
           case TypeData::Qualified:
                 return new QualifiedType( buildQualifiers( td ), typebuild( td->qualified.parent ), typebuild( td->qualified.child ) );
+                        return new GlobalScopeType();
+                case TypeData::Qualified:
+                        return new QualifiedType( buildQualifiers( td ), typebuild( td->qualified.parent ), typebuild( td->qualified.child ) );
           case TypeData::Symbolic:
           case TypeData::Enum:
           case TypeData::Aggregate:
                 assert( false );
+                        assert( false );
         } // switch
 …
         AggregateDecl * at;
         switch ( td->aggregate.kind ) {
+          case AggregateDecl::Struct:
+          case AggregateDecl::Coroutine:
+          case AggregateDecl::Generator:
+          case AggregateDecl::Monitor:
+          case AggregateDecl::Thread:
+          case DeclarationNode::Struct:
+          case DeclarationNode::Coroutine:
+          case DeclarationNode::Monitor:
+          case DeclarationNode::Thread:
                 at = new StructDecl( *td->aggregate.name, td->aggregate.kind, attributes, linkage );
                 buildForall( td->aggregate.params, at->get_parameters() );
                 break;
           case AggregateDecl::Union:
+          case DeclarationNode::Union:
                 at = new UnionDecl( *td->aggregate.name, attributes, linkage );
                 buildForall( td->aggregate.params, at->get_parameters() );
                 break;
           case AggregateDecl::Trait:
+          case DeclarationNode::Trait:
                 at = new TraitDecl( *td->aggregate.name, attributes, linkage );
                 buildList( td->aggregate.params, at->get_parameters() );
 …
                           AggregateDecl * typedecl = buildAggregate( type, attributes, linkage );
                           switch ( type->aggregate.kind ) {
                                 case AggregateDecl::Struct:
                                 case AggregateDecl::Coroutine:
                                 case AggregateDecl::Monitor:
                                 case AggregateDecl::Thread:
+                                case DeclarationNode::Struct:
+                                case DeclarationNode::Coroutine:
+                                case DeclarationNode::Monitor:
+                                case DeclarationNode::Thread:
                                   ret = new StructInstType( buildQualifiers( type ), (StructDecl *)typedecl );
                                   break;
                                 case AggregateDecl::Union:
+                                case DeclarationNode::Union:
                                   ret = new UnionInstType( buildQualifiers( type ), (UnionDecl *)typedecl );
                                   break;
                                 case AggregateDecl::Trait:
+                                case DeclarationNode::Trait:
                                   assert( false );
                                   //ret = new TraitInstType( buildQualifiers( type ), (TraitDecl *)typedecl );
 …
                   } else {
                           switch ( type->aggregate.kind ) {
                                 case AggregateDecl::Struct:
                                 case AggregateDecl::Coroutine:
                                 case AggregateDecl::Monitor:
                                 case AggregateDecl::Thread:
+                                case DeclarationNode::Struct:
+                                case DeclarationNode::Coroutine:
+                                case DeclarationNode::Monitor:
+                                case DeclarationNode::Thread:
                                   ret = new StructInstType( buildQualifiers( type ), *type->aggregate.name );
                                   break;
                                 case AggregateDecl::Union:
+                                case DeclarationNode::Union:
                                   ret = new UnionInstType( buildQualifiers( type ), *type->aggregate.name );
                                   break;
                                 case AggregateDecl::Trait:
+                                case DeclarationNode::Trait:
                                   ret = new TraitInstType( buildQualifiers( type ), *type->aggregate.name );
                                   break;
 …
           case TypeData::Aggregate: {
                   switch ( type->aggregate.kind ) {
                         case AggregateDecl::Struct:
                         case AggregateDecl::Coroutine:
                         case AggregateDecl::Monitor:
                         case AggregateDecl::Thread:
+                        case DeclarationNode::Struct:
+                        case DeclarationNode::Coroutine:
+                        case DeclarationNode::Monitor:
+                        case DeclarationNode::Thread:
                           ret = new StructInstType( buildQualifiers( type ), *type->aggregate.name );
                           break;
                         case AggregateDecl::Union:
+                        case DeclarationNode::Union:
                           ret = new UnionInstType( buildQualifiers( type ), *type->aggregate.name );
                           break;
                         case AggregateDecl::Trait:
+                        case DeclarationNode::Trait:
                           ret = new TraitInstType( buildQualifiers( type ), *type->aggregate.name );
                           break;

src/Parser/TypeData.h

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 15:18:36 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:42:35 2019
 // Update Count     : 199
+// Last Modified On : Thu Nov  1 20:56:46 2018
+// Update Count     : 196
 //
 …
 #include "ParseNode.h"           // for DeclarationNode, DeclarationNode::Ag...
 #include "SynTree/LinkageSpec.h" // for Spec
+#include "Parser/LinkageSpec.h"  // for Spec
 #include "SynTree/Type.h"        // for Type, ReferenceToType (ptr only)
 #include "SynTree/SynTree.h"     // for Visitor Nodes
 …
         struct Aggregate_t {
                 AggregateDecl::Aggregate kind;
+                DeclarationNode::Aggregate kind;
                 const std::string * name = nullptr;
                 DeclarationNode * params = nullptr;

src/Parser/TypedefTable.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 15:20:13 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb 15 08:06:36 2020
 // Update Count     : 259
+// Last Modified On : Wed Jul 25 15:32:35 2018
+// Update Count     : 258
 //
 …
 } // TypedefTable::~TypedefTable
 bool TypedefTable::exists( const string & identifier ) const {
+bool TypedefTable::exists( const string & identifier ) {
         return kindTable.find( identifier ) != kindTable.end();
 } // TypedefTable::exists
 bool TypedefTable::existsCurr( const string & identifier ) const {
+bool TypedefTable::existsCurr( const string & identifier ) {
         return kindTable.findAt( kindTable.currentScope() - 1, identifier ) != kindTable.end();
 } // TypedefTable::exists

src/Parser/TypedefTable.h

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 15:24:36 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb 15 08:06:37 2020
 // Update Count     : 117
+// Last Modified On : Wed Jul 25 15:33:55 2018
+// Update Count     : 114
 //
 …
         ~TypedefTable();
         bool exists( const std::string & identifier ) const;
         bool existsCurr( const std::string & identifier ) const;
+        bool exists( const std::string & identifier );
+        bool existsCurr( const std::string & identifier );
         int isKind( const std::string & identifier ) const;
         void makeTypedef( const std::string & name, int kind = TYPEDEFname );

src/Parser/lex.ll

-              r7030dab
+              r71d6bd8
  * Created On       : Sat Sep 22 08:58:10 2001
  * Last Modified By : Peter A. Buhr
  * Last Modified On : Sat Feb 15 11:05:50 2020
  * Update Count     : 737
+ * Last Modified On : Sun Aug  4 20:53:47 2019
+ * Update Count     : 719
  */
 …
 #include "TypedefTable.h"
-string * build_postfix_name( string * name );
 char *yyfilename;
 string *strtext;                                                                                // accumulate parts of character and string constant value
 …
 #define FLOATXX(v) KEYWORD_RETURN(v);
 #else
 #define FLOATXX(v) IDENTIFIER_RETURN();
+#define FLOATXX(v) IDENTIFIER_RETURN();
 #endif // HAVE_KEYWORDS_FLOATXX
 …
 _Static_assert  { KEYWORD_RETURN(STATICASSERT); }               // C11
 struct                  { KEYWORD_RETURN(STRUCT); }
+suspend                 { KEYWORD_RETURN(SUSPEND); }                    // CFA
+        /* suspend                      { KEYWORD_RETURN(SUSPEND); }                    // CFA */
 switch                  { KEYWORD_RETURN(SWITCH); }
 thread                  { KEYWORD_RETURN(THREAD); }                             // C11
 …
                                 /* identifier */
 {identifier}    { IDENTIFIER_RETURN(); }
 "``"{identifier} {                                                                              // CFA
         yytext[yyleng] = '\0'; yytext += 2;                                     // SKULLDUGGERY: remove backquotes (ok to shorten?)
+"`"{identifier}"`" {                                                                    // CFA
+        yytext[yyleng - 1] = '\0'; yytext += 1;                         // SKULLDUGGERY: remove backquotes (ok to shorten?)
         IDENTIFIER_RETURN();
+}
 …
 "?"({op_unary_pre_post}|"()"|"[?]"|"{}") { IDENTIFIER_RETURN(); }
 "^?{}"                  { IDENTIFIER_RETURN(); }
+"?`"{identifier} {                                                                              // postfix operator
+        yylval.tok.str = new string( &yytext[2] );                      // remove ?`
+        yylval.tok.str = build_postfix_name( yylval.tok.str ); // add prefix
+        RETURN_LOCN( typedefTable.isKind( *yylval.tok.str ) );
+}
+"?`"{identifier} { IDENTIFIER_RETURN(); }                               // postfix operator
 "?"{op_binary_over}"?"  { IDENTIFIER_RETURN(); }                // binary
         /*

src/Parser/module.mk

-              r7030dab
+              r71d6bd8
 ## Created On       : Sat May 16 15:29:09 2015
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Sat Dec 14 07:34:47 2019
 ## Update Count     : 107
+## Last Modified On : Wed Jun 28 21:58:29 2017
+## Update Count     : 104
 ###############################################################################
 …
 AM_YFLAGS = -d -t -v
+SRC += \
+SRC += Parser/parser.yy \
+       Parser/lex.ll \
+       Parser/TypedefTable.cc \
+       Parser/ParseNode.cc \
        Parser/DeclarationNode.cc \
        Parser/ExpressionNode.cc \
+       Parser/StatementNode.cc \
        Parser/InitializerNode.cc \
-       Parser/ParseNode.cc \
-       Parser/StatementNode.cc \
        Parser/TypeData.cc \
+       Parser/TypedefTable.cc \
+       Parser/lex.ll \
+       Parser/parser.yy \
+       Parser/LinkageSpec.cc \
        Parser/parserutility.cc
+SRCDEMANGLE += \
+        Parser/LinkageSpec.cc
 MOSTLYCLEANFILES += Parser/lex.cc Parser/parser.cc Parser/parser.hh Parser/parser.output

src/Parser/parser.yy

-              r7030dab
+              r71d6bd8
 // Created On       : Sat Sep  1 20:22:55 2001
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Mar  6 17:26:45 2020
 // Update Count     : 4474
+// Last Modified On : Sun Aug  4 21:48:23 2019
+// Update Count     : 4364
 //
 …
 using namespace std;
-#include "SynTree/Declaration.h"
 #include "ParseNode.h"
 #include "TypedefTable.h"
 #include "TypeData.h"
 #include "SynTree/LinkageSpec.h"
+#include "LinkageSpec.h"
 #include "Common/SemanticError.h"                                               // error_str
 #include "Common/utility.h"                                                             // for maybeMoveBuild, maybeBuild, CodeLo...
 …
 } // rebindForall
+string * build_postfix_name( string * name ) {
+        *name = string("__postfix_func_") + *name;
+        return name;
+NameExpr * build_postfix_name( const string * name ) {
+        NameExpr * new_name = build_varref( new string( "?`" + *name ) );
+        delete name;
+        return new_name;
 } // build_postfix_name
 …
         } // if
 } // forCtrl
 bool forall = false, yyy = false;                                               // aggregate have one or more forall qualifiers ?
 …
         ExpressionNode * en;
         DeclarationNode * decl;
         AggregateDecl::Aggregate aggKey;
         TypeDecl::Kind tclass;
+        DeclarationNode::Aggregate aggKey;
+        DeclarationNode::TypeClass tclass;
         StatementNode * sn;
         WaitForStmt * wfs;
 …
 %token OTYPE FTYPE DTYPE TTYPE TRAIT                                    // CFA
 %token SIZEOF OFFSETOF
+// %token RESUME                                                                        // CFA
+%token SUSPEND                                                                  // CFA
+// %token SUSPEND RESUME                                                                        // CFA
 %token ATTRIBUTE EXTENSION                                                              // GCC
 %token IF ELSE SWITCH CASE DEFAULT DO WHILE FOR BREAK CONTINUE GOTO RETURN
 …
 %type<op> ptrref_operator                               unary_operator                          assignment_operator
 %type<en> primary_expression                    postfix_expression                      unary_expression
 %type<en> cast_expression_list                  cast_expression                         exponential_expression          multiplicative_expression       additive_expression
+%type<en> cast_expression                               exponential_expression          multiplicative_expression       additive_expression
 %type<en> shift_expression                              relational_expression           equality_expression
 %type<en> AND_expression                                exclusive_OR_expression         inclusive_OR_expression
 …
 %type<decl> abstract_parameter_declaration
 %type<aggKey> aggregate_key aggregate_data aggregate_control
+%type<aggKey> aggregate_key
 %type<decl> aggregate_type aggregate_type_nobody
 …
         | '(' compound_statement ')'                                            // GCC, lambda expression
                 { $$ = new ExpressionNode( new StmtExpr( dynamic_cast< CompoundStmt * >(maybeMoveBuild< Statement >($2) ) ) ); }
+        | constant '`' IDENTIFIER                                                       // CFA, postfix call
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $3 ) ), $1 ) ); }
+        | string_literal '`' IDENTIFIER                                         // CFA, postfix call
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $3 ) ), new ExpressionNode( $1 ) ) ); }
+        | IDENTIFIER '`' IDENTIFIER                                                     // CFA, postfix call
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $3 ) ), new ExpressionNode( build_varref( $1 ) ) ) ); }
+        | tuple '`' IDENTIFIER                                                          // CFA, postfix call
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $3 ) ), $1 ) ); }
+        | '(' comma_expression ')' '`' IDENTIFIER                       // CFA, postfix call
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $5 ) ), $2 ) ); }
         | type_name '.' identifier                                                      // CFA, nested type
                 { SemanticError( yylloc, "Qualified name is currently unimplemented." ); $$ = nullptr; }
 …
         | postfix_expression '(' argument_expression_list ')'
                 { $$ = new ExpressionNode( build_func( $1, $3 ) ); }
-        | postfix_expression '`' identifier                                     // CFA, postfix call
-                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_varref( build_postfix_name( $3 ) ) ), $1 ) ); }
-        | constant '`' identifier                                                       // CFA, postfix call
-                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_varref( build_postfix_name( $3 ) ) ), $1 ) ); }
-        | string_literal '`' identifier                                         // CFA, postfix call
-                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_varref( build_postfix_name( $3 ) ) ), new ExpressionNode( $1 ) ) ); }
         | postfix_expression '.' identifier
                 { $$ = new ExpressionNode( build_fieldSel( $1, build_varref( $3 ) ) ); }
 …
         | postfix_expression '.' '[' field_name_list ']'        // CFA, tuple field selector
                 { $$ = new ExpressionNode( build_fieldSel( $1, build_tuple( $4 ) ) ); }
-        | postfix_expression '.' aggregate_control
-                { $$ = new ExpressionNode( build_keyword_cast( $3, $1 ) ); }
         | postfix_expression ARROW identifier
                 { $$ = new ExpressionNode( build_pfieldSel( $1, build_varref( $3 ) ) ); }
 …
         | '(' type_no_function ')' '@' '{' initializer_list_opt comma_opt '}' // CFA, explicit C compound-literal
                 { $$ = new ExpressionNode( build_compoundLiteral( $2, (new InitializerNode( $6, true ))->set_maybeConstructed( false ) ) ); }
         | '^' primary_expression '{' argument_expression_list '}' // CFA, destructor call
+        | '^' primary_expression '{' argument_expression_list '}' // CFA
+                {
                         Token fn;
 …
         | argument_expression
         | argument_expression_list ',' argument_expression
                 { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+                { $$ = (ExpressionNode *)( $1->set_last( $3 )); }
+        ;
 …
 field_name_list:                                                                                // CFA, tuple field selector
         field
         | field_name_list ',' field                                     { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+        | field_name_list ',' field                                     { $$ = (ExpressionNode *)$1->set_last( $3 ); }
+        ;
 …
         | '(' type_no_function ')' cast_expression
                 { $$ = new ExpressionNode( build_cast( $2, $4 ) ); }
+        | '(' aggregate_control '&' ')' cast_expression         // CFA
+                { $$ = new ExpressionNode( build_keyword_cast( $2, $5 ) ); }
+                // keyword cast cannot be grouped because of reduction in aggregate_key
+        | '(' GENERATOR '&' ')' cast_expression                         // CFA
+                { $$ = new ExpressionNode( build_keyword_cast( KeywordCastExpr::Coroutine, $5 ) ); }
+        | '(' COROUTINE '&' ')' cast_expression                         // CFA
+                { $$ = new ExpressionNode( build_keyword_cast( KeywordCastExpr::Coroutine, $5 ) ); }
+        | '(' THREAD '&' ')' cast_expression                            // CFA
+                { $$ = new ExpressionNode( build_keyword_cast( KeywordCastExpr::Thread, $5 ) ); }
+        | '(' MONITOR '&' ')' cast_expression                           // CFA
+                { $$ = new ExpressionNode( build_keyword_cast( KeywordCastExpr::Monitor, $5 ) ); }
                 // VIRTUAL cannot be opt because of look ahead issues
         | '(' VIRTUAL ')' cast_expression                                       // CFA
 …
         conditional_expression
         | unary_expression assignment_operator assignment_expression
+                {
+                        if ( $2 == OperKinds::AtAssn ) {
+                                SemanticError( yylloc, "C @= assignment is currently unimplemented." ); $$ = nullptr;
+                        } else {
+                                $$ = new ExpressionNode( build_binary_val( $2, $1, $3 ) );
+                        } // if
+                }
+                { $$ = new ExpressionNode( build_binary_val( $2, $1, $3 ) ); }
         | unary_expression '=' '{' initializer_list_opt comma_opt '}'
                 { SemanticError( yylloc, "Initializer assignment is currently unimplemented." ); $$ = nullptr; }
 …
                 { $$ = new ExpressionNode( build_tuple( (ExpressionNode *)(new ExpressionNode( nullptr ) )->set_last( $3 ) ) ); }
         | '[' push assignment_expression pop ',' tuple_expression_list ']'
                 { $$ = new ExpressionNode( build_tuple( (ExpressionNode *)($3->set_last( $6 ) ) )); }
+                { $$ = new ExpressionNode( build_tuple( (ExpressionNode *)$3->set_last( $6 ) ) ); }
+        ;
 …
         assignment_expression_opt
         | tuple_expression_list ',' assignment_expression_opt
                 { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+                { $$ = (ExpressionNode *)$1->set_last( $3 ); }
+        ;
 …
                 { $$ = forCtrl( $1, new string( DeclarationNode::anonymous.newName() ), new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
                                                 OperKinds::LThan, $1->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
-        | '=' comma_expression                                                                  // CFA
-                { $$ = forCtrl( $2, new string( DeclarationNode::anonymous.newName() ), new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
-                                                OperKinds::LEThan, $2->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
         | comma_expression inclexcl comma_expression            // CFA
                 { $$ = forCtrl( $1, new string( DeclarationNode::anonymous.newName() ), $1->clone(), $2, $3, new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
 …
                 { $$ = forCtrl( $3, $1, new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
                                                 OperKinds::LThan, $3->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
-        | comma_expression ';' '=' comma_expression                             // CFA
-                { $$ = forCtrl( $4, $1, new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
-                                                OperKinds::LEThan, $4->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
         | comma_expression ';' comma_expression inclexcl comma_expression // CFA
                 { $$ = forCtrl( $3, $1, $3->clone(), $4, $5, new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
 …
         | RETURN '{' initializer_list_opt comma_opt '}' ';'
                 { SemanticError( yylloc, "Initializer return is currently unimplemented." ); $$ = nullptr; }
+        | SUSPEND ';'
+                { $$ = new StatementNode( build_suspend( nullptr ) ); }
+        | SUSPEND compound_statement
+                { $$ = new StatementNode( build_suspend( $2 ) ); }
+        | SUSPEND COROUTINE ';'
+                { $$ = new StatementNode( build_suspend( nullptr, SuspendStmt::Coroutine ) ); }
+        | SUSPEND COROUTINE compound_statement
+                { $$ = new StatementNode( build_suspend( $3, SuspendStmt::Coroutine ) ); }
+        | SUSPEND GENERATOR ';'
+                { $$ = new StatementNode( build_suspend( nullptr, SuspendStmt::Generator ) ); }
+        | SUSPEND GENERATOR compound_statement
+                { $$ = new StatementNode( build_suspend( $3, SuspendStmt::Generator ) ); }
+        // | SUSPEND ';'
+        //      { SemanticError( yylloc, "Suspend expression is currently unimplemented." ); $$ = nullptr; }
+        // | SUSPEND compound_statement ';'
+        //      { SemanticError( yylloc, "Suspend expression is currently unimplemented." ); $$ = nullptr; }
         | THROW assignment_expression_opt ';'                           // handles rethrow
                 { $$ = new StatementNode( build_throw( $2 ) ); }
 …
         WAITFOR '(' cast_expression ')'
                 { $$ = $3; }
+//      | WAITFOR '(' cast_expression ',' argument_expression_list ')'
+//              { $$ = (ExpressionNode *)$3->set_last( $5 ); }
+        | WAITFOR '(' cast_expression_list ':' argument_expression_list ')'
+                { $$ = (ExpressionNode *)($3->set_last( $5 )); }
+        ;
+cast_expression_list:
+        cast_expression
+        | cast_expression_list ',' cast_expression
+                { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+        | WAITFOR '(' cast_expression ',' argument_expression_list ')'
+                { $$ = (ExpressionNode *)$3->set_last( $5 ); }
+        ;
 …
         asm_operand
         | asm_operands_list ',' asm_operand
                 { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+                { $$ = (ExpressionNode *)$1->set_last( $3 ); }
+        ;
 …
                 { $$ = new ExpressionNode( $1 ); }
         | asm_clobbers_list_opt ',' string_literal
+                { $$ = (ExpressionNode *)($1->set_last( new ExpressionNode( $3 ) )); }
+                // set_last returns ParseNode *
+                { $$ = (ExpressionNode *)$1->set_last( new ExpressionNode( $3 ) ); }
+        ;
 …
                 // type_specifier can resolve to just TYPEDEFname (e.g., typedef int T; int f( T );). Therefore this must be
                 // flattened to allow lookahead to the '(' without having to reduce identifier_or_type_name.
         cfa_abstract_tuple identifier_or_type_name '(' push cfa_parameter_ellipsis_list_opt pop ')' attribute_list_opt
+        cfa_abstract_tuple identifier_or_type_name '(' push cfa_parameter_ellipsis_list_opt pop ')'
                 // To obtain LR(1 ), this rule must be factored out from function return type (see cfa_abstract_declarator).
                 { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 )->addQualifiers( $8 ); }
         | cfa_function_return identifier_or_type_name '(' push cfa_parameter_ellipsis_list_opt pop ')' attribute_list_opt
                 { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 )->addQualifiers( $8 ); }
+                { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 ); }
+        | cfa_function_return identifier_or_type_name '(' push cfa_parameter_ellipsis_list_opt pop ')'
+                { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 ); }
+        ;
 …
 aggregate_key:
-        aggregate_data
-        | aggregate_control
+        ;
-aggregate_data:
         STRUCT
                 { yyy = true; $$ = AggregateDecl::Struct; }
+                { yyy = true; $$ = DeclarationNode::Struct; }
         | UNION
+                { yyy = true; $$ = AggregateDecl::Union; }
+        | EXCEPTION                                                                                     // CFA
+                { yyy = true; $$ = AggregateDecl::Exception; }
+        ;
+aggregate_control:                                                                              // CFA
+        GENERATOR
+                { yyy = true; $$ = AggregateDecl::Generator; }
+        | MONITOR GENERATOR
+                { SemanticError( yylloc, "monitor generator is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
+                { yyy = true; $$ = DeclarationNode::Union; }
+        | EXCEPTION
+                { yyy = true; $$ = DeclarationNode::Exception; }
+        | GENERATOR
+                { yyy = true; $$ = DeclarationNode::Coroutine; }
         | COROUTINE
                 { yyy = true; $$ = AggregateDecl::Coroutine; }
+                { yyy = true; $$ = DeclarationNode::Coroutine; }
         | MONITOR
+                { yyy = true; $$ = AggregateDecl::Monitor; }
+        | MONITOR COROUTINE
+                { SemanticError( yylloc, "monitor coroutine is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
+                { yyy = true; $$ = DeclarationNode::Monitor; }
         | THREAD
+                { yyy = true; $$ = AggregateDecl::Thread; }
+        | MONITOR THREAD
+                { SemanticError( yylloc, "monitor thread is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
+                { yyy = true; $$ = DeclarationNode::Thread; }
+        ;
 …
                         distInl( $3 );
+                }
-        | INLINE aggregate_control ';'                                          // CFA
-                { SemanticError( yylloc, "INLINE aggregate control currently unimplemented." ); $$ = nullptr; }
         | typedef_declaration ';'                                                       // CFA
         | cfa_field_declaring_list ';'                                          // CFA, new style field declaration
 …
         | initializer_list_opt ',' initializer          { $$ = (InitializerNode *)( $1->set_last( $3 ) ); }
         | initializer_list_opt ',' designation initializer
                 { $$ = (InitializerNode *)($1->set_last( $4->set_designators( $3 ) )); }
+                { $$ = (InitializerNode *)( $1->set_last( $4->set_designators( $3 ) ) ); }
+        ;
 …
         designator
         | designator_list designator
                 { $$ = (ExpressionNode *)($1->set_last( $2 )); }
+                { $$ = (ExpressionNode *)( $1->set_last( $2 ) ); }
         //| designator_list designator                                          { $$ = new ExpressionNode( $1, $2 ); }
+        ;
 …
         | type_specifier identifier_parameter_declarator
         | assertion_list
                 { $$ = DeclarationNode::newTypeParam( TypeDecl::Dtype, new string( DeclarationNode::anonymous.newName() ) )->addAssertions( $1 ); }
+                { $$ = DeclarationNode::newTypeParam( DeclarationNode::Dtype, new string( DeclarationNode::anonymous.newName() ) )->addAssertions( $1 ); }
+        ;
 type_class:                                                                                             // CFA
         OTYPE
                 { $$ = TypeDecl::Otype; }
+                { $$ = DeclarationNode::Otype; }
         | DTYPE
                 { $$ = TypeDecl::Dtype; }
+                { $$ = DeclarationNode::Dtype; }
         | FTYPE
                 { $$ = TypeDecl::Ftype; }
+                { $$ = DeclarationNode::Ftype; }
         | TTYPE
                 { $$ = TypeDecl::Ttype; }
+                { $$ = DeclarationNode::Ttype; }
+        ;
 …
                 { SemanticError( yylloc, toString("Expression generic parameters are currently unimplemented: ", $1->build()) ); $$ = nullptr; }
         | type_list ',' type
                 { $$ = (ExpressionNode *)($1->set_last( new ExpressionNode( new TypeExpr( maybeMoveBuildType( $3 ) ) ) )); }
+                { $$ = (ExpressionNode *)( $1->set_last( new ExpressionNode( new TypeExpr( maybeMoveBuildType( $3 ) ) ) ) ); }
         | type_list ',' assignment_expression
                 { SemanticError( yylloc, toString("Expression generic parameters are currently unimplemented: ", $3->build()) ); $$ = nullptr; }
 …
+                {
                         linkageStack.push( linkage );                           // handle nested extern "C"/"Cforall"
                         linkage = LinkageSpec::update( yylloc, linkage, $2 );
+                        linkage = LinkageSpec::linkageUpdate( yylloc, linkage, $2 );
+                }
           '{' up external_definition_list_opt down '}'

src/ResolvExpr/AdjustExprType.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sat May 16 23:41:42 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 21:43:56 2019
 // Update Count     : 6
+// Last Modified On : Wed Mar  2 17:34:53 2016
+// Update Count     : 4
 //
 …
                         // replace known function-type-variables with pointer-to-function
                         if ( const ast::EqvClass * eqvClass = tenv.lookup( inst->name ) ) {
                                 if ( eqvClass->data.kind == ast::TypeDecl::Ftype ) {
+                                if ( eqvClass->data.kind == ast::TypeVar::Ftype ) {
                                         return new ast::PointerType{ inst };
+                                }
                         } else if ( const ast::NamedTypeDecl * ntDecl = symtab.lookupType( inst->name ) ) {
                                 if ( auto tyDecl = dynamic_cast< const ast::TypeDecl * >( ntDecl ) ) {
                                         if ( tyDecl->kind == ast::TypeDecl::Ftype ) {
+                                        if ( tyDecl->kind == ast::TypeVar::Ftype ) {
                                                 return new ast::PointerType{ inst };
+                                        }

src/ResolvExpr/AlternativeFinder.cc

-              r7030dab
+              r71d6bd8
                 void postvisit( CastExpr * castExpr );
                 void postvisit( VirtualCastExpr * castExpr );
-                void postvisit( KeywordCastExpr * castExpr );
                 void postvisit( UntypedMemberExpr * memberExpr );
                 void postvisit( MemberExpr * memberExpr );
 …
+        }
-        void AlternativeFinder::Finder::postvisit( KeywordCastExpr * castExpr ) {
-                assertf( castExpr->get_result(), "Cast target should have been set in Validate." );
-                auto ref = dynamic_cast<ReferenceType*>(castExpr->get_result());
-                assert(ref);
-                auto inst = dynamic_cast<StructInstType*>(ref->base);
-                assert(inst);
-                auto target = inst->baseStruct;
-                AlternativeFinder finder( indexer, env );
-                auto pick_alternatives = [target, this](AltList & found, bool expect_ref) {
-                        for(auto & alt : found) {
-                                Type * expr = alt.expr->get_result();
-                                if(expect_ref) {
-                                        auto res = dynamic_cast<ReferenceType*>(expr);
-                                        if(!res) { continue; }
-                                        expr = res->base;
+                                }
-                                if(auto insttype = dynamic_cast<TypeInstType*>(expr)) {
-                                        auto td = alt.env.lookup(insttype->name);
-                                        if(!td) { continue; }
-                                        expr = td->type;
+                                }
-                                if(auto base = dynamic_cast<StructInstType*>(expr)) {
-                                        if(base->baseStruct == target) {
-                                                alternatives.push_back(
-                                                        std::move(alt)
-                                                );
+                                        }
+                                }
+                        }
-                };
-                try {
-                        // Attempt 1 : turn (thread&)X into ($thread&)X.__thrd
-                        // Clone is purely for memory management
-                        std::unique_ptr<Expression> tech1 { new UntypedMemberExpr(new NameExpr(castExpr->concrete_target.field), castExpr->arg->clone()) };
-                        // don't prune here, since it's guaranteed all alternatives will have the same type
-                        finder.findWithoutPrune( tech1.get() );
-                        pick_alternatives(finder.alternatives, false);
-                        return;
-                } catch(SemanticErrorException & ) {}
-                // Fallback : turn (thread&)X into ($thread&)get_thread(X)
-                std::unique_ptr<Expression> fallback { UntypedExpr::createDeref( new UntypedExpr(new NameExpr(castExpr->concrete_target.getter), { castExpr->arg->clone() })) };
-                // don't prune here, since it's guaranteed all alternatives will have the same type
-                finder.findWithoutPrune( fallback.get() );
-                pick_alternatives(finder.alternatives, true);
-                // Whatever happens here, we have no more fallbacks
+        }
         namespace {
                 /// Gets name from untyped member expression (member must be NameExpr)

src/ResolvExpr/PtrsCastable.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sun May 17 11:48:00 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 21:48:33 2019
 // Update Count     : 9
+// Last Modified On : Wed Mar  2 17:36:18 2016
+// Update Count     : 8
 //
 …
                         if ( const ast::NamedTypeDecl * named = symtab.lookupType( inst->name ) ) {
                                 if ( auto tyDecl = dynamic_cast< const ast::TypeDecl * >( named ) ) {
                                         if ( tyDecl->kind == ast::TypeDecl::Ftype ) {
+                                        if ( tyDecl->kind == ast::TypeVar::Ftype ) {
                                                 return -1;
+                                        }
+                                }
                         } else if ( const ast::EqvClass * eqvClass = env.lookup( inst->name ) ) {
                                 if ( eqvClass->data.kind == ast::TypeDecl::Ftype ) {
+                                if ( eqvClass->data.kind == ast::TypeVar::Ftype ) {
                                         return -1;
+                                }

src/ResolvExpr/Resolver.cc

-              r7030dab
+              r71d6bd8
 // Author           : Aaron B. Moss
 // Created On       : Sun May 17 12:17:01 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Fri Mar 27 11:58:00 2020
 // Update Count     : 242
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Wed May 29 11:00:00 2019
+// Update Count     : 241
 //
 …
                 void previsit( ThrowStmt * throwStmt );
                 void previsit( CatchStmt * catchStmt );
-                void postvisit( CatchStmt * catchStmt );
                 void previsit( WaitForStmt * stmt );
 …
                 // TODO: Replace *exception type with &exception type.
                 if ( throwStmt->get_expr() ) {
                         const StructDecl * exception_decl = indexer.lookupStruct( "__cfaehm_base_exception_t" );
+                        const StructDecl * exception_decl = indexer.lookupStruct( "__cfaabi_ehm__base_exception_t" );
                         assert( exception_decl );
                         Type * exceptType = new PointerType( noQualifiers, new StructInstType( noQualifiers, const_cast<StructDecl *>(exception_decl) ) );
 …
         void Resolver_old::previsit( CatchStmt * catchStmt ) {
-                // Until we are very sure this invarent (ifs that move between passes have thenPart)
-                // holds, check it. This allows a check for when to decode the mangling.
-                if ( IfStmt * ifStmt = dynamic_cast<IfStmt *>( catchStmt->body ) ) {
-                        assert( ifStmt->thenPart );
+                }
-                // Encode the catchStmt so the condition can see the declaration.
                 if ( catchStmt->cond ) {
+                        IfStmt * ifStmt = new IfStmt( catchStmt->cond, nullptr, catchStmt->body );
+                        catchStmt->cond = nullptr;
+                        catchStmt->body = ifStmt;
+                }
+        }
+        void Resolver_old::postvisit( CatchStmt * catchStmt ) {
+                // Decode the catchStmt so everything is stored properly.
+                IfStmt * ifStmt = dynamic_cast<IfStmt *>( catchStmt->body );
+                if ( nullptr != ifStmt && nullptr == ifStmt->thenPart ) {
+                        assert( ifStmt->condition );
+                        assert( ifStmt->elsePart );
+                        catchStmt->cond = ifStmt->condition;
+                        catchStmt->body = ifStmt->elsePart;
+                        ifStmt->condition = nullptr;
+                        ifStmt->elsePart = nullptr;
+                        delete ifStmt;
+                        findSingleExpression( catchStmt->cond, new BasicType( noQualifiers, BasicType::Bool ), indexer );
+                }
+        }
 …
                 if ( throwStmt->expr ) {
                         const ast::StructDecl * exceptionDecl =
                                 symtab.lookupStruct( "__cfaehm_base_exception_t" );
+                                symtab.lookupStruct( "__cfaabi_ehm__base_exception_t" );
                         assert( exceptionDecl );
                         ast::ptr< ast::Type > exceptType =
 …
         const ast::CatchStmt * Resolver_new::previsit( const ast::CatchStmt * catchStmt ) {
-                // TODO: This will need a fix for the decl/cond scoping problem.
                 if ( catchStmt->cond ) {
                         ast::ptr< ast::Type > boolType = new ast::BasicType{ ast::BasicType::Bool };

src/ResolvExpr/Unify.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Sun May 17 12:27:10 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:43:05 2019
 // Update Count     : 46
+// Last Modified By : Andrew Beach
+// Last Modified On : Wed Sep  4 10:00:00 2019
+// Update Count     : 44
 //
 …
 #include "Common/PassVisitor.h"     // for PassVisitor
 #include "FindOpenVars.h"           // for findOpenVars
 #include "SynTree/LinkageSpec.h"    // for C
+#include "Parser/LinkageSpec.h"     // for C
 #include "SynTree/Constant.h"       // for Constant
 #include "SynTree/Declaration.h"    // for TypeDecl, TypeDecl::Data, Declarati...
 …
                                 if ( const ast::EqvClass * clz = tenv.lookup( typeInst->name ) ) {
                                         // expand ttype parameter into its actual type
                                         if ( clz->data.kind == ast::TypeDecl::Ttype && clz->bound ) {
+                                        if ( clz->data.kind == ast::TypeVar::Ttype && clz->bound ) {
                                                 return clz->bound;
+                                        }

src/SymTab/Autogen.h

-              r7030dab
+              r71d6bd8
 // Created On       : Sun May 17 21:53:34 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 16:38:06 2019
 // Update Count     : 16
+// Last Modified On : Sat Jul 22 09:50:25 2017
+// Update Count     : 15
 //
 …
 #include "SynTree/Expression.h"   // for NameExpr, ConstantExpr, UntypedExpr...
 #include "SynTree/Type.h"         // for Type, ArrayType, Type::Qualifiers
-#include "SynTree/Statement.h"    // for CompoundStmt, DeclStmt, ExprStmt
 class CompoundStmt;

src/SymTab/Demangle.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Thu Jul 19 12:52:41 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb 11 15:09:18 2020
 // Update Count     : 10
+// Last Modified On : Tue Jul 30 13:46:33 2019
+// Update Count     : 3
 //
 …
 #include "CodeGen/GenType.h"
 #include "Common/PassVisitor.h"
-#include "Common/utility.h"                                                             // isPrefix
 #include "Mangler.h"
 #include "SynTree/Type.h"
 …
                                 // type variable types
                                 for (size_t k = 0; k < TypeDecl::NUMBER_OF_KINDS; ++k) {
                                         static const std::string typeVariableNames[] = { "DT", "OT", "FT", "TT", };
+                                        static const std::string typeVariableNames[] = { "DT", "FT", "TT", };
                                         static_assert(
                                                 sizeof(typeVariableNames)/sizeof(typeVariableNames[0]) == TypeDecl::NUMBER_OF_KINDS,
 …
                         bool StringView::isPrefix(const std::string & pref) {
+                                // if ( pref.size() > str.size()-idx ) return false;
+                                // auto its = std::mismatch( pref.begin(), pref.end(), std::next(str.begin(), idx) );
+                                // if (its.first == pref.end()) {
+                                //      idx += pref.size();
+                                //      return true;
+                                // }
+                                // This update is untested because there are no tests for this code.
+                                if ( ::isPrefix( str, pref, idx ) ) {
+                                if ( pref.size() > str.size()-idx ) return false;
+                                auto its = std::mismatch( pref.begin(), pref.end(), std::next(str.begin(), idx) );
+                                if (its.first == pref.end()) {
                                         idx += pref.size();
                                         return true;
 …
                                 PRINT( std::cerr << "====== " << str.size() << " " << str << std::endl; )
                                 if (str.size() < 2+Encoding::manglePrefix.size()) return false; // +2 for at least _1 suffix
                                 if ( ! isPrefix(Encoding::manglePrefix) || ! isdigit(str.back() ) ) return false;
+                                if (! isPrefix(Encoding::manglePrefix) || ! isdigit(str.back())) return false;
                                 // get name

src/SymTab/Indexer.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Sun May 17 21:37:33 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:43:19 2019
 // Update Count     : 22
+// Last Modified By : Aaron B. Moss
+// Last Modified On : Fri Mar  8 13:55:00 2019
+// Update Count     : 21
 //
 …
 #include "InitTweak/InitTweak.h"   // for isConstructor, isCopyFunction, isC...
 #include "Mangler.h"               // for Mangler
+#include "Parser/LinkageSpec.h"    // for isMangled, isOverridable, Spec
 #include "ResolvExpr/typeops.h"    // for typesCompatible
-#include "SynTree/LinkageSpec.h"   // for isMangled, isOverridable, Spec
 #include "SynTree/Constant.h"      // for Constant
 #include "SynTree/Declaration.h"   // for DeclarationWithType, FunctionDecl

src/SymTab/Mangler.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sun May 17 21:40:29 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb 15 13:55:12 2020
 // Update Count     : 33
+// Last Modified On : Tue Jul 30 13:46:10 2019
+// Update Count     : 26
 //
 #include "Mangler.h"
 …
 #include "Common/SemanticError.h"        // for SemanticError
 #include "Common/utility.h"              // for toString
+#include "Parser/LinkageSpec.h"          // for Spec, isOverridable, AutoGen, Int...
 #include "ResolvExpr/TypeEnvironment.h"  // for TypeEnvironment
-#include "SynTree/LinkageSpec.h"         // for Spec, isOverridable, AutoGen, Int...
 #include "SynTree/Declaration.h"         // for TypeDecl, DeclarationWithType
 #include "SynTree/Expression.h"          // for TypeExpr, Expression, operator<<
 …
                                 } // if
                                 mangleName << Encoding::manglePrefix;
                                 const CodeGen::OperatorInfo * opInfo = CodeGen::operatorLookup( declaration->get_name() );
                                 if ( opInfo ) {
                                         mangleName << opInfo->outputName.size() << opInfo->outputName;
+                                CodeGen::OperatorInfo opInfo;
+                                if ( operatorLookup( declaration->get_name(), opInfo ) ) {
+                                        mangleName << opInfo.outputName.size() << opInfo.outputName;
                                 } else {
                                         mangleName << declaration->name.size() << declaration->name;
 …
                         } // if
                         mangleName << Encoding::manglePrefix;
                         const CodeGen::OperatorInfo * opInfo = CodeGen::operatorLookup( decl->name );
                         if ( opInfo ) {
                                 mangleName << opInfo->outputName.size() << opInfo->outputName;
+                        CodeGen::OperatorInfo opInfo;
+                        if ( operatorLookup( decl->name, opInfo ) ) {
+                                mangleName << opInfo.outputName.size() << opInfo.outputName;
                         } else {
                                 mangleName << decl->name.size() << decl->name;
 …
                         // aside from the assert false.
                         assertf(false, "Mangler_new should not visit typedecl: %s", toCString(decl));
                         assertf( decl->kind < ast::TypeDecl::Kind::NUMBER_OF_KINDS, "Unhandled type variable kind: %d", decl->kind );
+                        assertf( decl->kind < ast::TypeVar::Kind::NUMBER_OF_KINDS, "Unhandled type variable kind: %d", decl->kind );
                         mangleName << Encoding::typeVariables[ decl->kind ] << ( decl->name.length() ) << decl->name;
+                }
 …
                                         for ( const ast::TypeDecl * decl : ptype->forall ) {
                                                 switch ( decl->kind ) {
                                                 case ast::TypeDecl::Kind::Dtype:
+                                                case ast::TypeVar::Kind::Dtype:
                                                         dcount++;
                                                         break;
                                                 case ast::TypeDecl::Kind::Ftype:
+                                                case ast::TypeVar::Kind::Ftype:
                                                         fcount++;
                                                         break;
                                                 case ast::TypeDecl::Kind::Ttype:
+                                                case ast::TypeVar::Kind::Ttype:
                                                         vcount++;
                                                         break;

src/SymTab/ManglerCommon.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Sun May 17 21:44:03 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 14:54:38 2019
 // Update Count     : 28
+// Last Modified On : Thu Feb 14 17:06:37 2019
+// Update Count     : 26
 //
 …
                         const std::string typeVariables[] = {
                                 "BD", // dtype
-                                "BO", // otype
                                 "BF", // ftype
                                 "BT", // ttype

src/SymTab/Validate.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Sun May 17 21:50:04 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:43:34 2019
 // Update Count     : 363
+// Last Modified By : Andrew Beach
+// Last Modified On : Wed Aug  7 6:42:00 2019
+// Update Count     : 360
 //
 …
 #include "InitTweak/GenInit.h"         // for fixReturnStatements
 #include "InitTweak/InitTweak.h"       // for isCtorDtorAssign
+#include "Parser/LinkageSpec.h"        // for C
 #include "ResolvExpr/typeops.h"        // for typesCompatible
 #include "ResolvExpr/Resolver.h"       // for findSingleExpression
 #include "ResolvExpr/ResolveTypeof.h"  // for resolveTypeof
 #include "SymTab/Autogen.h"            // for SizeType
-#include "SynTree/LinkageSpec.h"       // for C
 #include "SynTree/Attribute.h"         // for noAttributes, Attribute
 #include "SynTree/Constant.h"          // for Constant
 …
                         Stats::Heap::newPass("validate-A");
                         Stats::Time::BlockGuard guard("validate-A");
-                        VerifyCtorDtorAssign::verify( translationUnit );  // must happen before autogen, because autogen examines existing ctor/dtors
                         acceptAll( translationUnit, hoistDecls );
                         ReplaceTypedef::replaceTypedef( translationUnit );
 …
                         Stats::Time::BlockGuard guard("validate-C");
                         acceptAll( translationUnit, genericParams );  // check as early as possible - can't happen before LinkReferenceToTypes_old
+                        VerifyCtorDtorAssign::verify( translationUnit );  // must happen before autogen, because autogen examines existing ctor/dtors
                         ReturnChecker::checkFunctionReturns( translationUnit );
                         InitTweak::fixReturnStatements( translationUnit ); // must happen before autogen
 …
                         Stats::Heap::newPass("validate-F");
                         Stats::Time::BlockGuard guard("validate-F");
+                        Stats::Time::TimeCall("Fix Object Type",
+                                FixObjectType::fix, translationUnit);
+                        Stats::Time::TimeCall("Array Length",
+                                ArrayLength::computeLength, translationUnit);
+                        Stats::Time::TimeCall("Find Special Declarations",
+                                Validate::findSpecialDecls, translationUnit);
+                        Stats::Time::TimeCall("Fix Label Address",
+                                mutateAll<LabelAddressFixer>, translationUnit, labelAddrFixer);
+                        Stats::Time::TimeCall("Handle Attributes",
+                                Validate::handleAttributes, translationUnit);
+                        Stats::Time::TimeBlock("Fix Object Type", [&]() {
+                                FixObjectType::fix( translationUnit );
+                        });
+                        Stats::Time::TimeBlock("Array Length", [&]() {
+                                ArrayLength::computeLength( translationUnit );
+                        });
+                        Stats::Time::TimeBlock("Find Special Declarations", [&]() {
+                                Validate::findSpecialDecls( translationUnit );
+                        });
+                        Stats::Time::TimeBlock("Fix Label Address", [&]() {
+                                mutateAll( translationUnit, labelAddrFixer );
+                        });
+                        Stats::Time::TimeBlock("Handle Attributes", [&]() {
+                                Validate::handleAttributes( translationUnit );
+                        });
+                }
+        }
 …
                 Type * designatorType = tyDecl->base->stripDeclarator();
                 if ( StructInstType * aggDecl = dynamic_cast< StructInstType * >( designatorType ) ) {
                         declsToAddBefore.push_back( new StructDecl( aggDecl->name, AggregateDecl::Struct, noAttributes, tyDecl->linkage ) );
+                        declsToAddBefore.push_back( new StructDecl( aggDecl->name, DeclarationNode::Struct, noAttributes, tyDecl->linkage ) );
                 } else if ( UnionInstType * aggDecl = dynamic_cast< UnionInstType * >( designatorType ) ) {
                         declsToAddBefore.push_back( new UnionDecl( aggDecl->name, noAttributes, tyDecl->linkage ) );
 …
                 if ( CodeGen::isCtorDtorAssign( funcDecl->get_name() ) ) { // TODO: also check /=, etc.
                         if ( params.size() == 0 ) {
                                 SemanticError( funcDecl->location, "Constructors, destructors, and assignment functions require at least one parameter." );
+                                SemanticError( funcDecl, "Constructors, destructors, and assignment functions require at least one parameter " );
+                        }
                         ReferenceType * refType = dynamic_cast< ReferenceType * >( params.front()->get_type() );
                         if ( ! refType ) {
                                 SemanticError( funcDecl->location, "First parameter of a constructor, destructor, or assignment function must be a reference." );
+                                SemanticError( funcDecl, "First parameter of a constructor, destructor, or assignment function must be a reference " );
+                        }
                         if ( CodeGen::isCtorDtor( funcDecl->get_name() ) && returnVals.size() != 0 ) {
+                                if(!returnVals.front()->get_type()->isVoid()) {
+                                        SemanticError( funcDecl->location, "Constructors and destructors cannot have explicit return values." );
+                                }
+                                SemanticError( funcDecl, "Constructors and destructors cannot have explicit return values " );
+                        }
+                }

src/SynTree/AggregateDecl.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Sun May 17 23:56:39 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Dec 16 15:07:20 2019
 // Update Count     : 31
+// Last Modified By : Andrew Beach
+// Last Modified On : Fri Aug  4 14:22:00 2017
+// Update Count     : 22
 //
 …
 #include "Common/utility.h"      // for printAll, cloneAll, deleteAll
 #include "Declaration.h"         // for AggregateDecl, TypeDecl, Declaration
+#include "Initializer.h"
+#include "LinkageSpec.h"         // for Spec, linkageName, Cforall
+#include "Parser/LinkageSpec.h"  // for Spec, linkageName, Cforall
 #include "Type.h"                // for Type, Type::StorageClasses
-// These must harmonize with the corresponding AggregateDecl::Aggregate enumerations.
-static const char * aggregateNames[] = { "struct", "union", "enum", "exception", "trait", "generator", "coroutine", "monitor", "thread", "NoAggregateName" };
-const char * AggregateDecl::aggrString( AggregateDecl::Aggregate aggr ) {
-        return aggregateNames[aggr];
+}
 AggregateDecl::AggregateDecl( const std::string &name, const std::list< Attribute * > & attributes, LinkageSpec::Spec linkage ) : Parent( name, Type::StorageClasses(), linkage ), body( false ), attributes( attributes ) {
 …
         os << typeString() << " " << name << ":";
         if ( get_linkage() != LinkageSpec::Cforall ) {
                 os << " " << LinkageSpec::name( linkage );
+                os << " " << LinkageSpec::linkageName( linkage );
         } // if
         os << " with body " << has_body();
 …
+}
 const char * StructDecl::typeString() const { return aggrString( kind ); }
+std::string StructDecl::typeString() const { return "struct"; }
 const char * UnionDecl::typeString() const { return aggrString( Union ); }
+std::string UnionDecl::typeString() const { return "union"; }
 const char * EnumDecl::typeString() const { return aggrString( Enum ); }
+std::string EnumDecl::typeString() const { return "enum"; }
 const char * TraitDecl::typeString() const { return aggrString( Trait ); }
+std::string TraitDecl::typeString() const { return "trait"; }
 bool EnumDecl::valueOf( Declaration * enumerator, long long int & value ) {

src/SynTree/Attribute.h

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Feb 13 21:34:08 2020
 // Update Count     : 40
+// Last Modified On : Sat Jul 22 09:54:14 2017
+// Update Count     : 39
 //
 …
         virtual ~Attribute();
         const std::string & get_name() const { return name; }
+        std::string get_name() const { return name; }
         void set_name( const std::string & newValue ) { name = newValue; }
         std::list< Expression * > & get_parameters() { return parameters; }

src/SynTree/Declaration.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 16:39:56 2019
 // Update Count     : 36
+// Last Modified By : Andrew Beach
+// Last Modified On : Wed Aug  9 14:38:00 2017
+// Update Count     : 25
 //
 …
 #include "SynTree/Statement.h"       // for AsmStmt
 #include "SynTree/SynTree.h"         // for UniqueId
-#include "SynTree/Expression.h"
 #include "Type.h"                    // for Type, Type::StorageClasses
-// To canonicalize declarations
 static UniqueId lastUniqueId = 0;

src/SynTree/Declaration.h

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:11:22 2019
 // Update Count     : 157
+// Last Modified By : Andrew Beach
+// Last Modified On : Thr May  2 10:47:00 2019
+// Update Count     : 135
 //
 …
 #include "BaseSyntaxNode.h"      // for BaseSyntaxNode
 #include "Mutator.h"             // for Mutator
+#include "LinkageSpec.h"         // for Spec, Cforall
+#include "Parser/LinkageSpec.h"  // for Spec, Cforall
+#include "Parser/ParseNode.h"    // for DeclarationNode, DeclarationNode::Ag...
 #include "SynTree.h"             // for UniqueId
 #include "SynTree/Type.h"        // for Type, Type::StorageClasses, Type::Fu...
 …
         bool extension = false;
         Declaration( const std::string & name, Type::StorageClasses scs, LinkageSpec::Spec linkage );
         Declaration( const Declaration & other );
+        Declaration( const std::string &name, Type::StorageClasses scs, LinkageSpec::Spec linkage );
+        Declaration( const Declaration &other );
         virtual ~Declaration();
         const std::string & get_name() const { return name; }
+        const std::string &get_name() const { return name; }
         void set_name( std::string newValue ) { name = newValue; }
 …
         bool get_extension() const { return extension; }
         Declaration * set_extension( bool exten ) { extension = exten; return this; }
+        Declaration *set_extension( bool exten ) { extension = exten; return this; }
         void fixUniqueId( void );
         virtual Declaration * clone() const override = 0;
+        virtual Declaration *clone() const override = 0;
         virtual void accept( Visitor & v ) override = 0;
         virtual void accept( Visitor & v ) const override = 0;
         virtual Declaration * acceptMutator( Mutator & m ) override = 0;
         virtual void print( std::ostream & os, Indenter indent = {} ) const override = 0;
         virtual void printShort( std::ostream & os, Indenter indent = {} ) const = 0;
+        virtual Declaration *acceptMutator( Mutator &m ) override = 0;
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override = 0;
+        virtual void printShort( std::ostream &os, Indenter indent = {} ) const = 0;
         UniqueId uniqueId;
 …
         int scopeLevel = 0;
         Expression * asmName;
+        Expression *asmName;
         std::list< Attribute * > attributes;
         bool isDeleted = false;
         DeclarationWithType( const std::string & name, Type::StorageClasses scs, LinkageSpec::Spec linkage, const std::list< Attribute * > & attributes, Type::FuncSpecifiers fs );
         DeclarationWithType( const DeclarationWithType & other );
+        DeclarationWithType( const std::string &name, Type::StorageClasses scs, LinkageSpec::Spec linkage, const std::list< Attribute * > & attributes, Type::FuncSpecifiers fs );
+        DeclarationWithType( const DeclarationWithType &other );
         virtual ~DeclarationWithType();
 …
         DeclarationWithType * set_scopeLevel( int newValue ) { scopeLevel = newValue; return this; }
         Expression * get_asmName() const { return asmName; }
         DeclarationWithType * set_asmName( Expression * newValue ) { asmName = newValue; return this; }
+        Expression *get_asmName() const { return asmName; }
+        DeclarationWithType * set_asmName( Expression *newValue ) { asmName = newValue; return this; }
         std::list< Attribute * >& get_attributes() { return attributes; }
 …
         //void set_functionSpecifiers( Type::FuncSpecifiers newValue ) { fs = newValue; }
         virtual DeclarationWithType * clone() const override = 0;
         virtual DeclarationWithType * acceptMutator( Mutator & m )  override = 0;
+        virtual DeclarationWithType *clone() const override = 0;
+        virtual DeclarationWithType *acceptMutator( Mutator &m )  override = 0;
         virtual Type * get_type() const = 0;
 …
         typedef DeclarationWithType Parent;
   public:
         Type * type;
         Initializer * init;
         Expression * bitfieldWidth;
         ObjectDecl( const std::string & name, Type::StorageClasses scs, LinkageSpec::Spec linkage, Expression * bitfieldWidth, Type * type, Initializer * init,
+        Type *type;
+        Initializer *init;
+        Expression *bitfieldWidth;
+        ObjectDecl( const std::string &name, Type::StorageClasses scs, LinkageSpec::Spec linkage, Expression *bitfieldWidth, Type *type, Initializer *init,
                                 const std::list< Attribute * > attributes = std::list< Attribute * >(), Type::FuncSpecifiers fs = Type::FuncSpecifiers() );
         ObjectDecl( const ObjectDecl & other );
+        ObjectDecl( const ObjectDecl &other );
         virtual ~ObjectDecl();
         virtual Type * get_type() const override { return type; }
         virtual void set_type(Type * newType) override { type = newType; }
         Initializer * get_init() const { return init; }
         void set_init( Initializer * newValue ) { init = newValue; }
         Expression * get_bitfieldWidth() const { return bitfieldWidth; }
         void set_bitfieldWidth( Expression * newValue ) { bitfieldWidth = newValue; }
+        virtual void set_type(Type *newType) override { type = newType; }
+        Initializer *get_init() const { return init; }
+        void set_init( Initializer *newValue ) { init = newValue; }
+        Expression *get_bitfieldWidth() const { return bitfieldWidth; }
+        void set_bitfieldWidth( Expression *newValue ) { bitfieldWidth = newValue; }
         static ObjectDecl * newObject( const std::string & name, Type * type, Initializer * init );
         virtual ObjectDecl * clone() const override { return new ObjectDecl( *this ); }
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual DeclarationWithType * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
         virtual void print( std::ostream & os, Indenter indent = {} ) const override;
         virtual void printShort( std::ostream & os, Indenter indent = {} ) const override;
+        virtual ObjectDecl *clone() const override { return new ObjectDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual DeclarationWithType *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override;
+        virtual void printShort( std::ostream &os, Indenter indent = {} ) const override;
 };
 …
         typedef DeclarationWithType Parent;
   public:
         FunctionType * type;
         CompoundStmt * statements;
+        FunctionType *type;
+        CompoundStmt *statements;
         std::list< Expression * > withExprs;
         FunctionDecl( const std::string & name, Type::StorageClasses scs, LinkageSpec::Spec linkage, FunctionType * type, CompoundStmt * statements,
+        FunctionDecl( const std::string &name, Type::StorageClasses scs, LinkageSpec::Spec linkage, FunctionType *type, CompoundStmt *statements,
                                   const std::list< Attribute * > attributes = std::list< Attribute * >(), Type::FuncSpecifiers fs = Type::FuncSpecifiers() );
         FunctionDecl( const FunctionDecl & other );
+        FunctionDecl( const FunctionDecl &other );
         virtual ~FunctionDecl();
 …
         FunctionType * get_functionType() const { return type; }
         void set_functionType( FunctionType * newValue ) { type = newValue; }
         CompoundStmt * get_statements() const { return statements; }
         void set_statements( CompoundStmt * newValue ) { statements = newValue; }
+        void set_functionType( FunctionType *newValue ) { type = newValue; }
+        CompoundStmt *get_statements() const { return statements; }
+        void set_statements( CompoundStmt *newValue ) { statements = newValue; }
         bool has_body() const { return NULL != statements; }
         static FunctionDecl * newFunction( const std::string & name, FunctionType * type, CompoundStmt * statements );
         virtual FunctionDecl * clone() const override { return new FunctionDecl( *this ); }
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual DeclarationWithType * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
         virtual void print( std::ostream & os, Indenter indent = {} ) const override;
         virtual void printShort( std::ostream & os, Indenter indent = {} ) const override;
+        virtual FunctionDecl *clone() const override { return new FunctionDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual DeclarationWithType *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override;
+        virtual void printShort( std::ostream &os, Indenter indent = {} ) const override;
 };
 …
         typedef Declaration Parent;
   public:
         Type * base;
         std::list< TypeDecl * > parameters;
         std::list< DeclarationWithType * > assertions;
         NamedTypeDecl( const std::string & name, Type::StorageClasses scs, Type * type );
         NamedTypeDecl( const NamedTypeDecl & other );
+        Type *base;
+        std::list< TypeDecl* > parameters;
+        std::list< DeclarationWithType* > assertions;
+        NamedTypeDecl( const std::string &name, Type::StorageClasses scs, Type *type );
+        NamedTypeDecl( const NamedTypeDecl &other );
         virtual ~NamedTypeDecl();
         Type * get_base() const { return base; }
         void set_base( Type * newValue ) { base = newValue; }
         std::list< TypeDecl* > & get_parameters() { return parameters; }
         std::list< DeclarationWithType * >& get_assertions() { return assertions; }
         virtual const char * typeString() const = 0;
         virtual NamedTypeDecl * clone() const override = 0;
         virtual void print( std::ostream & os, Indenter indent = {} ) const override;
         virtual void printShort( std::ostream & os, Indenter indent = {} ) const override;
+        Type *get_base() const { return base; }
+        void set_base( Type *newValue ) { base = newValue; }
+        std::list< TypeDecl* >& get_parameters() { return parameters; }
+        std::list< DeclarationWithType* >& get_assertions() { return assertions; }
+        virtual std::string typeString() const = 0;
+        virtual NamedTypeDecl *clone() const override = 0;
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override;
+        virtual void printShort( std::ostream &os, Indenter indent = {} ) const override;
 };
 …
         typedef NamedTypeDecl Parent;
   public:
         enum Kind { Dtype, Otype, Ftype, Ttype, NUMBER_OF_KINDS };
         Kind kind;
+        enum Kind { Dtype, Ftype, Ttype, NUMBER_OF_KINDS };
+        Type * init;
         bool sized;
-        Type * init;
         /// Data extracted from a type decl
         struct Data {
                 Kind kind;
+                TypeDecl::Kind kind;
                 bool isComplete;
                 Data() : kind( NUMBER_OF_KINDS ), isComplete( false ) {}
                 Data( const TypeDecl * typeDecl ) : Data( typeDecl->get_kind(), typeDecl->isComplete() ) {}
+                Data() : kind( (TypeDecl::Kind)-1 ), isComplete( false ) {}
+                Data( TypeDecl * typeDecl ) : Data( typeDecl->get_kind(), typeDecl->isComplete() ) {}
                 Data( Kind kind, bool isComplete ) : kind( kind ), isComplete( isComplete ) {}
                 Data( const Data & d1, const Data & d2 )
                         : kind( d1.kind ), isComplete( d1.isComplete || d2.isComplete ) {}
                 bool operator==( const Data & other ) const { return kind == other.kind && isComplete == other.isComplete; }
                 bool operator!=( const Data & other ) const { return !(*this == other);}
+                Data( const Data& d1, const Data& d2 )
+                : kind( d1.kind ), isComplete ( d1.isComplete || d2.isComplete ) {}
+                bool operator==(const Data & other) const { return kind == other.kind && isComplete == other.isComplete; }
+                bool operator!=(const Data & other) const { return !(*this == other);}
         };
         TypeDecl( const std::string & name, Type::StorageClasses scs, Type * type, Kind kind, bool sized, Type * init = nullptr );
         TypeDecl( const TypeDecl & other );
+        TypeDecl( const std::string &name, Type::StorageClasses scs, Type *type, Kind kind, bool sized, Type * init = nullptr );
+        TypeDecl( const TypeDecl &other );
         virtual ~TypeDecl();
 …
         TypeDecl * set_sized( bool newValue ) { sized = newValue; return this; }
+        virtual const char * typeString() const override;
+        virtual const char * genTypeString() const;
+        virtual TypeDecl * clone() const override { return new TypeDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
+        virtual void print( std::ostream & os, Indenter indent = {} ) const override;
+        virtual std::string typeString() const override;
+        virtual std::string genTypeString() const;
+        virtual TypeDecl *clone() const override { return new TypeDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override;
+        Kind kind;
 };
 …
         typedef NamedTypeDecl Parent;
   public:
         TypedefDecl( const std::string & name, CodeLocation location, Type::StorageClasses scs, Type * type, LinkageSpec::Spec spec = LinkageSpec::Cforall )
+        TypedefDecl( const std::string &name, CodeLocation location, Type::StorageClasses scs, Type *type, LinkageSpec::Spec spec = LinkageSpec::Cforall )
                 : Parent( name, scs, type ) { set_linkage( spec ); this->location = location; }
         TypedefDecl( const TypedefDecl & other ) : Parent( other ) {}
         virtual const char * typeString() const override;
         virtual TypedefDecl * clone() const override { return new TypedefDecl( *this ); }
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual Declaration * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
+        TypedefDecl( const TypedefDecl &other ) : Parent( other ) {}
+        virtual std::string typeString() const override;
+        virtual TypedefDecl *clone() const override { return new TypedefDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
   private:
 };
 …
         typedef Declaration Parent;
   public:
-        enum Aggregate { Struct, Union, Enum, Exception, Trait, Generator, Coroutine, Monitor, Thread, NoAggregate };
-        static const char * aggrString( Aggregate aggr );
         std::list<Declaration*> members;
         std::list<TypeDecl*> parameters;
 …
         AggregateDecl * parent = nullptr;
         AggregateDecl( const std::string & name, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall );
         AggregateDecl( const AggregateDecl & other );
+        AggregateDecl( const std::string &name, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall );
+        AggregateDecl( const AggregateDecl &other );
         virtual ~AggregateDecl();
 …
         AggregateDecl * set_body( bool body ) { AggregateDecl::body = body; return this; }
         virtual void print( std::ostream & os, Indenter indent = {} ) const override final;
         virtual void printShort( std::ostream & os, Indenter indent = {} ) const override;
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override final;
+        virtual void printShort( std::ostream &os, Indenter indent = {} ) const override;
   protected:
         virtual const char * typeString() const = 0;
+        virtual std::string typeString() const = 0;
 };
 …
         typedef AggregateDecl Parent;
   public:
+        StructDecl( const std::string & name, Aggregate kind = Struct, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall ) : Parent( name, attributes, linkage ), kind( kind ) {}
+        StructDecl( const StructDecl & other ) : Parent( other ), kind( other.kind ) {}
+        bool is_coroutine() { return kind == Coroutine; }
+        bool is_generator() { return kind == Generator; }
+        bool is_monitor  () { return kind == Monitor  ; }
+        bool is_thread   () { return kind == Thread   ; }
+        virtual StructDecl * clone() const override { return new StructDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
+        Aggregate kind;
+  private:
+        virtual const char * typeString() const override;
+        StructDecl( const std::string &name, DeclarationNode::Aggregate kind = DeclarationNode::Struct, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall ) : Parent( name, attributes, linkage ), kind( kind ) {}
+        StructDecl( const StructDecl &other ) : Parent( other ), kind( other.kind ) {}
+        bool is_coroutine() { return kind == DeclarationNode::Coroutine; }
+        bool is_monitor() { return kind == DeclarationNode::Monitor; }
+        bool is_thread() { return kind == DeclarationNode::Thread; }
+        virtual StructDecl *clone() const override { return new StructDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+        DeclarationNode::Aggregate kind;
+  private:
+        virtual std::string typeString() const override;
 };
 …
         typedef AggregateDecl Parent;
   public:
         UnionDecl( const std::string & name, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall ) : Parent( name, attributes, linkage ) {}
         UnionDecl( const UnionDecl & other ) : Parent( other ) {}
         virtual UnionDecl * clone() const override { return new UnionDecl( *this ); }
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual Declaration * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
   private:
         virtual const char * typeString() const override;
+        UnionDecl( const std::string &name, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall ) : Parent( name, attributes, linkage ) {}
+        UnionDecl( const UnionDecl &other ) : Parent( other ) {}
+        virtual UnionDecl *clone() const override { return new UnionDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+  private:
+        virtual std::string typeString() const override;
 };
 …
         typedef AggregateDecl Parent;
   public:
         EnumDecl( const std::string & name, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall ) : Parent( name, attributes, linkage ) {}
         EnumDecl( const EnumDecl & other ) : Parent( other ) {}
+        EnumDecl( const std::string &name, const std::list< Attribute * > & attributes = std::list< class Attribute * >(), LinkageSpec::Spec linkage = LinkageSpec::Cforall ) : Parent( name, attributes, linkage ) {}
+        EnumDecl( const EnumDecl &other ) : Parent( other ) {}
         bool valueOf( Declaration * enumerator, long long int & value );
         virtual EnumDecl * clone() const override { return new EnumDecl( *this ); }
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual Declaration * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
+        virtual EnumDecl *clone() const override { return new EnumDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
   private:
         std::unordered_map< std::string, long long int > enumValues;
         virtual const char * typeString() const override;
+        virtual std::string typeString() const override;
 };
 …
         typedef AggregateDecl Parent;
   public:
         TraitDecl( const std::string & name, const std::list< Attribute * > & attributes, LinkageSpec::Spec linkage ) : Parent( name, attributes, linkage ) {
+        TraitDecl( const std::string &name, const std::list< Attribute * > & attributes, LinkageSpec::Spec linkage ) : Parent( name, attributes, linkage ) {
                 assertf( attributes.empty(), "attribute unsupported for traits" );
+        }
         TraitDecl( const TraitDecl & other ) : Parent( other ) {}
         virtual TraitDecl * clone() const override { return new TraitDecl( *this ); }
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual Declaration * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
   private:
         virtual const char * typeString() const override;
+        TraitDecl( const TraitDecl &other ) : Parent( other ) {}
+        virtual TraitDecl *clone() const override { return new TraitDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual Declaration *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+  private:
+        virtual std::string typeString() const override;
 };
 …
 class AsmDecl : public Declaration {
   public:
         AsmStmt * stmt;
         AsmDecl( AsmStmt * stmt );
         AsmDecl( const AsmDecl & other );
+        AsmStmt *stmt;
+        AsmDecl( AsmStmt *stmt );
+        AsmDecl( const AsmDecl &other );
         virtual ~AsmDecl();
         AsmStmt * get_stmt() { return stmt; }
         void set_stmt( AsmStmt * newValue ) { stmt = newValue; }
         virtual AsmDecl * clone() const override { return new AsmDecl( *this ); }
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual AsmDecl * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
         virtual void print( std::ostream & os, Indenter indent = {} ) const override;
         virtual void printShort( std::ostream & os, Indenter indent = {} ) const override;
+        AsmStmt *get_stmt() { return stmt; }
+        void set_stmt( AsmStmt *newValue ) { stmt = newValue; }
+        virtual AsmDecl *clone() const override { return new AsmDecl( *this ); }
+        virtual void accept( Visitor & v ) override { v.visit( this ); }
+        virtual void accept( Visitor & v ) const override { v.visit( this ); }
+        virtual AsmDecl *acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override;
+        virtual void printShort( std::ostream &os, Indenter indent = {} ) const override;
 };
 …
         virtual void accept( Visitor & v ) override { v.visit( this ); }
         virtual void accept( Visitor & v ) const override { v.visit( this ); }
         virtual StaticAssertDecl * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
         virtual void print( std::ostream & os, Indenter indent = {} ) const override;
         virtual void printShort( std::ostream & os, Indenter indent = {} ) const override;
+        virtual StaticAssertDecl * acceptMutator( Mutator &m )  override { return m.mutate( this ); }
+        virtual void print( std::ostream &os, Indenter indent = {} ) const override;
+        virtual void printShort( std::ostream &os, Indenter indent = {} ) const override;
 };

src/SynTree/DeclarationWithType.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:45:16 2019
 // Update Count     : 26
+// Last Modified On : Thu Mar 16 08:34:35 2017
+// Update Count     : 25
 //
 …
 #include "Common/utility.h"      // for cloneAll, deleteAll, maybeClone
 #include "Declaration.h"         // for DeclarationWithType, Declaration
 #include "LinkageSpec.h"         // for Spec
 #include "Expression.h"          // for ConstantExpr
+#include "Parser/LinkageSpec.h"  // for Spec
+#include "SynTree/Expression.h"  // for ConstantExpr
 #include "Type.h"                // for Type, Type::FuncSpecifiers, Type::St...

src/SynTree/Expression.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 07:55:15 2019
 // Update Count     : 70
+// Last Modified By : Andrew Beach
+// Last Modified On : Thr Aug 15 13:43:00 2019
+// Update Count     : 64
 //
 …
 #include "Common/utility.h"          // for maybeClone, cloneAll, deleteAll
+#include "Declaration.h"             // for ObjectDecl, DeclarationWithType
 #include "Expression.h"              // for Expression, ImplicitCopyCtorExpr
 #include "InitTweak/InitTweak.h"     // for getCallArg, getPointerBase
 …
+}
 KeywordCastExpr::KeywordCastExpr( Expression * arg, AggregateDecl::Aggregate target ) : Expression(), arg(arg), target( target ) {
+KeywordCastExpr::KeywordCastExpr( Expression * arg, Target target ) : Expression(), arg(arg), target( target ) {
+}
 …
+}
+const char * KeywordCastExpr::targetString() const {
+        return AggregateDecl::aggrString( target );
+const std::string & KeywordCastExpr::targetString() const {
+        static const std::string targetStrs[] = {
+                "coroutine", "thread", "monitor"
+        };
+        static_assert(
+                (sizeof(targetStrs) / sizeof(targetStrs[0])) == ((unsigned long)NUMBER_OF_TARGETS),
+                "Each KeywordCastExpr::Target should have a corresponding string representation"
+        );
+        return targetStrs[(unsigned long)target];
+}

src/SynTree/Expression.h

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 11 16:50:19 2019
 // Update Count     : 60
+// Last Modified By : Andrew Beach
+// Last Modified On : Thr Aug 15 13:46:00 2019
+// Update Count     : 54
 //
 …
 #include "Label.h"                // for Label
 #include "Mutator.h"              // for Mutator
-#include "Declaration.h"          // for Aggregate
 #include "SynTree.h"              // for UniqueId
 #include "Visitor.h"              // for Visitor
 …
 public:
         Expression * arg;
+        struct Concrete {
+                std::string field;
+                std::string getter;
+        };
+        AggregateDecl::Aggregate target;
+        Concrete concrete_target;
+        KeywordCastExpr( Expression * arg, AggregateDecl::Aggregate target );
+        enum Target {
+                Coroutine, Thread, Monitor, NUMBER_OF_TARGETS
+        } target;
+        KeywordCastExpr( Expression * arg, Target target );
         KeywordCastExpr( const KeywordCastExpr & other );
         virtual ~KeywordCastExpr();
         const char * targetString() const;
+        const std::string & targetString() const;
         virtual KeywordCastExpr * clone() const override { return new KeywordCastExpr( * this ); }

src/SynTree/FunctionDecl.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Dec 16 15:11:20 2019
 // Update Count     : 77
+// Last Modified On : Thu Mar 16 08:33:41 2017
+// Update Count     : 74
 //
 …
 #include "Common/utility.h"      // for maybeClone, printAll
 #include "Declaration.h"         // for FunctionDecl, FunctionDecl::Parent
+#include "Expression.h"
+#include "LinkageSpec.h"         // for Spec, linkageName, Cforall
+#include "Parser/LinkageSpec.h"  // for Spec, linkageName, Cforall
 #include "Statement.h"           // for CompoundStmt
 #include "Type.h"                // for Type, FunctionType, Type::FuncSpecif...
 …
         } // if
         if ( linkage != LinkageSpec::Cforall ) {
                 os << LinkageSpec::name( linkage ) << " ";
+                os << LinkageSpec::linkageName( linkage ) << " ";
         } // if

src/SynTree/Mutator.h

r7030dab	r71d6bd8
51	51	virtual Statement * mutate( CatchStmt * catchStmt ) = 0;
52	52	virtual Statement * mutate( FinallyStmt * catchStmt ) = 0;
53		~~virtual Statement * mutate( SuspendStmt * suspendStmt ) = 0;~~
54	53	virtual Statement * mutate( WaitForStmt * waitforStmt ) = 0;
55	54	virtual Declaration * mutate( WithStmt * withStmt ) = 0;

src/SynTree/NamedTypeDecl.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Dec 16 15:11:40 2019
 // Update Count     : 17
+// Last Modified By : Andrew Beach
+// Last Modified On : Wed Aug  9 13:28:00 2017
+// Update Count     : 14
 //
 …
 #include "Common/utility.h"      // for printAll, cloneAll, deleteAll, maybe...
 #include "Declaration.h"         // for NamedTypeDecl, DeclarationWithType
 #include "LinkageSpec.h"         // for Spec, Cforall, linkageName
+#include "Parser/LinkageSpec.h"  // for Spec, Cforall, linkageName
 #include "Type.h"                // for Type, Type::StorageClasses
 …
         if ( linkage != LinkageSpec::Cforall ) {
                 os << LinkageSpec::name( linkage ) << " ";
+                os << LinkageSpec::linkageName( linkage ) << " ";
         } // if
         get_storageClasses().print( os );
 …
+}
 const char * TypedefDecl::typeString() const { return "typedef"; }
+std::string TypedefDecl::typeString() const { return "typedef"; }
 // Local Variables: //

src/SynTree/ObjectDecl.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Dec 16 15:12:03 2019
 // Update Count     : 61
+// Last Modified On : Thu Mar 16 08:34:27 2017
+// Update Count     : 59
 //
 …
 #include "Expression.h"          // for Expression
 #include "Initializer.h"         // for Initializer
 #include "LinkageSpec.h"         // for Spec, linkageName, Cforall
+#include "Parser/LinkageSpec.h"  // for Spec, linkageName, Cforall
 #include "Type.h"                // for Type, Type::StorageClasses, Type::Fu...
 …
         if ( linkage != LinkageSpec::Cforall ) {
                 os << LinkageSpec::name( linkage ) << " ";
+                os << LinkageSpec::linkageName( linkage ) << " ";
         } // if

src/SynTree/Statement.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Mon Jan 20 16:03:00 2020
 // Update Count     : 71
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Sun Sep  3 20:46:44 2017
+// Update Count     : 68
 //
 …
 Statement::~Statement() {}
 ExprStmt::ExprStmt( Expression * expr ) : Statement(), expr( expr ) {}
 ExprStmt::ExprStmt( const ExprStmt & other ) : Statement( other ), expr( maybeClone( other.expr ) ) {}
+ExprStmt::ExprStmt( Expression *expr ) : Statement(), expr( expr ) {}
+ExprStmt::ExprStmt( const ExprStmt &other ) : Statement( other ), expr( maybeClone( other.expr ) ) {}
 ExprStmt::~ExprStmt() {
 …
+}
 void ExprStmt::print( std::ostream & os, Indenter indent ) const {
+void ExprStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Expression Statement:" << endl << indent+1;
         expr->print( os, indent+1 );
 …
 AsmStmt::AsmStmt( bool voltile, Expression * instruction, std::list<Expression *> output, std::list<Expression *> input, std::list<ConstantExpr *> clobber, std::list<Label> gotolabels ) : Statement(), voltile( voltile ), instruction( instruction ), output( output ), input( input ), clobber( clobber ), gotolabels( gotolabels ) {}
+AsmStmt::AsmStmt( bool voltile, Expression *instruction, std::list<Expression *> output, std::list<Expression *> input, std::list<ConstantExpr *> clobber, std::list<Label> gotolabels ) : Statement(), voltile( voltile ), instruction( instruction ), output( output ), input( input ), clobber( clobber ), gotolabels( gotolabels ) {}
 AsmStmt::AsmStmt( const AsmStmt & other ) : Statement( other ), voltile( other.voltile ), instruction( maybeClone( other.instruction ) ), gotolabels( other.gotolabels ) {
 …
+}
 void AsmStmt::print( std::ostream & os, Indenter indent ) const {
+void AsmStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Assembler Statement:" << endl;
         os << indent+1 << "instruction: " << endl << indent;
 …
 DirectiveStmt::DirectiveStmt( const std::string & directive ) : Statement(), directive( directive ) {}
 void DirectiveStmt::print( std::ostream & os, Indenter ) const {
+void DirectiveStmt::print( std::ostream &os, Indenter ) const {
         os << "GCC Directive:" << directive << endl;
+}
+const char * BranchStmt::brType[] = {
+        "Goto", "Break", "Continue", "Fall Through", "Fall Through Default",
+};
+const char *BranchStmt::brType[] = { "Goto", "Break", "Continue" };
 BranchStmt::BranchStmt( Label target, Type type ) throw ( SemanticErrorException ) :
 …
+}
 BranchStmt::BranchStmt( Expression * computedTarget, Type type ) throw ( SemanticErrorException ) :
+BranchStmt::BranchStmt( Expression *computedTarget, Type type ) throw ( SemanticErrorException ) :
         Statement(), computedTarget( computedTarget ), type( type ) {
         if ( type != BranchStmt::Goto || computedTarget == nullptr ) {
 …
+}
+void BranchStmt::print( std::ostream & os, Indenter indent ) const {
+        assert(type < 5);
+void BranchStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Branch (" << brType[type] << ")" << endl ;
         if ( target != "" ) os << indent+1 << "with target: " << target << endl;
 …
+}
 ReturnStmt::ReturnStmt( Expression * expr ) : Statement(), expr( expr ) {}
+ReturnStmt::ReturnStmt( Expression *expr ) : Statement(), expr( expr ) {}
 ReturnStmt::ReturnStmt( const ReturnStmt & other ) : Statement( other ), expr( maybeClone( other.expr ) ) {}
 …
+}
 void ReturnStmt::print( std::ostream & os, Indenter indent ) const {
+void ReturnStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Return Statement, returning: ";
         if ( expr != nullptr ) {
 …
+}
 IfStmt::IfStmt( Expression * condition, Statement * thenPart, Statement * elsePart, std::list<Statement *> initialization ):
+IfStmt::IfStmt( Expression *condition, Statement *thenPart, Statement *elsePart, std::list<Statement *> initialization ):
         Statement(), condition( condition ), thenPart( thenPart ), elsePart( elsePart ), initialization( initialization ) {}
 …
+}
 void IfStmt::print( std::ostream & os, Indenter indent ) const {
+void IfStmt::print( std::ostream &os, Indenter indent ) const {
         os << "If on condition: " << endl;
         os << indent+1;
 …
         thenPart->print( os, indent+1 );
         if ( elsePart != nullptr ) {
+        if ( elsePart != 0 ) {
                 os << indent << "... else: " << endl;
                 os << indent+1;
 …
+}
 SwitchStmt::SwitchStmt( Expression * condition, const std::list<Statement *> & statements ):
+SwitchStmt::SwitchStmt( Expression * condition, const std::list<Statement *> &statements ):
         Statement(), condition( condition ), statements( statements ) {
+}
 …
+}
 void SwitchStmt::print( std::ostream & os, Indenter indent ) const {
+void SwitchStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Switch on condition: ";
         condition->print( os );
 …
+}
 CaseStmt::CaseStmt( Expression * condition, const std::list<Statement *> & statements, bool deflt ) throw ( SemanticErrorException ) :
+CaseStmt::CaseStmt( Expression *condition, const std::list<Statement *> &statements, bool deflt ) throw ( SemanticErrorException ) :
         Statement(), condition( condition ), stmts( statements ), _isDefault( deflt ) {
         if ( isDefault() && condition != nullptr ) SemanticError( condition, "default case with condition: " );
+        if ( isDefault() && condition != 0 ) SemanticError( condition, "default case with condition: " );
+}
 …
+}
 void CaseStmt::print( std::ostream & os, Indenter indent ) const {
+void CaseStmt::print( std::ostream &os, Indenter indent ) const {
         if ( isDefault() ) os << indent << "Default ";
         else {
 …
+}
 WhileStmt::WhileStmt( Expression * condition, Statement * body, std::list< Statement * > & initialization, bool isDoWhile ):
+WhileStmt::WhileStmt( Expression *condition, Statement *body, std::list< Statement * > & initialization, bool isDoWhile ):
         Statement(), condition( condition), body( body), initialization( initialization ), isDoWhile( isDoWhile) {
+}
 …
+}
 void WhileStmt::print( std::ostream & os, Indenter indent ) const {
+void WhileStmt::print( std::ostream &os, Indenter indent ) const {
         os << "While on condition: " << endl ;
         condition->print( os, indent+1 );
 …
         os << indent << "... with body: " << endl;
         if ( body != nullptr ) body->print( os, indent+1 );
+}
 ForStmt::ForStmt( std::list<Statement *> initialization, Expression * condition, Expression * increment, Statement * body ):
+        if ( body != 0 ) body->print( os, indent+1 );
+}
+ForStmt::ForStmt( std::list<Statement *> initialization, Expression *condition, Expression *increment, Statement *body ):
         Statement(), initialization( initialization ), condition( condition ), increment( increment ), body( body ) {
+}
 …
+}
 void ForStmt::print( std::ostream & os, Indenter indent ) const {
+void ForStmt::print( std::ostream &os, Indenter indent ) const {
         Statement::print( os, indent ); // print labels
 …
+        }
         if ( body != nullptr ) {
+        if ( body != 0 ) {
                 os << "\n" << indent << "... with body: \n" << indent+1;
                 body->print( os, indent+1 );
 …
+}
 ThrowStmt::ThrowStmt( const ThrowStmt & other ) :
+ThrowStmt::ThrowStmt( const ThrowStmt &other ) :
         Statement ( other ), kind( other.kind ), expr( maybeClone( other.expr ) ), target( maybeClone( other.target ) ) {
+}
 …
+}
 void ThrowStmt::print( std::ostream & os, Indenter indent) const {
+void ThrowStmt::print( std::ostream &os, Indenter indent) const {
         if ( target ) os << "Non-Local ";
         os << "Throw Statement, raising: ";
 …
+}
 TryStmt::TryStmt( CompoundStmt * tryBlock, std::list<CatchStmt *> & handlers, FinallyStmt * finallyBlock ) :
+TryStmt::TryStmt( CompoundStmt *tryBlock, std::list<CatchStmt *> &handlers, FinallyStmt *finallyBlock ) :
         Statement(), block( tryBlock ),  handlers( handlers ), finallyBlock( finallyBlock ) {
+}
 TryStmt::TryStmt( const TryStmt & other ) : Statement( other ), block( maybeClone( other.block ) ), finallyBlock( maybeClone( other.finallyBlock ) ) {
+TryStmt::TryStmt( const TryStmt &other ) : Statement( other ), block( maybeClone( other.block ) ), finallyBlock( maybeClone( other.finallyBlock ) ) {
         cloneAll( other.handlers, handlers );
+}
 …
+}
 void TryStmt::print( std::ostream & os, Indenter indent ) const {
+void TryStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Try Statement" << endl;
         os << indent << "... with block:" << endl << indent+1;
 …
         // finally block
         if ( finallyBlock != nullptr ) {
+        if ( finallyBlock != 0 ) {
                 os << indent << "... and finally:" << endl << indent+1;
                 finallyBlock->print( os, indent+1 );
 …
+}
 CatchStmt::CatchStmt( Kind kind, Declaration * decl, Expression * cond, Statement * body ) :
+CatchStmt::CatchStmt( Kind kind, Declaration *decl, Expression *cond, Statement *body ) :
         Statement(), kind ( kind ), decl ( decl ), cond ( cond ), body( body ) {
                 assertf( decl, "Catch clause must have a declaration." );
 …
+}
 void CatchStmt::print( std::ostream & os, Indenter indent ) const {
+void CatchStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Catch " << ((Terminate == kind) ? "Terminate" : "Resume") << " Statement" << endl;
 …
 FinallyStmt::FinallyStmt( CompoundStmt * block ) : Statement(), block( block ) {
+FinallyStmt::FinallyStmt( CompoundStmt *block ) : Statement(), block( block ) {
+}
 …
+}
 void FinallyStmt::print( std::ostream & os, Indenter indent ) const {
+void FinallyStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Finally Statement" << endl;
         os << indent << "... with block:" << endl << indent+1;
         block->print( os, indent+1 );
+}
-SuspendStmt::SuspendStmt( const SuspendStmt & other )
-        : Statement( other )
-        , then( maybeClone(other.then) )
-{}
-SuspendStmt::~SuspendStmt() {
-        delete then;
+}
-void SuspendStmt::print( std::ostream & os, Indenter indent ) const {
-        os << "Suspend Statement";
-        switch (type) {
-                case None     : os << " with implicit target"; break;
-                case Generator: os << " for generator"       ; break;
-                case Coroutine: os << " for coroutine"       ; break;
+        }
-        os << endl;
-        indent += 1;
-        if(then) {
-                os << indent << " with post statement :" << endl;
-                then->print( os, indent + 1);
+        }
+}
 …
+}
 void WaitForStmt::print( std::ostream & os, Indenter indent ) const {
+void WaitForStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Waitfor Statement" << endl;
         indent += 1;
 …
+}
 void NullStmt::print( std::ostream & os, Indenter indent ) const {
+void NullStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Null Statement" << endl;
         Statement::print( os, indent );
 …
+}
 void ImplicitCtorDtorStmt::print( std::ostream & os, Indenter indent ) const {
+void ImplicitCtorDtorStmt::print( std::ostream &os, Indenter indent ) const {
         os << "Implicit Ctor Dtor Statement" << endl;
         os << indent << "... with Ctor/Dtor: ";

src/SynTree/Statement.h

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Jan 10 14:13:24 2020
 // Update Count     : 85
+// Last Modified On : Tue Mar 12 09:01:53 2019
+// Update Count     : 83
 //
 …
         Statement * body;
         ForStmt( std::list<Statement *> initialization, Expression * condition = nullptr, Expression * increment = nullptr, Statement * body = nullptr );
+        ForStmt( std::list<Statement *> initialization, Expression * condition = 0, Expression * increment = 0, Statement * body = 0 );
         ForStmt( const ForStmt & other );
         virtual ~ForStmt();
 …
         FinallyStmt * finallyBlock;
         TryStmt( CompoundStmt * tryBlock, std::list<CatchStmt *> & handlers, FinallyStmt * finallyBlock = nullptr );
+        TryStmt( CompoundStmt * tryBlock, std::list<CatchStmt *> & handlers, FinallyStmt * finallyBlock = 0 );
         TryStmt( const TryStmt & other );
         virtual ~TryStmt();
 …
 };
-class SuspendStmt : public Statement {
-  public:
-        CompoundStmt * then = nullptr;
-        enum Type { None, Coroutine, Generator } type = None;
-        SuspendStmt() = default;
-        SuspendStmt( const SuspendStmt & );
-        virtual ~SuspendStmt();
-        virtual SuspendStmt * clone() const override { return new SuspendStmt( *this ); }
-        virtual void accept( Visitor & v ) override { v.visit( this ); }
-        virtual void accept( Visitor & v ) const override { v.visit( this ); }
-        virtual Statement * acceptMutator( Mutator & m )  override { return m.mutate( this ); }
-        virtual void print( std::ostream & os, Indenter indent = {} ) const override;
-};
 class WaitForStmt : public Statement {
   public:

src/SynTree/SynTree.h

r7030dab	r71d6bd8
54	54	class CatchStmt;
55	55	class FinallyStmt;
56		~~class SuspendStmt;~~
57	56	class WaitForStmt;
58	57	class WithStmt;

src/SynTree/TupleType.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:44:38 2019
 // Update Count     : 4
+// Last Modified On : Wed Feb  1 17:10:58 2017
+// Update Count     : 3
 //
 …
 #include "Declaration.h"         // for Declaration, ObjectDecl
 #include "Initializer.h"         // for ListInit
 #include "LinkageSpec.h"         // for Cforall
+#include "Parser/LinkageSpec.h"  // for Cforall
 #include "Type.h"                // for TupleType, Type, Type::Qualifiers

src/SynTree/Type.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Dec 15 16:52:37 2019
 // Update Count     : 49
+// Last Modified On : Sun Aug  4 21:05:07 2019
+// Update Count     : 45
 //
 #include "Type.h"
 …
 using namespace std;
-// GENERATED START, DO NOT EDIT
-// GENERATED BY BasicTypes-gen.cc
 const char * BasicType::typeNames[] = {
         "_Bool",
 …
         "float",
         "float _Complex",
+        //"float _Imaginary",
         "_Float32x",
         "_Float32x _Complex",
 …
         "double",
         "double _Complex",
+        //"double _Imaginary",
         "_Float64x",
         "_Float64x _Complex",
 …
         "long double",
         "long double _Complex",
+        //"long double _Imaginary",
         "_Float128x",
         "_Float128x _Complex",
 };
+// GENERATED END
+static_assert(
+        sizeof(BasicType::typeNames) / sizeof(BasicType::typeNames[0]) == BasicType::NUMBER_OF_BASIC_TYPES,
+        "Each basic type name should have a corresponding kind enum value"
+);
 Type::Type( const Qualifiers &tq, const std::list< Attribute * > & attributes ) : tq( tq ), attributes( attributes ) {}

src/SynTree/TypeDecl.cc

-              r7030dab
+              r71d6bd8
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 15:26:14 2019
 // Update Count     : 21
+// Last Modified By : Andrew Beach
+// Last Modified On : Wed Aug  9 14:35:00 2017
+// Update Count     : 6
 //
 …
 #include "Type.h"            // for Type, Type::StorageClasses
 TypeDecl::TypeDecl( const std::string & name, Type::StorageClasses scs, Type * type, Kind kind, bool sized, Type * init ) : Parent( name, scs, type ), kind( kind ), sized( kind == Ttype || sized ), init( init ) {
+TypeDecl::TypeDecl( const std::string &name, Type::StorageClasses scs, Type *type, Kind kind, bool sized, Type * init ) : Parent( name, scs, type ), init( init ), sized( kind == Ttype || sized ), kind( kind ) {
+}
 TypeDecl::TypeDecl( const TypeDecl & other ) : Parent( other ), kind( other.kind ), sized( other.sized ), init( maybeClone( other.init ) ) {
+TypeDecl::TypeDecl( const TypeDecl &other ) : Parent( other ), init( maybeClone( other.init ) ), sized( other.sized ), kind( other.kind ) {
+}
 TypeDecl::~TypeDecl() {
         delete init;
+  delete init;
+}
 const char * TypeDecl::typeString() const {
         static const char * kindNames[] = { "sized data type", "sized object type", "sized function type", "sized tuple type" };
         static_assert( sizeof(kindNames)/sizeof(kindNames[0]) == TypeDecl::NUMBER_OF_KINDS, "typeString: kindNames is out of sync." );
         assertf( kind < TypeDecl::NUMBER_OF_KINDS, "TypeDecl kind is out of bounds." );
         return isComplete() ? kindNames[ kind ] : &kindNames[ kind ][ sizeof("sized") ]; // sizeof includes '\0'
+std::string TypeDecl::typeString() const {
+        static const std::string kindNames[] = { "object type", "function type", "tuple type" };
+        assertf( sizeof(kindNames)/sizeof(kindNames[0]) == DeclarationNode::NoTypeClass-1, "typeString: kindNames is out of sync." );
+        assertf( kind < sizeof(kindNames)/sizeof(kindNames[0]), "TypeDecl's kind is out of bounds." );
+        return (isComplete() ? "sized " : "") + kindNames[ kind ];
+}
 const char * TypeDecl::genTypeString() const {
         static const char * kindNames[] = { "dtype", "otype", "ftype", "ttype" };
         static_assert( sizeof(kindNames)/sizeof(kindNames[0]) == TypeDecl::NUMBER_OF_KINDS, "genTypeString: kindNames is out of sync." );
         assertf( kind < TypeDecl::NUMBER_OF_KINDS, "TypeDecl kind is out of bounds." );
+std::string TypeDecl::genTypeString() const {
+        static const std::string kindNames[] = { "dtype", "ftype", "ttype" };
+        assertf( sizeof(kindNames)/sizeof(kindNames[0]) == DeclarationNode::NoTypeClass-1, "genTypeString: kindNames is out of sync." );
+        assertf( kind < sizeof(kindNames)/sizeof(kindNames[0]), "TypeDecl's kind is out of bounds." );
         return kindNames[ kind ];
+}
 void TypeDecl::print( std::ostream &os, Indenter indent ) const {
         NamedTypeDecl::print( os, indent );
         if ( init ) {
                 os << std::endl << indent << "with type initializer: ";
                 init->print( os, indent + 1 );
+        } // if
+  NamedTypeDecl::print( os, indent );
+  if ( init ) {
+    os << std::endl << indent << "with type initializer: ";
+    init->print( os, indent + 1 );
+  }
+}
 std::ostream & operator<<( std::ostream & os, const TypeDecl::Data & data ) {
         return os << data.kind << ", " << data.isComplete;
+  return os << data.kind << ", " << data.isComplete;
+}

src/SynTree/Visitor.h

r7030dab	r71d6bd8
78	78	virtual void visit( FinallyStmt * node ) { visit( const_cast<const FinallyStmt *>(node) ); }
79	79	virtual void visit( const FinallyStmt * finallyStmt ) = 0;
80		~~virtual void visit( SuspendStmt * node ) { visit( const_cast<const SuspendStmt *>(node) ); }~~
81		~~virtual void visit( const SuspendStmt * suspendStmt ) = 0;~~
82	80	virtual void visit( WaitForStmt * node ) { visit( const_cast<const WaitForStmt *>(node) ); }
83	81	virtual void visit( const WaitForStmt * waitforStmt ) = 0;

src/SynTree/module.mk

-              r7030dab
+              r71d6bd8
 ## Created On       : Mon Jun  1 17:49:17 2015
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Sat Dec 14 07:26:43 2019
 ## Update Count     : 2
+## Last Modified On : Mon Jun  1 17:54:09 2015
+## Update Count     : 1
 ###############################################################################
 SRC_SYNTREE = \
+      SynTree/Type.cc \
+      SynTree/VoidType.cc \
+      SynTree/BasicType.cc \
+      SynTree/PointerType.cc \
+      SynTree/ArrayType.cc \
+      SynTree/ReferenceType.cc \
+      SynTree/FunctionType.cc \
+      SynTree/ReferenceToType.cc \
+      SynTree/TupleType.cc \
+      SynTree/TypeofType.cc \
+      SynTree/AttrType.cc \
+      SynTree/VarArgsType.cc \
+      SynTree/ZeroOneType.cc \
+      SynTree/Constant.cc \
+      SynTree/Expression.cc \
+      SynTree/TupleExpr.cc \
+      SynTree/CommaExpr.cc \
+      SynTree/TypeExpr.cc \
+      SynTree/ApplicationExpr.cc \
       SynTree/AddressExpr.cc \
+      SynTree/AggregateDecl.cc \
+      SynTree/ApplicationExpr.cc \
+      SynTree/ArrayType.cc \
+      SynTree/AttrType.cc \
+      SynTree/Attribute.cc \
+      SynTree/BasicType.cc \
+      SynTree/CommaExpr.cc \
+      SynTree/Statement.cc \
       SynTree/CompoundStmt.cc \
-      SynTree/Constant.cc \
-      SynTree/DeclReplacer.cc \
       SynTree/DeclStmt.cc \
       SynTree/Declaration.cc \
       SynTree/DeclarationWithType.cc \
       SynTree/Expression.cc \
+      SynTree/ObjectDecl.cc \
       SynTree/FunctionDecl.cc \
+      SynTree/FunctionType.cc \
+      SynTree/AggregateDecl.cc \
+      SynTree/NamedTypeDecl.cc \
+      SynTree/TypeDecl.cc \
       SynTree/Initializer.cc \
-      SynTree/LinkageSpec.cc \
-      SynTree/NamedTypeDecl.cc \
-      SynTree/ObjectDecl.cc \
-      SynTree/PointerType.cc \
-      SynTree/ReferenceToType.cc \
-      SynTree/ReferenceType.cc \
-      SynTree/Statement.cc \
-      SynTree/TupleExpr.cc \
-      SynTree/TupleType.cc \
-      SynTree/Type.cc \
-      SynTree/TypeDecl.cc \
-      SynTree/TypeExpr.cc \
       SynTree/TypeSubstitution.cc \
+      SynTree/TypeofType.cc \
+      SynTree/VarArgsType.cc \
+      SynTree/VoidType.cc \
+      SynTree/ZeroOneType.cc
+      SynTree/Attribute.cc \
+      SynTree/DeclReplacer.cc
 SRC += $(SRC_SYNTREE)

src/Tuples/TupleAssignment.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:45:33 2019
 // Update Count     : 9
+// Last Modified On : Fri Mar 17 09:43:03 2017
+// Update Count     : 8
 //
 …
 #include "InitTweak/GenInit.h"             // for genCtorInit
 #include "InitTweak/InitTweak.h"           // for getPointerBase, isAssignment
+#include "Parser/LinkageSpec.h"            // for Cforall
 #include "ResolvExpr/Alternative.h"        // for AltList, Alternative
 #include "ResolvExpr/AlternativeFinder.h"  // for AlternativeFinder, simpleC...
 …
 #include "ResolvExpr/TypeEnvironment.h"    // for TypeEnvironment
 #include "ResolvExpr/typeops.h"            // for combos
-#include "SynTree/LinkageSpec.h"           // for Cforall
 #include "SynTree/Declaration.h"           // for ObjectDecl
 #include "SynTree/Expression.h"            // for Expression, CastExpr, Name...

src/Tuples/TupleExpansion.cc

-              r7030dab
+              r71d6bd8
 // Author           : Rodolfo G. Esteves
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:45:51 2019
 // Update Count     : 24
+// Last Modified By : Andrew Beach
+// Last Modified On : Fri Oct  4 15:38:00 2019
+// Update Count     : 23
 //
 …
 #include "Common/utility.h"       // for CodeLocation
 #include "InitTweak/InitTweak.h"  // for getFunction
 #include "SynTree/LinkageSpec.h"  // for Spec, C, Intrinsic
+#include "Parser/LinkageSpec.h"   // for Spec, C, Intrinsic
 #include "SynTree/Constant.h"     // for Constant
 #include "SynTree/Declaration.h"  // for StructDecl, DeclarationWithType
 …
         const ast::TypeInstType * isTtype( const ast::Type * type ) {
                 if ( const ast::TypeInstType * inst = dynamic_cast< const ast::TypeInstType * >( type ) ) {
                         if ( inst->base && inst->base->kind == ast::TypeDecl::Ttype ) {
+                        if ( inst->base && inst->base->kind == ast::TypeVar::Ttype ) {
                                 return inst;
+                        }

src/cfa.make

-              r7030dab
+              r71d6bd8
 CFACOMPILE = $(CFACC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) $(AM_CFLAGS) $(CFLAGS)
 LTCFACOMPILE = $(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) \
         $(LIBTOOLFLAGS) --mode=compile $(CFACC) $(DEFS) \
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(AM_CFLAGS) $(CFAFLAGS) $(CFLAGS)
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) \
+        $(AM_CFLAGS) $(CFLAGS)
 AM_V_CFA = $(am__v_CFA_@AM_V@)
 …
         $(am__mv) $$depbase.Tpo $$depbase.Plo
+AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
+am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
+am__v_JAVAC_0 = @echo "  JAVAC   " $@;
+am__v_JAVAC_1 =
+AM_V_GOC = $(am__v_GOC_@AM_V@)
+am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
+am__v_GOC_0 = @echo "  GOC     " $@;
+am__v_GOC_1 =
 UPPCC = u++
 UPPCOMPILE = $(UPPCC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_UPPFLAGS) $(UPPFLAGS) $(AM_CXXFLAGS) $(CXXFLAGS) $(AM_CFLAGS) $(CFLAGS)
 …
 am__v_UPP_0 = @echo "  UPP     " $@;
 am__v_UPP_1 =
-AM_V_GOC = $(am__v_GOC_@AM_V@)
-am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
-am__v_GOC_0 = @echo "  GOC     " $@;
-am__v_GOC_1 =
-AM_V_PY = $(am__v_PY_@AM_V@)
-am__v_PY_ = $(am__v_PY_@AM_DEFAULT_V@)
-am__v_PY_0 = @echo "  PYTHON  " $@;
-am__v_PY_1 =
-AM_V_RUST = $(am__v_RUST_@AM_V@)
-am__v_RUST_ = $(am__v_RUST_@AM_DEFAULT_V@)
-am__v_RUST_0 = @echo "  RUST    " $@;
-am__v_RUST_1 =
-AM_V_NODEJS = $(am__v_NODEJS_@AM_V@)
-am__v_NODEJS_ = $(am__v_NODEJS_@AM_DEFAULT_V@)
-am__v_NODEJS_0 = @echo "  NODEJS  " $@;
-am__v_NODEJS_1 =
-AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
-am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
-am__v_JAVAC_0 = @echo "  JAVAC   " $@;
-am__v_JAVAC_1 =

src/main.cc

-              r7030dab
+              r71d6bd8
 // Created On       : Fri May 15 23:12:02 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb  8 08:33:50 2020
 // Update Count     : 633
+// Last Modified On : Fri Aug 23 06:50:08 2019
+// Update Count     : 607
 //
 …
 #include <cstdio>                           // for fopen, FILE, fclose, stdin
 #include <cstdlib>                          // for exit, free, abort, EXIT_F...
 #include <csignal>                          // for signal, SIGABRT, SIGSEGV
+#include <csignal>                         // for signal, SIGABRT, SIGSEGV
 #include <cstring>                          // for index
 #include <fstream>                          // for ofstream
 …
 #include <list>                             // for list
 #include <string>                           // for char_traits, operator<<
-using namespace std;
 #include "AST/Convert.hpp"
 …
 #include "InitTweak/GenInit.h"              // for genInit
 #include "MakeLibCfa.h"                     // for makeLibCfa
+#include "Parser/LinkageSpec.h"             // for Spec, Cforall, Intrinsic
 #include "Parser/ParseNode.h"               // for DeclarationNode, buildList
 #include "Parser/TypedefTable.h"            // for TypedefTable
 …
 #include "ResolvExpr/Resolver.h"            // for resolve
 #include "SymTab/Validate.h"                // for validate
-#include "SynTree/LinkageSpec.h"            // for Spec, Cforall, Intrinsic
 #include "SynTree/Declaration.h"            // for Declaration
 #include "SynTree/Visitor.h"                // for acceptAll
 …
 #include "Virtual/ExpandCasts.h"            // for expandCasts
+using namespace std;
 static void NewPass( const char * const name ) {
 …
 static bool waiting_for_gdb = false;                                    // flag to set cfa-cpp to wait for gdb on start
 static string PreludeDirector = "";
+static std::string PreludeDirector = "";
 static void parse_cmdline( int argc, char *argv[] );
 …
 static void backtrace( int start ) {                                    // skip first N stack frames
         enum { Frames = 50, };                                                          // maximum number of stack frames
+        enum { Frames = 50 };
         void * array[Frames];
         size_t size = ::backtrace( array, Frames );
+        int size = ::backtrace( array, Frames );
         char ** messages = ::backtrace_symbols( array, size ); // does not demangle names
 …
         // skip last 2 stack frames after main
         for ( unsigned int i = start; i < size - 2 && messages != nullptr; i += 1 ) {
+        for ( int i = start; i < size - 2 && messages != nullptr; i += 1 ) {
                 char * mangled_name = nullptr, * offset_begin = nullptr, * offset_end = nullptr;
+                for ( char * p = messages[i]; *p; p += 1 ) {    // find parantheses and +offset
+                for ( char *p = messages[i]; *p; ++p ) {        // find parantheses and +offset
                         if ( *p == '(' ) {
                                 mangled_name = p;
 …
 } // backtrace
+#define SIGPARMS int sig __attribute__(( unused )), siginfo_t * sfp __attribute__(( unused )), ucontext_t * cxt __attribute__(( unused ))
+static void Signal( int sig, void (*handler)(SIGPARMS), int flags ) {
+        struct sigaction act;
+        act.sa_sigaction = (void (*)(int, siginfo_t *, void *))handler;
+        act.sa_flags = flags;
+        if ( sigaction( sig, &act, nullptr ) == -1 ) {
+            cerr << "*CFA runtime error* problem installing signal handler, error(" << errno << ") " << strerror( errno ) << endl;
+            _exit( EXIT_FAILURE );
+        } // if
+} // Signal
+static void sigSegvBusHandler( SIGPARMS ) {
+        if ( sfp->si_addr == nullptr ) {
+                cerr << "Null pointer (nullptr) dereference." << endl;
+        } else {
+                cerr << (sig == SIGSEGV ? "Segment fault" : "Bus error") << " at memory location " << sfp->si_addr << "." << endl
+                         << "Possible cause is reading outside the address space or writing to a protected area within the address space with an invalid pointer or subscript." << endl;
+        } // if
+static void sigSegvBusHandler( int sig_num ) {
+        cerr << "*CFA runtime error* program cfa-cpp terminated with "
+                 <<     (sig_num == SIGSEGV ? "segment fault" : "bus error")
+                 << "." << endl;
         backtrace( 2 );                                                                         // skip first 2 stack frames
+        //_exit( EXIT_FAILURE );
         abort();                                                                                        // cause core dump for debugging
 } // sigSegvBusHandler
+static void sigFpeHandler( SIGPARMS ) {
+        const char * msg;
+        switch ( sfp->si_code ) {
+          case FPE_INTDIV: case FPE_FLTDIV: msg = "divide by zero"; break;
+          case FPE_FLTOVF: msg = "overflow"; break;
+          case FPE_FLTUND: msg = "underflow"; break;
+          case FPE_FLTRES: msg = "inexact result"; break;
+          case FPE_FLTINV: msg = "invalid operation"; break;
+          default: msg = "unknown";
+        } // choose
+        cerr << "Computation error " << msg << " at location " << sfp->si_addr << endl
+                 << "Possible cause is constant-expression evaluation invalid." << endl;
+        backtrace( 2 );                                                                         // skip first 2 stack frames
+        abort();                                                                                        // cause core dump for debugging
+} // sigFpeHandler
+static void sigAbortHandler( SIGPARMS ) {
+static void sigAbortHandler( __attribute__((unused)) int sig_num ) {
         backtrace( 6 );                                                                         // skip first 6 stack frames
         Signal( SIGABRT, (void (*)(SIGPARMS))SIG_DFL, SA_SIGINFO );     // reset default signal handler
+        signal( SIGABRT, SIG_DFL);                                                      // reset default signal handler
         raise( SIGABRT );                                                                       // reraise SIGABRT
 } // sigAbortHandler
 …
         list< Declaration * > translationUnit;
+        Signal( SIGSEGV, sigSegvBusHandler, SA_SIGINFO );
+        Signal( SIGBUS, sigSegvBusHandler, SA_SIGINFO );
+        Signal( SIGFPE, sigFpeHandler, SA_SIGINFO );
+        Signal( SIGABRT, sigAbortHandler, SA_SIGINFO );
+        // cout << "main" << endl;
+        signal( SIGSEGV, sigSegvBusHandler );
+        signal( SIGBUS, sigSegvBusHandler );
+        signal( SIGABRT, sigAbortHandler );
+        // std::cout << "main" << std::endl;
         // for ( int i = 0; i < argc; i += 1 ) {
         //      cout << '\t' << argv[i] << endl;
+        //      std::cout << '\t' << argv[i] << std::endl;
         // } // for
 …
         if ( waiting_for_gdb ) {
                 cerr << "Waiting for gdb" << endl;
                 cerr << "run :" << endl;
                 cerr << "  gdb attach " << getpid() << endl;
+                std::cerr << "Waiting for gdb" << std::endl;
+                std::cerr << "run :" << std::endl;
+                std::cerr << "  gdb attach " << getpid() << std::endl;
                 raise(SIGSTOP);
         } // if
 …
                 return EXIT_FAILURE;
         } catch ( ... ) {
                 exception_ptr eptr = current_exception();
+                std::exception_ptr eptr = std::current_exception();
                 try {
                         if (eptr) {
                                 rethrow_exception(eptr);
+                                std::rethrow_exception(eptr);
                         } else {
                                 cerr << "Exception Uncaught and Unknown" << endl;
                         } // if
                 } catch(const exception& e) {
                         cerr << "Uncaught Exception \"" << e.what() << "\"\n";
+                                std::cerr << "Exception Uncaught and Unknown" << std::endl;
+                        } // if
+                } catch(const std::exception& e) {
+                        std::cerr << "Uncaught Exception \"" << e.what() << "\"\n";
                 } // try
                 return EXIT_FAILURE;
 …
 static const char optstring[] = ":c:ghlLmNnpP:S:twW:D:";
+static const char optstring[] = ":hlLmNnpP:S:tgwW:D:";
 enum { PreludeDir = 128 };
 static struct option long_opts[] = {
-        { "colors", required_argument, nullptr, 'c' },
-        { "gdb", no_argument, nullptr, 'g' },
         { "help", no_argument, nullptr, 'h' },
         { "libcfa", no_argument, nullptr, 'l' },
 …
         { "statistics", required_argument, nullptr, 'S' },
         { "tree", no_argument, nullptr, 't' },
+        { "gdb", no_argument, nullptr, 'g' },
         { "", no_argument, nullptr, 0 },                                        // -w
         { "", no_argument, nullptr, 0 },                                        // -W
 …
 static const char * description[] = {
+        "diagnostic color: never, always, or auto.",          // -c
+        "wait for gdb to attach",                             // -g
+        "print help message",                                 // -h
+        "generate libcfa.c",                                  // -l
+        "generate line marks",                                // -L
+        "do not replace main",                                // -m
+        "do not generate line marks",                         // -N
+        "do not read prelude",                                // -n
+        "print help message",                                                           // -h
+        "generate libcfa.c",                                                            // -l
+        "generate line marks",                                                          // -L
+        "do not replace main",                                                          // -m
+        "do not generate line marks",                                           // -N
+        "do not read prelude",                                                          // -n
         "generate prototypes for prelude functions",            // -p
         "print",                                              // -P
+        "print",                                                                                        // -P
         "<directory> prelude directory for debug/nodebug",      // no flag
         "<option-list> enable profiling information:\n          counters,heap,time,all,none", // -S
+        "building cfa standard lib",                          // -t
+        "",                                                   // -w
+        "",                                                   // -W
+        "",                                                   // -D
+        "building cfa standard lib",                                                                    // -t
+        "wait for gdb to attach",                                                                       // -g
+        "",                                                                                                     // -w
+        "",                                                                                                     // -W
+        "",                                                                                                     // -D
 }; // description
 …
         while ( (c = getopt_long( argc, argv, optstring, long_opts, nullptr )) != -1 ) {
                 switch ( c ) {
-                  case 'c':                                                                             // diagnostic colors
-                        if ( strcmp( optarg, "always" ) == 0 ) {
-                                ErrorHelpers::colors = ErrorHelpers::Colors::Always;
-                        } else if ( strcmp( optarg, "never" ) == 0 ) {
-                                ErrorHelpers::colors = ErrorHelpers::Colors::Never;
-                        } else if ( strcmp( optarg, "auto" ) == 0 ) {
-                                ErrorHelpers::colors = ErrorHelpers::Colors::Auto;
-                        } // if
-                        break;
                   case 'h':                                                                             // help message
                         usage( argv );                                                          // no return

tests/.expect/alloc-ERROR.txt

-              r7030dab
+              r71d6bd8
 alloc.cfa:362:1 error: No reasonable alternatives for expression Applying untyped:
+alloc.cfa:311:1 error: No reasonable alternatives for expression Applying untyped:
   Name: ?=?
 ...to:
   Name: ip
+  Name: p
   Applying untyped:
     Name: realloc
 …
 alloc.cfa:363:1 error: No reasonable alternatives for expression Applying untyped:
+alloc.cfa:312:1 error: No reasonable alternatives for expression Applying untyped:
   Name: ?=?
 ...to:
+  Name: ip
+  Name: p
+  Applying untyped:
+    Name: alloc
+  ...to:
+    Name: stp
+    Applying untyped:
+      Name: ?*?
+    ...to:
+      Name: dim
+      Sizeof Expression on: Applying untyped:
+          Name: *?
+        ...to:
+          Name: stp
+alloc.cfa:313:1 error: No reasonable alternatives for expression Applying untyped:
+  Name: ?=?
+...to:
+  Name: p
   Applying untyped:
     Name: memset
 …
 alloc.cfa:364:1 error: No reasonable alternatives for expression Applying untyped:
+alloc.cfa:314:1 error: No reasonable alternatives for expression Applying untyped:
   Name: ?=?
 ...to:
   Name: ip
+  Name: p
   Applying untyped:
     Name: memcpy

tests/.expect/alloc.txt

-              r7030dab
+              r71d6bd8
 CFA malloc 0xdeadbeef
 CFA alloc 0xdeadbeef
+CFA array alloc, fill 0xde
 CFA alloc, fill dededede
 CFA alloc, fill 3
 …
 xefefefef 0xefefefef 0xefefefef 0xefefefef 0xefefefef 0xefefefef 0xefefefef 0xefefefef 0xefefefef 0xefefefef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101
 CFA realloc array alloc
+CFA resize array alloc
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef
 CFA realloc array alloc
+CFA resize array alloc
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101
 CFA realloc array alloc
+CFA resize array alloc
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef
 CFA realloc array alloc, fill
+CFA resize array alloc, fill
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
 CFA realloc array alloc, fill
+CFA resize array alloc, fill
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef
+CFA realloc array alloc, fill
+xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
+CFA realloc array alloc, 5
+xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
+CFA realloc array alloc, 5
+xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef
+CFA realloc array alloc, 5
+xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff 0xffffffff
+CFA resize array alloc, fill
+xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
 C   memalign 42 42.5

tests/.expect/loopctrl.txt

-              r7030dab
+              r71d6bd8
+A
 A A A A A A A A A A
-A A A A A A A A A A A
 B B B B B
 C C C C C
 …
 1 2 3 4 5 6 7 8 9
-1 2 3 4 5 6 7 8 9 10
 3 5 7 9
 8 6 4 2
 …
 N N N N N N N N N N
 1 2 3 4 5 6 7 8 9
-1 2 3 4 5 6 7 8 9 10
 9 8 7 6 5 4 3 2 1

tests/.expect/nested-types-ERR1.txt

r7030dab	r71d6bd8
1		nested-types.cfa:83:1 error: Use of undefined type T
	1	nested-types.cfa:70:1 error: Use of undefined type T

tests/.expect/nested-types-ERR2.txt

-              r7030dab
+              r71d6bd8
 nested-types.cfa:86:1 error: Use of undefined global type Z
 nested-types.cfa:87:1 error: Qualified type requires an aggregate on the left, but has: signed int
 nested-types.cfa:88:1 error: Undefined type in qualified type: Qualified Type:
+nested-types.cfa:73:1 error: Use of undefined global type Z
+nested-types.cfa:74:1 error: Qualified type requires an aggregate on the left, but has: signed int
+nested-types.cfa:75:1 error: Undefined type in qualified type: Qualified Type:
   instance of struct S with body 1
   instance of type Z (not function type)

tests/.expect/rational.txt

r7030dab	r71d6bd8
1	1	constructor
2		3/1 4/1 0/1 ~~0/1 1/1~~
	2	3/1 4/1 0/1
3	3	1/2 5/7
4	4	2/3 -3/2

tests/.expect/references.txt

r7030dab	r71d6bd8
36	36	3
37	37	3 9 { 1., 7. }, [1, 2, 3]
38		4
39	38	Destructing a Y
40	39	Destructing a Y

tests/.expect/time.txt

r7030dab	r71d6bd8
18	18	Dividing that by 2 gives 2403.5 seconds
19	19	4807 seconds is 1 hours, 20 minutes, 7 seconds
20		~~2020 Jan 5 14:01:40 (GMT)~~
21		~~1970 Jan 5 14:00:00 (GMT)~~
22		~~1973 Jan 2 06:59:00 (GMT)~~

tests/Makefile.am

-              r7030dab
+              r71d6bd8
 # adjust CC to current flags
 CC = $(if $(DISTCC_CFA_PATH),distcc $(DISTCC_CFA_PATH) ${ARCH_FLAGS},$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
+CC = $(if $(DISTCC_CFA_PATH),distcc $(DISTCC_CFA_PATH),$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
 CFACC = $(CC)
 …
 # adjusted CC but without the actual distcc call
 CFACCLOCAL = $(if $(DISTCC_CFA_PATH),$(DISTCC_CFA_PATH) ${ARCH_FLAGS},$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
+CFACCLOCAL = $(if $(DISTCC_CFA_PATH),$(DISTCC_CFA_PATH),$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
 PRETTY_PATH=mkdir -p $(dir $(abspath ${@})) && cd ${srcdir} &&

tests/Makefile.in

-              r7030dab
+              r71d6bd8
 # adjust CC to current flags
 CC = $(if $(DISTCC_CFA_PATH),distcc $(DISTCC_CFA_PATH) ${ARCH_FLAGS},$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
+CC = $(if $(DISTCC_CFA_PATH),distcc $(DISTCC_CFA_PATH),$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
 CCAS = @CCAS@
 CCASDEPMODE = @CCASDEPMODE@
 …
 LTCFACOMPILE = $(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) \
         $(LIBTOOLFLAGS) --mode=compile $(CFACC) $(DEFS) \
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(AM_CFLAGS) $(CFAFLAGS) $(CFLAGS)
+        $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) \
+        $(AM_CFLAGS) $(CFLAGS)
 AM_V_CFA = $(am__v_CFA_@AM_V@)
 …
 am__v_CFA_0 = @echo "  CFA     " $@;
 am__v_CFA_1 =
+AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
+am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
+am__v_JAVAC_0 = @echo "  JAVAC   " $@;
+am__v_JAVAC_1 =
+AM_V_GOC = $(am__v_GOC_@AM_V@)
+am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
+am__v_GOC_0 = @echo "  GOC     " $@;
+am__v_GOC_1 =
 UPPCC = u++
 UPPCOMPILE = $(UPPCC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_UPPFLAGS) $(UPPFLAGS) $(AM_CXXFLAGS) $(CXXFLAGS) $(AM_CFLAGS) $(CFLAGS)
 …
 am__v_UPP_0 = @echo "  UPP     " $@;
 am__v_UPP_1 =
-AM_V_GOC = $(am__v_GOC_@AM_V@)
-am__v_GOC_ = $(am__v_GOC_@AM_DEFAULT_V@)
-am__v_GOC_0 = @echo "  GOC     " $@;
-am__v_GOC_1 =
-AM_V_PY = $(am__v_PY_@AM_V@)
-am__v_PY_ = $(am__v_PY_@AM_DEFAULT_V@)
-am__v_PY_0 = @echo "  PYTHON  " $@;
-am__v_PY_1 =
-AM_V_RUST = $(am__v_RUST_@AM_V@)
-am__v_RUST_ = $(am__v_RUST_@AM_DEFAULT_V@)
-am__v_RUST_0 = @echo "  RUST    " $@;
-am__v_RUST_1 =
-AM_V_NODEJS = $(am__v_NODEJS_@AM_V@)
-am__v_NODEJS_ = $(am__v_NODEJS_@AM_DEFAULT_V@)
-am__v_NODEJS_0 = @echo "  NODEJS  " $@;
-am__v_NODEJS_1 =
-AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
-am__v_JAVAC_ = $(am__v_JAVAC_@AM_DEFAULT_V@)
-am__v_JAVAC_0 = @echo "  JAVAC   " $@;
-am__v_JAVAC_1 =
 debug = yes
 installed = no
 …
 # adjusted CC but without the actual distcc call
 CFACCLOCAL = $(if $(DISTCC_CFA_PATH),$(DISTCC_CFA_PATH) ${ARCH_FLAGS},$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
+CFACCLOCAL = $(if $(DISTCC_CFA_PATH),$(DISTCC_CFA_PATH),$(TARGET_CFA) ${DEBUG_FLAGS} ${ARCH_FLAGS})
 PRETTY_PATH = mkdir -p $(dir $(abspath ${@})) && cd ${srcdir} &&
 avl_test_SOURCES = avltree/avl_test.cfa avltree/avl0.cfa avltree/avl1.cfa avltree/avl2.cfa avltree/avl3.cfa avltree/avl4.cfa avltree/avl-private.cfa

tests/alloc.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Feb  3 07:56:22 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Apr  1 10:58:35 2020
 // Update Count     : 424
+// Last Modified On : Sun Oct 20 21:45:21 2019
+// Update Count     : 391
 //
 …
         size_t dim = 10;
         char fill = '\xde';
         int * ip, * ip1;
+        int * p, * p1;
         // allocation, non-array types
+        ip = (int *)malloc( sizeof(*ip) );                                      // C malloc, type unsafe
+        *ip = 0xdeadbeef;
+        printf( "C   malloc %#x\n", *ip );
+        free( ip );
+        ip = malloc();                                                                          // CFA malloc, type safe
+        *ip = 0xdeadbeef;
+        printf( "CFA malloc %#x\n", *ip );
+        free( ip );
+        ip = alloc();                                                                           // CFA alloc, type safe
+        *ip = 0xdeadbeef;
+        printf( "CFA alloc %#x\n", *ip );
+        free( ip );
+        ip = alloc_set( fill );                                                         // CFA alloc, fill
+        printf( "CFA alloc, fill %08x\n", *ip );
+        free( ip );
+        ip = alloc_set( 3 );                                                            // CFA alloc, fill
+        printf( "CFA alloc, fill %d\n", *ip );
+        free( ip );
+        p = (int *)malloc( sizeof(*p) );                                        // C malloc, type unsafe
+        *p = 0xdeadbeef;
+        printf( "C   malloc %#x\n", *p );
+        free( p );
+        p = malloc();                                       // CFA malloc, type safe
+        *p = 0xdeadbeef;
+        printf( "CFA malloc %#x\n", *p );
+        free( p );
+        p = alloc();                                        // CFA alloc, type safe
+        *p = 0xdeadbeef;
+        printf( "CFA alloc %#x\n", *p );
+        free( p );
+        p = alloc_set( fill );                                                          // CFA alloc, fill
+        printf( "CFA array alloc, fill %#hhx\n", fill );
+        printf( "CFA alloc, fill %08x\n", *p );
+        free( p );
+        p = alloc_set( 3 );                                                                     // CFA alloc, fill
+        printf( "CFA alloc, fill %d\n", *p );
+        free( p );
 …
         printf( "\n" );
         ip = (int *)calloc( dim, sizeof( *ip ) );                       // C array calloc, type unsafe
+        p = (int *)calloc( dim, sizeof( *p ) );                         // C array calloc, type unsafe
         printf( "C   array calloc, fill 0\n" );
         for ( i; dim ) { printf( "%#x ", ip[i] ); }
         printf( "\n" );
         free( ip );
         ip = calloc( dim );                                                                     // CFA array calloc, type safe
+        for ( i; dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        free( p );
+        p = calloc( dim );                                  // CFA array calloc, type safe
         printf( "CFA array calloc, fill 0\n" );
         for ( i; dim ) { printf( "%#x ", ip[i] ); }
         printf( "\n" );
         free( ip );
         ip = alloc( dim );                                                                      // CFA array alloc, type safe
         for ( i; dim ) { ip[i] = 0xdeadbeef; }
+        for ( i; dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        free( p );
+        p = alloc( dim );                                   // CFA array alloc, type safe
+        for ( i; dim ) { p[i] = 0xdeadbeef; }
         printf( "CFA array alloc, no fill\n" );
         for ( i; dim ) { printf( "%#x ", ip[i] ); }
         printf( "\n" );
         free( ip );
         ip = alloc_set( 2 * dim, fill );                                        // CFA array alloc, fill
+        for ( i; dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        free( p );
+        p = alloc_set( 2 * dim, fill );                                         // CFA array alloc, fill
         printf( "CFA array alloc, fill %#hhx\n", fill );
         for ( i; 2 * dim ) { printf( "%#x ", ip[i] ); }
         printf( "\n" );
         free( ip );
         ip = alloc_set( 2 * dim, 0xdeadbeef );                          // CFA array alloc, fill
+        for ( i; 2 * dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        free( p );
+        p = alloc_set( 2 * dim, 0xdeadbeef );                           // CFA array alloc, fill
         printf( "CFA array alloc, fill %#hhx\n", 0xdeadbeef );
         for ( i; 2 * dim ) { printf( "%#x ", ip[i] ); }
         printf( "\n" );
         // do not free
         ip1 = alloc_set( 2 * dim, ip );                                         // CFA array alloc, fill
+        for ( i; 2 * dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
+        p1 = alloc_set( 2 * dim, p );                                           // CFA array alloc, fill
         printf( "CFA array alloc, fill from array\n" );
         for ( i; 2 * dim ) { printf( "%#x %#x, ", ip[i], ip1[i] ); }
         free( ip1 );
         printf( "\n" );
         // realloc, non-array types
         printf( "\n" );
         ip = (int *)realloc( ip, dim * sizeof(*ip) );           // C realloc
+        for ( i; 2 * dim ) { printf( "%#x %#x, ", p[i], p1[i] ); }
+        free( p1 );
+        printf( "\n" );
+        // resize, non-array types
+        printf( "\n" );
+        p = (int *)realloc( p, dim * sizeof(*p) );                      // C realloc
         printf( "C realloc\n" );
         for ( i; dim ) { printf( "%#x ", ip[i] ); }
         printf( "\n" );
         // do not free
         ip = realloc( ip, 2 * dim * sizeof(*ip) );                      // CFA realloc
         for ( i; dim ~ 2 * dim ) { ip[i] = 0x1010101; }
+        for ( i; dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
+        p = realloc( p, 2 * dim * sizeof(*p) );             // CFA realloc
+        for ( i; dim ~ 2 * dim ) { p[i] = 0x1010101; }
         printf( "CFA realloc\n" );
+        for ( i; 2 * dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        // realloc, array types
+        printf( "\n" );
+        ip = alloc( ip, dim );                                                          // CFA realloc array alloc
+        for ( i; dim ) { ip[i] = 0xdeadbeef; }
+        printf( "CFA realloc array alloc\n" );
+        for ( i; dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        ip = alloc( ip, 2 * dim );                                                      // CFA realloc array alloc
+        for ( i; dim ~ 2 * dim ) { ip[i] = 0x1010101; }         // fill upper part
+        printf( "CFA realloc array alloc\n" );
+        for ( i; 2 * dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        ip = alloc( ip, dim );                                                          // CFA realloc array alloc
+        printf( "CFA realloc array alloc\n" );
+        for ( i; dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        ip = alloc_set( ip, 3 * dim, fill );                            // CFA realloc array alloc, fill
+        printf( "CFA realloc array alloc, fill\n" );
+        for ( i; 3 * dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        ip = alloc_set( ip, dim, fill );                                        // CFA realloc array alloc, fill
+        printf( "CFA realloc array alloc, fill\n" );
+        for ( i; dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        ip = alloc_set( ip, 3 * dim, fill );                            // CFA realloc array alloc, fill
+        printf( "CFA realloc array alloc, fill\n" );
+        for ( i; 3 * dim ) { printf( "%#x ", ip[i] );; }
+        printf( "\n" );
+        // do not free
+        ip = alloc_set( ip, 3 * dim, 5 );                                       // CFA realloc array alloc, 5
+        printf( "CFA realloc array alloc, 5\n" );
+        for ( i; 3 * dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        ip = alloc_set( ip, dim, 5 );                                           // CFA realloc array alloc, 5
+        printf( "CFA realloc array alloc, 5\n" );
+        for ( i; dim ) { printf( "%#x ", ip[i] ); }
+        printf( "\n" );
+        // do not free
+        ip = alloc_set( ip, 3 * dim, 5 );                                       // CFA realloc array alloc, 5
+        printf( "CFA realloc array alloc, 5\n" );
+        for ( i; 3 * dim ) { printf( "%#x ", ip[i] );; }
+        printf( "\n" );
+        free( ip );
+        // resize, non-array types
+        struct S {
+                int a[5];
+        };
+    ip = alloc();
+        *ip = 5;
+    double * dp = alloc( ip );
+        *dp = 5.5;
+    S * sp = alloc( dp );
+        *sp = (S){ {0, 1, 2, 3, 4} };
+    ip = alloc( sp );
+        *ip = 3;
+    free( ip );
+        for ( i; 2 * dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
         // resize, array types
+    ip = alloc( 5 );
+        for ( i; 5 ) { ip[i] = 5; }
+    dp = alloc( ip, 5 );
+        for ( i; 5 ) { dp[i] = 5.5; }
+    sp = alloc( dp, 5 );
+        for ( i; 5 ) { sp[i] = (S){ {0, 1, 2, 3, 4} }; }
+    ip = alloc( sp, 3 );
+        for ( i; 3 ) { ip[i] = 3; }
+    ip = alloc( ip, 7 );
+        for ( i; 7 ) { ip[i] = 7; }
+    ip = alloc( ip, 7, false );
+        for ( i; 7 ) { ip[i] = 7; }
+    free( ip );
+        printf( "\n" );
+        p = alloc( p, dim );                                // CFA resize array alloc
+        for ( i; dim ) { p[i] = 0xdeadbeef; }
+        printf( "CFA resize array alloc\n" );
+        for ( i; dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
+        p = alloc( p, 2 * dim );                            // CFA resize array alloc
+        for ( i; dim ~ 2 * dim ) { p[i] = 0x1010101; }
+        printf( "CFA resize array alloc\n" );
+        for ( i; 2 * dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
+        p = alloc( p, dim );                                // CFA resize array alloc
+        printf( "CFA resize array alloc\n" );
+        for ( i; dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
+        p = alloc_set( p, 3 * dim, fill );                                      // CFA resize array alloc, fill
+        printf( "CFA resize array alloc, fill\n" );
+        for ( i; 3 * dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
+        p = alloc_set( p, dim, fill );                                          // CFA resize array alloc, fill
+        printf( "CFA resize array alloc, fill\n" );
+        for ( i; dim ) { printf( "%#x ", p[i] ); }
+        printf( "\n" );
+        // do not free
+        p = alloc_set( p, 3 * dim, fill );                                      // CFA resize array alloc, fill
+        printf( "CFA resize array alloc, fill\n" );
+        for ( i; 3 * dim ) { printf( "%#x ", p[i] );; }
+        printf( "\n" );
+        free( p );
 …
         free( stp );
         stp = &(*memalign( Alignment )){ 42, 42.5 };            // CFA memalign
+        stp = &(*memalign( Alignment )){ 42, 42.5 };          // CFA memalign
         assert( (uintptr_t)stp % Alignment == 0 );
         printf( "CFA memalign %d %g\n", stp->x, stp->y );
 …
         free( fp - 1 );
         ip = foo( bar( baz( malloc(), 0 ), 0 ), 0 );
         *ip = 0xdeadbeef;
         printf( "CFA deep malloc %#x\n", *ip );
         free( ip );
+        p = foo( bar( baz( malloc(), 0 ), 0 ), 0 );
+        *p = 0xdeadbeef;
+        printf( "CFA deep malloc %#x\n", *p );
+        free( p );
 #ifdef ERR1
         stp = malloc();
         printf( "\nSHOULD FAIL\n" );
+        ip = realloc( stp, dim * sizeof( *stp ) );
+        ip = memset( stp, 10 );
+        ip = memcpy( &st1, &st );
+        p = realloc( stp, dim * sizeof( *stp ) );
+        p = alloc( stp, dim * sizeof( *stp ) );
+        p = memset( stp, 10 );
+        p = memcpy( &st1, &st );
 #endif
 } // main

tests/builtins/sync.cfa

-              r7030dab
+              r71d6bd8
 void foo() {
         volatile _Bool * vpB = 0; _Bool vB = 0;
+        volatile char * vpc = 0; char * rpc = 0; char vc = 0;
+        volatile short * vps = 0; short * rps = 0; short vs = 0;
+        volatile int * vpi = 0; int * rpi = 0; int vi = 0;
+        volatile long int * vpl = 0; long int * rpl = 0; long int vl = 0;
+        volatile long long int * vpll = 0; long long int * rpll = 0; long long int vll = 0;
+        #if defined(__SIZEOF_INT128__)
+        volatile __int128 * vplll = 0; __int128 * rplll = 0; __int128 vlll = 0;
+        volatile char * vp1 = 0; char * rp1 = 0; char v1 = 0;
+        volatile short * vp2 = 0; short * rp2 = 0; short v2 = 0;
+        volatile int * vp4 = 0; int * rp4 = 0; int v4 = 0;
+        volatile long long int * vp8 = 0; long long int * rp8 = 0; long long int v8 = 0;
+        #if defined(__SIZEOF_INT128__)
+        volatile __int128 * vp16 = 0; __int128 * rp16 = 0; __int128 v16 = 0;
         #endif
         struct type * volatile * vpp = 0; struct type ** rpp = 0; struct type * vp = 0;
+        { char ret; ret = __sync_fetch_and_add(vpc, vc); }
+        { short ret; ret = __sync_fetch_and_add(vps, vs); }
+        { int ret; ret = __sync_fetch_and_add(vpi, vi); }
+        { long int ret; ret = __sync_fetch_and_add(vpl, vl); }
+        { long long int ret; ret = __sync_fetch_and_add(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_add(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_fetch_and_sub(vpc, vc); }
+        { short ret; ret = __sync_fetch_and_sub(vps, vs); }
+        { int ret; ret = __sync_fetch_and_sub(vpi, vi); }
+        { long int ret; ret = __sync_fetch_and_sub(vpl, vl); }
+        { long long int ret; ret = __sync_fetch_and_sub(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_sub(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_fetch_and_or(vpc, vc); }
+        { short ret; ret = __sync_fetch_and_or(vps, vs); }
+        { int ret; ret = __sync_fetch_and_or(vpi, vi); }
+        { long int ret; ret = __sync_fetch_and_or(vpl, vl); }
+        { long long int ret; ret = __sync_fetch_and_or(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_or(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_fetch_and_and(vpc, vc); }
+        { short ret; ret = __sync_fetch_and_and(vps, vs); }
+        { int ret; ret = __sync_fetch_and_and(vpi, vi); }
+        { long int ret; ret = __sync_fetch_and_and(vpl, vl); }
+        { long long int ret; ret = __sync_fetch_and_and(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_and(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_fetch_and_xor(vpc, vc); }
+        { short ret; ret = __sync_fetch_and_xor(vps, vs); }
+        { int ret; ret = __sync_fetch_and_xor(vpi, vi); }
+        { long int ret; ret = __sync_fetch_and_xor(vpl, vl); }
+        { long long int ret; ret = __sync_fetch_and_xor(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_xor(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_fetch_and_nand(vpc, vc); }
+        { short ret; ret = __sync_fetch_and_nand(vps, vs); }
+        { int ret; ret = __sync_fetch_and_nand(vpi, vi); }
+        { long int ret; ret = __sync_fetch_and_nand(vpl, vl); }
+        { long long int ret; ret = __sync_fetch_and_nand(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_nand(vplll, vlll); }
+        { __int128 ret; ret = __sync_fetch_and_nand_16(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_add_and_fetch(vpc, vc); }
+        { short ret; ret = __sync_add_and_fetch(vps, vs); }
+        { int ret; ret = __sync_add_and_fetch(vpi, vi); }
+        { long int ret; ret = __sync_add_and_fetch(vpl, vl); }
+        { long long int ret; ret = __sync_add_and_fetch(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_add_and_fetch(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_sub_and_fetch(vpc, vc); }
+        { short ret; ret = __sync_sub_and_fetch(vps, vs); }
+        { int ret; ret = __sync_sub_and_fetch(vpi, vi); }
+        { long int ret; ret = __sync_sub_and_fetch(vpl, vl); }
+        { long long int ret; ret = __sync_sub_and_fetch(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_sub_and_fetch(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_or_and_fetch(vpc, vc); }
+        { short ret; ret = __sync_or_and_fetch(vps, vs); }
+        { int ret; ret = __sync_or_and_fetch(vpi, vi); }
+        { long int ret; ret = __sync_or_and_fetch(vpl, vl); }
+        { long long int ret; ret = __sync_or_and_fetch(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_or_and_fetch(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_and_and_fetch(vpc, vc); }
+        { short ret; ret = __sync_and_and_fetch(vps, vs); }
+        { int ret; ret = __sync_and_and_fetch(vpi, vi); }
+        { long int ret; ret = __sync_and_and_fetch(vpl, vl); }
+        { long long int ret; ret = __sync_and_and_fetch(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_and_and_fetch(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_xor_and_fetch(vpc, vc); }
+        { short ret; ret = __sync_xor_and_fetch(vps, vs); }
+        { int ret; ret = __sync_xor_and_fetch(vpi, vi); }
+        { long int ret; ret = __sync_xor_and_fetch(vpl, vl); }
+        { long long int ret; ret = __sync_xor_and_fetch(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_xor_and_fetch(vplll, vlll); }
+        #endif
+        { char ret; ret = __sync_nand_and_fetch(vpc, vc); }
+        { short ret; ret = __sync_nand_and_fetch(vps, vs); }
+        { int ret; ret = __sync_nand_and_fetch(vpi, vi); }
+        { long int ret; ret = __sync_nand_and_fetch(vpl, vl); }
+        { long long int ret; ret = __sync_nand_and_fetch(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_nand_and_fetch(vplll, vlll); }
+        #endif
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vpc, vc, vc); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vps, vs, vs); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vpi, vi, vi); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vpl, vl, vl); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vpll, vll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vplll, vlll, vlll); }
+        { char ret; ret = __sync_fetch_and_add(vp1, v1); }
+        { char ret; ret = __sync_fetch_and_add_1(vp1, v1); }
+        { short ret; ret = __sync_fetch_and_add(vp2, v2); }
+        { short ret; ret = __sync_fetch_and_add_2(vp2, v2); }
+        { int ret; ret = __sync_fetch_and_add(vp4, v4); }
+        { int ret; ret = __sync_fetch_and_add_4(vp4, v4); }
+        { long long int ret; ret = __sync_fetch_and_add(vp8, v8); }
+        { long long int ret; ret = __sync_fetch_and_add_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_add(vp16, v16); }
+        { __int128 ret; ret = __sync_fetch_and_add_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_fetch_and_sub(vp1, v1); }
+        { char ret; ret = __sync_fetch_and_sub_1(vp1, v1); }
+        { short ret; ret = __sync_fetch_and_sub(vp2, v2); }
+        { short ret; ret = __sync_fetch_and_sub_2(vp2, v2); }
+        { int ret; ret = __sync_fetch_and_sub(vp4, v4); }
+        { int ret; ret = __sync_fetch_and_sub_4(vp4, v4); }
+        { long long int ret; ret = __sync_fetch_and_sub(vp8, v8); }
+        { long long int ret; ret = __sync_fetch_and_sub_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_sub(vp16, v16); }
+        { __int128 ret; ret = __sync_fetch_and_sub_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_fetch_and_or(vp1, v1); }
+        { char ret; ret = __sync_fetch_and_or_1(vp1, v1); }
+        { short ret; ret = __sync_fetch_and_or(vp2, v2); }
+        { short ret; ret = __sync_fetch_and_or_2(vp2, v2); }
+        { int ret; ret = __sync_fetch_and_or(vp4, v4); }
+        { int ret; ret = __sync_fetch_and_or_4(vp4, v4); }
+        { long long int ret; ret = __sync_fetch_and_or(vp8, v8); }
+        { long long int ret; ret = __sync_fetch_and_or_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_or(vp16, v16); }
+        { __int128 ret; ret = __sync_fetch_and_or_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_fetch_and_and(vp1, v1); }
+        { char ret; ret = __sync_fetch_and_and_1(vp1, v1); }
+        { short ret; ret = __sync_fetch_and_and(vp2, v2); }
+        { short ret; ret = __sync_fetch_and_and_2(vp2, v2); }
+        { int ret; ret = __sync_fetch_and_and(vp4, v4); }
+        { int ret; ret = __sync_fetch_and_and_4(vp4, v4); }
+        { long long int ret; ret = __sync_fetch_and_and(vp8, v8); }
+        { long long int ret; ret = __sync_fetch_and_and_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_and(vp16, v16); }
+        { __int128 ret; ret = __sync_fetch_and_and_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_fetch_and_xor(vp1, v1); }
+        { char ret; ret = __sync_fetch_and_xor_1(vp1, v1); }
+        { short ret; ret = __sync_fetch_and_xor(vp2, v2); }
+        { short ret; ret = __sync_fetch_and_xor_2(vp2, v2); }
+        { int ret; ret = __sync_fetch_and_xor(vp4, v4); }
+        { int ret; ret = __sync_fetch_and_xor_4(vp4, v4); }
+        { long long int ret; ret = __sync_fetch_and_xor(vp8, v8); }
+        { long long int ret; ret = __sync_fetch_and_xor_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_xor(vp16, v16); }
+        { __int128 ret; ret = __sync_fetch_and_xor_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_fetch_and_nand(vp1, v1); }
+        { char ret; ret = __sync_fetch_and_nand_1(vp1, v1); }
+        { short ret; ret = __sync_fetch_and_nand(vp2, v2); }
+        { short ret; ret = __sync_fetch_and_nand_2(vp2, v2); }
+        { int ret; ret = __sync_fetch_and_nand(vp4, v4); }
+        { int ret; ret = __sync_fetch_and_nand_4(vp4, v4); }
+        { long long int ret; ret = __sync_fetch_and_nand(vp8, v8); }
+        { long long int ret; ret = __sync_fetch_and_nand_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_fetch_and_nand(vp16, v16); }
+        { __int128 ret; ret = __sync_fetch_and_nand_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_add_and_fetch(vp1, v1); }
+        { char ret; ret = __sync_add_and_fetch_1(vp1, v1); }
+        { short ret; ret = __sync_add_and_fetch(vp2, v2); }
+        { short ret; ret = __sync_add_and_fetch_2(vp2, v2); }
+        { int ret; ret = __sync_add_and_fetch(vp4, v4); }
+        { int ret; ret = __sync_add_and_fetch_4(vp4, v4); }
+        { long long int ret; ret = __sync_add_and_fetch(vp8, v8); }
+        { long long int ret; ret = __sync_add_and_fetch_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_add_and_fetch(vp16, v16); }
+        { __int128 ret; ret = __sync_add_and_fetch_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_sub_and_fetch(vp1, v1); }
+        { char ret; ret = __sync_sub_and_fetch_1(vp1, v1); }
+        { short ret; ret = __sync_sub_and_fetch(vp2, v2); }
+        { short ret; ret = __sync_sub_and_fetch_2(vp2, v2); }
+        { int ret; ret = __sync_sub_and_fetch(vp4, v4); }
+        { int ret; ret = __sync_sub_and_fetch_4(vp4, v4); }
+        { long long int ret; ret = __sync_sub_and_fetch(vp8, v8); }
+        { long long int ret; ret = __sync_sub_and_fetch_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_sub_and_fetch(vp16, v16); }
+        { __int128 ret; ret = __sync_sub_and_fetch_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_or_and_fetch(vp1, v1); }
+        { char ret; ret = __sync_or_and_fetch_1(vp1, v1); }
+        { short ret; ret = __sync_or_and_fetch(vp2, v2); }
+        { short ret; ret = __sync_or_and_fetch_2(vp2, v2); }
+        { int ret; ret = __sync_or_and_fetch(vp4, v4); }
+        { int ret; ret = __sync_or_and_fetch_4(vp4, v4); }
+        { long long int ret; ret = __sync_or_and_fetch(vp8, v8); }
+        { long long int ret; ret = __sync_or_and_fetch_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_or_and_fetch(vp16, v16); }
+        { __int128 ret; ret = __sync_or_and_fetch_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_and_and_fetch(vp1, v1); }
+        { char ret; ret = __sync_and_and_fetch_1(vp1, v1); }
+        { short ret; ret = __sync_and_and_fetch(vp2, v2); }
+        { short ret; ret = __sync_and_and_fetch_2(vp2, v2); }
+        { int ret; ret = __sync_and_and_fetch(vp4, v4); }
+        { int ret; ret = __sync_and_and_fetch_4(vp4, v4); }
+        { long long int ret; ret = __sync_and_and_fetch(vp8, v8); }
+        { long long int ret; ret = __sync_and_and_fetch_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_and_and_fetch(vp16, v16); }
+        { __int128 ret; ret = __sync_and_and_fetch_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_xor_and_fetch(vp1, v1); }
+        { char ret; ret = __sync_xor_and_fetch_1(vp1, v1); }
+        { short ret; ret = __sync_xor_and_fetch(vp2, v2); }
+        { short ret; ret = __sync_xor_and_fetch_2(vp2, v2); }
+        { int ret; ret = __sync_xor_and_fetch(vp4, v4); }
+        { int ret; ret = __sync_xor_and_fetch_4(vp4, v4); }
+        { long long int ret; ret = __sync_xor_and_fetch(vp8, v8); }
+        { long long int ret; ret = __sync_xor_and_fetch_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_xor_and_fetch(vp16, v16); }
+        { __int128 ret; ret = __sync_xor_and_fetch_16(vp16, v16); }
+        #endif
+        { char ret; ret = __sync_nand_and_fetch(vp1, v1); }
+        { char ret; ret = __sync_nand_and_fetch_1(vp1, v1); }
+        { short ret; ret = __sync_nand_and_fetch(vp2, v2); }
+        { short ret; ret = __sync_nand_and_fetch_2(vp2, v2); }
+        { int ret; ret = __sync_nand_and_fetch(vp4, v4); }
+        { int ret; ret = __sync_nand_and_fetch_4(vp4, v4); }
+        { long long int ret; ret = __sync_nand_and_fetch(vp8, v8); }
+        { long long int ret; ret = __sync_nand_and_fetch_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_nand_and_fetch(vp16, v16); }
+        { __int128 ret; ret = __sync_nand_and_fetch_16(vp16, v16); }
+        #endif
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vp1, v1, v1); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap_1(vp1, v1, v1); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vp2, v2, v2); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap_2(vp2, v2, v2); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vp4, v4, v4); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap_4(vp4, v4, v4); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vp8, v8, v8); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap_8(vp8, v8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { _Bool ret; ret = __sync_bool_compare_and_swap(vp16, v16, v16); }
+        { _Bool ret; ret = __sync_bool_compare_and_swap_16(vp16, v16,v16); }
         #endif
         { _Bool ret; ret = __sync_bool_compare_and_swap(vpp, vp, vp); }
+        { char ret; ret = __sync_val_compare_and_swap(vpc, vc, vc); }
+        { short ret; ret = __sync_val_compare_and_swap(vps, vs, vs); }
+        { int ret; ret = __sync_val_compare_and_swap(vpi, vi, vi); }
+        { long int ret; ret = __sync_val_compare_and_swap(vpl, vl, vl); }
+        { long long int ret; ret = __sync_val_compare_and_swap(vpll, vll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_val_compare_and_swap(vplll, vlll, vlll); }
+        { char ret; ret = __sync_val_compare_and_swap(vp1, v1, v1); }
+        { char ret; ret = __sync_val_compare_and_swap_1(vp1, v1, v1); }
+        { short ret; ret = __sync_val_compare_and_swap(vp2, v2, v2); }
+        { short ret; ret = __sync_val_compare_and_swap_2(vp2, v2, v2); }
+        { int ret; ret = __sync_val_compare_and_swap(vp4, v4, v4); }
+        { int ret; ret = __sync_val_compare_and_swap_4(vp4, v4, v4); }
+        { long long int ret; ret = __sync_val_compare_and_swap(vp8, v8, v8); }
+        { long long int ret; ret = __sync_val_compare_and_swap_8(vp8, v8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_val_compare_and_swap(vp16, v16, v16); }
+        { __int128 ret; ret = __sync_val_compare_and_swap_16(vp16, v16,v16); }
         #endif
         { struct type * ret; ret = __sync_val_compare_and_swap(vpp, vp, vp); }
+        { char ret; ret = __sync_lock_test_and_set(vpc, vc); }
+        { short ret; ret = __sync_lock_test_and_set(vps, vs); }
+        { int ret; ret = __sync_lock_test_and_set(vpi, vi); }
+        { long int ret; ret = __sync_lock_test_and_set(vpl, vl); }
+        { long long int ret; ret = __sync_lock_test_and_set(vpll, vll); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_lock_test_and_set(vplll, vlll); }
+        #endif
+        { __sync_lock_release(vpc); }
+        { __sync_lock_release(vps); }
+        { __sync_lock_release(vpi); }
+        { __sync_lock_release(vpl); }
+        { __sync_lock_release(vpll); }
+        #if defined(__SIZEOF_INT128__)
+        { __sync_lock_release(vplll); }
+        { char ret; ret = __sync_lock_test_and_set(vp1, v1); }
+        { char ret; ret = __sync_lock_test_and_set_1(vp1, v1); }
+        { short ret; ret = __sync_lock_test_and_set(vp2, v2); }
+        { short ret; ret = __sync_lock_test_and_set_2(vp2, v2); }
+        { int ret; ret = __sync_lock_test_and_set(vp4, v4); }
+        { int ret; ret = __sync_lock_test_and_set_4(vp4, v4); }
+        { long long int ret; ret = __sync_lock_test_and_set(vp8, v8); }
+        { long long int ret; ret = __sync_lock_test_and_set_8(vp8, v8); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __sync_lock_test_and_set(vp16, v16); }
+        { __int128 ret; ret = __sync_lock_test_and_set_16(vp16, v16); }
+        #endif
+        { __sync_lock_release(vp1); }
+        { __sync_lock_release_1(vp1); }
+        { __sync_lock_release(vp2); }
+        { __sync_lock_release_2(vp2); }
+        { __sync_lock_release(vp4); }
+        { __sync_lock_release_4(vp4); }
+        { __sync_lock_release(vp8); }
+        { __sync_lock_release_8(vp8); }
+        #if defined(__SIZEOF_INT128__)
+        { __sync_lock_release(vp16); }
+        { __sync_lock_release_16(vp16); }
         #endif
 …
         { _Bool ret; ret = __atomic_test_and_set(vpB, vB); }
         { _Bool ret; ret = __atomic_test_and_set(vpc, vc); }
+        { _Bool ret; ret = __atomic_test_and_set(vp1, v1); }
         { __atomic_clear(vpB, vB); }
+        { __atomic_clear(vpc, vc); }
+        { char ret; ret = __atomic_exchange_n(vpc, vc, __ATOMIC_SEQ_CST); }
+        { char ret; __atomic_exchange(vpc, &vc, &ret, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_exchange_n(vps, vs, __ATOMIC_SEQ_CST); }
+        { short ret; __atomic_exchange(vps, &vs, &ret, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_exchange_n(vpi, vi, __ATOMIC_SEQ_CST); }
+        { int ret; __atomic_exchange(vpi, &vi, &ret, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_exchange_n(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long int ret; __atomic_exchange(vpl, &vl, &ret, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_exchange_n(vpll, vll, __ATOMIC_SEQ_CST); }
+        { long long int ret; __atomic_exchange(vpll, &vll, &ret, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_exchange_n(vplll, vlll, __ATOMIC_SEQ_CST); }
+        { __int128 ret; __atomic_exchange(vplll, &vlll, &ret, __ATOMIC_SEQ_CST); }
+        { __atomic_clear(vp1, v1); }
+        { char ret; ret = __atomic_exchange_n(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_exchange_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; __atomic_exchange(vp1, &v1, &ret, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_exchange_n(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_exchange_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; __atomic_exchange(vp2, &v2, &ret, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_exchange_n(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_exchange_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; __atomic_exchange(vp4, &v4, &ret, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_exchange_n(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_exchange_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; __atomic_exchange(vp8, &v8, &ret, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_exchange_n(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_exchange_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; __atomic_exchange(vp16, &v16, &ret, __ATOMIC_SEQ_CST); }
         #endif
         { struct type * ret; ret = __atomic_exchange_n(vpp, vp, __ATOMIC_SEQ_CST); }
         { struct type * ret; __atomic_exchange(vpp, &vp, &ret, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_load_n(vpc, __ATOMIC_SEQ_CST); }
+        { char ret; __atomic_load(vpc, &ret, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_load_n(vps, __ATOMIC_SEQ_CST); }
+        { short ret; __atomic_load(vps, &ret, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_load_n(vpi, __ATOMIC_SEQ_CST); }
+        { int ret; __atomic_load(vpi, &ret, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_load_n(vpl, __ATOMIC_SEQ_CST); }
+        { long int ret; __atomic_load(vpl, &ret, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_load_n(vpll, __ATOMIC_SEQ_CST); }
+        { long long int ret; __atomic_load(vpll, &ret, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_load_n(vplll, __ATOMIC_SEQ_CST); }
+        { __int128 ret; __atomic_load(vplll, &ret, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_load_n(vp1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_load_1(vp1, __ATOMIC_SEQ_CST); }
+        { char ret; __atomic_load(vp1, &ret, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_load_n(vp2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_load_2(vp2, __ATOMIC_SEQ_CST); }
+        { short ret; __atomic_load(vp2, &ret, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_load_n(vp4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_load_4(vp4, __ATOMIC_SEQ_CST); }
+        { int ret; __atomic_load(vp4, &ret, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_load_n(vp8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_load_8(vp8, __ATOMIC_SEQ_CST); }
+        { long long int ret; __atomic_load(vp8, &ret, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_load_n(vp16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_load_16(vp16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; __atomic_load(vp16, &ret, __ATOMIC_SEQ_CST); }
         #endif
         { struct type * ret; ret = __atomic_load_n(vpp, __ATOMIC_SEQ_CST); }
         { struct type * ret; __atomic_load(vpp, &ret, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vpc, rpc, vc, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vpc, rpc, &vc, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vps, rps, vs, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vps, rps, &vs, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vpi, rpi, vi, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vpi, rpi, &vi, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vpl, rpl, vl, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vpl, rpl, &vl, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vpll, rpll, vll, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vpll, rpll, &vll, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { _Bool ret; ret = __atomic_compare_exchange_n(vplll, rplll, vlll, 0, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vplll, rplll, &vlll, 0, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vp1, rp1, v1, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_1(vp1, rp1, v1, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vp1, rp1, &v1, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vp2, rp2, v2, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_2(vp2, rp2, v2, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vp2, rp2, &v2, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vp4, rp4, v4, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_4(vp4, rp4, v4, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vp4, rp4, &v4, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_n(vp8, rp8, v8, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_8(vp8, rp8, v8, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vp8, rp8, &v8, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { _Bool ret; ret = __atomic_compare_exchange_n(vp16, rp16, v16, 0, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange_16(vp16, rp16, v16, 0, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { _Bool ret; ret = __atomic_compare_exchange(vp16, rp16, &v16, 0, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
         #endif
         { _Bool ret; ret = __atomic_compare_exchange_n(vpp, rpp, vp, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
         { _Bool ret; ret = __atomic_compare_exchange(vpp, rpp, &vp, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vpc, vc, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vpc, &vc, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vps, vs, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vps, &vs, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vpi, vi, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vpi, &vi, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vpl, vl, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vpl, &vl, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vpll, vll, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vpll, &vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __atomic_store_n(vplll, vlll, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vplll, &vlll, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vp1, v1, __ATOMIC_SEQ_CST); }
+        { __atomic_store_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vp1, &v1, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vp2, v2, __ATOMIC_SEQ_CST); }
+        { __atomic_store_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vp2, &v2, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vp4, v4, __ATOMIC_SEQ_CST); }
+        { __atomic_store_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vp4, &v4, __ATOMIC_SEQ_CST); }
+        { __atomic_store_n(vp8, v8, __ATOMIC_SEQ_CST); }
+        { __atomic_store_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vp8, &v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __atomic_store_n(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __atomic_store_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __atomic_store(vp16, &v16, __ATOMIC_SEQ_CST); }
         #endif
         { __atomic_store_n(vpp, vp, __ATOMIC_SEQ_CST); }
         { __atomic_store(vpp, &vp, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_add_fetch(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_add_fetch(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_add_fetch(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_add_fetch(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_add_fetch(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_add_fetch(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_sub_fetch(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_sub_fetch(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_sub_fetch(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_sub_fetch(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_sub_fetch(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_sub_fetch(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_and_fetch(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_and_fetch(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_and_fetch(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_and_fetch(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_and_fetch(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_and_fetch(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_nand_fetch(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_nand_fetch(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_nand_fetch(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_nand_fetch(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_nand_fetch(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_nand_fetch(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_xor_fetch(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_xor_fetch(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_xor_fetch(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_xor_fetch(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_xor_fetch(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_xor_fetch(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_or_fetch(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_or_fetch(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_or_fetch(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_or_fetch(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_or_fetch(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_or_fetch(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_add(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_add(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_add(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_fetch_add(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_add(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_add(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_sub(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_sub(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_sub(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_fetch_sub(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_sub(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_sub(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_and(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_and(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_and(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_fetch_and(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_and(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_and(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_nand(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_nand(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_nand(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_fetch_nand(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_nand(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_nand(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_xor(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_xor(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_xor(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_fetch_xor(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_xor(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_xor(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_or(vpc, vc, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_or(vps, vs, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_or(vpi, vi, __ATOMIC_SEQ_CST); }
+        { long int ret; ret = __atomic_fetch_or(vpl, vl, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_or(vpll, vll, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_or(vplll, vlll, __ATOMIC_SEQ_CST); }
+        #endif
+        { _Bool ret; ret = __atomic_always_lock_free(sizeof(int), vpi); }
+        { _Bool ret; ret = __atomic_is_lock_free(sizeof(int), vpi); }
+        { char ret; ret = __atomic_add_fetch(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_add_fetch_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_add_fetch(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_add_fetch_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_add_fetch(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_add_fetch_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_add_fetch(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_add_fetch_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_add_fetch(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_add_fetch_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_sub_fetch(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_sub_fetch_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_sub_fetch(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_sub_fetch_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_sub_fetch(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_sub_fetch_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_sub_fetch(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_sub_fetch_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_sub_fetch(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_sub_fetch_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_and_fetch(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_and_fetch_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_and_fetch(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_and_fetch_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_and_fetch(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_and_fetch_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_and_fetch(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_and_fetch_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_and_fetch(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_and_fetch_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_nand_fetch(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_nand_fetch_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_nand_fetch(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_nand_fetch_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_nand_fetch(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_nand_fetch_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_nand_fetch(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_nand_fetch_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_nand_fetch(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_nand_fetch_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_xor_fetch(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_xor_fetch_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_xor_fetch(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_xor_fetch_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_xor_fetch(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_xor_fetch_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_xor_fetch(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_xor_fetch_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_xor_fetch(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_xor_fetch_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_or_fetch(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_or_fetch_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_or_fetch(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_or_fetch_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_or_fetch(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_or_fetch_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_or_fetch(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_or_fetch_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_or_fetch(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_or_fetch_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_add(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_fetch_add_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_add(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_add_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_add(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_add_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_add(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_add_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_add(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_fetch_add_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_sub(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_fetch_sub_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_sub(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_sub_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_sub(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_sub_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_sub(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_sub_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_sub(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_fetch_sub_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_and(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_fetch_and_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_and(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_and_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_and(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_and_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_and(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_and_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_and(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_fetch_and_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_nand(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_fetch_nand_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_nand(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_nand_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_nand(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_nand_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_nand(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_nand_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_nand(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_fetch_nand_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_xor(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_fetch_xor_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_xor(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_xor_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_xor(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_xor_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_xor(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_xor_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_xor(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_fetch_xor_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { char ret; ret = __atomic_fetch_or(vp1, v1, __ATOMIC_SEQ_CST); }
+        { char ret; ret = __atomic_fetch_or_1(vp1, v1, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_or(vp2, v2, __ATOMIC_SEQ_CST); }
+        { short ret; ret = __atomic_fetch_or_2(vp2, v2, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_or(vp4, v4, __ATOMIC_SEQ_CST); }
+        { int ret; ret = __atomic_fetch_or_4(vp4, v4, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_or(vp8, v8, __ATOMIC_SEQ_CST); }
+        { long long int ret; ret = __atomic_fetch_or_8(vp8, v8, __ATOMIC_SEQ_CST); }
+        #if defined(__SIZEOF_INT128__)
+        { __int128 ret; ret = __atomic_fetch_or(vp16, v16, __ATOMIC_SEQ_CST); }
+        { __int128 ret; ret = __atomic_fetch_or_16(vp16, v16, __ATOMIC_SEQ_CST); }
+        #endif
+        { _Bool ret; ret = __atomic_always_lock_free(sizeof(int), vp4); }
+        { _Bool ret; ret = __atomic_is_lock_free(sizeof(int), vp4); }
         { __atomic_thread_fence(__ATOMIC_SEQ_CST); }
         { __atomic_signal_fence(__ATOMIC_SEQ_CST); }

tests/concurrent/coroutineYield.cfa

r7030dab	r71d6bd8
33	33	sout \| "Coroutine 2";
34	34	#endif
35		suspend;
	35	suspend();
36	36	}
37	37	}

tests/concurrent/examples/.expect/datingService.txt

-              r7030dab
+              r71d6bd8
+Girl:17 is dating Boy at 2 with ccode 17
+ Boy:2 is dating Girl 17 with ccode 17
+ Boy:14 is dating Girl 5 with ccode 5
+Girl:5 is dating Boy at 14 with ccode 5
+ Boy:9 is dating Girl 10 with ccode 10
+Girl:10 is dating Boy at 9 with ccode 10
+ Boy:1 is dating Girl 18 with ccode 18
+Girl:18 is dating Boy at 1 with ccode 18
+ Boy:16 is dating Girl 3 with ccode 3
+Girl:3 is dating Boy at 16 with ccode 3
+ Boy:5 is dating Girl 14 with ccode 14
+Girl:14 is dating Boy at 5 with ccode 14
+ Boy:15 is dating Girl 4 with ccode 4
+Girl:4 is dating Boy at 15 with ccode 4
+Girl:0 is dating Boy at 19 with ccode 0
+ Boy:19 is dating Girl 0 with ccode 0
+Girl:9 is dating Boy at 10 with ccode 9
+ Boy:10 is dating Girl 9 with ccode 9
+Girl:11 is dating Boy at 8 with ccode 11
+ Boy:8 is dating Girl 11 with ccode 11
+ Boy:12 is dating Girl 7 with ccode 7
+Girl:7 is dating Boy at 12 with ccode 7
+ Boy:11 is dating Girl 8 with ccode 8
+Girl:8 is dating Boy at 11 with ccode 8
+Girl:16 is dating Boy at 3 with ccode 16
+ Boy:3 is dating Girl 16 with ccode 16
+Girl:15 is dating Boy at 4 with ccode 15
+ Boy:4 is dating Girl 15 with ccode 15
+Girl:19 is dating Boy at 0 with ccode 19
+ Boy:0 is dating Girl 19 with ccode 19
+Girl:2 is dating Boy at 17 with ccode 2
+ Boy:17 is dating Girl 2 with ccode 2
+ Boy:13 is dating Girl 6 with ccode 6
+Girl:6 is dating Boy at 13 with ccode 6
+ Boy:7 is dating Girl 12 with ccode 12
+Girl:12 is dating Boy at 7 with ccode 12
+Girl:13 is dating Boy at 6 with ccode 13
+ Boy:6 is dating Girl 13 with ccode 13
+Girl:1 is dating Boy at 18 with ccode 1
+ Boy:18 is dating Girl 1 with ccode 1

tests/concurrent/examples/boundedBufferEXT.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Apr 18 22:52:12 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Jan 16 22:36:34 2020
 // Update Count     : 15
+// Last Modified On : Fri Jun 21 08:19:20 2019
+// Update Count     : 14
 //
 …
         void insert( Buffer(T) & mutex buffer, T elem ) with( buffer ) {
                 if ( count == BufferSize ) waitfor( remove : buffer );
+                if ( count == BufferSize ) waitfor( remove, buffer );
                 elements[back] = elem;
                 back = ( back + 1 ) % BufferSize;
 …
         T remove( Buffer(T) & mutex buffer ) with( buffer ) {
                 if ( count == 0 ) waitfor( insert : buffer );
+                if ( count == 0 ) waitfor( insert, buffer );
                 T elem = elements[front];
                 front = ( front + 1 ) % BufferSize;

tests/concurrent/examples/boundedBufferTHREAD.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Apr 18 22:52:12 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Jan 16 23:09:43 2020
 // Update Count     : 25
+// Last Modified On : Fri Jun 21 11:50:12 2019
+// Update Count     : 24
 //
 …
         void main( Buffer & buffer ) with( buffer ) {
                 for () {
                         waitfor( ^?{} : buffer ) {
+                        waitfor( ^?{}, buffer ) {
                                 break;
                         } or when ( count != 20 ) waitfor( insert : buffer ) {
+                        } or when ( count != 20 ) waitfor( insert, buffer ) {
                                 back = (back + 1) % 20;
                                 count += 1;
                         } or when ( count != 0 ) waitfor( remove : buffer ) {
+                        } or when ( count != 0 ) waitfor( remove, buffer ) {
                                 front = (front + 1) % 20;
                                 count -= 1;

tests/concurrent/examples/datingService.cfa

-              r7030dab
+              r71d6bd8
 //
 // Cforall Version 1.0.0 Copyright (C) 2017 University of Waterloo
 //
+//
 // The contents of this file are covered under the licence agreement in the
 // file "LICENCE" distributed with Cforall.
 …
                 signal_block( Boys[ccode] );                                    // restart boy to set phone number
         } // if
         //sout | "Girl:" | PhoneNo | "is dating Boy at" | BoyPhoneNo | "with ccode" | ccode;
+        sout | "Girl:" | PhoneNo | "is dating Boy at" | BoyPhoneNo | "with ccode" | ccode;
         return BoyPhoneNo;
 } // DatingService girl
 …
                 signal_block( Girls[ccode] );                                   // restart girl to set phone number
         } // if
         //sout | " Boy:" | PhoneNo | "is dating Girl" | GirlPhoneNo | "with ccode" | ccode;
+        sout | " Boy:" | PhoneNo | "is dating Girl" | GirlPhoneNo | "with ccode" | ccode;
         return GirlPhoneNo;
 } // DatingService boy

tests/concurrent/examples/gortn.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Feb 20 08:02:37 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Jan 16 22:43:40 2020
 // Update Count     : 5
+// Last Modified On : Fri Jun 21 08:25:03 2019
+// Update Count     : 4
 //
 …
 void main( GoRtn & gortn ) with( gortn ) {
         for () {
                 waitfor( mem1 : gortn ) sout | i;
                 or waitfor( mem2 : gortn ) sout | f;
                 or waitfor( mem3 : gortn ) sout | m.i | m.j;
                 or waitfor( ^?{} : gortn ) break;
+                waitfor( mem1, gortn ) sout | i;
+                or waitfor( mem2, gortn ) sout | f;
+                or waitfor( mem3, gortn ) sout | m.i | m.j;
+                or waitfor( ^?{}, gortn ) break;
+        }
+}

tests/concurrent/examples/quickSort.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Dec  6 12:15:52 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Feb 12 18:24:47 2020
 // Update Count     : 177
+// Last Modified On : Thu Oct 10 13:58:18 2019
+// Update Count     : 176
 //
 …
 void ?{}( Quicksort & qs, int values[], int size, int depth ) {
         qs.[values, low, high, depth] = [values, 0, size, depth];
+        qs.values = values;  qs.low = 0;  qs.high = size;  qs.depth = depth;
 } // Quicksort
 …
                         values[counter] = size - counter;                       // descending values
                 } // for
                 for ( i; 200 ) {                                                                // random shuffle a few values
+                for ( int i = 0; i < 200; i +=1 ) {                             // random shuffle a few values
                         swap( values[rand() % size], values[rand() % size] );
                 } // for

tests/concurrent/multi-monitor.cfa

-              r7030dab
+              r71d6bd8
 void increment( monitor_t & mutex p1, monitor_t & mutex p2, int & value ) {
-        assert(active_thread() == get_monitor(p1)->owner);
-        assert(active_thread() == get_monitor(p2)->owner);
         value += 1;
-        assert(active_thread() == get_monitor(p1)->owner);
-        assert(active_thread() == get_monitor(p2)->owner);
+}

tests/concurrent/preempt.cfa

r7030dab	r71d6bd8
36	36	if( (counter % 7) == this.value ) {
37	37	__cfaabi_check_preemption();
38		int next = __atomic_add_fetch~~( &counter, 1, __ATOMIC_SEQ_CST~~ );
	38	int next = __atomic_add_fetch_4(&counter, 1, __ATOMIC_SEQ_CST);
39	39	__cfaabi_check_preemption();
40	40	if( (next % 100) == 0 ) printf("%d\n", (int)next);

tests/concurrent/signal/block.cfa

-              r7030dab
+              r71d6bd8
 monitor global_data_t {
         $thread * last_thread;
         $thread * last_signaller;
+        thread_desc * last_thread;
+        thread_desc * last_signaller;
 };
 …
         if( !is_empty( cond ) ) {
                 $thread * next = front( cond );
+                thread_desc * next = front( cond );
                 if( ! signal_block( cond ) ) {

tests/concurrent/signal/wait.cfa

r7030dab	r71d6bd8
98	98	}
99	99
100		__~~atomic_fetch_sub( &waiter_left, 1, __ATOMIC_SEQ_CST~~ );
	100	__sync_fetch_and_sub_4( &waiter_left, 1);
101	101	}
102	102
…	…
109	109	}
110	110
111		__~~atomic_fetch_sub( &waiter_left, 1, __ATOMIC_SEQ_CST~~ );
	111	__sync_fetch_and_sub_4( &waiter_left, 1);
112	112	}
113	113
…	…
120	120	}
121	121
122		__~~atomic_fetch_sub( &waiter_left, 1, __ATOMIC_SEQ_CST~~ );
	122	__sync_fetch_and_sub_4( &waiter_left, 1);
123	123	}
124	124
…	…
131	131	}
132	132
133		__~~atomic_fetch_sub( &waiter_left, 1, __ATOMIC_SEQ_CST~~ );
	133	__sync_fetch_and_sub_4( &waiter_left, 1);
134	134	}
135	135

tests/concurrent/thread.cfa

-              r7030dab
+              r71d6bd8
 thread Second { semaphore* lock; };
 void ?{}( First  & this, semaphore & lock ) { ((thread&)this){ "Thread 1" }; this.lock = &lock; }
 void ?{}( Second & this, semaphore & lock ) { ((thread&)this){ "Thread 2" }; this.lock = &lock; }
+void ?{}( First  & this, semaphore & lock ) { ((thread&)this){"Thread 1"}; this.lock = &lock; }
+void ?{}( Second & this, semaphore & lock ) { ((thread&)this){"Thread 2"}; this.lock = &lock; }
 void main(First& this) {

tests/concurrent/waitfor/barge.cfa

r7030dab	r71d6bd8
65	65	yield(random( 10 ));
66	66	this.state = WAITFOR;
67		waitfor(do_call : this) {
	67	waitfor(do_call, this) {
68	68	sout \| i;
69	69	}

tests/concurrent/waitfor/dtor.cfa

r7030dab	r71d6bd8
47	47	yield(random( 10 ));
48	48	set_state( this, MAIN );
49		waitfor( ^?{} : this ) {
	49	waitfor( ^?{}, this ) {
50	50	set_state( this, AFTER );
51	51	}

tests/concurrent/waitfor/else.cfa

-              r7030dab
+              r71d6bd8
         sout | "Starting";
         when( false ) waitfor( notcalled : m );
+        when( false ) waitfor( notcalled, m );
         sout | "Step" | i++;
         waitfor( notcalled : m ); or else {
+        waitfor( notcalled, m ); or else {
                 sout | "else called";
+        }
 …
         sout | "Step" | i++;
         when( true ) waitfor( notcalled : m ); or when( true ) else {
+        when( true ) waitfor( notcalled, m ); or when( true ) else {
                 sout | "else called";
+        }
 …
         sout | "Step" | i++;
         when( false ) waitfor( notcalled : m ); or when( true ) else {
+        when( false ) waitfor( notcalled, m ); or when( true ) else {
                 sout | "else called";
+        }
 …
         sout | "Step" | i++;
         when( false ) waitfor( notcalled : m ); or when( false ) else {
+        when( false ) waitfor( notcalled, m ); or when( false ) else {
                 sout | "else called";
+        }

tests/concurrent/waitfor/parse.cfa

-              r7030dab
+              r71d6bd8
         //---------------------------------------
         waitfor( f1 : a ) {
+        waitfor( f1, a ) {
 ;
+        }
         //---------------------------------------
         waitfor( f1 : a ) {
+        waitfor( f1, a ) {
 ;
+        }
         waitfor( f2 : a ) {
+        waitfor( f2, a ) {
 ;
+        }
         //---------------------------------------
         when( 1 < 3 ) waitfor( f2 : a, a ) {
+        when( 1 < 3 ) waitfor( f2, a, a ) {
 ;
+        }
 …
         //---------------------------------------
         when( 2 < 3 ) waitfor( f3 : a ) {
+        when( 2 < 3 ) waitfor( f3, a ) {
 ;
+        }
 …
         //---------------------------------------
         when( 3 < 3 ) waitfor( f3 : a, a ) {
+        when( 3 < 3 ) waitfor( f3, a, a ) {
 ;
+        }
 …
         //---------------------------------------
         when( 6 < 3 ) waitfor( f3 : a, a, a ) {
+        when( 6 < 3 ) waitfor( f3, a, a, a ) {
 ;
+        }
         or when( 7 < 3 ) waitfor( f1 : a  ) {
+        or when( 7 < 3 ) waitfor( f1, a  ) {
 ;
+        }
 …
         //---------------------------------------
         when( 8 < 3 ) waitfor( f3 : a, a ) {
+        when( 8 < 3 ) waitfor( f3, a, a ) {
 ;
+        }
         or waitfor( f1 : a  ) {
+        or waitfor( f1, a  ) {
 ;
+        }
 …
         //---------------------------------------
         when( 10 < 3 ) waitfor( f1 : a ) {
+        when( 10 < 3 ) waitfor( f1, a ) {
 ;
+        }
         or waitfor( f2 : a, a ) {
+        or waitfor( f2, a, a ) {
 ;
+        }
 …
+}
+int main() {}
+int main() {
+}

tests/concurrent/waitfor/parse2.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Aug 30 17:53:29 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Jan 16 23:13:37 2020
 // Update Count     : 6
+// Last Modified On : Fri Mar 22 13:42:11 2019
+// Update Count     : 3
 //
 …
+        }
         waitfor( x : z ) {
+        waitfor( x, z ) {
+        }
 …
         or waitfor( y );
         waitfor( x : z );
+        waitfor( x, z );
         or waitfor( y );
 …
         or when( true ) waitfor( y );
         when( true ) waitfor( x : z );
+        when( true ) waitfor( x, z );
         or when( true ) waitfor( y );
 …
+        }
         waitfor( x : z ) {
+        waitfor( x, z ) {
         } or waitfor( y ) {
+        }
 …
         or else;
         when( true ) waitfor( x : z );
+        when( true ) waitfor( x, z );
         or else;
 …
+        }
         when( true ) waitfor( x : z );
+        when( true ) waitfor( x, z );
         or else {
+        }
 …
         or when( true ) else;
         when( true ) waitfor( x : z );
+        when( true ) waitfor( x, z );
         or when( true ) else;
 …
+        }
         when( true ) waitfor( x : z );
+        when( true ) waitfor( x, z );
         or when( true ) else {
+        }
 …
         or timeout( 3 );
         waitfor( x : z );
+        waitfor( x, z );
         or timeout( 3 );
 …
+        }
         when( true ) waitfor( x : z ) {
+        when( true ) waitfor( x, z ) {
         } or timeout( 3 ) {
+        }
 …
+        }
         when( true ) waitfor( x : z ) {
+        when( true ) waitfor( x, z ) {
         } or when ( true ) timeout( 3 ) {
+        }
 …
         int or, timeout;
         waitfor( timeout : 7 ) 3;
         waitfor( timeout : 7 ) 3; or waitfor( timeout : 7 ) 3;
         when( or ) waitfor( or : ) { 4; } or timeout( 1 ) 3;
         when( 3 ) waitfor( or : 2 ) 4; or else 4;
         when( 3 ) waitfor( or : 3 ) 4; or when( or ) timeout( or ) 4; or when( or ) else timeout;
         when( 3 ) waitfor( or : or ) 3; or when( or ) waitfor( or : timeout ) 4; or else 4;
         when( 3 ) waitfor( or : or ) 3; or waitfor( or : 9 ) 4; or when( or ) timeout( timeout ) 4;
         when( 3 ) waitfor( or : 3 ) 3; or waitfor( or : 7 ) or; or timeout( 1 ) or; or when( 3 ) else or;
+        waitfor( timeout, 7 ) 3;
+        waitfor( timeout, 7 ) 3; or waitfor( timeout, 7 ) 3;
+        when( or ) waitfor( or, ) { 4; } or timeout( 1 ) 3;
+        when( 3 ) waitfor( or, 2 ) 4; or else 4;
+        when( 3 ) waitfor( or, 3 ) 4; or when( or ) timeout( or ) 4; or when( or ) else timeout;
+        when( 3 ) waitfor( or, or ) 3; or when( or ) waitfor( or, timeout ) 4; or else 4;
+        when( 3 ) waitfor( or, or ) 3; or waitfor( or, 9 ) 4; or when( or ) timeout( timeout ) 4;
+        when( 3 ) waitfor( or, 3 ) 3; or waitfor( or, 7 ) or; or timeout( 1 ) or; or when( 3 ) else or;
         // test else selection

tests/concurrent/waitfor/recurse.cfa

-              r7030dab
+              r71d6bd8
         rand_yield();
         waitfor( call4 : this );
+        waitfor( call4, this );
         rand_yield();
 …
         rand_yield();
         waitfor( call3 : this );
+        waitfor( call3, this );
         rand_yield();
 …
         rand_yield();
         waitfor( call2 : this );
+        waitfor( call2, this );
         rand_yield();

tests/concurrent/waitfor/statment.cfa

-              r7030dab
+              r71d6bd8
         while( !done ) {
                    waitfor( get_index : this );
                 or waitfor( call1 : this ) { sout | "Statement"; if( this.last_val != 1 ) { serr | "Incorrect index: expected" | 1 | "got" | this.last_val; } }
                 or waitfor( call2 : this ) { sout | "Statement"; if( this.last_val != 2 ) { serr | "Incorrect index: expected" | 2 | "got" | this.last_val; } }
                 or waitfor( call3 : this ) { sout | "Statement"; if( this.last_val != 3 ) { serr | "Incorrect index: expected" | 3 | "got" | this.last_val; } }
                 or waitfor( call4 : this ) { sout | "Statement"; if( this.last_val != 4 ) { serr | "Incorrect index: expected" | 4 | "got" | this.last_val; } }
                 or waitfor( call5 : this ) { sout | "Statement"; if( this.last_val != 5 ) { serr | "Incorrect index: expected" | 5 | "got" | this.last_val; } }
                 or waitfor( call6 : this ) { sout | "Statement"; if( this.last_val != 6 ) { serr | "Incorrect index: expected" | 6 | "got" | this.last_val; } }
                 or waitfor( call7 : this ) { sout | "Statement"; if( this.last_val != 7 ) { serr | "Incorrect index: expected" | 7 | "got" | this.last_val; } }
+                   waitfor( get_index, this );
+                or waitfor( call1, this ) { sout | "Statement"; if( this.last_val != 1 ) { serr | "Incorrect index: expected" | 1 | "got" | this.last_val; } }
+                or waitfor( call2, this ) { sout | "Statement"; if( this.last_val != 2 ) { serr | "Incorrect index: expected" | 2 | "got" | this.last_val; } }
+                or waitfor( call3, this ) { sout | "Statement"; if( this.last_val != 3 ) { serr | "Incorrect index: expected" | 3 | "got" | this.last_val; } }
+                or waitfor( call4, this ) { sout | "Statement"; if( this.last_val != 4 ) { serr | "Incorrect index: expected" | 4 | "got" | this.last_val; } }
+                or waitfor( call5, this ) { sout | "Statement"; if( this.last_val != 5 ) { serr | "Incorrect index: expected" | 5 | "got" | this.last_val; } }
+                or waitfor( call6, this ) { sout | "Statement"; if( this.last_val != 6 ) { serr | "Incorrect index: expected" | 6 | "got" | this.last_val; } }
+                or waitfor( call7, this ) { sout | "Statement"; if( this.last_val != 7 ) { serr | "Incorrect index: expected" | 7 | "got" | this.last_val; } }
                 done = true;

tests/concurrent/waitfor/when.cfa

-              r7030dab
+              r71d6bd8
 void arbiter( global_t & mutex this ) {
         for( int i = 0; i < N; i++ ) {
                    when( this.last_call == 6 ) waitfor( call1 : this ) { if( this.last_call != 1) { serr | "Expected last_call to be 1 got" | this.last_call; } }
                 or when( this.last_call == 1 ) waitfor( call2 : this ) { if( this.last_call != 2) { serr | "Expected last_call to be 2 got" | this.last_call; } }
                 or when( this.last_call == 2 ) waitfor( call3 : this ) { if( this.last_call != 3) { serr | "Expected last_call to be 3 got" | this.last_call; } }
                 or when( this.last_call == 3 ) waitfor( call4 : this ) { if( this.last_call != 4) { serr | "Expected last_call to be 4 got" | this.last_call; } }
                 or when( this.last_call == 4 ) waitfor( call5 : this ) { if( this.last_call != 5) { serr | "Expected last_call to be 5 got" | this.last_call; } }
                 or when( this.last_call == 5 ) waitfor( call6 : this ) { if( this.last_call != 6) { serr | "Expected last_call to be 6 got" | this.last_call; } }
+                   when( this.last_call == 6 ) waitfor( call1, this ) { if( this.last_call != 1) { serr | "Expected last_call to be 1 got" | this.last_call; } }
+                or when( this.last_call == 1 ) waitfor( call2, this ) { if( this.last_call != 2) { serr | "Expected last_call to be 2 got" | this.last_call; } }
+                or when( this.last_call == 2 ) waitfor( call3, this ) { if( this.last_call != 3) { serr | "Expected last_call to be 3 got" | this.last_call; } }
+                or when( this.last_call == 3 ) waitfor( call4, this ) { if( this.last_call != 4) { serr | "Expected last_call to be 4 got" | this.last_call; } }
+                or when( this.last_call == 4 ) waitfor( call5, this ) { if( this.last_call != 5) { serr | "Expected last_call to be 5 got" | this.last_call; } }
+                or when( this.last_call == 5 ) waitfor( call6, this ) { if( this.last_call != 6) { serr | "Expected last_call to be 6 got" | this.last_call; } }
                 sout | this.last_call;

tests/coroutine/.expect/fmtLines.txt

-              r7030dab
+              r71d6bd8
 {                                                         // f  or n  ewli
 ne c  hara  cter  s                                     su
 spen  d;                                        i  f (   fmt.
 ch !  = '\  n' )   bre  ak;
         //   igno  re n  ewli  ne
                 }   // f  or                            so  ut |
  fmt  .ch;                                                      /  / pr
 int   char  acte  r                       } //
  for                    s  out   | "    ";
                                                         /  / pr  int   bloc
 k se  para  tor         } /  / fo
 r               s  out   | nl  ;
                 //   pri  nt g  roup   sep
 arat  or        }   //   for}   //
 main  void   prt  ( Fo  rmat
  & f  mt,   char   ch   ) {
    f  mt.c  h =   ch;      r
 esum  e( f  mt )  ;} /  / pr
 tint   mai  n()   {     Fo  rmat
  fmt  ; ch  ar c  h;    f  or (
  ;;   ) {               sin   | c  h;
                                                                   // r  ead   one
 char  acte  r       if (   eof
 ( si  n )   ) br  eak;
                         /  / eo  f ?            prt  ( fm
 t, c  h );      } /  / fo  r} /
 / ma  in//   Loc  al V  aria
 bles  : //  // t  ab-w  idth
 : 4   ////   com  pile  -com
 mand  : "c  fa f  mtLi  nes.
 cfa"   ///  / En  d: /  /
+spen  d();                                      if   ( fm
+t.ch   !=   '\n'   ) b  reak
+;               /  / ig  nore   new  line
+                                  } //   for                              sout
+ | f  mt.c  h;                                                  //
+prin  t ch  arac  ter                   }
+// f  or                        sou  t |   "  "
+;                                                               //   prin  t bl
+ock   sepa  rato  r             }   //
+for             sou  t |   nl;
+                                  // p  rint   gro  up s
+epar  ator      } /  / fo  r} /
+/ ma  invo  id p  rt(   Form
+at &   fmt  , ch  ar c  h )
+{      fmt  .ch   = ch  ;
+ res  ume(   fmt   );}   //
+prti  nt m  ain(  ) {     Form
+at f  mt;         char   ch;    for
+ ( ;  ; )   {           s  in |   ch;
+                                                                                //   rea  d on
+e ch  arac  ter     if   ( e
+of(   sin   ) )   brea  k;
+                                        //   eof   ?            p  rt(
+fmt,   ch   );  }   //   for}
+ //   main  // L  ocal   Var
+iabl  es:   ////   tab  -wid
+th:   4 //  // c  ompi  le-c
+omma  nd:   "cfa   fmt  Line
+s.cf  a" /  ///   End:   //

tests/coroutine/.in/fmtLines.txt

r7030dab	r71d6bd8
35	35	for ( fmt.b = 0; fmt.b < 4; fmt.b += 1 ) { // blocks of 4 characters
36	36	for ( ;; ) { // for newline characters
37		suspend;
	37	suspend();
38	38	if ( fmt.ch != '\n' ) break; // ignore newline
39	39	} // for

tests/coroutine/cntparens.cfa

-              r7030dab
+              r71d6bd8
 //
+//
 // Cforall Version 1.0.0 Copyright (C) 2017 University of Waterloo
 //
 // The contents of this file are covered under the licence agreement in the
 // file "LICENCE" distributed with Cforall.
 //
+//
 // cntparens.cfa -- match left/right parenthesis
 //
+//
 // Author           : Peter A. Buhr
 // Created On       : Sat Apr 20 11:04:45 2019
 …
 // Last Modified On : Sat Apr 20 11:06:21 2019
 // Update Count     : 1
 //
+//
 #include <fstream.hfa>
 …
 void main( CntParens & cpns ) with( cpns ) {
         for ( ; ch == '('; cnt += 1 ) {                                         // left parenthesis
                 suspend;
+                suspend();
+        }
         for ( ; ch == ')' && cnt > 1; cnt -= 1 ) {                      // right parenthesis
                 suspend;
+                suspend();
+        }
         status = ch == ')' ? Match : Error;
 } // main
 void ?{}( CntParens & cpns ) with( cpns ) { status = Cont; cnt = 0; }

tests/coroutine/devicedriver.cfa

-              r7030dab
+              r71d6bd8
 //
+//
 // Cforall Version 1.0.0 Copyright (C) 2017 University of Waterloo
 //
 // The contents of this file are covered under the licence agreement in the
 // file "LICENCE" distributed with Cforall.
 //
 // devicedriver.cfa --
 //
+//
+// devicedriver.cfa --
+//
 // Author           : Peter A. Buhr
 // Created On       : Sat Mar 16 15:30:34 2019
 …
 // Last Modified On : Sat Apr 20 09:07:19 2019
 // Update Count     : 90
 //
+//
 #include <fstream.hfa>
 …
 void checkCRC( Driver & d, unsigned int sum ) with( d ) {
         suspend;
+        suspend();
         unsigned short int crc = byte << 8;                                     // sign extension over written
         suspend;
+        suspend();
         // prevent sign extension for signed char
         status = (crc | (unsigned char)byte) == sum ? MSG : ECRC;
 …
                 status = CONT;
                 unsigned int lnth = 0, sum = 0;
                 while ( byte != STX ) suspend;
+                while ( byte != STX ) suspend();
           emsg: for () {
                         suspend;
+                        suspend();
                         choose ( byte ) {                                                       // process byte
                           case STX:
                                 status = ESTX; suspend; continue msg;
+                                status = ESTX; suspend(); continue msg;
                           case ETX:
                                 break emsg;
                           case ESC:
                                 suspend;
+                                suspend();
                         } // choose
                         if ( lnth >= MaxMsg ) {                                         // buffer full ?
                                 status = ELNTH; suspend; continue msg;
+                                status = ELNTH; suspend(); continue msg;
                         } // if
                         msg[lnth++] = byte;
 …
                 msg[lnth] = '\0';                                                               // terminate string
                 checkCRC( d, sum );                                                             // refactor CRC check
                 suspend;
+                suspend();
         } // for
 } // main

tests/coroutine/fibonacci.cfa

-              r7030dab
+              r71d6bd8
         int fn1, fn2;                                                                           // retained between resumes
         fn = 0;  fn1 = fn;                                                                      // 1st case
         suspend;                                                                                        // restart last resume
+        suspend();                                                                                      // restart last resume
         fn = 1;  fn2 = fn1;  fn1 = fn;                                          // 2nd case
         suspend;                                                                                        // restart last resume
+        suspend();                                                                                      // restart last resume
         for () {
                 fn = fn1 + fn2;  fn2 = fn1;  fn1 = fn;                  // general case
                 suspend;                                                                                // restart last resume
+                suspend();                                                                              // restart last resume
         } // for
+}

tests/coroutine/fibonacci_1.cfa

-              r7030dab
+              r71d6bd8
 // Last Modified On : Thu Mar 21 08:10:45 2019
 // Update Count     : 25
 //
+//
 #include <fstream.hfa>
 …
         [fn1, fn] = [0, 1];                                                                     // precompute first two states
         for () {
                 suspend;                                                                                // restart last resume
+                suspend();                                                                              // restart last resume
                 [fn1, fn] = [fn, fn1 + fn];                                             // general case
         } // for

tests/coroutine/fmtLines.cfa

r7030dab	r71d6bd8
27	27	for ( b = 0; b < 4; b += 1 ) { // blocks of 4 characters
28	28	for () { // for newline characters
29		suspend;
	29	suspend();
30	30	if ( ch != '\n' ) break; // ignore newline
31	31	} // for

tests/coroutine/raii.cfa

r7030dab	r71d6bd8
39	39	Raii raii = { "Coroutine" };
40	40	sout \| "Before Suspend";
41		suspend;
	41	suspend();
42	42	sout \| "After Suspend";
43	43	}

tests/coroutine/runningTotal.cfa

r7030dab	r71d6bd8
25	25	void update( RunTotal & rntl, int input ) with( rntl ) { // helper
26	26	total += input; // remember between activations
27		suspend; // inactivate on stack
	27	suspend(); // inactivate on stack
28	28	}
29	29

tests/coroutine/suspend_then.cfa

-              r7030dab
+              r71d6bd8
 #include <fstream.hfa>
+#include <coroutine.hfa>
+generator Fibonacci {
+        int fn;                                                                         // used for communication
+        int fn1, fn2;                                                           // retained between resumes
+};
+void then() {
+        sout | "Then!";
+}
+coroutine Fibonacci { int fn; };                                                // used for communication
 void main( Fibonacci & fib ) with( fib ) {                              // called on first resume
+        int fn1, fn2;                                                           // retained between resumes
         fn = 0;  fn1 = fn;                                                      // 1st case
         suspend { sout | "Then!"; }                                             // restart last resume
+        suspend_then(then);                                                     // restart last resume
         fn = 1;  fn2 = fn1;  fn1 = fn;                                  // 2nd case
         suspend { sout | "Then!"; }                                             // restart last resume
+        suspend_then(then);                                                     // restart last resume
         for () {
                 fn = fn1 + fn2;  fn2 = fn1;  fn1 = fn;                  // general case
                 suspend { sout | "Then!"; }                                     // restart last resume
+                suspend_then(then);                                             // restart last resume
         } // for
+}

tests/expression.cfa

-              r7030dab
+              r71d6bd8
-struct S { int i; };
-void ?{}( S & s, int i ) { s.i = i; }
-int ?`mary( int );
-int ?`mary( S );
-[int] ?`mary( [int, int] );
-int & ?`jane( int & );
-int jack( int );
 int main() {
+    int a[3] = { 0, 0, 0 };
+    S s = { 3 }, * ps = &s;
+    [int] t = { 3 };
+    * [int] pt = &t;
+    int i = 1, j = 2;
+    struct s { int i; } x, *p = &x;
+    int i = 3;
     // operators
     !i;
+    ! i;
     ~i;
     +i;
     -i;
     *ps;
     ++ps;
     --ps;
     ps++;
     ps--;
+    *p;
+    ++p;
+    --p;
+    p++;
+    p--;
     i + j;
     i - j;
     i * j;
+    i+i;
+    i-i;
+    i*i;
     i / j;
     i % j;
     i ^ j;
     i & j;
     i | j;
     i < j;
     i > j;
     i = j;
+    i/i;
+    i%i;
+    i^i;
+    i&i;
+    i|i;
+    i<i;
+    i>i;
+    i=i;
+    i == j;
+    i != j;
+    i << j;
+    i >> j;
+    i <= j;
+    i >= j;
+    i && j;
+    i || j;
+    ps->i;
+    i==i;
+    i!=i;
+    i<<i;
+    i>>i;
+    i<=i;
+    i>=i;
+    i&&i;
+    i||i;
+    p->i;
+    i*=i;
+    i/=i;
+    i%=i;
+    i+=i;
+    i-=i;
+    i&=i;
+    i|=i;
+    i^=i;
+    i<<=i;
+    i>>=i;
+    i *= j;
+    i /= j;
+    i %= j;
+    i += j;
+    i -= j;
+    i &= j;
+    i |= j;
+    i ^= j;
+    i <<= j;
+    i >>= j;
+    i ? i : j;
+    // postfix function call
+    (3 + 4)`mary;
+    ({3 + 4;})`mary;
+    [3, 4]`mary;
+`mary;
+    a[0]`mary;
+    a[0]`mary`mary;
+    s{0}`mary;
+    a[3]`jane++;
+    jack(3)`mary;
+    s.i`mary;
+    t.0`mary;
+    s.[i]`mary;
+    ps->i`mary;
+    pt->0`mary;
+    ps->[i]`mary;
+    i++`mary;
+    i--`mary;
+    (S){2}`mary;
+    (S)@{2}`mary;
+    i?i:i;
 } // main

tests/heap.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Nov  6 17:54:56 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Nov 24 12:34:51 2019
 // Update Count     : 28
+// Last Modified On : Fri Jul 19 08:22:34 2019
+// Update Count     : 19
 //
 …
         enum { NoOfAllocs = 5000, NoOfMmaps = 10 };
         char * locns[NoOfAllocs];
+        size_t amount;
+        enum { limit = 64 * 1024 };                                                     // check alignments up to here
+        int i;
         // check alloc/free
 …
                 size_t s = (i + 1) * 20;
                 char * area = (char *)malloc( s );
                 if ( area == 0p ) abort( "malloc/free out of memory" );
+                if ( area == 0 ) abort( "malloc/free out of memory" );
                 area[0] = '\345'; area[s - 1] = '\345';                 // fill first/last
                 area[malloc_usable_size( area ) - 1] = '\345';  // fill ultimate byte
 …
                 size_t s = i + 1;                                                               // +1 to make initialization simpler
                 locns[i] = (char *)malloc( s );
                 if ( locns[i] == 0p ) abort( "malloc/free out of memory" );
+                if ( locns[i] == 0 ) abort( "malloc/free out of memory" );
                 locns[i][0] = '\345'; locns[i][s - 1] = '\345'; // fill first/last
                 locns[i][malloc_usable_size( locns[i] ) - 1] = '\345'; // fill ultimate byte
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 char * area = (char *)malloc( s );
                 if ( area == 0p ) abort( "malloc/free out of memory" );
+                if ( area == 0 ) abort( "malloc/free out of memory" );
                 area[0] = '\345'; area[s - 1] = '\345';                 // fill first/last
                 area[malloc_usable_size( area ) - 1] = '\345';  // fill ultimate byte
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 locns[i] = (char *)malloc( s );
                 if ( locns[i] == 0p ) abort( "malloc/free out of memory" );
+                if ( locns[i] == 0 ) abort( "malloc/free out of memory" );
                 locns[i][0] = '\345'; locns[i][s - 1] = '\345'; // fill first/last
                 locns[i][malloc_usable_size( locns[i] ) - 1] = '\345'; // fill ultimate byte
 …
                 size_t s = (i + 1) * 20;
                 char * area = (char *)calloc( 5, s );
                 if ( area == 0p ) abort( "calloc/free out of memory" );
+                if ( area == 0 ) abort( "calloc/free out of memory" );
                 if ( area[0] != '\0' || area[s - 1] != '\0' ||
                          area[malloc_usable_size( area ) - 1] != '\0' ||
 …
                 size_t s = i + 1;
                 locns[i] = (char *)calloc( 5, s );
                 if ( locns[i] == 0p ) abort( "calloc/free out of memory" );
+                if ( locns[i] == 0 ) abort( "calloc/free out of memory" );
                 if ( locns[i][0] != '\0' || locns[i][s - 1] != '\0' ||
                          locns[i][malloc_usable_size( locns[i] ) - 1] != '\0' ||
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 char * area = (char *)calloc( 1, s );
                 if ( area == 0p ) abort( "calloc/free out of memory" );
+                if ( area == 0 ) abort( "calloc/free out of memory" );
                 if ( area[0] != '\0' || area[s - 1] != '\0' ) abort( "calloc/free corrupt storage4.1" );
                 if ( area[malloc_usable_size( area ) - 1] != '\0' ) abort( "calloc/free corrupt storage4.2" );
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 locns[i] = (char *)calloc( 1, s );
                 if ( locns[i] == 0p ) abort( "calloc/free out of memory" );
+                if ( locns[i] == 0 ) abort( "calloc/free out of memory" );
                 if ( locns[i][0] != '\0' || locns[i][s - 1] != '\0' ||
                          locns[i][malloc_usable_size( locns[i] ) - 1] != '\0' ||
 …
         // check memalign/free (sbrk)
+        enum { limit = 64 * 1024 };                                                     // check alignments up to here
         for ( a; libAlign() ~= limit ~ a ) {                            // generate powers of 2
                 //sout | alignments[a];
                 for ( s; 1 ~ NoOfAllocs ) {                                             // allocation of size 0 can return null
                         char * area = (char *)memalign( a, s );
                         if ( area == 0p ) abort( "memalign/free out of memory" );
                         //sout | i | area;
+                        if ( area == 0 ) abort( "memalign/free out of memory" );
+                        //sout | i | " " | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
                                 abort( "memalign/free bad alignment : memalign(%d,%d) = %p", (int)a, s, area );
                         } // if
                         area[0] = '\345'; area[s - 1] = '\345';         // fill first/last byte
+                        area[0] = '\345'; area[s - 1] = '\345'; // fill first/last byte
                         area[malloc_usable_size( area ) - 1] = '\345'; // fill ultimate byte
                         free( area );
 …
                         size_t s = i + default_mmap_start();            // cross over point
                         char * area = (char *)memalign( a, s );
                         if ( area == 0p ) abort( "memalign/free out of memory" );
                         //sout | i | area;
+                        if ( area == 0 ) abort( "memalign/free out of memory" );
+                        //sout | i | " " | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
                                 abort( "memalign/free bad alignment : memalign(%d,%d) = %p", (int)a, (int)s, area );
 …
                 // initial N byte allocation
                 char * area = (char *)calloc( 5, i );
                 if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
+                if ( area == 0 ) abort( "calloc/realloc/free out of memory" );
                 if ( area[0] != '\0' || area[i - 1] != '\0' ||
                          area[malloc_usable_size( area ) - 1] != '\0' ||
 …
                 for ( s; i ~ 256 * 1024 ~ 26 ) {                                // start at initial memory request
                         area = (char *)realloc( area, s );                      // attempt to reuse storage
                         if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
+                        if ( area == 0 ) abort( "calloc/realloc/free out of memory" );
                         if ( area[0] != '\0' || area[s - 1] != '\0' ||
                                  area[malloc_usable_size( area ) - 1] != '\0' ||
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 char * area = (char *)calloc( 1, s );
                 if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
+                if ( area == 0 ) abort( "calloc/realloc/free out of memory" );
                 if ( area[0] != '\0' || area[s - 1] != '\0' ||
                          area[malloc_usable_size( area ) - 1] != '\0' ||
 …
                 for ( r; i ~ 256 * 1024 ~ 26 ) {                                // start at initial memory request
                         area = (char *)realloc( area, r );                      // attempt to reuse storage
                         if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
+                        if ( area == 0 ) abort( "calloc/realloc/free out of memory" );
                         if ( area[0] != '\0' || area[r - 1] != '\0' ||
                                  area[malloc_usable_size( area ) - 1] != '\0' ||
 …
         // check memalign/realloc/free
         amount = 2;
+        size_t amount = 2;
         for ( a; libAlign() ~= limit ~ a ) {                            // generate powers of 2
                 // initial N byte allocation
                 char * area = (char *)memalign( a, amount );    // aligned N-byte allocation
                 if ( area == 0p ) abort( "memalign/realloc/free out of memory" ); // no storage ?
                 //sout | alignments[a] | area;
+                if ( area == 0 ) abort( "memalign/realloc/free out of memory" ); // no storage ?
+                //sout | alignments[a] | " " | area;
                 if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
                         abort( "memalign/realloc/free bad alignment : memalign(%d,%d) = %p", (int)a, (int)amount, area );
 …
                         if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "memalign/realloc/free corrupt storage" );
                         area = (char *)realloc( area, s );                      // attempt to reuse storage
                         if ( area == 0p ) abort( "memalign/realloc/free out of memory" ); // no storage ?
                         //sout | i | area;
+                        if ( area == 0 ) abort( "memalign/realloc/free out of memory" ); // no storage ?
+                        //sout | i | " " | area;
                         if ( (size_t)area % a != 0 ) {                          // check for initial alignment
                                 abort( "memalign/realloc/free bad alignment %p", area );
 …
                 for ( s; 1 ~ limit ) {                                                  // allocation of size 0 can return null
                         char * area = (char *)cmemalign( a, 1, s );
                         if ( area == 0p ) abort( "cmemalign/free out of memory" );
                         //sout | i | area;
+                        if ( area == 0 ) abort( "cmemalign/free out of memory" );
+                        //sout | i | " " | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
                                 abort( "cmemalign/free bad alignment : cmemalign(%d,%d) = %p", (int)a, s, area );
 …
                 // initial N byte allocation
                 char * area = (char *)cmemalign( a, 1, amount ); // aligned N-byte allocation
                 if ( area == 0p ) abort( "cmemalign/realloc/free out of memory" ); // no storage ?
                 //sout | alignments[a] | area;
+                if ( area == 0 ) abort( "cmemalign/realloc/free out of memory" ); // no storage ?
+                //sout | alignments[a] | " " | area;
                 if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
                         abort( "cmemalign/realloc/free bad alignment : cmemalign(%d,%d) = %p", (int)a, (int)amount, area );
 …
                         if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "cmemalign/realloc/free corrupt storage2" );
                         area = (char *)realloc( area, s );                      // attempt to reuse storage
                         if ( area == 0p ) abort( "cmemalign/realloc/free out of memory" ); // no storage ?
                         //sout | i | area;
+                        if ( area == 0 ) abort( "cmemalign/realloc/free out of memory" ); // no storage ?
+                        //sout | i | " " | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
                                 abort( "cmemalign/realloc/free bad alignment %p", area );
 …
                 free( area );
         } // for
-        // check memalign/realloc with align/free
-        amount = 2;
-        for ( a; libAlign() ~= limit ~ a ) {                            // generate powers of 2
-                // initial N byte allocation
-                char * area = (char *)memalign( a, amount );    // aligned N-byte allocation
-                if ( area == 0p ) abort( "memalign/realloc with align/free out of memory" ); // no storage ?
-                //sout | alignments[a] | area | endl;
-                if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
-                        abort( "memalign/realloc with align/free bad alignment : memalign(%d,%d) = %p", (int)a, (int)amount, area );
-                } // if
-                area[0] = '\345'; area[amount - 2] = '\345';    // fill first/penultimate byte
-                // Do not start this loop index at 0 because realloc of 0 bytes frees the storage.
-                for ( s; amount ~ 256 * 1024 ) {                                // start at initial memory request
-                        if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "memalign/realloc/free corrupt storage" );
-                        area = (char *)realloc( area, a * 2, s );       // attempt to reuse storage
-                        if ( area == 0p ) abort( "memalign/realloc with align/free out of memory" ); // no storage ?
-                        //sout | i | area | endl;
-                        if ( (size_t)area % a * 2 != 0 ) {                      // check for initial alignment
-                                abort( "memalign/realloc with align/free bad alignment %p", area );
-                        } // if
-                        area[s - 1] = '\345';                                           // fill last byte
-                } // for
-                free( area );
-        } // for
-        // check cmemalign/realloc with align/free
-        amount = 2;
-        for ( size_t a = libAlign() + libAlign(); a <= limit; a += a ) { // generate powers of 2
-                // initial N byte allocation
-                char *area = (char *)cmemalign( a, 1, amount ); // aligned N-byte allocation
-                if ( area == 0p ) abort( "cmemalign/realloc with align/free out of memory" ); // no storage ?
-                //sout | alignments[a] | area | endl;
-                if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
-                        abort( "cmemalign/realloc with align/free bad alignment : cmemalign(%d,%d) = %p", (int)a, (int)amount, area );
-                } // if
-                if ( area[0] != '\0' || area[amount - 1] != '\0' ||
-                         area[malloc_usable_size( area ) - 1] != '\0' ||
-                         ! malloc_zero_fill( area ) ) abort( "cmemalign/realloc with align/free corrupt storage1" );
-                area[0] = '\345'; area[amount - 2] = '\345';    // fill first/penultimate byte
-                // Do not start this loop index at 0 because realloc of 0 bytes frees the storage.
-                for ( int s = amount; s < 256 * 1024; s += 1 ) { // start at initial memory request
-                        if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "cmemalign/realloc with align/free corrupt storage2" );
-                        area = (char *)realloc( area, a * 2, s );       // attempt to reuse storage
-                        if ( area == 0p ) abort( "cmemalign/realloc with align/free out of memory" ); // no storage ?
-                        //sout | i | area | endl;
-                        if ( (size_t)area % a * 2 != 0 || malloc_alignment( area ) != a * 2 ) { // check for initial alignment
-                                abort( "cmemalign/realloc with align/free bad alignment %p %jd %jd", area, malloc_alignment( area ), a * 2 );
-                        } // if
-                        if ( area[s - 1] != '\0' || area[s - 1] != '\0' ||
-                                 area[malloc_usable_size( area ) - 1] != '\0' ||
-                                 ! malloc_zero_fill( area ) ) abort( "cmemalign/realloc/free corrupt storage3" );
-                        area[s - 1] = '\345';                                           // fill last byte
-                } // for
-                free( area );
-        } // for
         //sout | "worker" | thisTask() | "successful completion";
 } // Worker main

tests/labelledExit.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Aug 10 07:29:39 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Feb  5 16:49:48 2020
 // Update Count     : 9
+// Last Modified On : Fri Oct 25 17:41:51 2019
+// Update Count     : 7
 //
 …
         // computed goto
+        {
                 void *array[] = { &&foo, &&bar, &&hack };
           foo: bar: hack:
                 &&foo;
                 &&bar;
                 goto *array[i];
+        }
+        // {
+        //      void *array[] = { &&foo, &&bar, &&hack };
+        //   foo: bar: hack:
+        //      &&foo;
+        //      &&bar;
+        //      goto *array[i];
+        // }
   Q: if ( i > 5 ) {

tests/linking/withthreads.cfa

r7030dab	r71d6bd8
34	34	// Local Variables: //
35	35	// tab-width: 4 //
36		// compile-command: "cfa ~~with~~threads.cfa" //
	36	// compile-command: "cfa nothreads.cfa" //
37	37	// End: //

tests/loopctrl.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed Aug  8 18:32:59 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Dec 12 17:55:26 2019
 // Update Count     : 108
+// Last Modified On : Fri Jul 12 12:05:05 2019
+// Update Count     : 106
 //
 …
         for ( 1 ) { sout | "A"; }                                                       sout | nl;
         for ( 10 ) { sout | "A"; }                                                      sout | nl;
-        for ( = 10 ) { sout | "A"; }                                            sout | nl;
         for ( 1 ~= 10 ~ 2 ) { sout | "B"; }                                     sout | nl;
         for ( 10 -~= 1 ~ 2 ) { sout | "C"; }                            sout | nl;
 …
         for ( i; 10 ) { sout | i; }                                                     sout | nl;
-        for ( i; = 10 ) { sout | i; }                                           sout | nl;
         for ( i; 1 ~= 10 ~ 2 ) { sout | i; }                            sout | nl;
         for ( i; 10 -~= 1 ~ 2 ) { sout | i; }                           sout | nl;
 …
         for ( N ) { sout | "N"; }                                                       sout | nl;
         for ( i; N ) { sout | i; }                                                      sout | nl;
-        for ( i; = N ) { sout | i; }                                            sout | nl;
         for ( i; N -~ 0 ) { sout | i; }                                         sout | nl | nl;

tests/nested-types.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Jul 9 10:20:03 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Feb 12 18:21:15 2020
 // Update Count     : 3
+// Last Modified On : Tue Nov  6 17:59:40 2018
+// Update Count     : 2
 //
 …
 //   double d;
 // };
-// struct S {
-//     enum C { R, G, B };
-//     int i;
-//     struct T {
-//      int i;
-//     };
-//     T t;
-// };
-// S s;
-// S.C c;
-// S.T t;
 int main() {

tests/pybin/settings.py

-              r7030dab
+              r71d6bd8
                         print("updated to %s" % self.target)
+        def filter(self, tests):
+                return [test for test in tests if not test.arch or self.target == test.arch]
+        def match(self, arch):
                 return True if not arch else self.target == arch
 …
 def init( options ):
-        global all_arch
-        global all_debug
-        global all_install
         global arch
         global archive
-        global continue_
         global debug
+        global distcc
         global dry_run
         global generating
 …
         global output_width
         global timeout
-        global timeout2gdb
+        all_arch     = [Architecture(o) for o in list(dict.fromkeys(options.arch   ))]
+        all_debug    = [Debug(o)        for o in list(dict.fromkeys(options.debug  ))]
+        all_install  = [Install(o)      for o in list(dict.fromkeys(options.install))]
+        arch         = Architecture(options.arch)
         archive      = os.path.abspath(os.path.join(original_path, options.archive_errors)) if options.archive_errors else None
         continue_    = options.continue_
+        debug        = Debug(options.debug)
         dry_run      = options.dry_run # must be called before tools.config_hash()
+        distcc       = "DISTCC_CFA_PATH=~/.cfadistcc/%s/cfa" % tools.config_hash()
         generating   = options.regenerate_expected
+        install      = Install(options.install)
         make         = ['make']
         output_width = 24
         timeout      = Timeouts(options.timeout, options.global_timeout)
-        timeout2gdb  = options.timeout_with_gdb
         # if we distribute, distcc errors will fail tests, use log file for distcc
 …
 def validate():
-        """Validate the current configuration and update globals"""
-        global distcc
-        distcc       = "DISTCC_CFA_PATH=~/.cfadistcc/%s/cfa" % tools.config_hash()
         errf = os.path.join(BUILDDIR, ".validate.err")
         make_ret, out = tools.make( ".validate", error_file = errf, output_file=subprocess.DEVNULL, error=subprocess.DEVNULL )

tests/pybin/tools.py

-              r7030dab
+              r71d6bd8
                                         return proc.returncode, out.decode("utf-8") if out else None
                                 except subprocess.TimeoutExpired:
+                                        if settings.timeout2gdb:
+                                                print("Process {} timeout".format(proc.pid))
+                                                proc.communicate()
+                                                return 124, str(None)
+                                        else:
+                                                proc.send_signal(signal.SIGABRT)
+                                                proc.communicate()
+                                                return 124, str(None)
+                                        proc.send_signal(signal.SIGABRT)
+                                        proc.communicate()
+                                        return 124, str(None)
         except Exception as ex:
 …
 def which(program):
+        fpath, fname = os.path.split(program)
+        if fpath:
+                if is_exe(program):
+                        return program
+        else:
+                for path in os.environ["PATH"].split(os.pathsep):
+                        exe_file = os.path.join(path, program)
+                        if is_exe(exe_file):
+                                return exe_file
+        return None
+    fpath, fname = os.path.split(program)
+    if fpath:
+        if is_exe(program):
+            return program
+    else:
+        for path in os.environ["PATH"].split(os.pathsep):
+            exe_file = os.path.join(path, program)
+            if is_exe(exe_file):
+                return exe_file
+    return None
 @contextlib.contextmanager
 …
         raise argparse.ArgumentTypeError(msg)
-# Convert a function that converts a string to one that converts comma separated string.
-def comma_separated(elements):
-    return lambda string: [elements(part) for part in string.split(',')]
 def fancy_print(text):
         column = which('column')
 …
 class Timed:
         def __enter__(self):
                 self.start = time.time()
                 return self
         def __exit__(self, *args):
                 self.end = time.time()
                 self.duration = self.end - self.start
+    def __enter__(self):
+        self.start = time.time()
+        return self
+    def __exit__(self, *args):
+        self.end = time.time()
+        self.duration = self.end - self.start
 def timed(src, timeout):
         expire = time.time() + timeout
         i = iter(src)
+        with contextlib.suppress(StopIteration):
+                while True:
+                        yield i.next(max(expire - time.time(), 0))
+        while True:
+                yield i.next(max(expire - time.time(), 0))

tests/quotedKeyword.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:07:07 2020
 // Update Count     : 25
+// Last Modified On : Tue Dec  4 21:45:53 2018
+// Update Count     : 23
 //
 …
 struct {
         int ``otype;
         int ``struct;
+        int `otype`;
+        int `struct`;
 } st = { 10, 10 };
 typedef int ``forall;
 ``forall xxx = 10;
+typedef int `forall`;
+`forall` xxx = 10;
 int ``_Alignas, ``_Alignof, ``__alignof, ``__alignof__, ``asm, ``__asm, ``__asm__, ``_At, ``_Atomic, ``__attribute,
         ``__attribute__, ``auto, ``_Bool, ``break, ``case, ``catch, ``catchResume, ``char, ``choose, ``_Complex, ``__complex,
         ``__complex__, ``const, ``__const, ``__const__, ``continue, ``default, ``disable, ``do, ``double, ``dtype, ``else,
         ``enable, ``enum, ``__extension__, ``extern, ``fallthru, ``finally, ``float, ``__float128, ``for, ``forall, ``fortran,
         ``ftype, ``_Generic, ``goto, ``if, ``_Imaginary, ``__imag, ``__imag__, ``inline, ``__inline, ``__inline__, ``int,
         ``__int128, ``__label__, ``long, ``lvalue, ``_Noreturn, ``__builtin_offsetof, ``otype, ``register, ``restrict,
         ``__restrict, ``__restrict__, ``return, ``short, ``signed, ``__signed, ``__signed__, ``sizeof, ``static,
         ``_Static_assert, ``struct, ``switch, ``_Thread_local, ``throw, ``throwResume, ``trait, ``try, ``typedef,
         ``typeof, ``__typeof, ``__typeof__, ``union, ``unsigned, ``__builtin_va_list, ``void, ``volatile, ``__volatile,
         ``__volatile__, ``while;
+int `_Alignas`, `_Alignof`, `__alignof`, `__alignof__`, `asm`, `__asm`, `__asm__`, `_At`, `_Atomic`, `__attribute`,
+        `__attribute__`, `auto`, `_Bool`, `break`, `case`, `catch`, `catchResume`, `char`, `choose`, `_Complex`, `__complex`,
+        `__complex__`, `const`, `__const`, `__const__`, `continue`, `default`, `disable`, `do`, `double`, `dtype`, `else`,
+        `enable`, `enum`, `__extension__`, `extern`, `fallthru`, `finally`, `float`, `__float128`, `for`, `forall`, `fortran`,
+        `ftype`, `_Generic`, `goto`, `if`, `_Imaginary`, `__imag`, `__imag__`, `inline`, `__inline`, `__inline__`, `int`,
+        `__int128`, `__label__`, `long`, `lvalue`, `_Noreturn`, `__builtin_offsetof`, `otype`, `register`, `restrict`,
+        `__restrict`, `__restrict__`, `return`, `short`, `signed`, `__signed`, `__signed__`, `sizeof`, `static`,
+        `_Static_assert`, `struct`, `switch`, `_Thread_local`, `throw`, `throwResume`, `trait`, `try`, `typedef`,
+        `typeof`, `__typeof`, `__typeof__`, `union`, `unsigned`, `__builtin_va_list`, `void`, `volatile`, `__volatile`,
+        `__volatile__`, `while`;
 int main() {
         int ``if = 0;
         ``catch = 1;
         st.``otype = 2;
         st.``struct = 3;
         ``throw = 4;
         sout | ``catch + st.``otype + st.``struct + ``throw;
+        int `if` = 0;
+        `catch` = 1;
+        st.`otype` = 2;
+        st.`struct` = 3;
+        `throw` = 4;
+        sout | `catch` + st.`otype` + st.`struct` + `throw`;
+}

tests/raii/dtor-early-exit.cfa

-              r7030dab
+              r71d6bd8
+}
-void i() {
-        // potential loop
-        for() {
-                if(true) continue;
-                int t = 0;
+        }
+}
 // TODO: implement __label__ and uncomment these lines
 void computedGoto() {

tests/rational.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Mon Mar 28 08:43:12 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb  8 18:46:23 2020
 // Update Count     : 86
+// Last Modified On : Wed Mar 27 07:37:17 2019
+// Update Count     : 80
 //
 …
 #include <fstream.hfa>
+typedef Rational(int) RatInt;
+double convert( int i ) { return (double)i; }                   // used by narrow/widen
+double convert( int i ) { return (double)i; }
 int convert( double d ) { return (int)d; }
 int main() {
         sout | "constructor";
         RatInt a = { 3 }, b = { 4 }, c, d = 0, e = 1;
         sout | a | b | c | d | e;
+        Rational(int) a = { 3 }, b = { 4 }, c;
+        sout | a | b | c;
         a = (RatInt){ 4, 8 };
         b = (RatInt){ 5, 7 };
+        a = (Rational(int)){ 4, 8 };
+        b = (Rational(int)){ 5, 7 };
         sout | a | b;
         a = (RatInt){ -2, -3 };
         b = (RatInt){ 3, -2 };
+        a = (Rational(int)){ -2, -3 };
+        b = (Rational(int)){ 3, -2 };
         sout | a | b;
         a = (RatInt){ -2, 3 };
         b = (RatInt){ 3, 2 };
+        a = (Rational(int)){ -2, 3 };
+        b = (Rational(int)){ 3, 2 };
         sout | a | b;
         sout | "logical";
         a = (RatInt){ -2 };
         b = (RatInt){ -3, 2 };
+        a = (Rational(int)){ -2 };
+        b = (Rational(int)){ -3, 2 };
         sout | a | b;
 //      sout | a == 1; // FIX ME
 …
         sout | "conversion";
         a = (RatInt){ 3, 4 };
+        a = (Rational(int)){ 3, 4 };
         sout | widen( a );
         a = (RatInt){ 1, 7 };
+        a = (Rational(int)){ 1, 7 };
         sout | widen( a );
         a = (RatInt){ 355, 113 };
+        a = (Rational(int)){ 355, 113 };
         sout | widen( a );
         sout | narrow( 0.75, 4 );
 …
         sout | "more tests";
         RatInt x = { 1, 2 }, y = { 2 };
+        Rational(int) x = { 1, 2 }, y = { 2 };
         sout | x - y;
         sout | x > y;
 …
         sout | y | denominator( y, -2 ) | y;
         RatInt z = { 0, 5 };
+        Rational(int) z = { 0, 5 };
         sout | z;
         sout | x | numerator( x, 0 ) | x;
         x = (RatInt){ 1, MAX } + (RatInt){ 1, MAX };
+        x = (Rational(int)){ 1, MAX } + (Rational(int)){ 1, MAX };
         sout | x;
         x = (RatInt){ 3, MAX } + (RatInt){ 2, MAX };
+        x = (Rational(int)){ 3, MAX } + (Rational(int)){ 2, MAX };
         sout | x;

tests/references.cfa

-              r7030dab
+              r71d6bd8
                 f( 3, a + b, (S){ 1.0, 7.0 }, (int [3]){ 1, 2, 3 } ); // two rvalue to reference
+        }
+        {
-                int a = 3;
-                int *p = &a;
-                asm (
-                        "incl %[p]\n\t"
-                        : [p] "+m" (*p)
-                );
-                printf("%d\n", a);
+        }
+}

tests/test.py

-              r7030dab
+              r71d6bd8
 import argparse
-import itertools
 import re
 import sys
 import tempfile
 import time
-import os
-import psutil
-import signal
 ################################################################################
 …
                         test.path = match.group(1)
                         test.arch = match.group(3)[1:] if match.group(3) else None
+                        expected.append(test)
+                        if settings.arch.match(test.arch):
+                                expected.append(test)
         path_walk( match_test )
 …
                         x.target().startswith( tuple(excludes) )
+                ]
-        # sort the test alphabetically for convenience
-        test_list.sort(key=lambda t: ('~' if t.arch else '') + t.target() + (t.arch if t.arch else ''))
         return test_list
 …
                         if test :
                                 tests.extend( test )
+                                tests.append( test[0] )
                         else :
                                 print('ERROR: No expected file for test %s, ignoring it' % testname, file=sys.stderr)
 …
         # create a parser with the arguments for the tests script
         parser = argparse.ArgumentParser(description='Script which runs cforall tests')
+        parser.add_argument('--debug', help='Run all tests in debug or release', type=comma_separated(yes_no), default='yes')
+        parser.add_argument('--install', help='Run all tests based on installed binaries or tree binaries', type=comma_separated(yes_no), default='no')
+        parser.add_argument('--arch', help='Test for specific architecture', type=comma_separated(str), default='')
+        parser.add_argument('--continue', help='When multiple specifications are passed (debug/install/arch), sets whether or not to continue if the last specification failed', type=yes_no, default='yes', dest='continue_')
+        parser.add_argument('--debug', help='Run all tests in debug or release', type=yes_no, default='yes')
+        parser.add_argument('--install', help='Run all tests based on installed binaries or tree binaries', type=yes_no, default='no')
+        parser.add_argument('--arch', help='Test for specific architecture', type=str, default='')
         parser.add_argument('--timeout', help='Maximum duration in seconds after a single test is considered to have timed out', type=int, default=60)
         parser.add_argument('--global-timeout', help='Maximum cumulative duration in seconds after the ALL tests are considered to have timed out', type=int, default=7200)
-        parser.add_argument('--timeout-with-gdb', help='Instead of killing the command when it times out, orphan it and print process id to allow gdb to attach', type=yes_no, default="no")
         parser.add_argument('--dry-run', help='Don\'t run the tests, only output the commands', action='store_true')
         parser.add_argument('--list', help='List all test available', action='store_true')
 …
         # run everything in a temp directory to make sure core file are handled properly
         with tempdir():
                 # if the make command succeeds continue otherwise skip to diff
+                # if the make command succeds continue otherwise skip to diff
                 if success(make_ret):
                         with Timed() as run_dur:
 …
         make('clean', output_file=subprocess.DEVNULL, error=subprocess.DEVNULL)
-        # since python prints stacks by default on a interrupt, redo the interrupt handling to be silent
-        def worker_init():
-                def sig_int(signal_num, frame):
-                        pass
-                signal.signal(signal.SIGINT, sig_int)
         # create the executor for our jobs and handle the signal properly
         pool = multiprocessing.Pool(jobs, worker_init)
+        pool = multiprocessing.Pool(jobs)
         failed = False
-        def stop(x, y):
-                print("Tests interrupted by user", file=sys.stderr)
-                sys.exit(1)
-        signal.signal(signal.SIGINT, stop)
         # for each test to run
 …
         make('clean', output_file=subprocess.DEVNULL, error=subprocess.DEVNULL)
         return failed
+        return 1 if failed else 0
 …
         settings.init( options )
+        # fetch the liest of all valid tests
+        all_tests = list_tests( options.include, options.exclude )
+        # if user wants all tests than no other treatement of the test list is required
+        if options.all or options.list or options.list_comp or options.include :
+                tests = all_tests
+        #otherwise we need to validate that the test list that was entered is valid
+        else :
+                tests = valid_tests( options )
+        # make sure we have at least some test to run
+        if not tests :
+                print('ERROR: No valid test to run', file=sys.stderr)
+                sys.exit(1)
+        # sort the test alphabetically for convenience
+        tests.sort(key=lambda t: (t.arch if t.arch else '') + t.target())
         # users may want to simply list the tests
         if options.list_comp :
+                # fetch the liest of all valid tests
+                tests = list_tests( None, None )
+                # print the possible options
+                print("-h --help --debug --dry-run --list --arch --all --regenerate-expected --archive-errors --install --timeout --global-timeout --timeout-with-gdb -j --jobs -I --include -E --exclude --continue ", end='')
+                print("-h --help --debug --dry-run --list --arch --all --regenerate-expected --archive-errors --install --timeout --global-timeout -j --jobs ", end='')
                 print(" ".join(map(lambda t: "%s" % (t.target()), tests)))
         elif options.list :
+                # fetch the liest of all valid tests
+                tests = list_tests( options.include, options.exclude )
+                # print the available tests
+                print("Listing for %s:%s"% (settings.arch.string, settings.debug.string))
                 fancy_print("\n".join(map(lambda t: t.toString(), tests)))
         else :
+                # fetch the liest of all valid tests
+                all_tests = list_tests( options.include, options.exclude )
+                # if user wants all tests than no other treatement of the test list is required
+                if options.all or options.include :
+                        tests = all_tests
+                #otherwise we need to validate that the test list that was entered is valid
+                else :
+                        tests = valid_tests( options )
+                # make sure we have at least some test to run
+                if not tests :
+                        print('ERROR: No valid test to run', file=sys.stderr)
+                        sys.exit(1)
+                # prep invariants
+                # check the build configuration works
                 settings.prep_output(tests)
+                failed = 0
+                # for each build configurations, run the test
+                for arch, debug, install in itertools.product(settings.all_arch, settings.all_debug, settings.all_install):
+                        settings.arch    = arch
+                        settings.debug   = debug
+                        settings.install = install
+                        # filter out the tests for a different architecture
+                        # tests are the same across debug/install
+                        local_tests = settings.arch.filter( tests )
+                        options.jobs, forceJobs = job_count( options, local_tests )
+                        settings.update_make_cmd(forceJobs, options.jobs)
+                        # check the build configuration works
+                        settings.validate()
+                        # print configuration
+                        print('%s %i tests on %i cores (%s:%s)' % (
+                                'Regenerating' if settings.generating else 'Running',
+                                len(local_tests),
+                                options.jobs,
+                                settings.arch.string,
+                                settings.debug.string
+                        ))
+                        # otherwise run all tests and make sure to return the correct error code
+                        failed = run_tests(local_tests, options.jobs)
+                        if failed:
+                                result = 1
+                                if not settings.continue_:
+                                        break
+                sys.exit( failed )
+                settings.validate()
+                options.jobs, forceJobs = job_count( options, tests )
+                settings.update_make_cmd(forceJobs, options.jobs)
+                print('%s %i tests on %i cores (%s:%s)' % (
+                        'Regenerating' if settings.generating else 'Running',
+                        len(tests),
+                        options.jobs,
+                        settings.arch.string,
+                        settings.debug.string
+                ))
+                # otherwise run all tests and make sure to return the correct error code
+                sys.exit( run_tests(tests, options.jobs) )

tests/time.cfa

-              r7030dab
+              r71d6bd8
 // Created On       : Tue Mar 27 17:24:56 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Jan  5 18:27:37 2020
 // Update Count     : 34
+// Last Modified On : Thu Dec 20 23:09:21 2018
+// Update Count     : 23
 //
 …
         Duration d1 = 3`h, d2 = 2`s, d3 = 3.375`s, d4 = 12`s, d5 = 1`s + 10_000`ns;
         sout | d1 | d2 | d3 | d4 | d5;
+        int i;
         d1 = 0;
         sout | d1 | d2 | d3;
 …
         sout | t;
         t = t + d1;
         sout | t | t`ns;
+        sout | t | t.tv;
         Time t1 = (timespec){ 104_414, 10_000_000 };
         sout | t1 | t1`ns;
         sout | t - t  | t + d5 | t`ns;
         char buf[64];
+        sout | t1 | t1.tv;
+        sout | t - t  | t + d5 | t.tv;
+        char buf[16];
         sout | "yy/mm/dd" | [t, buf]`ymd | nonl;                        // shared buf => separate calls
         sout | "mm/dd/yy" | mm_dd_yy( t, buf ) | nonl;
 …
         sout | "dd/yy/mm" | [t, buf]`dmy;
         Time t2 = { 2001, 7, 4, 0, 0, 1, 0 }, t3 = (timeval){ 994_219_201 };
         sout | t2 | t2`ns | nl | t3 | t3`ns;
+        sout | t2 | t2.tv | nl | t3 | t3.tv;
         sout | nl;
 …
         sout | "Dividing that by 2 gives" | s / 2 | "seconds";
         sout | s | "seconds is" | s`h | "hours," | (s % 1`h)`m | "minutes," | (s % 1`m)`s | "seconds";
-    t1 = (Time){ 2020, 1, 5, 9, 0, 0, 100000000000LL };
-    t2 = (Time){ 1969, 13, 5, 9 };
-    t3 = (Time){ 1970, 25, 366, 48, 120, -120, 60000000000LL };
-    strftime( buf, 128, "%Y %b %e %H:%M:%S (GMT)", t1 );
-    sout | buf;
-    strftime( buf, 128, "%Y %b %e %H:%M:%S (GMT)", t2 );
-    sout | buf;
-    strftime( buf, 128, "%Y %b %e %H:%M:%S (GMT)", t3 );
-    sout | buf;
 } // main

tests/userLiterals.cfa

-              r7030dab
+              r71d6bd8
 // file "LICENCE" distributed with Cforall.
 //
 // userLiterals.cfa --
+// user_literals.cfa --
 //
 // Author           : Peter A. Buhr
 // Created On       : Wed Sep  6 21:40:50 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Feb 19 07:48:45 2020
 // Update Count     : 74
+// Last Modified On : Tue Dec  4 22:03:10 2018
+// Update Count     : 56
 //
 …
 int ?`__thingy_( int x ) { sout | "_thingy_" | x; return x; }
 int ?`s( const char * s ) { sout | "s" | s; return 0; }
 int ?`m( const char16_t * m ) { sout | "m" | m; return 0;}
 int ?`h( const char32_t * h ) { sout | "h" | h; return 0; }
+int ?`s( const char * s ) { sout | "secs" | s; return 0; }
+int ?`m( const char16_t * m ) { sout | "mins" | m; return 0;}
+int ?`h( const char32_t * h ) { sout | "hours" | h; return 0; }
 int ?`_A_( const wchar_t * str ) { sout | "_A_" | str; return 0; }
 int ?`__thingy_( const char * str ) { sout | "_thingy_" | str; return 0; }
 …
         return (Weight){ l.stones + r.stones };
+}
+ofstream & ?|?( ofstream & os, Weight w ) { return os | wd(1,1, w.stones); }
+void ?|?( ofstream & os, Weight w ) { (ofstream)(os | w); ends( os ); }
+ofstream & ?|?( ofstream & os, Weight w ) { return os | w.stones; }
 Weight ?`st( double w ) { return (Weight){ w }; }               // backquote for user literals
 …
         sout | w;
 `s;
+//      0`secs;
 `s;
 `s;
 …
         "abc"`s;
+        // FIX ME: requires char16_t, char32_t, and wchar_t be unique types
+        // u"abc"`m;
+        // U_"abc"`h;
+        // L"abc"`_A_;
+//      u"abc"`m;
+//      U_"abc"`h;
+//      L"abc"`_A_;
         u8_"abc"`__thingy_;
 } // main
 …
 // Local Variables: //
 // tab-width: 4 //
 // compile-command: "cfa userLiterals.cfa" //
+// compile-command: "cfa user_literals.cfa" //
 // End: //

tools/catchsig.c

-              r7030dab
+              r71d6bd8
         printf("Starting...\n");
         sig(SIGHUP);
         sig(SIGINT);
         sig(SIGQUIT);
         sig(SIGILL);
         sig(SIGABRT);
         sig(SIGFPE);
         sig(SIGSEGV);
         sig(SIGPIPE);
         sig(SIGALRM);
         sig(SIGTERM);
         sig(SIGUSR1);
         sig(SIGUSR2);
         sig(SIGCHLD);
         sig(SIGCONT);
         sig(SIGTSTP);
         sig(SIGTTIN);
         sig(SIGTTOU);
+      sig(SIGINT);
+      sig(SIGQUIT);
+      sig(SIGILL);
+      sig(SIGABRT);
+      sig(SIGFPE);
+      sig(SIGSEGV);
+      sig(SIGPIPE);
+      sig(SIGALRM);
+      sig(SIGTERM);
+      sig(SIGUSR1);
+      sig(SIGUSR2);
+      sig(SIGCHLD);
+      sig(SIGCONT);
+      sig(SIGTSTP);
+      sig(SIGTTIN);
+      sig(SIGTTOU);
         while(1);
         return 0;

tools/cfa.nanorc

-              r7030dab
+              r71d6bd8
 # Declarations
 color brightgreen "\<(struct|union|typedef|trait|coroutine|generator)\>"
 color brightgreen "\<(monitor|thread|with)\>"
+color brightgreen "\<(struct|union|typedef|trait|coroutine|monitor|thread)\>"
+color brightgreen "\<(with)\>"
 # Control Flow Structures
 color brightyellow "\<(if|else|while|do|for|switch|choose|case|default)\>"
 color brightyellow "\<(disable|enable|waitfor|when|timeout|suspend)\>"
+color brightyellow "\<(disable|enable|waitfor|when|timeout)\>"
 color brightyellow "\<(try|catch(Resume)?|finally)\>"
 …
 # Escaped Keywords, now Identifiers.
 color white "``\w+"
+color white "`\w+`"
 # Operator Names
 …
 ## Update/Redistribute
 # GCC builtins
 color cyan "__attribute__[[:space:]]*\(\(([^)]|[^)]\))*\)\)"
+color cyan "__attribute__[[:space:]]*\(\([^()]*(\([^()]*\)[^()]*)*\)\)"
 ##color cyan "__(aligned|asm|builtin|hidden|inline|packed|restrict|section|typeof|weak)__"

tools/stat.py

r7030dab	r71d6bd8
17	17	avg = numpy.mean (content)
18	18	std = numpy.std (content)
19		print "median {0:.1f} avg {1:.1f} stddev {2:.1f}".format( med, avg, std )
	19	print "median {0:.1f} avg {1:.1f} stddev {2:.2f}".format( med, avg, std )
20	20
21	21

tools/vscode/uwaterloo.cforall-0.1.0/package.json

-              r7030dab
+              r71d6bd8
         "name": "cforall",
         "version": "0.1.0",
         "displayName": "C∀ (C-for-all) Language Support",
+        "displayName": "Cforall Language Support",
         "description": "Cforall - colorizer, grammar and snippets.",
         "publisher": "uwaterloo",
 …
                 "vscode": "^1.5.0"
         },
         "icon": "images/icon.png",
+        "icon": "images/icon.svg",
         "categories": [
                 "Programming Languages",
+                "Languages",
                 "Linters",
                 "Other"
         ],
-        "activationEvents": [
-                "onLanguage:cforall"
-        ],
-        "main": "./client/main.js",
         "contributes": {
                 "languages": [
 …
                                 "aliases": [
                                         "C∀",
+                                        "Cforall",
                                         "CForAll",
-                                        "Cforall",
                                         "cforall"
                                 ],
                                 "extensions": [
+                                        ".cfa",
+                                        ".hfa",
+                                        ".ifa"
+                                        ".cf"
                                 ],
                                 "configuration": "./cforall.configuration.json"
 …
+                        {
                                 "language": "cforall",
                                 "scopeName": "source.cfa",
                                 "path": "./syntaxes/cfa.tmLanguage.json"
+                                "scopeName": "source.cf",
+                                "path": "./syntaxes/cfa.tmLanguage"
+                        }
+                ],
+                "configuration": {
+                        "type": "object",
+                        "title": "Example configuration",
+                        "properties": {
+                                "cforall.maxNumberOfProblems": {
+                                        "scope": "resource",
+                                        "type": "number",
+                                        "default": 100,
+                                        "description": "Controls the maximum number of problems produced by the server."
+                                },
+                                "cforall.trace.server": {
+                                        "scope": "window",
+                                        "type": "string",
+                                        "enum": [
+                                                "off",
+                                                "messages",
+                                                "verbose"
+                                        ],
+                                        "default": "off",
+                                        "description": "Traces the communication between VS Code and the language server."
+                                }
+                        }
+                }
+        },
+        "dependencies": {
+                "vscode-languageclient": "^4.1.4"
+        },
+        "devDependencies": {
+                "vscode-languageclient": "^4.1.4"
+                ]
+        }
+}

Context Navigation

Changes in / [7030dab:71d6bd8]

Legend: