Changeset 04e6f93

Jenkinsfile

-              r41efd33
+              r04e6f93
                 //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, 'signal'        , groupConcurrency, false, 'Internal and External Scheduling')
                 do_plot(Settings.RunBenchmark && Settings.Publish, 'signal.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, 'scheduling'     , groupConcurrency, false, 'Internal and External Scheduling')
+                do_plot(Settings.RunBenchmark && Settings.Publish, 'scheduling.diff', groupConcurrency, true , 'Internal and External Scheduling (relative)')
+        }
+}

benchmark/Makefile.am

-              r41efd33
+              r04e6f93
 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_UPP_verbose = $(AM_V_UPP)
 __bench_v_GOC_verbose = $(AM_V_GOC)
+__bench_v_RUSTC_verbose = $(AM_V_RUSTC)
+__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)
 …
         echo "int main() { return 0; }" > ${@}
+#.SILENT:               # do not print recipe
+.ONESHELL:              # use one shell to execute recipe
+.SILENT:                # do not print recipe
 .NOTPARALLEL:
 .PHONY: compile.csv basic.csv ctxswitch.csv mutex.csv schedint.csv
+.PHONY: jenkins cleancsv
 ## =========================================================================================================
 …
 FIX_NEW_LINES = cat $@ | tr "\n" "\t" | sed -r 's/\t,/,/' | tr "\t" "\n" > $@
+jenkins$(EXEEXT):
+cleancsv:
+        rm -f compile.csv basic.csv ctxswitch.csv mutex.csv scheduling.csv
+jenkins$(EXEEXT): cleancsv
 @DOifskipcompile@
         +make compile.csv
         -+make compile.diff.csv
 @DOendif@
-        +make basic.csv
-        -+make basic.diff.csv
         +make ctxswitch.csv
         -+make ctxswitch.diff.csv
         +make mutex.csv
         -+make mutex.diff.csv
         +make schedint.csv
         -+make schedint.diff.csv
+        +make scheduling.csv
+        -+make scheduling.diff.csv
 @DOifskipcompile@
         cat compile.csv
         -cat compile.diff.csv
 @DOendif@
-        cat basic.csv
-        -cat basic.diff.csv
         cat ctxswitch.csv
         -cat ctxswitch.diff.csv
         cat mutex.csv
         -cat mutex.diff.csv
         cat schedint.csv
         -cat schedint.diff.csv
+        cat scheduling.csv
+        -cat scheduling.diff.csv
 compile.csv:
+        echo "building $@"
         echo "array,attributes,empty,expression,io,monitor,operators,typeof" > $@
         +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-array.make >> $@
 …
         $(srcdir)/fixcsv.sh $@
-basic.csv:
-        echo "generator,coroutine,thread" > $@
-        +make basic-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
-        +make basic-cfa_coroutine.runquiet >> $@ && echo -n ',' >> $@
-        +make basic-cfa_thread.runquiet >> $@
-        $(srcdir)/fixcsv.sh $@
 ctxswitch.csv:
+        echo "building $@"
         echo "generator,coroutine,thread" > $@
         +make ctxswitch-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
 …
 mutex.csv:
+        echo "building $@"
         echo "1-monitor,2-monitor" > $@
         +make mutex-cfa1.runquiet >> $@ && echo -n ',' >> $@
 …
         $(srcdir)/fixcsv.sh $@
+schedint.csv:
+scheduling.csv:
+        echo "building $@"
         echo "schedint-1,schedint-2,schedext-1,schedext-2" > $@
         +make schedint-cfa1.runquiet >> $@ && echo -n ',' >> $@
 …
 ctxswitch-python_coroutine$(EXEEXT):
         echo "#!/bin/sh" > a.out
+        $(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):
         echo "#!/bin/sh" > a.out
+        $(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):
         echo "#!/bin/sh" > a.out
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
         echo "nodejs $(srcdir)/ctxswitch/node_await.js" >> a.out
         chmod a+x a.out
 …
 creation-python_coroutine$(EXEEXT):
         echo "#!/bin/sh" > a.out
+        $(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):
         echo "#!/bin/sh" > a.out
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
         echo "nodejs $(srcdir)/creation/node_cor.js" >> a.out
         chmod a+x a.out
 …
 compile-typeof$(EXEEXT):
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/typeof.cfa
+## =========================================================================================================
+size$(EXEEXT) : size-cfa.runquiet
+size-cfa$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/size/size.cfa

benchmark/Makefile.in

-              r41efd33
+              r04e6f93
 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_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_0 = @echo "  NODEJS  " $@;
 am__v_NODEJS_1 =
 AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)
 …
 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_UPP_verbose = $(AM_V_UPP)
 __bench_v_GOC_verbose = $(AM_V_GOC)
+__bench_v_RUSTC_verbose = $(AM_V_RUSTC)
+__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)
 …
         echo "int main() { return 0; }" > ${@}
+#.SILENT:               # do not print recipe
+.ONESHELL:              # use one shell to execute recipe
+.SILENT:                # do not print recipe
 .NOTPARALLEL:
 .PHONY: compile.csv basic.csv ctxswitch.csv mutex.csv schedint.csv
+.PHONY: jenkins cleancsv
 all : basic$(EXEEXT) ctxswitch$(EXEEXT) mutex$(EXEEXT) schedint$(EXEEXT) schedext$(EXEEXT) creation$(EXEEXT)
 …
         +make -C ${abs_top_builddir}/tools repeat
+jenkins$(EXEEXT):
+cleancsv:
+        rm -f compile.csv basic.csv ctxswitch.csv mutex.csv scheduling.csv
+jenkins$(EXEEXT): cleancsv
 @DOifskipcompile@
         +make compile.csv
         -+make compile.diff.csv
 @DOendif@
-        +make basic.csv
-        -+make basic.diff.csv
         +make ctxswitch.csv
         -+make ctxswitch.diff.csv
         +make mutex.csv
         -+make mutex.diff.csv
         +make schedint.csv
         -+make schedint.diff.csv
+        +make scheduling.csv
+        -+make scheduling.diff.csv
 @DOifskipcompile@
         cat compile.csv
         -cat compile.diff.csv
 @DOendif@
-        cat basic.csv
-        -cat basic.diff.csv
         cat ctxswitch.csv
         -cat ctxswitch.diff.csv
         cat mutex.csv
         -cat mutex.diff.csv
         cat schedint.csv
         -cat schedint.diff.csv
+        cat scheduling.csv
+        -cat scheduling.diff.csv
 compile.csv:
+        echo "building $@"
         echo "array,attributes,empty,expression,io,monitor,operators,typeof" > $@
         +make TIME_FORMAT='%e,' PRINT_FORMAT='' compile-array.make >> $@
 …
         $(srcdir)/fixcsv.sh $@
-basic.csv:
-        echo "generator,coroutine,thread" > $@
-        +make basic-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
-        +make basic-cfa_coroutine.runquiet >> $@ && echo -n ',' >> $@
-        +make basic-cfa_thread.runquiet >> $@
-        $(srcdir)/fixcsv.sh $@
 ctxswitch.csv:
+        echo "building $@"
         echo "generator,coroutine,thread" > $@
         +make ctxswitch-cfa_generator.runquiet >> $@ && echo -n ',' >> $@
 …
 mutex.csv:
+        echo "building $@"
         echo "1-monitor,2-monitor" > $@
         +make mutex-cfa1.runquiet >> $@ && echo -n ',' >> $@
 …
         $(srcdir)/fixcsv.sh $@
+schedint.csv:
+scheduling.csv:
+        echo "building $@"
         echo "schedint-1,schedint-2,schedext-1,schedext-2" > $@
         +make schedint-cfa1.runquiet >> $@ && echo -n ',' >> $@
 …
 ctxswitch-python_coroutine$(EXEEXT):
         echo "#!/bin/sh" > a.out
+        $(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):
         echo "#!/bin/sh" > a.out
+        $(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):
         echo "#!/bin/sh" > a.out
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
         echo "nodejs $(srcdir)/ctxswitch/node_await.js" >> a.out
         chmod a+x a.out
 …
 creation-python_coroutine$(EXEEXT):
         echo "#!/bin/sh" > a.out
+        $(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):
         echo "#!/bin/sh" > a.out
+        $(BENCH_V_NODEJS)echo "#!/bin/sh" > a.out
         echo "nodejs $(srcdir)/creation/node_cor.js" >> a.out
         chmod a+x a.out
 …
         $(CFACOMPILE) -fsyntax-only -w $(testdir)/typeof.cfa
+size$(EXEEXT) : size-cfa.runquiet
+size-cfa$(EXEEXT):
+        $(BENCH_V_CFA)$(CFACOMPILE) $(srcdir)/size/size.cfa
 # Tell versions [3.59,3.63) of GNU make to not export all variables.
 # Otherwise a system limit (for SysV at least) may be exceeded.

benchmark/baselines/x64/schedint.csv

r41efd33	r04e6f93
1		~~waitfor-2,signal-1,signal-2,waitfor~~-1
	1	schedext-2,schedint-1,schedint-2,schedext-1
2	2	393.69606249999987,325.99158333333327,409.01025000000004,319.90975000000003

benchmark/baselines/x86/schedint.csv

r41efd33	r04e6f93
1		s~~ignal-2,waitfor-1,waitfor-2,signal~~-1
	1	schedint-2,schedext-1,schedext-2,schedint-1
2	2	532.5297959183672,413.3084897959184,506.7579591836735,423.78826530612247

doc/bibliography/pl.bib

-              r41efd33
+              r04e6f93
 %    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},
+    keywords    = {Scala, polymorphism, subtyping, type inference},
     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, Minnesota, U.S.A.},
+    note        = {OOPSLA'00, Oct. 15--19, 2000, Minneapolis, Minn., 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

-              r41efd33
+              r04e6f93
 \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{.}{\protect\ETAL}%
                 {\protect\ETAL.\xspace}%
+        \@ifnextchar{.}{\ETAL}%
+                {\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, __restrict, __restrict__, __signed, __signed__, _Static_assert, thread,
+                otype, restrict, resume, __restrict, __restrict__, __signed, __signed__, _Static_assert, suspend, 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{0.4ex}{\rule{0.8ex}{0.1ex}}}}1 {^}{\raisebox{0.6ex}{$\scriptstyle\land\,$}}1
+        literate={-}{\makebox[1ex][c]{\raisebox{0.5ex}{\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{#1}}
+{\lstset{language=uC++,moredelim=**[is][\protect\color{red}]{`}{`},#1}\lstset{#1}}
 {}
 \lstnewenvironment{Go}[1][]
 …
 \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 features 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 approaches 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}
+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.
+\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".
 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 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 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 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, 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 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 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, \ie 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, \eg 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 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.
+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.
 } (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.
+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.
+(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.
 The main contributions of this work are:
 \begin{itemize}[topsep=3pt,itemsep=1pt]
+\begin{itemize}[topsep=3pt,itemsep=0pt]
 \item
+language-level generators, coroutines and user-level threading, which respect the expectations of C programmers.
+a set of fundamental execution properties that dictate which language-level control-flow features need to be supported,
 \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.
+integration of these language-level control-flow features, while respecting the style and expectations of C programmers,
 \item
+providing statically type-safe interfaces that integrate with the \CFA polymorphic type-system and other language features.
+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,
 % \item
 % library extensions for executors, futures, and actors built on the basic mechanisms.
 \item
+a runtime system with no spurious wakeup.
+a runtime system without spurious wake-up and no performance loss,
 \item
+a dynamic partitioning mechanism to segregate the execution environment for specialized requirements.
+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).
 % \item
 % a non-blocking I/O library
 \item
 experimental results showing comparable performance of the new features with similar mechanisms in other programming languages.
+experimental results showing comparable performance of the \CFA features with similar mechanisms in other languages.
 \end{itemize}
+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: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: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 OpenJDK-9, Go 1.12.6 and \uC 7.0.0.
+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.
 …
 \label{s:StatefulFunction}
+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.
+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|.
+}%
 \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 fn1, fn;  void * `next`;
+        int `restart`, fn1, fn;
 } Fib;
 #define FibCtor { 1, 0, NULL }
+#define FibCtor { `0`, 1, 0 }
 Fib * comain( Fib * f ) {
+        if ( f->next ) goto *f->next;
+        f->next = &&s1;
+        `static void * states[] = {&&s0, &&s1};`
+        `goto *states[f->restart];`
+  s0: f->`restart` = 1;
         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 asymmetric generator]{\label{f:CFibonacci}\usebox\myboxA}
+\subfloat[C]{\label{f:CFibonacci}\usebox\myboxA}
 \hspace{3pt}
 \vrule
 \hspace{3pt}
 \subfloat[\CFA asymmetric generator]{\label{f:CFAFibonacciGen}\usebox\myboxB}
+\subfloat[\CFA]{\label{f:CFAFibonacciGen}\usebox\myboxB}
 \hspace{3pt}
 \vrule
 \hspace{3pt}
 \subfloat[C generator implementation]{\label{f:CFibonacciSim}\usebox\myboxC}
+\subfloat[C generated code for \CFA version]{\label{f:CFibonacciSim}\usebox\myboxC}
 \caption{Fibonacci (output) asymmetric generator}
 \label{f:FibonacciAsymmetricGenerator}
 …
 };
 void ?{}( Fmt & fmt ) { `resume(fmt);` } // constructor
 void ^?{}( Fmt & f ) with(f) { $\C[1.75in]{// destructor}$
+void ^?{}( Fmt & f ) with(f) { $\C[2.25in]{// destructor}$
         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 ) { $\C{// blocks}$
                                 `suspend;` $\C{// wait for character}$
                                 while ( ch == '\n' ) `suspend;` // ignore
                                 sout | ch;                                              // newline
                         } sout | " ";  // block spacer
                 } sout | nl; // group newline
+                                do { `suspend;` $\C{// wait for character}$
+                                while ( ch == '\n' ); // ignore newline
+                                sout | ch;                      $\C{// print character}$
+                        } sout | " ";  $\C{// block separator}$
+                } sout | nl; $\C{// group separator}$
+        }
+}
 …
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
 typedef struct {
         void * next;
+        int `restart`, g, b;
         char ch;
-        int g, b;
 } Fmt;
 void comain( Fmt * f ) {
+        if ( f->next ) goto *f->next;
+        f->next = &&s1;
+        `static void * states[] = {&&s0, &&s1};`
+        `goto *states[f->restart];`
+  s0: f->`restart` = 1;
         for ( ;; ) {
                 for ( f->g = 0; f->g < 5; f->g += 1 ) {
                         for ( f->b = 0; f->b < 4; f->b += 1 ) {
                                 return;
                           s1:;  while ( f->ch == '\n' ) return;
+                                do { return;  s1: ;
+                                } while ( f->ch == '\n' );
                                 printf( "%c", f->ch );
                         } printf( " " );
 …
+}
 int main() {
         Fmt fmt = { NULL };  comain( &fmt ); // prime
+        Fmt fmt = { `0` };  comain( &fmt ); // prime
         for ( ;; ) {
                 scanf( "%c", &fmt.ch );
 …
 \end{lrbox}
 \subfloat[\CFA asymmetric generator]{\label{f:CFAFormatGen}\usebox\myboxA}
 \hspace{3pt}
+\subfloat[\CFA]{\label{f:CFAFormatGen}\usebox\myboxA}
+\hspace{35pt}
 \vrule
 \hspace{3pt}
 \subfloat[C generator simulation]{\label{f:CFormatSim}\usebox\myboxB}
+\subfloat[C generated code for \CFA version]{\label{f:CFormatGenImpl}\usebox\myboxB}
 \hspace{3pt}
 \caption{Formatter (input) asymmetric generator}
 …
 \end{figure}
+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.
+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.
 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.
+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.
+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.
 The generator main contains @suspend@ statements that suspend execution without ending the generator versus @return@.
+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.}
+For the Fibonacci generator-main,
 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 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()|.
+}%
+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.
 \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()`; } $\C{// use function-call interface to skip N values}$
+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
+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();                                         $\C{// alternative interfaces}\CRT$
 \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.
+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.
+\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.
 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 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.
+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.
 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: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:
+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:
 \begin{center}
 \ldots\, STX \ldots\, message \ldots\, ESC ETX \ldots\, message \ldots\, ETX 2-byte crc \ldots
 \end{center}
 is a network message beginning with the control character STX, ending with an ETX, and followed by a 2-byte cyclic-redundancy check.
+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.
 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 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.
+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.
 \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;
         unsigned char byte, * msg; // communication
         unsigned int lnth, sum;      // local state
         unsigned short int crc;
+        char byte, * msg; // communication
+        int lnth, sum;      // local state
+        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.
 The steps for symmetric control-flow are creating, executing, and terminating the cycle.
+Figure~\ref{f:PingPongFullCoroutineSteps} shows 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 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).
+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).
 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, 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@.
+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.
 \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) {name,N,0,NULL,NULL}
+#define PPCtor(name, N) {0, N, 0, name, NULL}
 void comain( PingPong * pp ) {
+        if ( pp->next ) goto *pp->next;
+        pp->next = &&cycle;
+        static void * states[] = {&&s0, &&s1};
+        goto *states[pp->restart];
+  s0: pp->restart = 1;
         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( "popq %rbx" );
+                asm( "add  $16, %rsp" );
+                asm( "popq %rbp" );
                 asm( "jmp  comain" );
           cycle: ;
+          s1: ;
+        }
+}
 …
 \end{figure}
+Finally, part of this generator work was inspired by the recent \CCtwenty generator proposal~\cite{C++20Coroutine19} (which they call coroutines).
+\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.
 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 extend generator semantics, \ie there is an implicit closure and @suspend@ may appear in a helper function called from the coroutine main.
+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.
 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, execution, due to context switching among stacks, and terminating, due to possible stack unwinding and dynamic stack deallocation.
+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.
 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{description}
+\item[Fibonacci]
+Move the declaration of @fn1@ to the start of 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{cfa}[xleftmargin=0pt]
 void main( Fib & fib ) with(fib) {
+void main( Fib & fib ) ...
         `int fn1;`
+\end{cfa}
+\item[Formatter]
+Move the declaration of @g@ and @b@ to the for loops in the coroutine main.
+\end{cfa}
+&
 \begin{cfa}[xleftmargin=0pt]
 for ( `g`; 5 ) {
         for ( `b`; 4 ) {
+\end{cfa}
+\item[Device Driver]
+Move the declaration of @lnth@ and @sum@ to their points of initialization.
+\end{cfa}
+&
 \begin{cfa}[xleftmargin=0pt]
+        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.
+status = CONT;
+`int lnth = 0, sum = 0;`
+...
+`short int crc = byte << 8;`
+\end{cfa}
+&
 \begin{cfa}[xleftmargin=0pt]
 void main( PingPong & pp ) with(pp) {
+void main( PingPong & pp ) ...
         for ( `i`; N ) {
+\end{cfa}
+\end{description}
+\end{cfa}
+\end{tabular}
+\end{center}
 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}
 unsigned int Crc() {
+int Crc() {
         `suspend;`
         unsigned short int crc = byte << 8;
+        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 @next@.
+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@.
 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 @next@, takes a Fibonacci instance and context switches to it using @resume@;
+The interface function @restart@, 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;                       // communication
+        Cons & c;                       $\C[1.5in]{// communication}$
         int N, money, receipt;
 };
 void main( Prod & prod ) with( prod ) {
+        // 1st resume starts here
+        for ( i; N ) {
+        for ( i; N ) {          $\C{// 1st resume}\CRT$
                 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;                       // communication
+        Prod & p;                       $\C[1.5in]{// communication}$
         int p1, p2, status;
         bool done;
 };
 void ?{}( Cons & cons, Prod & p ) {
         &cons.p = &p; // reassignable reference
+        &cons.p = &p;           $\C{// reassignable reference}$
         cons.[status, done ] = [0, false];
+}
 void main( Cons & cons ) with( cons ) {
+        // 1st resume starts here
+        int money = 1, receipt;
+        int money = 1, receipt; $\C{// 1st resume}\CRT$
         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 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 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 producer call to @delivery@ transfers values into the consumer's communication variables, resumes the consumer, and returns the consumer status.
+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.
+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.
 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 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.
+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?
 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 context switch back to the first resumer (starter) once the coroutine ends.
+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}).
 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.
+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 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 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.
+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.
+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.
 …
 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.
+\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.
 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 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.
+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.
 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 task is similar).
+Figure~\ref{f:CoroutineMemoryLayout} shows different memory-layout options for a coroutine (where a thread 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 VLS\footnote{
 We are examining variable-sized structures (VLS), where fields can be variable-sized structures or arrays.
+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.
 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 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.
+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.
 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 may be impossible.
+Split-stack allocation is under development but requires recompilation of legacy code, which is not always possible.
 \begin{figure}
 …
 Concurrency is nondeterministic scheduling of independent sequential execution paths (threads), where each thread has its own stack.
 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.
+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.
 (It is \emph{impossible} to generate a concurrency error when coroutining.)
+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}.
+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}, 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.
+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.
 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}, which restrict nondeterminism for correctness;
+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;
 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 either be a stackless or stackful.
+A scheduler can also be 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 needs the ability to start a thread and wait for its completion.
+Threading (Table~\ref{t:ExecutionPropertyComposition} case 11) needs the ability to start a thread and wait for its completion.
 A common API for this ability is @fork@ and @join@.
+\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(...);
+\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(...);
 `t.start();` // start
 // concurrency
 …
+&
 \begin{cfa}
 class MyTask { ... } // functor
 MyTask mytask;
 `thread t( mytask, ... );` // start
+class MyThread { ... } // functor
+MyThread mythread;
+`thread t( mythread, ... );` // start
 // concurrency
 `t.join();` // wait
 …
 \end{cfa}
 \end{tabular}
+\end{cquote}
+\vspace{1pt}
+\par\noindent
 \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 MyTask {};
 void main( MyTask & this ) { ... }
+thread MyThread {};
+void main( MyThread & this ) { ... }
 int main() {
         MyTask team`[10]`; $\C[2.5in]{// allocate stack-based threads, implicit start after construction}$
+        MyThread 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}
 MyTask * factory( int N ) { ... return `anew( N )`; } $\C{// allocate heap-based threads, implicit start after construction}$
+MyThread * factory( int N ) { ... return `anew( N )`; } $\C{// allocate heap-based threads, implicit start after construction}$
 int main() {
         MyTask * team = factory( 10 );
+        MyThread * 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 task-interface functions.
+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.
 \begin{cquote}
 \begin{tabular}{@{}c@{\hspace{3\parindentlnth}}c@{}}
 …
 \label{s:MutualExclusionSynchronization}
 Unrestricted nondeterminism is meaningless as there is no way to know when the result is completed without synchronization.
+Unrestricted nondeterminism is meaningless as there is no way to know when a result is completed and safe to access.
 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}.
+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).
+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).
 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 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.
+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.
 …
 \label{s:Monitor}
+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).
+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).
 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}[morekeywords=nomutex]
+\begin{cfa}
 `monitor` Aint { int cnt; }; $\C[4.25in]{// atomic integer counter}$
+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@.
+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.
 \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} mechanisms like \CC RAII;
+monitor \emph{implicit} safety ensures no programmer usage errors.
+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.
 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.
+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@.
+\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 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 at most one level of indirection, excluding pointers as it is impossible to statically determine the size.
+Therefore, \CFA only acquires one monitor per parameter with exactly one level of indirection, and exclude pointer types to unknown sized arrays.
 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 acquisition order is consistent across calls to @mutex@ functions using the same monitors as arguments, so acquiring multiple monitors is safe from deadlock.
+\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.
 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 );`
+        `scoped_lock lock( my.m, your.m );` // bulk acquire
         deposit( my, -me2you ); // debit
         deposit( your, me2you ); // credit
 …
 \end{figure}
 Users can still force the acquiring order by using @mutex@/\lstinline[morekeywords=nomutex]@nomutex@.
+Users can still force the acquiring order by using or not using @mutex@.
 \begin{cfa}
 void foo( M & mutex m1, M & mutex m2 ); $\C{// acquire m1 and m2}$
 void bar( M & mutex m1, M & /* nomutex */ m2 ) { $\C{// acquire m1}$
+void bar( M & mutex m1, M & m2 ) { $\C{// only acquire m1}$
         ... foo( m1, m2 ); ... $\C{// acquire m2}$
+}
 void baz( M & /* nomutex */ m1, M & mutex m2 ) { $\C{// acquire m2}$
+void baz( M & m1, M & mutex m2 ) { $\C{// only acquire m2}$
         ... 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 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.
+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.
 Synchronization is generally achieved with internal~\cite{Hoare74} or external~\cite[\S~2.9.2]{uC++} scheduling.
+\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}.
+\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}.
 % \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 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.
+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.
 (Java conceptually moves the signalled thread to the calling queue, and hence, allows barging.)
+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.
+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.
 \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) & buffer ) with(buffer) {
+        void ?{}( Buffer(T) & buf ) with(buf) {
                 front = back = count = 0;
+        }
+        void insert( Buffer(T) & mutex buffer, T elem )
                                 with(buffer) {
                 if ( count == 10 ) `wait( empty )`;
                 // insert elem into buffer
+        void insert(Buffer(T) & mutex buf, T elm) with(buf){
+                if ( count == 10 ) `wait( empty )`; // full ?
+                // insert elm into buf
                 `signal( full )`;
+        }
         T remove( Buffer(T) & mutex buffer ) with(buffer) {
                 if ( count == 0 ) `wait( full )`;
                 // remove elem from buffer
+        T remove( Buffer(T) & mutex buf ) with(buf) {
+                if ( count == 0 ) `wait( full )`; // empty ?
+                // remove elm from buf
                 `signal( empty )`;
                 return elem;
+                return elm;
+        }
+}
 \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
+        int rcnt, wcnt; // readers/writer using resource
+        `condition RWers;`
 };
 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 ) `waitfor( EndWrite, rw );`
+        if ( wcnt !=0 || ! empty( RWers ) )
+                `wait( RWers, READER )`;
         rcnt += 1;
+        if ( ! empty(RWers) && `front(RWers) == READER` )
+                `signal( RWers )`;  // daisy-chain signalling
+}
 void StartWrite( ReadersWriter & mutex rw ) with(rw) {
         if ( wcnt > 0 ) `waitfor( EndWrite, rw );`
+        else while ( rcnt > 0 ) `waitfor( EndRead, rw );`
+        if ( wcnt != 0 || rcnt != 0 ) `wait( RWers, WRITER )`;
         wcnt = 1;
+}
 \end{cfa}
 \end{lrbox}
+\subfloat[Generic bounded buffer, internal scheduling]{\label{f:BBInt}\usebox\myboxA}
+\hspace{3pt}
+\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}
 \vrule
 \hspace{3pt}
 \subfloat[Readers / writer lock, external scheduling]{\label{f:RWExt}\usebox\myboxB}
 \caption{Internal / external scheduling}
 \label{f:InternalExternalScheduling}
+\subfloat[External scheduling]{\label{f:RWExt}\usebox\myboxB}
+\caption{Readers / writer lock}
+\label{f:ReadersWriterLock}
 \end{figure}
+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.
+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}.)
 \begin{figure}
 …
 };
 int girl( DS & mutex ds, int phNo, int ccode ) {
         if ( is_empty( Boys[ccode] ) ) {
+        if ( empty( Boys[ccode] ) ) {
                 wait( Girls[ccode] );
                 GirlPhNo = phNo;
 …
 };
 int girl( DS & mutex ds, int phNo, int ccode ) {
         if ( is_empty( Boys[ccode] ) ) { // no compatible
+        if ( 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}
 \label{f:DatingService}
+\caption{Dating service Monitor}
+\label{f:DatingServiceMonitor}
 \end{figure}
+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.
+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.
 Both internal and external scheduling extend to multiple monitors in a natural way.
 \begin{cquote}
 \begin{tabular}{@{}l@{\hspace{3\parindentlnth}}l@{}}
+\begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
 \begin{cfa}
 monitor M { `condition e`; ... };
 …
+&
 \begin{cfa}
 void rtn$\(_1\)$( M & mutex m1, M & mutex m2 );
+void rtn$\(_1\)$( M & mutex m1, M & mutex m2 ); // overload rtn
 void rtn$\(_2\)$( M & mutex m1 );
 void bar( M & mutex m1, M & mutex m2 ) {
         ... waitfor( `rtn` ); ...       // $\LstCommentStyle{waitfor( rtn\(_1\), m1, m2 )}$
         ... waitfor( `rtn, m1` ); ... // $\LstCommentStyle{waitfor( rtn\(_2\), m1 )}$
+        ... waitfor( `rtn`${\color{red}\(_1\)}$ ); ...       // $\LstCommentStyle{waitfor( rtn\(_1\) : m1, m2 )}$
+        ... waitfor( `rtn${\color{red}\(_2\)}$ : 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 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.
+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.
 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.
+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
 \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 monitors, not being perfectly composable.
+While deadlock can occur with multiple/nesting acquisition, this is a consequence of locks, and by extension monitor locking is not 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.
+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 )@.
 \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 restructred 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 restructured 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.
+Implementing a fast subset check for the necessary released monitors is important and discussed in the following sections.
 % 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{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.
+\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.
+(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 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.)
+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 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.
+% 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.
 \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; };
 thread GoRtn { int i;  float f;  Msg m; };
+monitor 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;
+                `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
+        }
 …
 \hspace{3pt}
 \subfloat[Go]{\label{f:Gochannel}\usebox\myboxB}
+\caption{Direct communication}
+\label{f:DirectCommunication}
+\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}
 \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{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}
+\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.
 …
 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 in the \CFA runtime, but we strongly advocate using high-level mechanisms whenever possible.
+Some of these low-level mechanism are used to build the \CFA runtime, but we always 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 threads and virtual processors (abstract kernel-thread) that execute the (user) threads from its own ready queue (like an OS executing kernel threads).
+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).
 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, affinityFreebsd, affinityNetbsd, affinityMacosx}, which is used when caching issues occur or for heterogeneous hardware processors.)
+(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
 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 applicaton workload do next.
+balancing the workload with processors is difficult because it requires future knowledge, \ie what will the application 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 Microsoft 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 Windows 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 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.
+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.
 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 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}
+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.
 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 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.
+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.
 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 \CFA/\uC are compiled with gcc 6.5.
+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.
 All benchmarks are run using the following harness. (The Java harness is augmented to circumvent JIT issues.)
 \begin{cfa}
+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.
+#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.
 All omitted tests for other languages are functionally identical to the \CFA tests and available online~\cite{CforallBenchMarks}.
+% 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}
+% tar --exclude-ignore=exclude -cvhf benchmark.tar benchmark
+\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 eliminate this cost.)
 Similarly, when modularization extends to coroutines/tasks, the time for a context switch becomes a relevant factor.
+(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.
 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} only shows the \CFA code for coroutines/threads (other systems are similar) with all results in Table~\ref{tab:ctx-switch}.
+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.
 \begin{multicols}{2}
 …
 \begin{cfa}[aboveskip=0pt,belowskip=0pt]
 @coroutine@ C {} c;
 void main( C & ) { for ( ;; ) { @suspend;@ } }
+void main( C & ) { while () { @suspend;@ } }
 int main() { // coroutine test
         BENCH( for ( N ) { @resume( c );@ } )
         sout | result`ns;
+}
 int main() { // task test
+        sout | result;
+}
+int main() { // thread test
         BENCH( for ( N ) { @yield();@ } )
         sout | result`ns;
+        sout | result;
+}
 \end{cfa}
 …
 \vspace*{-16pt}
 \captionof{table}{Context switch comparison (nanoseconds)}
 \label{tab:ctx-switch}
+\label{t: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   & 0.01  \\
+\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
+C function                      & 1.8           & 1.8           & 0.0   \\
+\CFA generator          & 1.8           & 1.8           & 0.1   \\
+\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
 \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.
+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}.
+\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}.
 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 __attribute__((noinline))
+do_call( M & @mutex m/*, 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;
+}
 int main() {
+        BENCH(
+                for( N ) do_call( m1/*, m2, m3, m4*/ );
+        )
+        sout | result`ns;
+}
+\end{cfa}
+\captionof{figure}{\CFA acquire/release mutex benchmark}
+\label{f:mutex}
+        T t;
+        wait( m1/*, m2, m3, m4*/ );
+}
+\end{cfa}
+\captionof{figure}{\CFA Internal-scheduling benchmark}
+\label{f:schedint}
 \columnbreak
 \vspace*{-16pt}
+\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
+\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
 \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:ext-sched} shows the code for \CFA, with results in Table~\ref{tab:ext-sched}.
+Figure~\ref{f:schedext} shows the code for \CFA with results in Table~\ref{t:schedext}.
 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}
+volatile int go = 0;
+@monitor@ M {} m;
+@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*/ );@ }
+}
 thread T {};
-void __attribute__((noinline))
-do_call( M & @mutex@ ) {}
 void main( T & ) {
+        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;
+        BENCH( for ( N ) { call( m1/*, m2, m3, m4*/ ); } )
+        sout | result;
+}
 int main() {
         T t;
         do_wait( m );
+        wait( m1/*, m2, m3, m4*/ );
+}
 \end{cfa}
 \captionof{figure}{\CFA external-scheduling benchmark}
 \label{f:ext-sched}
+\label{f:schedext}
 \columnbreak
 …
 \vspace*{-16pt}
 \captionof{table}{External-scheduling comparison (nanoseconds)}
 \label{tab:ext-sched}
+\label{t:schedext}
 \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@     & 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
+\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
 \end{tabular}
 \end{multicols}
+\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.
+\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.
 \begin{multicols}{2}
 \lstset{language=CFA,moredelim=**[is][\color{red}]{@}{@},deletedelim=**[is][]{`}{`}}
 \begin{cfa}
+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;
+}
+@monitor@ M {} m1/*, m2, m3, m4*/;
+call( M & @mutex p1/*, p2, p3, p4*/@ ) {}
 int main() {
         T t;
         do_wait( m );
+}
 \end{cfa}
 \captionof{figure}{\CFA Internal-scheduling benchmark}
 \label{f:int-sched}
+        BENCH( for( N ) call( m1/*, m2, m3, m4*/ ); )
+        sout | result;
+}
+\end{cfa}
+\captionof{figure}{\CFA acquire/release mutex benchmark}
+\label{f:mutex}
 \columnbreak
 \vspace*{-16pt}
+\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
+\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
 \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 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@.
+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@.
 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 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.
+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.
 …
 \label{futur:nbio}
 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).
+Many modern workloads are not bound by computation but IO operations, common cases 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 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.
+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.
 {%
 \fontsize{9bp}{12bp}\selectfont%
+\fontsize{9bp}{11.5bp}\selectfont%
 \bibliography{pl,local}
 }%

doc/papers/concurrency/examples/Fib.py

-              r41efd33
+              r04e6f93
         while True:
                 fn = fn1 + fn2; fn2 = fn1; fn1 = fn; yield fn
 f1 = Fib()
 …
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.5 Fib.py" #
+# compile-command: "python3.7 Fib.py" #
 # End: #

doc/papers/concurrency/examples/Fib2.c

-              r41efd33
+              r04e6f93
 #include <stdio.h>
-void mary() {
-        printf( "MARY\n" );
+}
 #define FIB_INIT { 0 }
 typedef struct { int next; int fn1, fn2; } Fib;
+typedef struct { int restart; int fn1, fn2; } Fib;
 int fib( Fib * f ) {
+        static void * states[] = { &&s1, &&s2, &&s3 };
+        goto *states[f->next];
+        static void * states[] = { &&s0, &&s1, &&s2 };
+        goto *states[f->restart];
+  s0:
+        f->fn1 = 0;
+        f->restart = 1;
+        return f->fn1;
   s1:
-        mary();
-        f->fn1 = 0;
-        f->next = 1;
-        return f->fn1;
-  s2:
-        mary();
         f->fn2 = f->fn1;
         f->fn1 = 1;
         f->next = 2;
+        f->restart = 2;
         return f->fn1;
+  s3:;
+        mary();
+  s2:;
         int fn = f->fn1 + f->fn2;
         f->fn2 = f->fn1;

doc/papers/concurrency/examples/Fib2.py

-              r41efd33
+              r04e6f93
 def Fib():
     fn1, fn = 0, 1
+    fn1, fn = 1, 0
     while True:
         yield fn1
+        yield fn
         fn1, fn = fn, fn1 + fn
 …
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.5 Fib2.py" #
+# compile-command: "python3.7 Fib2.py" #
 # End: #

doc/papers/concurrency/examples/Fib3.c

-              r41efd33
+              r04e6f93
 typedef struct {
+        int fn1, fn;
+        void * next;
+        int restart, fn1, fn;
 } Fib;
 #define FibCtor { 1, 0, NULL }
+#define FibCtor { 0, 1, 0 }
 Fib * comain( Fib * f ) {
+        if ( __builtin_expect(f->next != 0, 1) ) goto *f->next;
+        f->next = &&s1;
+        static void * states[] = {&&s0, &&s1};
+        goto *states[f->restart];
+  s0: f->restart = 1;
         for ( ;; ) {
                 return f;

doc/papers/concurrency/examples/FibRefactor.py

r41efd33	r04e6f93
22	22	# Local Variables: #
23	23	# tab-width: 4 #
24		# compile-command: "python3.5 FibRefactor.py" #
	24	# compile-command: "python3.7 FibRefactor.py" #
25	25	# End: #

doc/papers/concurrency/examples/Format.c

-              r41efd33
+              r04e6f93
 typedef struct {
         void * next;
+        int restart, g, b;
         char ch;
-        int g, b;
 } Fmt;
 void comain( Fmt * f ) {
+        if ( __builtin_expect(f->next != 0, 1) ) goto *f->next;
+        f->next = &&s1;
+        static void * states[] = {&&s0, &&s1};
+        goto *states[f->restart];
+  s0: f->restart = 1;
         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:;  while ( f->ch == '\n' ) return;         // ignore
+                                do {
+                                        return;  s1: ;
+                                } while ( f->ch == '\n' );                              // ignore
                                 printf( "%c", f->ch );                                  // print character
+                        }
 …
 int main() {
         Fmt fmt = { NULL };
+        Fmt fmt = { 0 };
         comain( &fmt );                                                                         // prime
         for ( ;; ) {

doc/papers/concurrency/examples/Format.cc

-              r41efd33
+              r04e6f93
                         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 -nodebubg Format.cc" //
+// compile-command: "u++-work -O2 -nodebug Format.cc" //
 // End: //

doc/papers/concurrency/examples/Format.cfa

-              r41efd33
+              r04e6f93
                 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

r41efd33	r04e6f93
1		abcdefghijklmnopqrstuvwxyzxxxxxxxxxxxxxx
	1	abcdefghijklmnop
	2	qrstuvwxyzx
	3	xxxxxxxxxxxxx

doc/papers/concurrency/examples/Format.py

-              r41efd33
+              r04e6f93
                         for g in range( 5 ):    # groups of 5 blocks
                                 for b in range( 4 ): # blocks of 4 characters
+                                        print( (yield), end='' ) # receive from send
+                                        while True:
+                                                ch = (yield) # receive from send
+                                                if '\n' not in ch:
+                                                        break
+                                        print( ch, 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 range( 41 ):
         fmt.send( 'a' );                                # send to yield
+for i in input:
+        fmt.send( i );                          # send to yield
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.5 Format.py" #
+# compile-command: "python3.7 Format.py" #
 # End: #

doc/papers/concurrency/examples/Format1.c

-              r41efd33
+              r04e6f93
 typedef struct {
         void * next;
+        int restart, g, b;
         char ch;
-        int g, b;
 } Fmt;
 void format( Fmt * f ) {
+        if ( __builtin_expect(f->next != 0, 1) ) goto *f->next;
+        f->next = &&s1;
+        static void * states[] = {&&s0, &&s1};
+        goto *states[f->restart];
+  s0: f->restart = 1;
         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 = { NULL };
+        Fmt fmt = { 0 };
         format( &fmt );                                                                         // prime
+        for ( ;; ) {
+                scanf( "%c", &fmt.ch );                                                 // direct read into communication variable
+          if ( feof( stdin ) ) break;
+        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;
                 format( &fmt );
+        }
         fmt.ch = '\0';
+        fmt.ch = '\0';                                                                          // sentential (EOF)
         format( &fmt );
+}

doc/papers/concurrency/examples/PingPong.c

-              r41efd33
+              r04e6f93
 typedef struct PingPong {
+        int restart;                                                                            // style 1
+        int N, i;
         const char * name;
-        int N, i;
         struct PingPong * partner;
         void * next;
+        void * next;                                                                            // style 2
 } PingPong;
+#define PPCtor( name, N ) { name, N, 0, NULL, NULL }
+#define PPCtor( name, N ) { 0, N, 0, name, NULL, NULL }
 void comain( PingPong * pp ) __attribute__(( noinline ));
 void comain( PingPong * pp ) {
+#if 0
         if ( __builtin_expect(pp->next != 0, 1) ) goto *pp->next;
-#if 0
-        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-8 -g -DPRINT PingPong.c" //
+// compile-command: "gcc-9 -g -DPRINT PingPong.c" //
 // End: //

doc/papers/concurrency/examples/Pingpong.py

-              r41efd33
+              r04e6f93
 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
+        while True:
+                n = next( n )           # schedule coroutine
+        n = yield                                       # starting coroutine
+        try:
+                while True:
+                        n = next( n )           # schedule coroutine
+        except StopIteration:
+                pass
 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:
         print( "scheduler stop" )
+except StopIteration:                   # scheduler stopped
+        pass
 print( "stop" )
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.5 Pingpong.py" #
+# compile-command: "python3.7 Pingpong.py" #
 # End: #

doc/papers/concurrency/examples/ProdCons.py

-              r41efd33
+              r04e6f93
 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
+        while True:
+                n = next( n )           # schedule coroutine
+        n = yield                                       # starting coroutine
+        try:
+                while True:
+                        n = next( n )           # schedule coroutine
+        except StopIteration:
+                pass
 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:
         print( "scheduler stop" )
+except StopIteration:                   # scheduler stopped
+        pass
 print( "stop" )
 # Local Variables: #
 # tab-width: 4 #
 # compile-command: "python3.5 ProdCons.py" #
+# compile-command: "python3.7 ProdCons.py" #
 # End: #

doc/papers/concurrency/examples/RWMonitorEXT.cfa

-              r41efd33
+              r04e6f93
         int rcnt, wcnt;                                                                         // number of 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 );
+        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 );
+        if ( wcnt > 0 ) waitfor( EndWrite : rw );
+        else while ( rcnt > 0 ) waitfor( EndRead : rw );
         wcnt = 1;
+}
-void ?{}( ReadersWriter & rw ) with(rw) { rcnt = wcnt = 0; }
 int readers( ReadersWriter & rw ) { return rw.rcnt; }
 void Read( ReadersWriter & rw ) {
         StartRead( rw );
 …
         EndWrite( rw );
+}
 thread Worker {
         ReadersWriter &rw;
 …
         } // for
+}
 int main() {
         enum { MaxTask = 5 };
 …
 } // main
 // Local Variables: //
 // tab-width: 4 //
 // compile-command: "cfa -O2 RWMonitor.cfa" //
+// compile-command: "cfa -O2 RWMonitorEXT.cfa" //
 // End: //

doc/papers/concurrency/examples/Refactor.py

r41efd33	r04e6f93
26	26	# Local Variables: #
27	27	# tab-width: 4 #
28		# compile-command: "python3.5 Refactor.py" #
	28	# compile-command: "python3.7 Refactor.py" #
29	29	# End: #

doc/papers/concurrency/figures/FullCoroutinePhases.fig

-              r41efd33
+              r04e6f93
 -2
 2
 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 0 1 0 7 100 0 -1 0.000 0 0 1 0 5175.000 2437.500 4875 1875 5175 1800 5475 1875
 1 1.00 45.00 90.00
 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 0 1 0 7 100 0 -1 0.000 0 0 1 0 5175.000 1537.500 5475 2100 5175 2175 4875 2100
 1 1.00 45.00 90.00
 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 0 1 0 7 50 -1 -1 0.000 0 1 1 0 4807.500 1642.500 4725 1425 4575 1650 4800 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 0 100 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
+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
+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
+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 2 165 705 2100 1500 pgm main\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
+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

doc/papers/concurrency/figures/RunTimeStructure.fig

-              r41efd33
+              r04e6f93
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4500 3600 15 15 4500 3600 4515 3615
 -6
+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
+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
+-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
+-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
+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
+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
+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
+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
 -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
+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
+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
 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 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
+2775 6825 3750
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 4
 1 1.00 45.00 90.00
+3975 7875 2325 7200 2325 7200 2550
+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
 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

-              r41efd33
+              r04e6f93
 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

driver/cfa.cc

-              r41efd33
+              r04e6f93
                 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++] = "-ldl";

libcfa/src/bits/containers.hfa

-              r41efd33
+              r04e6f93
         static inline forall( dtype T | is_node(T) ) {
                 void ?{}( __queue(T) & this ) with( this ) {
                         head{ 0p };
+                        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 ) {
+                        if( head != 1p ) {
                                 this.head = get_next( *head );
                                 if( !get_next( *head ) ) {
+                                if( get_next( *head ) == 1p ) {
                                         this.tail = &this.head;
+                                }
                                 get_next( *head ) = 0p;
+                        }
+                        return head;
+                                verify(*this.tail == 1p);
+                                return head;
+                        }
+                        verify(*this.tail == 1p);
+                        return 0p;
+                }
 …
                         get_next( *val ) = 0p;
                         verify( (head == 0p) == (&head == tail) );
                         verify( *tail == 0p );
+                        verify( (head == 1p) == (&head == tail) );
+                        verify( *tail == 1p );
                         return val;
+                }

libcfa/src/bits/locks.hfa

-              r41efd33
+              r04e6f93
+        }
-        extern void yield( unsigned int );
         static inline void ?{}( __spinlock_t & this ) {
                 this.lock = 0;
 …
         // 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 ) {
-                        disable_interrupts();
                         __cfaabi_dbg_record( this, caller );
+                } else {
+                        enable_interrupts_noPoll();
+                }
                 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
+                }
-                disable_interrupts();
                 __cfaabi_dbg_record( this, caller );
+        }
         static inline void unlock( __spinlock_t & this ) {
+                __atomic_clear( &this.lock, __ATOMIC_RELEASE );
                 enable_interrupts_noPoll();
-                __atomic_clear( &this.lock, __ATOMIC_RELEASE );
+        }

libcfa/src/concurrency/CtxSwitch-arm.S

-              r41efd33
+              r04e6f93
         .text
         .align  2
         .global CtxSwitch
         .type   CtxSwitch, %function
+        .global __cfactx_switch
+        .type   __cfactx_switch, %function
 CtxSwitch:
+__cfactx_switch:
         @ 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 CtxInvokeStub
         .type   CtxInvokeStub, %function
+        .global __cfactx_invoke_stub
+        .type   __cfactx_invoke_stub, %function
 CtxInvokeStub:
+__cfactx_invoke_stub:
         ldmfd r13!, {r0-r1}
         mov r15, r1

libcfa/src/concurrency/CtxSwitch-i386.S

-              r41efd33
+              r04e6f93
         .text
         .align 2
         .globl CtxSwitch
         .type  CtxSwitch, @function
 CtxSwitch:
+        .globl __cfactx_switch
+        .type  __cfactx_switch, @function
+__cfactx_switch:
         // Copy the "from" context argument from the stack to register eax
 …
         ret
         .size  CtxSwitch, .-CtxSwitch
+        .size  __cfactx_switch, .-__cfactx_switch
 // Local Variables: //

libcfa/src/concurrency/CtxSwitch-x86_64.S

-              r41efd33
+              r04e6f93
         .text
         .align 2
         .globl CtxSwitch
         .type  CtxSwitch, @function
 CtxSwitch:
+        .globl __cfactx_switch
+        .type  __cfactx_switch, @function
+__cfactx_switch:
         // Save volatile registers on the stack.
 …
         ret
         .size  CtxSwitch, .-CtxSwitch
+        .size  __cfactx_switch, .-__cfactx_switch
 //-----------------------------------------------------------------------------
 …
         .text
         .align 2
         .globl CtxInvokeStub
         .type    CtxInvokeStub, @function
 CtxInvokeStub:
+        .globl __cfactx_invoke_stub
+        .type    __cfactx_invoke_stub, @function
+__cfactx_invoke_stub:
         movq %rbx, %rdi
         movq %r12, %rsi
         jmp *%r13
         .size  CtxInvokeStub, .-CtxInvokeStub
+        .size  __cfactx_invoke_stub, .-__cfactx_invoke_stub
 // Local Variables: //

libcfa/src/concurrency/alarm.cfa

r41efd33	r04e6f93
47	47	//=============================================================================================
48	48
49		void ?{}( alarm_node_t & this, ~~thread_desc~~ * thrd, Time alarm, Duration period ) with( this ) {
	49	void ?{}( alarm_node_t & this, $thread * thrd, Time alarm, Duration period ) with( this ) {
50	50	this.thrd = thrd;
51	51	this.alarm = alarm;

libcfa/src/concurrency/alarm.hfa

-              r41efd33
+              r04e6f93
 #include "time.hfa"
 struct thread_desc;
+struct $thread;
 struct processor;
 …
         union {
                 thread_desc * thrd;     // thrd who created event
+                $thread * thrd; // thrd who created event
                 processor * proc;               // proc who created event
         };
 …
 typedef alarm_node_t ** __alarm_it_t;
 void ?{}( alarm_node_t & this, thread_desc * thrd, Time alarm, Duration period );
+void ?{}( alarm_node_t & this, $thread * 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

-              r41efd33
+              r04e6f93
 extern "C" {
         void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct coroutine_desc *) __attribute__ ((__noreturn__));
+        void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct $coroutine *) __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_desc & this, const char name[], void * storage, size_t storageSize ) with( this ) {
+void ?{}( $coroutine & this, const char name[], void * storage, size_t storageSize ) with( this ) {
         (this.context){0p, 0p};
         (this.stack){storage, storageSize};
 …
+}
 void ^?{}(coroutine_desc& this) {
+void ^?{}($coroutine& this) {
         if(this.state != Halted && this.state != Start && this.state != Primed) {
                 coroutine_desc * src = TL_GET( this_thread )->curr_cor;
                 coroutine_desc * dst = &this;
+                $coroutine * src = TL_GET( this_thread )->curr_cor;
+                $coroutine * dst = &this;
                 struct _Unwind_Exception storage;
 …
+                }
                 CoroutineCtxSwitch( src, dst );
+                $ctx_switch( src, dst );
+        }
+}
 …
 forall(dtype T | is_coroutine(T))
 void prime(T& cor) {
         coroutine_desc* this = get_coroutine(cor);
+        $coroutine* this = get_coroutine(cor);
         assert(this->state == Start);
 …
 // is not inline (We can't inline Cforall in C)
 extern "C" {
         void __leave_coroutine( struct coroutine_desc * src ) {
                 coroutine_desc * starter = src->cancellation != 0 ? src->last : src->starter;
+        void __cfactx_cor_leave( struct $coroutine * src ) {
+                $coroutine * starter = src->cancellation != 0 ? src->last : src->starter;
                 src->state = Halted;
 …
                         src->name, src, starter->name, starter );
                 CoroutineCtxSwitch( src, starter );
+        }
         struct coroutine_desc * __finish_coroutine(void) {
                 struct coroutine_desc * cor = kernelTLS.this_thread->curr_cor;
+                $ctx_switch( src, starter );
+        }
+        struct $coroutine * __cfactx_cor_finish(void) {
+                struct $coroutine * cor = kernelTLS.this_thread->curr_cor;
                 if(cor->state == Primed) {

libcfa/src/concurrency/coroutine.hfa

-              r41efd33
+              r04e6f93
 trait is_coroutine(dtype T) {
       void main(T & this);
       coroutine_desc * get_coroutine(T & this);
+      $coroutine * get_coroutine(T & this);
 };
 #define DECL_COROUTINE(X) static inline coroutine_desc* get_coroutine(X& this) { return &this.__cor; } void main(X& this)
+#define DECL_COROUTINE(X) static inline $coroutine* get_coroutine(X& this) { return &this.__cor; } void main(X& this)
 //-----------------------------------------------------------------------------
 …
 // void ^?{}( coStack_t & this );
 void ?{}( coroutine_desc & this, const char name[], void * storage, size_t storageSize );
 void ^?{}( coroutine_desc & this );
+void  ?{}( $coroutine & this, const char name[], void * storage, size_t storageSize );
+void ^?{}( $coroutine & this );
 static inline void ?{}( coroutine_desc & this)                                       { this{ "Anonymous Coroutine", 0p, 0 }; }
 static inline void ?{}( coroutine_desc & this, size_t stackSize)                     { this{ "Anonymous Coroutine", 0p, 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, 0p, 0 }; }
 static inline void ?{}( coroutine_desc & this, const char name[], size_t stackSize ) { this{ name, 0p, stackSize }; }
+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 }; }
 //-----------------------------------------------------------------------------
 …
 void prime(T & cor);
 static inline struct coroutine_desc * active_coroutine() { return TL_GET( this_thread )->curr_cor; }
+static inline struct $coroutine * active_coroutine() { return TL_GET( this_thread )->curr_cor; }
 //-----------------------------------------------------------------------------
 …
 // Start coroutine routines
 extern "C" {
         void CtxInvokeCoroutine(void (*main)(void *), void * this);
+        void __cfactx_invoke_coroutine(void (*main)(void *), void * this);
         forall(dtype T)
         void CtxStart(void (*main)(T &), struct coroutine_desc * cor, T & this, void (*invoke)(void (*main)(void *), void *));
+        void __cfactx_start(void (*main)(T &), struct $coroutine * cor, T & this, void (*invoke)(void (*main)(void *), void *));
         extern void _CtxCoroutine_Unwind(struct _Unwind_Exception * storage, struct coroutine_desc *) __attribute__ ((__noreturn__));
+        extern void __cfactx_coroutine_unwind(struct _Unwind_Exception * storage, struct $coroutine *) __attribute__ ((__noreturn__));
         extern void CtxSwitch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("CtxSwitch");
+        extern void __cfactx_switch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("__cfactx_switch");
+}
 // Private wrappers for context switch and stack creation
 // Wrapper for co
 static inline void CoroutineCtxSwitch(coroutine_desc* src, coroutine_desc* dst) {
+static inline void $ctx_switch( $coroutine * src, $coroutine * dst ) __attribute__((nonnull (1, 2))) {
         // set state of current coroutine to inactive
         src->state = src->state == Halted ? Halted : Inactive;
 …
         // context switch to specified coroutine
         verify( dst->context.SP );
         CtxSwitch( &src->context, &dst->context );
         // when CtxSwitch returns we are back in the src coroutine
+        __cfactx_switch( &src->context, &dst->context );
+        // when __cfactx_switch returns we are back in the src coroutine
         // set state of new coroutine to active
 …
         if( unlikely(src->cancellation != 0p) ) {
                 _CtxCoroutine_Unwind(src->cancellation, src);
+                __cfactx_coroutine_unwind(src->cancellation, src);
+        }
+}
 …
         // will also migrate which means this value will
         // stay in syn with the TLS
         coroutine_desc * src = TL_GET( this_thread )->curr_cor;
+        $coroutine * src = TL_GET( this_thread )->curr_cor;
         assertf( src->last != 0,
 …
                 src->name, src, src->last->name, src->last );
         CoroutineCtxSwitch( src, src->last );
+        $ctx_switch( src, src->last );
+}
 …
         // will also migrate which means this value will
         // stay in syn with the TLS
         coroutine_desc * src = TL_GET( this_thread )->curr_cor;
         coroutine_desc * dst = get_coroutine(cor);
+        $coroutine * src = TL_GET( this_thread )->curr_cor;
+        $coroutine * dst = get_coroutine(cor);
         if( unlikely(dst->context.SP == 0p) ) {
                 TL_GET( this_thread )->curr_cor = dst;
                 __stack_prepare(&dst->stack, 65000);
                 CtxStart(main, dst, cor, CtxInvokeCoroutine);
+                __cfactx_start(main, dst, cor, __cfactx_invoke_coroutine);
                 TL_GET( this_thread )->curr_cor = src;
+        }
 …
         // always done for performance testing
         CoroutineCtxSwitch( src, dst );
+        $ctx_switch( src, dst );
         return cor;
+}
 static inline void resume(coroutine_desc * dst) {
+static inline void resume( $coroutine * dst ) __attribute__((nonnull (1))) {
         // 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_desc * src = TL_GET( this_thread )->curr_cor;
+        $coroutine * src = TL_GET( this_thread )->curr_cor;
         // not resuming self ?
 …
         // always done for performance testing
         CoroutineCtxSwitch( src, dst );
+        $ctx_switch( src, dst );
+}

libcfa/src/concurrency/invoke.c

-              r41efd33
+              r04e6f93
 // Called from the kernel when starting a coroutine or task so must switch back to user mode.
+extern void __leave_coroutine ( struct coroutine_desc * );
+extern struct coroutine_desc * __finish_coroutine(void);
+extern void __leave_thread_monitor();
+extern struct $coroutine * __cfactx_cor_finish(void);
+extern void __cfactx_cor_leave ( struct $coroutine * );
+extern void __cfactx_thrd_leave();
 extern void disable_interrupts() OPTIONAL_THREAD;
 extern void enable_interrupts( __cfaabi_dbg_ctx_param );
 void CtxInvokeCoroutine(
+void __cfactx_invoke_coroutine(
         void (*main)(void *),
         void *this
 ) {
         // Finish setting up the coroutine by setting its state
         struct coroutine_desc * cor = __finish_coroutine();
+        struct $coroutine * cor = __cfactx_cor_finish();
         // Call the main of the coroutine
 …
         //Final suspend, should never return
         __leave_coroutine( cor );
+        __cfactx_cor_leave( cor );
         __cabi_abort( "Resumed dead coroutine" );
+}
 static _Unwind_Reason_Code _CtxCoroutine_UnwindStop(
+static _Unwind_Reason_Code __cfactx_coroutine_unwindstop(
         __attribute((__unused__)) int version,
         _Unwind_Action actions,
 …
                 // We finished unwinding the coroutine,
                 // leave it
                 __leave_coroutine( param );
+                __cfactx_cor_leave( param );
                 __cabi_abort( "Resumed dead coroutine" );
+        }
 …
+}
 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 );
+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 );
         printf("UNWIND ERROR %d after force unwind\n", ret);
         abort();
+}
 void CtxInvokeThread(
+void __cfactx_invoke_thread(
         void (*main)(void *),
         void *this
 …
         // The order of these 4 operations is very important
         //Final suspend, should never return
         __leave_thread_monitor();
+        __cfactx_thrd_leave();
         __cabi_abort( "Resumed dead thread" );
+}
 void CtxStart(
+void __cfactx_start(
         void (*main)(void *),
         struct coroutine_desc * cor,
+        struct $coroutine * cor,
         void *this,
         void (*invoke)(void *)
 …
         fs->dummyReturn = NULL;
         fs->rturn = CtxInvokeStub;
+        fs->rturn = __cfactx_invoke_stub;
         fs->fixedRegisters[0] = main;
         fs->fixedRegisters[1] = this;
 …
         struct FakeStack *fs = (struct FakeStack *)cor->context.SP;
         fs->intRegs[8] = CtxInvokeStub;
+        fs->intRegs[8] = __cfactx_invoke_stub;
         fs->arg[0] = this;
         fs->arg[1] = invoke;

libcfa/src/concurrency/invoke.h

-              r41efd33
+              r04e6f93
         extern "Cforall" {
                 extern __attribute__((aligned(128))) thread_local struct KernelThreadData {
                         struct thread_desc    * volatile this_thread;
+                        struct $thread    * volatile this_thread;
                         struct processor      * volatile this_processor;
 …
         };
+        enum coroutine_state { Halted, Start, Inactive, Active, Primed };
+        struct coroutine_desc {
+                // context that is switch during a CtxSwitch
+        enum coroutine_state { Halted, Start, Primed, Inactive, Active, Rerun };
+        enum __Preemption_Reason { __NO_PREEMPTION, __ALARM_PREEMPTION, __POLL_PREEMPTION, __MANUAL_PREEMPTION };
+        struct $coroutine {
+                // context that is switch during a __cfactx_switch
                 struct __stack_context_t context;
 …
                 // first coroutine to resume this one
                 struct coroutine_desc * starter;
+                struct $coroutine * starter;
                 // last coroutine to resume this one
                 struct coroutine_desc * last;
+                struct $coroutine * last;
                 // If non-null stack must be unwound with this exception
 …
         };
         struct monitor_desc {
+        struct $monitor {
                 // spinlock to protect internal data
                 struct __spinlock_t lock;
                 // current owner of the monitor
                 struct thread_desc * owner;
+                struct $thread * owner;
                 // queue of threads that are blocked waiting for the monitor
                 __queue_t(struct thread_desc) entry_queue;
+                __queue_t(struct $thread) entry_queue;
                 // stack of conditions to run next once we exit the monitor
 …
         struct __monitor_group_t {
                 // currently held monitors
                 __cfa_anonymous_object( __small_array_t(monitor_desc*) );
+                __cfa_anonymous_object( __small_array_t($monitor*) );
                 // last function that acquired monitors
 …
         };
         struct thread_desc {
+        struct $thread {
                 // Core threading fields
                 // context that is switch during a CtxSwitch
+                // context that is switch during a __cfactx_switch
                 struct __stack_context_t context;
                 // current execution status for coroutine
+                enum coroutine_state state;
+                volatile int state;
+                enum __Preemption_Reason preempted;
                 //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_desc  self_cor;
+                struct $coroutine  self_cor;
                 // current active context
                 struct coroutine_desc * curr_cor;
+                struct $coroutine * curr_cor;
                 // monitor body used for mutual exclusion
                 struct monitor_desc    self_mon;
+                struct $monitor    self_mon;
                 // pointer to monitor with sufficient lifetime for current monitors
                 struct monitor_desc *  self_mon_p;
+                struct $monitor *  self_mon_p;
                 // pointer to the cluster on which the thread is running
 …
                 // Link lists fields
                 // instrusive link field for threads
                 struct thread_desc * next;
+                struct $thread * next;
                 struct {
                         struct thread_desc * next;
                         struct thread_desc * prev;
+                        struct $thread * next;
+                        struct $thread * prev;
                 } node;
         };
 …
         #ifdef __cforall
         extern "Cforall" {
                 static inline thread_desc *& get_next( thread_desc & this ) {
+                static inline $thread *& get_next( $thread & this ) __attribute__((const)) {
                         return this.next;
+                }
                 static inline [thread_desc *&, thread_desc *& ] __get( thread_desc & this ) {
+                static inline [$thread *&, $thread *& ] __get( $thread & this ) __attribute__((const)) {
                         return this.node.[next, prev];
+                }
 …
+                }
                 static inline void ?{}(__monitor_group_t & this, struct monitor_desc ** data, __lock_size_t size, fptr_t func) {
+                static inline void ?{}(__monitor_group_t & this, struct $monitor ** 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 ) {
+                static inline bool ?==?( const __monitor_group_t & lhs, const __monitor_group_t & rhs ) __attribute__((const)) {
                         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 CtxInvokeStub( void );
         extern void CtxSwitch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("CtxSwitch");
+        extern void __cfactx_invoke_stub( void );
+        extern void __cfactx_switch( struct __stack_context_t * from, struct __stack_context_t * to ) asm ("__cfactx_switch");
         // void CtxStore ( void * this ) asm ("CtxStore");
         // void CtxRet   ( void * dst  ) asm ("CtxRet");

libcfa/src/concurrency/kernel.cfa

-              r41efd33
+              r04e6f93
 //-----------------------------------------------------------------------------
 //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 ) ));
+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_desc,     mainThread);
+KERNEL_STORAGE($thread, mainThread);
 KERNEL_STORAGE(__stack_t,       mainThreadCtx);
 cluster     * mainCluster;
 processor   * mainProcessor;
 thread_desc * mainThread;
+$thread * mainThread;
 extern "C" {
 …
 // Main thread construction
 void ?{}( coroutine_desc & this, current_stack_info_t * info) with( this ) {
+void ?{}( $coroutine & this, current_stack_info_t * info) with( this ) {
         stack.storage = info->storage;
         with(*stack.storage) {
 …
+}
 void ?{}( thread_desc & this, current_stack_info_t * info) with( this ) {
+void ?{}( $thread & this, current_stack_info_t * info) with( this ) {
         state = Start;
         self_cor{ info };
 …
+}
+static void start(processor * this);
+static void * __invoke_processor(void * arg);
 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;
 …
         idleLock{};
+        start( &this );
+        __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);
+}
 …
 // Kernel Scheduling logic
 //=============================================================================================
 static void runThread(processor * this, thread_desc * dst);
 static void finishRunning(processor * this);
 static void halt(processor * this);
+static $thread * __next_thread(cluster * this);
+static void __run_thread(processor * this, $thread * dst);
+static void __halt(processor * this);
 //Main of the processor contexts
 …
                 __cfaabi_dbg_print_safe("Kernel : core %p started\n", this);
                 thread_desc * readyThread = 0p;
+                $thread * readyThread = 0p;
                 for( unsigned int spin_count = 0; ! __atomic_load_n(&this->do_terminate, __ATOMIC_SEQ_CST); spin_count++ ) {
                         readyThread = nextThread( this->cltr );
+                        readyThread = __next_thread( 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);
+                                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                                /* paranoid */ verifyf( readyThread->state == Inactive || readyThread->state == Start || readyThread->preempted != __NO_PREEMPTION, "state : %d, preempted %d\n", readyThread->state, readyThread->preempted);
+                                /* paranoid */ verifyf( readyThread->next == 0p, "Expected null got %p", readyThread->next );
+                                __run_thread(this, readyThread);
+                                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
                                 spin_count = 0;
                         } else {
                                 // spin(this, &spin_count);
                                 halt(this);
+                                __halt(this);
+                        }
+                }
 …
 // runThread runs a thread by context switching
 // from the processor coroutine to the target thread
+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;
+static void __run_thread(processor * this, $thread * thrd_dst) {
+        $coroutine * proc_cor = get_coroutine(this->runner);
         // Update global state
         kernelTLS.this_thread = thrd_dst;
+        // 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;
+        // set state of processor coroutine to inactive
+        verify(proc_cor->state == Active);
+        proc_cor->state = Inactive;
+        // 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 == Start || thrd_dst->state == Primed || thrd_dst->state == Inactive);
+                        thrd_dst->state = Active;
+                }
+                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                // 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 */ 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.
+                // 3 - Polite Racy case : the thread has blocked, someone has already tried to schedule it, but the thread is nice and wants to go through the ready-queue any way
+                // 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.
+                // In case 3, we lost the race but can just reschedule 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;
+                }
+                // 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, Inactive, __ATOMIC_SEQ_CST);
+                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 );
+                                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 Inactive/Start/Primed %d\n", old_state);
+                }
+        }
+        // Just before returning to the processor, set the processor coroutine to active
         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] );
+                }
+        case Release_Multi_Schedule:
+                for(int i = 0; i < lock_count; i++) {
+                        unlock( *locks[i] );
+                }
+                for(int i = 0; i < thrd_count; i++) {
+                        ScheduleThread( thrds[i] );
+                }
+        case Callback:
+                callback();
+        default:
+                abort("KERNEL ERROR: Unexpected action to run after thread");
+        }
+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 );
+}
 …
 // This is the entry point for processors (kernel threads)
 // It effectively constructs a coroutine by stealing the pthread stack
 static void * CtxInvokeProcessor(void * arg) {
+static void * __invoke_processor(void * arg) {
         processor * proc = (processor *) arg;
         kernelTLS.this_processor = proc;
 …
 } // Abort
 void * create_pthread( pthread_t * pthread, void * (*start)(void *), void * arg ) {
+void * __create_pthread( pthread_t * pthread, void * (*start)(void *), void * arg ) {
         pthread_attr_t attr;
 …
+}
-static void start(processor * this) {
-        __cfaabi_dbg_print_safe("Kernel : Starting core %p\n", this);
-        this->stack = create_pthread( &this->kernel_thread, CtxInvokeProcessor, (void *)this );
-        __cfaabi_dbg_print_safe("Kernel : core %p started\n", this);
+}
 // KERNEL_ONLY
 void kernel_first_resume( processor * this ) {
         thread_desc * src = mainThread;
         coroutine_desc * dst = get_coroutine(this->runner);
+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;
         __stack_prepare( &dst->stack, 65000 );
         CtxStart(main, dst, this->runner, CtxInvokeCoroutine);
+        __cfactx_start(main, dst, this->runner, __cfactx_invoke_coroutine);
         verify( ! kernelTLS.preemption_state.enabled );
 …
         // context switch to specified coroutine
         verify( dst->context.SP );
         CtxSwitch( &src->context, &dst->context );
         // when CtxSwitch returns we are back in the src coroutine
+        __cfactx_switch( &src->context, &dst->context );
+        // when __cfactx_switch returns we are back in the src coroutine
         mainThread->curr_cor = &mainThread->self_cor;
 …
 // KERNEL_ONLY
 void kernel_last_resume( processor * this ) {
         coroutine_desc * src = &mainThread->self_cor;
         coroutine_desc * dst = get_coroutine(this->runner);
+static void __kernel_last_resume( processor * this ) {
+        $coroutine * src = &mainThread->self_cor;
+        $coroutine * dst = get_coroutine(this->runner);
         verify( ! kernelTLS.preemption_state.enabled );
 …
         // context switch to the processor
         CtxSwitch( &src->context, &dst->context );
+        __cfactx_switch( &src->context, &dst->context );
+}
 //-----------------------------------------------------------------------------
 // Scheduler routines
 // KERNEL ONLY
+void ScheduleThread( thread_desc * thrd ) {
+        verify( thrd );
+        verify( thrd->state != Halted );
+        verify( ! kernelTLS.preemption_state.enabled );
+        verifyf( thrd->next == 0p, "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);
+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 == Inactive || 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 );
+        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);
+                }
                 else if( struct processor * idle = idles.head ) {
                         wake_fast(idle);
+                }
+        }
         verify( ! kernelTLS.preemption_state.enabled );
+                unlock    (proc_list_lock);
+        }
+        else if( struct processor * idle = idles.head ) {
+                wake_fast(idle);
+        }
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+}
 // KERNEL ONLY
+thread_desc * nextThread(cluster * this) with( *this ) {
+        verify( ! kernelTLS.preemption_state.enabled );
+static $thread * __next_thread(cluster * this) with( *this ) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
         lock( ready_queue_lock __cfaabi_dbg_ctx2 );
         thread_desc * head = pop_head( ready_queue );
+        $thread * head = pop_head( ready_queue );
         unlock( ready_queue_lock );
+        verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
         return head;
+}
+void BlockInternal() {
+void unpark( $thread * thrd ) {
+        if( !thrd ) return;
         disable_interrupts();
+        verify( ! kernelTLS.preemption_state.enabled );
+        static_assert(sizeof(thrd->state) == sizeof(int));
+        enum coroutine_state old_state = __atomic_exchange_n(&thrd->state, Rerun, __ATOMIC_SEQ_CST);
+        switch(old_state) {
+                case Active:
+                        // Wake won the race, the thread will reschedule/rerun itself
+                        break;
+                case Inactive:
+                        /* paranoid */ verify( ! thrd->preempted != __NO_PREEMPTION );
+                        // Wake lost the race,
+                        thrd->state = Inactive;
+                        __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( void ) {
+        /* paranoid */ verify( kernelTLS.preemption_state.enabled );
+        disable_interrupts();
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ verify( kernelTLS.this_thread->preempted == __NO_PREEMPTION );
         returnToKernel();
+        verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
         enable_interrupts( __cfaabi_dbg_ctx );
+}
+void BlockInternal( __spinlock_t * lock ) {
+        /* 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();
+        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
+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 );
+        $thread * thrd = kernelTLS.this_thread;
+        /* paranoid */ verify(thrd->state == Active || thrd->state == Rerun);
+        // SKULLDUGGERY: It is possible that we are preempting this thread just before
+        // it was going to park itself. If that is the case and it is already using the
+        // intrusive fields then we can't use them to preempt the thread
+        // If that is the case, abandon the preemption.
+        bool preempted = false;
+        if(thrd->next == 0p) {
+                preempted = true;
+                thrd->preempted = reason;
+                returnToKernel();
+        }
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        enable_interrupts_noPoll();
+        /* paranoid */ verify( kernelTLS.preemption_state.enabled );
+        return preempted;
+}
 …
 //-----------------------------------------------------------------------------
 // Kernel boot procedures
 static void kernel_startup(void) {
+static void __kernel_startup(void) {
         verify( ! kernelTLS.preemption_state.enabled );
         __cfaabi_dbg_print_safe("Kernel : Starting\n");
 …
         // SKULLDUGGERY: the mainThread steals the process main thread
         // which will then be scheduled by the mainProcessor normally
         mainThread = (thread_desc *)&storage_mainThread;
+        mainThread = ($thread *)&storage_mainThread;
         current_stack_info_t info;
         info.storage = (__stack_t*)&storage_mainThreadCtx;
 …
         // 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
         ScheduleThread(mainThread);
+        __schedule_thread(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 CtxInvokeThread, like all other threads. The trick here is that
+        // context. Hence, the main thread does not begin through __cfactx_invoke_thread, 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");
 …
         // 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;
 …
 // Kernel Quiescing
 //=============================================================================================
 static void halt(processor * this) with( *this ) {
+static void __halt(processor * this) with( *this ) {
         // verify( ! __atomic_load_n(&do_terminate, __ATOMIC_SEQ_CST) );
 …
 void kernel_abort_msg( void * kernel_data, char * abort_text, int abort_text_size ) {
         thread_desc * thrd = kernel_data;
+        $thread * thrd = kernel_data;
         if(thrd) {
 …
                 // atomically release spin lock and block
+                BlockInternal( &lock );
+                unlock( lock );
+                park();
+        }
         else {
 …
 void V(semaphore & this) with( this ) {
         thread_desc * thrd = 0p;
+        $thread * thrd = 0p;
         lock( lock __cfaabi_dbg_ctx2 );
         count += 1;
 …
         // make new owner
         WakeThread( thrd );
+        unpark( thrd );
+}
 …
+}
 void doregister( cluster * cltr, thread_desc & thrd ) {
+void doregister( cluster * cltr, $thread & thrd ) {
         lock      (cltr->thread_list_lock __cfaabi_dbg_ctx2);
         cltr->nthreads += 1;
 …
+}
 void unregister( cluster * cltr, thread_desc & thrd ) {
+void unregister( cluster * cltr, $thread & thrd ) {
         lock  (cltr->thread_list_lock __cfaabi_dbg_ctx2);
         remove(cltr->threads, thrd );
 …
 //-----------------------------------------------------------------------------
 // Debug
 bool threading_enabled(void) {
+bool threading_enabled(void) __attribute__((const)) {
         return true;
+}

libcfa/src/concurrency/kernel.hfa

-              r41efd33
+              r04e6f93
         __spinlock_t lock;
         int count;
         __queue_t(thread_desc) waiting;
+        __queue_t($thread) waiting;
 };
 …
 // 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 = 0p;
-        this.lock = 0p;
+}
-static inline void ^?{}(FinishAction &) {}
 // Processor
 …
         // RunThread data
         // Action to do after a thread is ran
         struct FinishAction finish;
+        $thread * destroyer;
         // Preemption data
 …
 static inline void  ?{}(processor & this, const char name[]) { this{name, *mainCluster }; }
+static inline [processor *&, processor *& ] __get( processor & this ) {
+        return this.node.[next, prev];
+}
+static inline [processor *&, processor *& ] __get( processor & this ) __attribute__((const)) { return this.node.[next, prev]; }
 //-----------------------------------------------------------------------------
 …
         // Ready queue for threads
         __queue_t(thread_desc) ready_queue;
+        __queue_t($thread) ready_queue;
         // Name of the cluster
 …
         // List of threads
         __spinlock_t thread_list_lock;
         __dllist_t(struct thread_desc) threads;
+        __dllist_t(struct $thread) threads;
         unsigned int nthreads;
 …
 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 [cluster *&, cluster *& ] __get( cluster & this ) __attribute__((const)) { return this.node.[next, prev]; }
 static inline struct processor * active_processor() { return TL_GET( this_processor ); } // UNSAFE

libcfa/src/concurrency/kernel_private.hfa

-              r41efd33
+              r04e6f93
+}
+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);
+void __schedule_thread( $thread * ) __attribute__((nonnull (1)));
 //Block current thread and release/wake-up the following resources
+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 __leave_thread() __attribute__((noreturn));
 //-----------------------------------------------------------------------------
 …
 void main(processorCtx_t *);
 void * create_pthread( pthread_t *, void * (*)(void *), void * );
+void * __create_pthread( pthread_t *, void * (*)(void *), void * );
 static inline void wake_fast(processor * this) {
 …
 // Threads
 extern "C" {
       void CtxInvokeThread(void (*main)(void *), void * this);
+      void __cfactx_invoke_thread(void (*main)(void *), void * this);
+}
-extern void ThreadCtxSwitch(coroutine_desc * src, coroutine_desc * dst);
 __cfaabi_dbg_debug_do(
         extern void __cfaabi_dbg_thread_register  ( thread_desc * thrd );
         extern void __cfaabi_dbg_thread_unregister( thread_desc * thrd );
+        extern void __cfaabi_dbg_thread_register  ( $thread * thrd );
+        extern void __cfaabi_dbg_thread_unregister( $thread * 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_desc & thrd );
 void unregister( struct cluster * cltr, struct thread_desc & thrd );
+void doregister( struct cluster * cltr, struct $thread & thrd );
+void unregister( struct cluster * cltr, struct $thread & thrd );
 void doregister( struct cluster * cltr, struct processor * proc );

libcfa/src/concurrency/monitor.cfa

-              r41efd33
+              r04e6f93
 // file "LICENCE" distributed with Cforall.
 //
 // monitor_desc.c --
+// $monitor.c --
 //
 // Author           : Thierry Delisle
 …
 //-----------------------------------------------------------------------------
 // Forward declarations
 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 __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 lock_all  ( __spinlock_t * locks [], __lock_size_t count );
 static inline void lock_all  ( monitor_desc * source [], __spinlock_t * /*out*/ locks [], __lock_size_t count );
+static inline void lock_all  ( $monitor * 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_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 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 $thread *        check_condition   ( __condition_criterion_t * );
 static inline void                 brand_condition   ( condition & );
 static inline [thread_desc *, int] search_entry_queue( const __waitfor_mask_t &, monitor_desc * monitors [], __lock_size_t count );
+static inline [$thread *, int] search_entry_queue( const __waitfor_mask_t &, $monitor * 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_desc * storage [], const __waitfor_mask_t & mask );
+static inline __lock_size_t aggregate    ( $monitor * storage [], const __waitfor_mask_t & mask );
 //-----------------------------------------------------------------------------
 …
 #define monitor_ctx( mons, cnt )                                /* Define that create the necessary struct for internal/external scheduling operations */ \
         monitor_desc ** monitors = mons;                          /* Save the targeted monitors                                                          */ \
+        $monitor ** 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
+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;
+                }
+// 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_print_safe( "Kernel : %10p Entered  mon %p\n", thrd, this);
+                // Release the lock and leave
+                /* 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 );
                 unlock( this->lock );
                 return;
+        }
+        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 ** 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
+                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 );
+                // Park current thread waiting
+                park();
+                // 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();
+                /* 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 );
+}
+// 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);
+                }
+                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 );
+                if( this->recursion != 1 ) {
+                        abort( "Destroyed monitor %p has %d outstanding nested calls.\n", this, this->recursion - 1);
+                }
+                __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 );
+                //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);
+                        }
+                )
+        }
+        )
+}
+extern "C" {
         // Leave the thread monitor
         // last routine called by a thread.
         // Should never return
         void __leave_thread_monitor() {
                 thread_desc * thrd = TL_GET( this_thread );
                 monitor_desc * this = &thrd->self_mon;
+        void __cfactx_thrd_leave() {
+                $thread * thrd = TL_GET( this_thread );
+                $monitor * this = &thrd->self_mon;
                 // Lock the monitor now
 …
                 disable_interrupts();
                 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 );
+                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 );
                 // Leaving a recursion level, decrement the counter
 …
                 // Fetch the next thread, can be null
+                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 );
+                $thread * new_owner = next_thread( this );
+                // Release the monitor lock
+                unlock( this->lock );
+                // Unpark the next owner if needed
+                /* paranoid */ verifyf( !new_owner || new_owner == this->owner, "Expected owner to be %p, got %p (m: %p)", new_owner, this->owner, this );
+                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                /* paranoid */ verify( ! kernelTLS.this_processor->destroyer );
+                /* paranoid */ verify( thrd->state == Halted );
+                kernelTLS.this_processor->destroyer = new_owner;
+                // Leave the thread
+                __leave_thread();
                 // Control flow should never reach here!
 …
 static inline void enter( __monitor_group_t monitors ) {
         for( __lock_size_t i = 0; i < monitors.size; i++) {
                 __enter_monitor_desc( monitors[i], monitors );
+                __enter( monitors[i], monitors );
+        }
+}
 …
 // Leave multiple monitor
 // relies on the monitor array being sorted
 static inline void leave(monitor_desc * monitors [], __lock_size_t count) {
+static inline void leave($monitor * monitors [], __lock_size_t count) {
         for( __lock_size_t i = count - 1; i >= 0; i--) {
                 __leave_monitor_desc( monitors[i] );
+                __leave( monitors[i] );
+        }
+}
 …
 // Ctor for monitor guard
 // Sorts monitors before entering
 void ?{}( monitor_guard_t & this, monitor_desc * m [], __lock_size_t count, fptr_t func ) {
         thread_desc * thrd = TL_GET( this_thread );
+void ?{}( monitor_guard_t & this, $monitor * m [], __lock_size_t count, fptr_t func ) {
+        $thread * thrd = TL_GET( this_thread );
         // Store current array
 …
 // Ctor for monitor guard
 // Sorts monitors before entering
 void ?{}( monitor_dtor_guard_t & this, monitor_desc * m [], fptr_t func ) {
+void ?{}( monitor_dtor_guard_t & this, $monitor * m [], fptr_t func ) {
         // optimization
         thread_desc * thrd = TL_GET( this_thread );
+        $thread * thrd = TL_GET( this_thread );
         // Store current array
 …
         (thrd->monitors){m, 1, func};
         __enter_monitor_dtor( this.m, func );
+        __dtor_enter( this.m, func );
+}
 …
 void ^?{}( monitor_dtor_guard_t & this ) {
         // Leave the monitors in order
         __leave_dtor_monitor_desc( this.m );
+        __dtor_leave( this.m );
         // Restore thread context
 …
 //-----------------------------------------------------------------------------
 // Internal scheduling types
 void ?{}(__condition_node_t & this, thread_desc * waiting_thread, __lock_size_t count, uintptr_t user_info ) {
+void ?{}(__condition_node_t & this, $thread * waiting_thread, __lock_size_t count, uintptr_t user_info ) {
         this.waiting_thread = waiting_thread;
         this.count = count;
 …
+}
 void ?{}(__condition_criterion_t & this, monitor_desc * target, __condition_node_t & owner ) {
+void ?{}(__condition_criterion_t & this, $monitor * target, __condition_node_t & owner ) {
         this.ready  = false;
         this.target = target;
 …
         // 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_desc * threads[ count ];
+        $thread * threads[ count ];
         __builtin_memset( threads, 0, sizeof( threads ) );
 …
         // Remove any duplicate threads
         for( __lock_size_t i = 0; i < count; i++) {
                 thread_desc * new_owner = next_thread( monitors[i] );
+                $thread * 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] );
+        }
         // Everything is ready to go to sleep
         BlockInternal( locks, count, threads, thread_count );
+        park();
         // We are back, restore the owners and recursions
 …
         //Some more checking in debug
         __cfaabi_dbg_debug_do(
                 thread_desc * this_thrd = TL_GET( this_thread );
+                $thread * 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 );
 …
         //Find the thread to run
+        thread_desc * signallee = pop_head( this.blocked )->waiting_thread;
+        set_owner( monitors, count, signallee );
+        $thread * signallee = pop_head( this.blocked )->waiting_thread;
+        /* paranoid */ verify( signallee->next == 0p );
+        __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 );
         //Everything is ready to go to sleep
         BlockInternal( locks, count, &signallee, 1 );
+        park();
 …
         // Create one!
         __lock_size_t max = count_max( mask );
         monitor_desc * mon_storage[max];
+        $monitor * 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_desc * next; int index;
+                $thread * 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_desc * mon2dtor = accepted[0];
+                                $monitor * 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 );
+                                // Everything is ready to go to sleep
+                                BlockInternal( locks, count, &next, 1 );
+                                __set_owner( monitors, count, next );
+                                // unlock all the monitors
+                                unlock_all( locks, count );
+                                // unpark the thread we signalled
+                                unpark( next );
+                                //Everything is ready to go to sleep
+                                park();
                                 // We are back, restore the owners and recursions
 …
+        }
+        // unlock all the monitors
+        unlock_all( locks, count );
         //Everything is ready to go to sleep
         BlockInternal( locks, count );
+        park();
 …
 // Utilities
 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 );
+static inline void __set_owner( $monitor * this, $thread * owner ) {
+        /* paranoid */ verify( this->lock.lock );
         //Pass the monitor appropriately
 …
+}
+static inline void set_owner( monitor_desc * monitors [], __lock_size_t count, thread_desc * owner ) {
+        monitors[0]->owner     = owner;
+        monitors[0]->recursion = 1;
+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;
         for( __lock_size_t i = 1; i < count; i++ ) {
+                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 ) {
+                /* 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 ) {
         for( __lock_size_t i = 0; i < count; i++) {
                 storage[i]->mask = mask;
 …
+}
 static inline void reset_mask( monitor_desc * this ) {
+static inline void reset_mask( $monitor * this ) {
         this->mask.accepted = 0p;
         this->mask.data = 0p;
 …
+}
 static inline thread_desc * next_thread( monitor_desc * this ) {
+static inline $thread * next_thread( $monitor * 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
+                set_owner( this,  urgent->owner->waiting_thread );
+                /* 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 );
                 return check_condition( urgent );
 …
         // No signaller thread
         // Get the next thread in the entry_queue
+        thread_desc * new_owner = pop_head( this->entry_queue );
+        set_owner( this, new_owner );
+        $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 );
         return new_owner;
+}
 static inline bool is_accepted( monitor_desc * this, const __monitor_group_t & group ) {
+static inline bool is_accepted( $monitor * 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_desc * monitors [], __condition_node_t & waiter, __condition_criterion_t criteria [] ) {
+static inline void init( __lock_size_t count, $monitor * 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_desc * 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 [] ) {
         for( __lock_size_t i = 0; i < count; i++) {
                 (criteria[i]){ monitors[i], waiter };
 …
+}
 static inline void lock_all( monitor_desc * source [], __spinlock_t * /*out*/ locks [], __lock_size_t count ) {
+static inline void lock_all( $monitor * 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_desc * locks [], __lock_size_t count ) {
+static inline void unlock_all( $monitor * locks [], __lock_size_t count ) {
         for( __lock_size_t i = 0; i < count; i++ ) {
                 unlock( locks[i]->lock );
 …
 static inline void save(
         monitor_desc * ctx [],
+        $monitor * ctx [],
         __lock_size_t count,
         __attribute((unused)) __spinlock_t * locks [],
 …
 static inline void restore(
         monitor_desc * ctx [],
+        $monitor * 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_desc * check_condition( __condition_criterion_t * target ) {
+static inline $thread * check_condition( __condition_criterion_t * target ) {
         __condition_node_t * node = target->owner;
         unsigned short count = node->count;
 …
 static inline void brand_condition( condition & this ) {
         thread_desc * thrd = TL_GET( this_thread );
+        $thread * thrd = TL_GET( this_thread );
         if( !this.monitors ) {
                 // __cfaabi_dbg_print_safe( "Branding\n" );
 …
                 this.monitor_count = thrd->monitors.size;
                 this.monitors = (monitor_desc **)malloc( this.monitor_count * sizeof( *this.monitors ) );
+                this.monitors = ($monitor **)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_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;
+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;
         // For each thread in the entry-queue
         for(    thread_desc ** thrd_it = &entry_queue.head;
                 *thrd_it;
+        for(    $thread ** thrd_it = &entry_queue.head;
+                *thrd_it != 1p;
                 thrd_it = &(*thrd_it)->next
         ) {
 …
+}
 static inline __lock_size_t aggregate( monitor_desc * storage [], const __waitfor_mask_t & mask ) {
+static inline __lock_size_t aggregate( $monitor * 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

-              r41efd33
+              r04e6f93
 trait is_monitor(dtype T) {
         monitor_desc * get_monitor( T & );
+        $monitor * get_monitor( T & );
         void ^?{}( T & mutex );
 };
 static inline void ?{}(monitor_desc & this) with( this ) {
+static inline void ?{}($monitor & this) with( this ) {
         lock{};
         entry_queue{};
 …
+}
 static inline void ^?{}(monitor_desc & ) {}
+static inline void ^?{}($monitor & ) {}
 struct monitor_guard_t {
         monitor_desc **         m;
+        $monitor **     m;
         __lock_size_t           count;
         __monitor_group_t prev;
 };
 void ?{}( monitor_guard_t & this, monitor_desc ** m, __lock_size_t count, void (*func)() );
+void ?{}( monitor_guard_t & this, $monitor ** m, __lock_size_t count, void (*func)() );
 void ^?{}( monitor_guard_t & this );
 struct monitor_dtor_guard_t {
         monitor_desc *    m;
+        $monitor *    m;
         __monitor_group_t prev;
 };
 void ?{}( monitor_dtor_guard_t & this, monitor_desc ** m, void (*func)() );
+void ?{}( monitor_dtor_guard_t & this, $monitor ** m, void (*func)() );
 void ^?{}( monitor_dtor_guard_t & this );
 …
         // The monitor this criterion concerns
         monitor_desc * target;
+        $monitor * 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_desc * waiting_thread;
+        $thread * waiting_thread;
         // Array of criteria (Criterions are contiguous in memory)
 …
+}
 void ?{}(__condition_node_t & this, thread_desc * waiting_thread, __lock_size_t count, uintptr_t user_info );
+void ?{}(__condition_node_t & this, $thread * waiting_thread, __lock_size_t count, uintptr_t user_info );
 void ?{}(__condition_criterion_t & this );
 void ?{}(__condition_criterion_t & this, monitor_desc * target, __condition_node_t * owner );
+void ?{}(__condition_criterion_t & this, $monitor * target, __condition_node_t * owner );
 struct condition {
 …
         // Array of monitor pointers (Monitors are NOT contiguous in memory)
         monitor_desc ** monitors;
+        $monitor ** monitors;
         // Number of monitors in the array
 …
               bool signal      ( condition & this );
               bool signal_block( condition & this );
 static inline bool is_empty    ( condition & this ) { return !this.blocked.head; }
+static inline bool is_empty    ( condition & this ) { return this.blocked.head == 1p; }
          uintptr_t front       ( condition & this );

libcfa/src/concurrency/mutex.cfa

-              r41efd33
+              r04e6f93
         if( is_locked ) {
                 append( blocked_threads, kernelTLS.this_thread );
+                BlockInternal( &lock );
+                unlock( lock );
+                park();
+        }
         else {
 …
         lock( this.lock __cfaabi_dbg_ctx2 );
         this.is_locked = (this.blocked_threads != 0);
         WakeThread(
+        unpark(
                 pop_head( this.blocked_threads )
         );
 …
         else {
                 append( blocked_threads, kernelTLS.this_thread );
+                BlockInternal( &lock );
+                unlock( lock );
+                park();
+        }
+}
 …
         recursion_count--;
         if( recursion_count == 0 ) {
                 thread_desc * thrd = pop_head( blocked_threads );
+                $thread * thrd = pop_head( blocked_threads );
                 owner = thrd;
                 recursion_count = (thrd ? 1 : 0);
                 WakeThread( thrd );
+                unpark( thrd );
+        }
         unlock( lock );
 …
 void notify_one(condition_variable & this) with(this) {
         lock( lock __cfaabi_dbg_ctx2 );
         WakeThread(
+        unpark(
                 pop_head( this.blocked_threads )
         );
 …
         lock( lock __cfaabi_dbg_ctx2 );
         while(this.blocked_threads) {
                 WakeThread(
+                unpark(
                         pop_head( this.blocked_threads )
                 );
 …
         lock( this.lock __cfaabi_dbg_ctx2 );
         append( this.blocked_threads, kernelTLS.this_thread );
+        BlockInternal( &this.lock );
+        unlock( this.lock );
+        park();
+}
 …
         lock( this.lock __cfaabi_dbg_ctx2 );
         append( this.blocked_threads, kernelTLS.this_thread );
+        void __unlock(void) {
+                unlock(l);
+                unlock(this.lock);
+        }
+        BlockInternal( __unlock );
+        unlock(l);
+        unlock(this.lock);
+        park();
         lock(l);
+}

libcfa/src/concurrency/mutex.hfa

-              r41efd33
+              r04e6f93
         // List of blocked threads
         __queue_t(struct thread_desc) blocked_threads;
+        __queue_t(struct $thread) blocked_threads;
         // Locked flag
 …
         // List of blocked threads
         __queue_t(struct thread_desc) blocked_threads;
+        __queue_t(struct $thread) blocked_threads;
         // Current thread owning the lock
         struct thread_desc * owner;
+        struct $thread * owner;
         // Number of recursion level
 …
         // List of blocked threads
         __queue_t(struct thread_desc) blocked_threads;
+        __queue_t(struct $thread) blocked_threads;
 };

libcfa/src/concurrency/preemption.cfa

-              r41efd33
+              r04e6f93
 // FwdDeclarations : timeout handlers
 static void preempt( processor   * this );
 static void timeout( thread_desc * this );
+static void timeout( $thread * this );
 // FwdDeclarations : Signal handlers
 …
         // Enable interrupts by decrementing the counter
         // If counter reaches 0, execute any pending CtxSwitch
+        // If counter reaches 0, execute any pending __cfactx_switch
         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;
                                         BlockInternal( thrd );
+                                        force_yield( __POLL_PREEMPTION );
+                                }
+                        }
 …
         // Disable interrupts by incrementint the counter
         // Don't execute any pending CtxSwitch even if counter reaches 0
+        // Don't execute any pending __cfactx_switch even if counter reaches 0
         void enable_interrupts_noPoll() {
                 unsigned short prev = kernelTLS.preemption_state.disable_count;
 …
 // reserved for future use
 static void timeout( thread_desc * this ) {
+static void timeout( $thread * this ) {
         //TODO : implement waking threads
+}
 // KERNEL ONLY
 // Check if a CtxSwitch signal handler shoud defer
+// Check if a __cfactx_switch 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 ); // CtxSwitch handler
+        __cfaabi_sigaction( SIGUSR1, sigHandler_ctxSwitch, SA_SIGINFO | SA_RESTART ); // __cfactx_switch handler
         signal_block( SIGALRM );
         alarm_stack = create_pthread( &alarm_thread, alarm_loop, 0p );
+        alarm_stack = __create_pthread( &alarm_thread, alarm_loop, 0p );
+}
 …
         // Preemption can occur here
         BlockInternal( kernelTLS.this_thread ); // Do the actual CtxSwitch
+        force_yield( __ALARM_PREEMPTION ); // Do the actual __cfactx_switch
+}

libcfa/src/concurrency/thread.cfa

-              r41efd33
+              r04e6f93
 #include "invoke.h"
-extern "C" {
-        #include <fenv.h>
-        #include <stddef.h>
+}
-//extern volatile thread_local processor * this_processor;
 //-----------------------------------------------------------------------------
 // Thread ctors and dtors
 void ?{}(thread_desc & this, const char * const name, cluster & cl, void * storage, size_t storageSize ) with( this ) {
+void ?{}($thread & this, const char * const name, cluster & cl, void * storage, size_t storageSize ) with( this ) {
         context{ 0p, 0p };
         self_cor{ name, storage, storageSize };
         state = Start;
+        preempted = __NO_PREEMPTION;
         curr_cor = &self_cor;
         self_mon.owner = &this;
 …
+}
 void ^?{}(thread_desc& this) with( this ) {
+void ^?{}($thread& 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 ) {
 …
+}
-//-----------------------------------------------------------------------------
-// Starting and stopping threads
-forall( dtype T | is_thread(T) )
-void __thrd_start( T & this, void (*main_p)(T &) ) {
-        thread_desc * this_thrd = get_thread(this);
-        thread_desc * curr_thrd = TL_GET( this_thread );
-        disable_interrupts();
-        CtxStart(main_p, get_coroutine(this), 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 );
+}
-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

-              r41efd33
+              r04e6f93
       void ^?{}(T& mutex this);
       void main(T& this);
       thread_desc* get_thread(T& this);
+      $thread* get_thread(T& this);
 };
+#define DECL_THREAD(X) thread_desc* get_thread(X& this) { return &this.__thrd; } void main(X& 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; }
 forall( dtype T | is_thread(T) )
+static inline coroutine_desc* get_coroutine(T & this) {
+        return &get_thread(this)->self_cor;
+}
+static inline $monitor  * get_monitor  (T & this) __attribute__((const)) { return &get_thread(this)->self_mon; }
+forall( dtype T | is_thread(T) )
+static inline monitor_desc* get_monitor(T & this) {
+        return &get_thread(this)->self_mon;
+}
+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; }
+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;
+}
+//-----------------------------------------------------------------------------
+// forward declarations needed for threads
 extern struct cluster * mainCluster;
 …
 //-----------------------------------------------------------------------------
 // Ctors and dtors
 void ?{}(thread_desc & this, const char * const name, struct cluster & cl, void * storage, size_t storageSize );
 void ^?{}(thread_desc & this);
+void ?{}($thread & this, const char * const name, struct cluster & cl, void * storage, size_t storageSize );
+void ^?{}($thread & this);
 static inline void ?{}(thread_desc & this)                                                                  { this{ "Anonymous Thread", *mainCluster, 0p, 65000 }; }
 static inline void ?{}(thread_desc & this, size_t stackSize )                                               { this{ "Anonymous Thread", *mainCluster, 0p, 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, 0p, 65000 }; }
 static inline void ?{}(thread_desc & this, struct cluster & cl, size_t stackSize )                          { this{ "Anonymous Thread", cl, 0p, 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, 0p, 65000 }; }
 static inline void ?{}(thread_desc & this, const char * const name, struct cluster & cl )                   { this{ name, cl, 0p, 65000 }; }
 static inline void ?{}(thread_desc & this, const char * const name, struct cluster & cl, size_t stackSize ) { this{ name, cl, 0p, stackSize }; }
+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 }; }
 //-----------------------------------------------------------------------------
 …
 void ^?{}( scoped(T)& this );
+void yield();
+void yield( unsigned times );
+//-----------------------------------------------------------------------------
+// Thread getters
+static inline struct $thread * active_thread () { return TL_GET( this_thread ); }
+static inline struct thread_desc * active_thread () { return TL_GET( this_thread ); }
+//-----------------------------------------------------------------------------
+// Scheduler API
+//----------
+// Park thread: block until corresponding call to unpark, won't block if unpark is already called
+void park( void );
+//----------
+// 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 );
+forall( dtype T | is_thread(T) )
+static inline void unpark( T & this ) { if(!&this) return; unpark( get_thread( this ) );}
+//----------
+// 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();
+        }
+}
 // Local Variables: //

libcfa/src/exception.c

-              r41efd33
+              r04e6f93
+}
+#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 // PIC
+#pragma GCC push_options
+#pragma GCC optimize("O0")
 // 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.
 …
         // Setup the personality routine and exception table.
+#ifdef __PIC__
+        asm volatile (".cfi_personality 0x9b,CFA.ref.__gcfa_personality_v0");
+        asm volatile (".cfi_lsda 0x1b, .LLSDACFA2");
+#else
         asm volatile (".cfi_personality 0x3,__gcfa_personality_v0");
         asm volatile (".cfi_lsda 0x3, .LLSDACFA2");
+#endif
         // Label which defines the start of the area for which the handler is setup.
 …
 // have a single call to the try routine.
+#ifdef __PIC__
+#if defined( __i386 ) || defined( __x86_64 )
+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)
+        // 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-__cfaabi_ehm__try_terminate\n"
+        "       .uleb128 .TRYEND-.TRYSTART\n"
+        "       .uleb128 .CATCH-__cfaabi_ehm__try_terminate\n"
+        "       .uleb128 1\n"
+        ".LLSDACSECFA2:\n"
+        // TABLE FOOTER
+        "       .text\n"
+        "       .size   __cfaabi_ehm__try_terminate, .-__cfaabi_ehm__try_terminate\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
+#error Exception Handling: unknown architecture for position independent code.
+#endif // __i386 || __x86_64
+#else // __PIC__
 #if defined( __i386 ) || defined( __x86_64 )
 asm (
 …
         "       .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"
+        "       .section        .note.GNU-stack,\"x\",@progbits\n"
 );
+#else
+#error Exception Handling: unknown architecture for position dependent code.
 #endif // __i386 || __x86_64
+#endif // PIC
+#endif // __PIC__
+#pragma GCC pop_options

libcfa/src/fstream.hfa

-              r41efd33
+              r04e6f93
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:00:51 2020
 // Update Count     : 174
+// Last Modified On : Mon Feb 17 08:29:23 2020
+// Update Count     : 175
 //
 …
 void close( ofstream & );
 ofstream & write( ofstream &, const char data[], size_t size );
 int fmt( ofstream &, const char format[], ... );
+int fmt( ofstream &, const char format[], ... ) __attribute__(( format(printf, 2, 3) ));
 void ?{}( ofstream & os );
 …
 ifstream & read( ifstream & is, char * data, size_t size );
 ifstream & ungetc( ifstream & is, char c );
 int fmt( ifstream &, const char format[], ... );
+int fmt( ifstream &, const char format[], ... ) __attribute__(( format(scanf, 2, 3) ));
 void ?{}( ifstream & is );

libcfa/src/interpose.cfa

-              r41efd33
+              r04e6f93
 // Created On       : Wed Mar 29 16:10:31 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb  8 08:40:34 2020
 // Update Count     : 163
+// Last Modified On : Mon Feb 17 10:18:53 2020
+// Update Count     : 166
 //
 …
         if ( fmt[strlen( fmt ) - 1] != '\n' ) {                         // add optional newline if missing at the end of the format text
                 __cfaabi_dbg_write( "\n", 1 );
+        }
+        } // if
         kernel_abort_msg( kernel_data, abort_text, abort_text_size );
+        __cfaabi_backtrace( signalAbort ? 4 : 3 );
+        __cfaabi_backtrace( signalAbort ? 4 : 2 );
         __cabi_libc.abort();                                                            // print stack trace in handler

libcfa/src/iostream.cfa

-              r41efd33
+              r04e6f93
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 18:48:38 2020
 // Update Count     : 825
+// Last Modified On : Thu Feb 20 15:53:23 2020
+// Update Count     : 829
 //
 …
 #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)));
 …
                 (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, ... ) \
 …
+\
                 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\n", &fmtstr[star] ); */ \
+                        /* printf( "%s %c %c\n", &fmtstr[star], f.base, CODE ); */ \
                         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) fmt ) { \
+        if ( fmt.val > UINT64_MAX ) { \
+                fmt.val /= P10_UINT64; \
+                base10_128( os, fmt ); /* recursive */ \
+                _Ostream_Manip(unsigned long long int) fmt2 @= { (uint64_t)(fmt.val % P10_UINT64), 0, 19, 'u', { .all : 0 } }; \
+                fmt2.flags.nobsdp = true; \
+                printf( "fmt2 %c %lld %d\n", fmt2.base, fmt2.val, fmt2.all );   \
+                sepOff( os ); \
+                (ostype &)(os | fmt2); \
+        } else { \
+                printf( "fmt %c %lld %d\n", fmt.base, fmt.val, fmt.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 == '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' ) { \
+                                        if ( f.wd > 64 ) fmt.wd = f.wd - 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; \
+                                        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;    \
+                                                fmt2.wd = 16;                                                   \
+                                        } /* if */ \
+                                        (ostype &)(os | fmt | "" | fmt2); \
+                                } /* if */ \
+                        } /* if */ \
+                } else { \
+                        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 ***********************************

libcfa/src/iostream.hfa

-              r41efd33
+              r04e6f93
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 17:53:52 2020
 // Update Count     : 336
+// Last Modified On : Thu Feb 20 15:30:56 2020
+// Update Count     : 337
 //
 …
         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 );
 …
 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 ***********************************

longrun_tests/Makefile.in

-              r41efd33
+              r04e6f93
 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_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_0 = @echo "  NODEJS  " $@;
 am__v_NODEJS_1 =
 AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)

src/CodeGen/CodeGenerator.cc

-              r41efd33
+              r04e6f93
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:13:28 2019
 // Update Count     : 508
+// Last Modified On : Sun Feb 16 08:32:48 2020
+// Update Count     : 532
 //
 #include "CodeGenerator.h"
 …
         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() ) ) {
                         OperatorInfo opInfo;
                         if ( varExpr->get_var()->get_linkage() == LinkageSpec::Intrinsic && operatorLookup( varExpr->get_var()->get_name(), opInfo ) ) {
+                        const OperatorInfo * opInfo;
+                        if ( varExpr->get_var()->get_linkage() == LinkageSpec::Intrinsic && ( opInfo = operatorLookup( varExpr->get_var()->get_name() ) ) ) {
                                 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 ) ) {
                         OperatorInfo opInfo;
                         if ( operatorLookup( nameExpr->name, opInfo ) ) {
+                        const OperatorInfo * opInfo = operatorLookup( nameExpr->name );
+                        if ( 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 );
                 OperatorInfo opInfo;
                 if ( operatorLookup( nameExpr->name, opInfo ) ) {
                         if ( opInfo.type == OT_CONSTANT ) {
                                 output << opInfo.symbol;
+                const OperatorInfo * opInfo = operatorLookup( nameExpr->name );
+                if ( opInfo ) {
+                        if ( opInfo->type == OT_CONSTANT ) {
+                                output << opInfo->symbol;
                         } else {
                                 output << opInfo.outputName;
+                                output << opInfo->outputName;
+                        }
                 } else {
 …
         void CodeGenerator::postvisit( VariableExpr * variableExpr ) {
                 extension( variableExpr );
                 OperatorInfo opInfo;
                 if ( variableExpr->get_var()->get_linkage() == LinkageSpec::Intrinsic && operatorLookup( variableExpr->get_var()->get_name(), opInfo ) && opInfo.type == OT_CONSTANT ) {
                         output << opInfo.symbol;
+                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;
                 } 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 ) {
                 CodeGen::OperatorInfo opInfo;
                 if ( operatorLookup( decl->get_name(), opInfo ) ) {
                         return opInfo.outputName;
+                const OperatorInfo * opInfo = operatorLookup( decl->get_name() );
+                if ( opInfo ) {
+                        return opInfo->outputName;
                 } else {
                         return decl->get_name();

src/CodeGen/CodeGenerator.h

-              r41efd33
+              r04e6f93
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Tue Apr 30 12:01:00 2019
 // Update Count     : 57
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Sun Feb 16 03:58:31 2020
+// Update Count     : 62
 //
 …
 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

-              r41efd33
+              r04e6f93
 // Created On       : Thr Jan 12 14:11:09 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:12:21 2019
 // Update Count     : 3
+// Last Modified On : Sun Feb 16 03:24:32 2020
+// Update Count     : 5
 //
 …
                 static std::unique_ptr<FunctionDecl> main_signature;
         };
 };
+} // namespace CodeGen

src/CodeGen/GenType.h

-              r41efd33
+              r04e6f93
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Tue Apr 30 11:47:00 2019
 // Update Count     : 3
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Sun Feb 16 04:11:40 2020
+// Update Count     : 5
 //
 …
         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

-              r41efd33
+              r04e6f93
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:38:56 2019
 // Update Count     : 8
+// Last Modified On : Sun Feb 16 03:01:51 2020
+// Update Count     : 9
 //
 #include "Generate.h"
 …
         void generate( BaseSyntaxNode * node, std::ostream & os ) {
                 if ( Type * type = dynamic_cast< Type * >( node ) ) {
                         os << CodeGen::genPrettyType( type, "" );
+                        os << genPrettyType( type, "" );
                 } else {
                         PassVisitor<CodeGenerator> cgv( os, true, false, false, false );

src/CodeGen/OperatorTable.cc

-              r41efd33
+              r04e6f93
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Jul 15 17:12:22 2017
 // Update Count     : 15
+// Last Modified On : Tue Feb 18 15:55:01 2020
+// Update Count     : 55
 //
 …
 #include <map>        // for map, _Rb_tree_const_iterator, map<>::const_iterator
 #include <utility>    // for pair
+using namespace std;
 #include "OperatorTable.h"
 …
 namespace CodeGen {
         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          },
                 };
+        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
                 const int numOps = sizeof( tableValues ) / sizeof( OperatorInfo );
+        std::map< std::string, OperatorInfo > CodeGen::table;
+                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
+        CodeGen::CodeGen() {
+                enum { numOps = sizeof( tableValues ) / sizeof( OperatorInfo ) };
+                for ( int i = 0; i < numOps; i += 1 ) {
+                        table[ tableValues[i].inputName ] = tableValues[i];
+                } // for
+        }
+        bool isOperator( const std::string & funcName ) {
+                OperatorInfo info;
+                return operatorLookup( funcName, info );
+        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;
+        }
+        /// 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;
+                }
+        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;
                 return false;
+        }
+        bool isConstructor( const std::string & funcName ) {
+                static OperatorType types[] = { OT_CTOR };
+                return isOperatorType( funcName, std::begin(types), std::end(types) );
+        bool isDestructor( const string & funcName ) {
+                const OperatorInfo * info = operatorLookup( funcName );
+                if ( info ) return info->type == OT_DTOR;
+                return false;
+        }
+        bool isDestructor( const std::string & funcName ) {
+                static OperatorType types[] = { OT_DTOR };
+                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 isAssignment( const std::string & funcName ) {
+                static OperatorType types[] = { OT_PREFIXASSIGN, OT_POSTFIXASSIGN, OT_INFIXASSIGN };
+                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 isCtorDtor( const std::string & funcName ) {
+                static OperatorType types[] = { OT_CTOR, OT_DTOR };
+                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 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) );
+        }
+        CodeGen codegen;                                                                        // initialize singleton package
 } // namespace CodeGen
 // Local Variables: //
 // tab-width: 4 //
-// mode: c++ //
-// compile-command: "make install" //
 // End: //

src/CodeGen/OperatorTable.h

-              r41efd33
+              r04e6f93
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Jul 21 22:17:11 2017
 // Update Count     : 6
+// Last Modified On : Sun Feb 16 08:13:34 2020
+// Update Count     : 26
 //
 …
 #include <string>
+#include <map>
 namespace CodeGen {
         enum OperatorType {
-                OT_INDEX,
                 OT_CTOR,
                 OT_DTOR,
+                OT_CALL,
+                OT_PREFIX,
+                OT_POSTFIX,
+                OT_INFIX,
+                OT_CONSTRUCTOR = OT_DTOR,
                 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 );
+        bool operatorLookup( const std::string & funcName, OperatorInfo & info );
+        const OperatorInfo * operatorLookup( const std::string & funcName );
+        std::string operatorFriendlyName( const std::string & funcName );
         bool isConstructor( const std::string & );

src/CodeGen/Options.h

-              r41efd33
+              r04e6f93
 // Author           : Andrew Beach
 // Created On       : Tue Apr 30 11:36:00 2019
 // Last Modified By : Andrew Beach
 // Last Modified On : Thr May  2 10:45:00 2019
 // Update Count     : 2
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Sat Feb 15 18:37:06 2020
+// Update Count     : 3
 //
 #pragma once
+namespace CodeGen {
+        struct Options {
+                // External Options: Same thoughout a pass.
+                bool pretty;
+                bool genC;
+                bool lineMarks;
+                bool printExprTypes;
+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/CodeTools/ResolvProtoDump.cc

-              r41efd33
+              r04e6f93
 // Author           : Aaron Moss
 // Created On       : Tue Sep 11 09:04:00 2018
 // Last Modified By : Aaron Moss
 // Last Modified On : Tue Sep 11 09:04:00 2018
 // Update Count     : 1
+// Last Modified By : Peter A. Buhr
+// Last Modified On : Sat Feb 15 13:50:11 2020
+// Update Count     : 3
 //
 …
                         // replace operator names
                         CodeGen::OperatorInfo info;
                         if ( CodeGen::operatorLookup( name, info ) ) {
+                        const CodeGen::OperatorInfo * opInfo = CodeGen::operatorLookup( name );
+                        if ( opInfo ) {
                                 ss << new_prefix(pre, "");
                                 op_name( info.outputName, ss );
+                                op_name( opInfo->outputName, ss );
                                 return;
+                        }

src/Common/SemanticError.h

-              r41efd33
+              r04e6f93
 struct WarningData {
         const char * const name;
+        const Severity default_severity;
         const char * const message;
-        const Severity default_severity;
 };
 constexpr WarningData WarningFormats[] = {
+        {"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},
+        {"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" },
 };
 …
         SuperfluousDecl,
         GccAttributes,
+        CppCopy,
         NUMBER_OF_WARNINGS, // This MUST be the last warning
 };

src/Common/utility.h

-              r41efd33
+              r04e6f93
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Jul 24 14:28:19 2019
 // Update Count     : 41
+// Last Modified On : Tue Feb 11 13:00:36 2020
+// Update Count     : 50
 //
 …
 #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 ) {
+// determines if pref is a prefix of str
+static inline bool isPrefix( const std::string & str, const std::string & pref, unsigned int start = 0 ) {
         if ( pref.size() > str.size() ) return false;
         auto its = std::mismatch( pref.begin(), pref.end(), str.begin() );
         return its.first == pref.end();
+    return 0 == memcmp( str.c_str() + start, pref.c_str(), pref.size() );
+        // return prefix == full.substr(0, prefix.size()); // for future, requires c++17
+}

src/Concurrency/Keywords.cc

-              r41efd33
+              r04e6f93
         //      int data;                                  int data;
         //      a_struct_t more_data;                      a_struct_t more_data;
         //                                =>             thread_desc __thrd_d;
+        //                                =>             $thread __thrd_d;
         // };                                        };
         //                                           static inline thread_desc * get_thread( MyThread * this ) { return &this->__thrd_d; }
+        //                                           static inline $thread * get_thread( MyThread * this ) { return &this->__thrd_d; }
         //
         class ThreadKeyword final : public ConcurrentSueKeyword {
 …
                 ThreadKeyword() : ConcurrentSueKeyword(
                         "thread_desc",
+                        "$thread",
                         "__thrd",
                         "get_thread",
 …
         //      int data;                                  int data;
         //      a_struct_t more_data;                      a_struct_t more_data;
         //                                =>             coroutine_desc __cor_d;
+        //                                =>             $coroutine __cor_d;
         // };                                        };
         //                                           static inline coroutine_desc * get_coroutine( MyCoroutine * this ) { return &this->__cor_d; }
+        //                                           static inline $coroutine * get_coroutine( MyCoroutine * this ) { return &this->__cor_d; }
         //
         class CoroutineKeyword final : public ConcurrentSueKeyword {
 …
                 CoroutineKeyword() : ConcurrentSueKeyword(
                         "coroutine_desc",
+                        "$coroutine",
                         "__cor",
                         "get_coroutine",
 …
         //      int data;                                  int data;
         //      a_struct_t more_data;                      a_struct_t more_data;
         //                                =>             monitor_desc __mon_d;
+        //                                =>             $monitor __mon_d;
         // };                                        };
         //                                           static inline monitor_desc * get_coroutine( MyMonitor * this ) { return &this->__cor_d; }
+        //                                           static inline $monitor * get_coroutine( MyMonitor * this ) { return &this->__cor_d; }
         //
         class MonitorKeyword final : public ConcurrentSueKeyword {
 …
                 MonitorKeyword() : ConcurrentSueKeyword(
                         "monitor_desc",
+                        "$monitor",
                         "__mon",
                         "get_monitor",
 …
         //Handles mutex routines definitions :
         // void foo( A * mutex a, B * mutex b,  int i ) {                  void foo( A * a, B * b,  int i ) {
         //                                                                       monitor_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
+        //                                                                       $monitor * __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_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
+        //                                                                       $monitor * __monitors[] = { get_monitor(a), get_monitor(b) };
         //                                                                       monitor_guard_t __guard = { __monitors, 2 };
         //    /*Some code*/                                       =>           /*Some code*/
 …
         Expression * ConcurrentSueKeyword::postmutate( KeywordCastExpr * cast ) {
                 if ( cast_target == cast->target ) {
                         // convert (thread &)t to (thread_desc &)*get_thread(t), etc.
+                        // convert (thread &)t to ($thread &)*get_thread(t), etc.
                         if( !type_decl ) SemanticError( cast, context_error );
                         if( !dtor_decl ) SemanticError( cast, context_error );
 …
                         get_type,
                         nullptr,
                         noAttributes,
+                        { new Attribute("const") },
                         Type::Inline
                 );
 …
         void MutexKeyword::postvisit(StructDecl* decl) {
                 if( decl->name == "monitor_desc" && decl->body ) {
+                if( decl->name == "$monitor" && decl->body ) {
                         assert( !monitor_decl );
                         monitor_decl = decl;
 …
                 );
                 //monitor_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
+                //$monitor * __monitors[] = { get_monitor(a), get_monitor(b) };
                 body->push_front( new DeclStmt( monitors) );
+        }
 …
                 );
                 //monitor_desc * __monitors[] = { get_monitor(a), get_monitor(b) };
+                //$monitor * __monitors[] = { get_monitor(a), get_monitor(b) };
                 body->push_front( new DeclStmt( monitors) );
+        }
 …
         //=============================================================================================
         void ThreadStarter::previsit( StructDecl * decl ) {
                 if( decl->name == "thread_desc" && decl->body ) {
+                if( decl->name == "$thread" && decl->body ) {
                         assert( !thread_decl );
                         thread_decl = decl;

src/Concurrency/Waitfor.cc

r41efd33	r04e6f93
244	244	decl_mask = decl;
245	245	}
246		else if( decl->name == "~~monitor_desc~~" ) {
	246	else if( decl->name == "$monitor" ) {
247	247	assert( !decl_monitor );
248	248	decl_monitor = decl;

src/ControlStruct/Mutate.cc

-              r41efd33
+              r04e6f93
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Aug  4 11:39:08 2016
 // Update Count     : 9
+// Last Modified On : Sun Feb 16 03:22:07 2020
+// Update Count     : 10
 //
 …
                 mutateAll( translationUnit, formut );
+        }
 } // namespace CodeGen
+} // namespace ControlStruct
 // Local Variables: //

src/InitTweak/FixInit.cc

-              r41efd33
+              r04e6f93
 // Created On       : Wed Jan 13 16:29:30 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:41:27 2019
 // Update Count     : 77
+// Last Modified On : Sun Feb 16 04:17:07 2020
+// Update Count     : 82
 //
 #include "FixInit.h"
 …
+                }
                 // 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

-              r41efd33
+              r04e6f93
 // Created On       : Wed Jan 13 16:29:30 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Jul 22 09:31:06 2017
 // Update Count     : 6
+// Last Modified On : Sun Feb 16 07:54:50 2020
+// Update Count     : 8
 //
 …
 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/MakeLibCfa.cc

-              r41efd33
+              r04e6f93
 // Created On       : Sat May 16 10:33:33 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:41:40 2019
 // Update Count     : 42
+// Last Modified On : Sun Feb 16 03:49:49 2020
+// Update Count     : 45
 //
 …
                         FunctionDecl *funcDecl = origFuncDecl->clone();
                         CodeGen::OperatorInfo opInfo;
                         bool lookResult = CodeGen::operatorLookup( funcDecl->get_name(), opInfo );
                         assert( lookResult );
+                        const CodeGen::OperatorInfo * opInfo;
+                        opInfo = CodeGen::operatorLookup( funcDecl->get_name() );
+                        assert( opInfo );
                         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/Parser/ParserTypes.h

-              r41efd33
+              r04e6f93
 // Created On       : Sat Sep 22 08:58:10 2001
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Jul 22 09:33:28 2017
 // Update Count     : 350
+// Last Modified On : Sat Feb 15 11:04:40 2020
+// Update Count     : 351
 //
 …
 // 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/TypedefTable.cc

-              r41efd33
+              r04e6f93
 // Created On       : Sat May 16 15:20:13 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Jul 25 15:32:35 2018
 // Update Count     : 258
+// Last Modified On : Sat Feb 15 08:06:36 2020
+// Update Count     : 259
 //
 …
 } // TypedefTable::~TypedefTable
 bool TypedefTable::exists( const string & identifier ) {
+bool TypedefTable::exists( const string & identifier ) const {
         return kindTable.find( identifier ) != kindTable.end();
 } // TypedefTable::exists
 bool TypedefTable::existsCurr( const string & identifier ) {
+bool TypedefTable::existsCurr( const string & identifier ) const {
         return kindTable.findAt( kindTable.currentScope() - 1, identifier ) != kindTable.end();
 } // TypedefTable::exists

src/Parser/TypedefTable.h

-              r41efd33
+              r04e6f93
 // Created On       : Sat May 16 15:24:36 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Jul 25 15:33:55 2018
 // Update Count     : 114
+// Last Modified On : Sat Feb 15 08:06:37 2020
+// Update Count     : 117
 //
 …
         ~TypedefTable();
         bool exists( const std::string & identifier );
         bool existsCurr( const std::string & identifier );
+        bool exists( const std::string & identifier ) const;
+        bool existsCurr( const std::string & identifier ) const;
         int isKind( const std::string & identifier ) const;
         void makeTypedef( const std::string & name, int kind = TYPEDEFname );

src/Parser/lex.ll

-              r41efd33
+              r04e6f93
  * Created On       : Sat Sep 22 08:58:10 2001
  * Last Modified By : Peter A. Buhr
  * Last Modified On : Fri Feb  7 19:02:43 2020
  * Update Count     : 725
+ * Last Modified On : Sat Feb 15 11:05:50 2020
+ * Update Count     : 737
  */
 …
 #include "ParseNode.h"
 #include "TypedefTable.h"
+string * build_postfix_name( string * name );
 char *yyfilename;
 …
 "?"({op_unary_pre_post}|"()"|"[?]"|"{}") { IDENTIFIER_RETURN(); }
 "^?{}"                  { IDENTIFIER_RETURN(); }
+"?`"{identifier} { IDENTIFIER_RETURN(); }                               // postfix operator
+"?`"{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 ) );
+}
 "?"{op_binary_over}"?"  { IDENTIFIER_RETURN(); }                // binary
         /*

src/Parser/parser.yy

-              r41efd33
+              r04e6f93
 // Created On       : Sat Sep  1 20:22:55 2001
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb  1 10:04:40 2020
 // Update Count     : 4440
+// Last Modified On : Wed Feb 26 14:27:39 2020
+// Update Count     : 4472
 //
 …
 } // rebindForall
+NameExpr * build_postfix_name( const string * name ) {
+        NameExpr * new_name = build_varref( new string( "?`" + *name ) );
+        delete name;
+        return new_name;
+string * build_postfix_name( string * name ) {
+        *name = string("__postfix_func_") + *name;
+        return name;
 } // build_postfix_name
 …
                 { $$ = new ExpressionNode( build_func( $1, $3 ) ); }
         | postfix_expression '`' identifier                                     // CFA, postfix call
                 { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $3 ) ), $1 ) ); }
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_varref( build_postfix_name( $3 ) ) ), $1 ) ); }
         | constant '`' identifier                                                       // CFA, postfix call
                 { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $3 ) ), $1 ) ); }
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_varref( build_postfix_name( $3 ) ) ), $1 ) ); }
         | string_literal '`' identifier                                         // CFA, postfix call
                 { $$ = new ExpressionNode( build_func( new ExpressionNode( build_postfix_name( $3 ) ), new ExpressionNode( $1 ) ) ); }
+                { $$ = new ExpressionNode( build_func( new ExpressionNode( build_varref( build_postfix_name( $3 ) ) ), new ExpressionNode( $1 ) ) ); }
         | postfix_expression '.' identifier
                 { $$ = new ExpressionNode( build_fieldSel( $1, build_varref( $3 ) ) ); }
 …
                 // 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 ')'
+        cfa_abstract_tuple identifier_or_type_name '(' push cfa_parameter_ellipsis_list_opt pop ')' attribute_list_opt
                 // To obtain LR(1 ), this rule must be factored out from function return type (see cfa_abstract_declarator).
                 { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 ); }
         | cfa_function_return identifier_or_type_name '(' push cfa_parameter_ellipsis_list_opt pop ')'
                 { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 ); }
+                { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 )->addQualifiers( $8 ); }
+        | cfa_function_return identifier_or_type_name '(' push cfa_parameter_ellipsis_list_opt pop ')' attribute_list_opt
+                { $$ = DeclarationNode::newFunction( $2, $1, $5, 0 )->addQualifiers( $8 ); }
+        ;
 …
 aggregate_control:                                                                              // CFA
         GENERATOR
+                { yyy = true; $$ = AggregateDecl::Coroutine; }
+                { SemanticError( yylloc, "generator is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
+        | MONITOR GENERATOR
+                { SemanticError( yylloc, "monitor generator is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
         | COROUTINE
                 { yyy = true; $$ = AggregateDecl::Coroutine; }
         | MONITOR
                 { yyy = true; $$ = AggregateDecl::Monitor; }
+        | MONITOR COROUTINE
+                { SemanticError( yylloc, "monitor coroutine is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
         | THREAD
                 { yyy = true; $$ = AggregateDecl::Thread; }
+        | MONITOR THREAD
+                { SemanticError( yylloc, "monitor thread is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
+        ;

src/ResolvExpr/AlternativeFinder.cc

-              r41efd33
+              r04e6f93
                 try {
                         // Attempt 1 : turn (thread&)X into (thread_desc&)X.__thrd
+                        // Attempt 1 : turn (thread&)X into ($thread&)X.__thrd
                         // Clone is purely for memory management
                         std::unique_ptr<Expression> tech1 { new UntypedMemberExpr(new NameExpr(castExpr->concrete_target.field), castExpr->arg->clone()) };
 …
                 } catch(SemanticErrorException & ) {}
                 // Fallback : turn (thread&)X into (thread_desc&)get_thread(X)
+                // Fallback : turn (thread&)X into ($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

src/ResolvExpr/Resolver.cc

-              r41efd33
+              r04e6f93
                 void previsit( ThrowStmt * throwStmt );
                 void previsit( CatchStmt * catchStmt );
+                void postvisit( CatchStmt * catchStmt );
                 void previsit( WaitForStmt * stmt );
 …
         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 ) {
+                        findSingleExpression( catchStmt->cond, new BasicType( noQualifiers, BasicType::Bool ), indexer );
+                        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;
+                }
+        }
 …
         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/SymTab/Demangle.cc

-              r41efd33
+              r04e6f93
 // Created On       : Thu Jul 19 12:52:41 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 14:54:15 2019
 // Update Count     : 4
+// Last Modified On : Tue Feb 11 15:09:18 2020
+// Update Count     : 10
 //
 …
 #include "CodeGen/GenType.h"
 #include "Common/PassVisitor.h"
+#include "Common/utility.h"                                                             // isPrefix
 #include "Mangler.h"
 #include "SynTree/Type.h"
 …
                         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()) {
+                                // 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 ) ) {
                                         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/Mangler.cc

-              r41efd33
+              r04e6f93
 // Created On       : Sun May 17 21:40:29 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:43:49 2019
 // Update Count     : 28
+// Last Modified On : Sat Feb 15 13:55:12 2020
+// Update Count     : 33
 //
 #include "Mangler.h"
 …
                                 } // if
                                 mangleName << Encoding::manglePrefix;
                                 CodeGen::OperatorInfo opInfo;
                                 if ( operatorLookup( declaration->get_name(), opInfo ) ) {
                                         mangleName << opInfo.outputName.size() << opInfo.outputName;
+                                const CodeGen::OperatorInfo * opInfo = CodeGen::operatorLookup( declaration->get_name() );
+                                if ( opInfo ) {
+                                        mangleName << opInfo->outputName.size() << opInfo->outputName;
                                 } else {
                                         mangleName << declaration->name.size() << declaration->name;
 …
                         } // if
                         mangleName << Encoding::manglePrefix;
                         CodeGen::OperatorInfo opInfo;
                         if ( operatorLookup( decl->name, opInfo ) ) {
                                 mangleName << opInfo.outputName.size() << opInfo.outputName;
+                        const CodeGen::OperatorInfo * opInfo = CodeGen::operatorLookup( decl->name );
+                        if ( opInfo ) {
+                                mangleName << opInfo->outputName.size() << opInfo->outputName;
                         } else {
                                 mangleName << decl->name.size() << decl->name;

src/SymTab/Validate.cc

-              r41efd33
+              r04e6f93
                         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
 …
                 if ( CodeGen::isCtorDtorAssign( funcDecl->get_name() ) ) { // TODO: also check /=, etc.
                         if ( params.size() == 0 ) {
                                 SemanticError( funcDecl, "Constructors, destructors, and assignment functions require at least one parameter " );
+                                SemanticError( funcDecl->location, "Constructors, destructors, and assignment functions require at least one parameter." );
+                        }
                         ReferenceType * refType = dynamic_cast< ReferenceType * >( params.front()->get_type() );
                         if ( ! refType ) {
                                 SemanticError( funcDecl, "First parameter of a constructor, destructor, or assignment function must be a reference " );
+                                SemanticError( funcDecl->location, "First parameter of a constructor, destructor, or assignment function must be a reference." );
+                        }
                         if ( CodeGen::isCtorDtor( funcDecl->get_name() ) && returnVals.size() != 0 ) {
+                                SemanticError( funcDecl, "Constructors and destructors cannot have explicit return values " );
+                                if(!returnVals.front()->get_type()->isVoid()) {
+                                        SemanticError( funcDecl->location, "Constructors and destructors cannot have explicit return values." );
+                                }
+                        }
+                }

src/SynTree/Attribute.h

-              r41efd33
+              r04e6f93
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Jul 22 09:54:14 2017
 // Update Count     : 39
+// Last Modified On : Thu Feb 13 21:34:08 2020
+// Update Count     : 40
 //
 …
         virtual ~Attribute();
         std::string get_name() const { return name; }
+        const std::string & get_name() const { return name; }
         void set_name( const std::string & newValue ) { name = newValue; }
         std::list< Expression * > & get_parameters() { return parameters; }

src/cfa.make

-              r41efd33
+              r04e6f93
 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_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_0 = @echo "  NODEJS  " $@;
 am__v_NODEJS_1 =

tests/.expect/alloc-ERROR.txt

-              r41efd33
+              r04e6f93
 alloc.cfa:311:1 error: No reasonable alternatives for expression Applying untyped:
+alloc.cfa:310:1 error: No reasonable alternatives for expression Applying untyped:
   Name: ?=?
 ...to:
 …
 alloc.cfa:312:1 error: No reasonable alternatives for expression Applying untyped:
+alloc.cfa:311:1 error: No reasonable alternatives for expression Applying untyped:
   Name: ?=?
 ...to:
 …
 alloc.cfa:313:1 error: No reasonable alternatives for expression Applying untyped:
+alloc.cfa:312:1 error: No reasonable alternatives for expression Applying untyped:
   Name: ?=?
 ...to:
 …
 alloc.cfa:314:1 error: No reasonable alternatives for expression Applying untyped:
+alloc.cfa:313:1 error: No reasonable alternatives for expression Applying untyped:
   Name: ?=?
 ...to:

tests/.expect/alloc.txt

r41efd33	r04e6f93
2	2	CFA malloc 0xdeadbeef
3	3	CFA alloc 0xdeadbeef
4		~~CFA array alloc, fill 0xde~~
5	4	CFA alloc, fill dededede
6	5	CFA alloc, fill 3

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

r41efd33	r04e6f93
1		nested-types.cfa:70:1 error: Use of undefined type T
	1	nested-types.cfa:83:1 error: Use of undefined type T

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

-              r41efd33
+              r04e6f93
 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:
+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:
   instance of struct S with body 1
   instance of type Z (not function type)

tests/Makefile.in

-              r41efd33
+              r04e6f93
 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_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_0 = @echo "  NODEJS  " $@;
 am__v_NODEJS_1 =
 AM_V_JAVAC = $(am__v_JAVAC_@AM_V@)

tests/alloc.cfa

-              r41efd33
+              r04e6f93
 // Created On       : Wed Feb  3 07:56:22 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Nov 22 15:34:19 2019
 // Update Count     : 404
+// Last Modified On : Sun Feb 16 09:21:13 2020
+// Update Count     : 405
 //
 …
         p = alloc_set( fill );                                                          // CFA alloc, fill
-        printf( "CFA array alloc, fill %#hhx\n", fill );
         printf( "CFA alloc, fill %08x\n", *p );
         free( p );

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

-              r41efd33
+              r04e6f93
-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/datingService.cfa

-              r41efd33
+              r04e6f93
 //
 // 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/quickSort.cfa

-              r41efd33
+              r04e6f93
 // Created On       : Wed Dec  6 12:15:52 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Oct 10 13:58:18 2019
 // Update Count     : 176
+// Last Modified On : Wed Feb 12 18:24:47 2020
+// Update Count     : 177
 //
 …
 void ?{}( Quicksort & qs, int values[], int size, int depth ) {
         qs.values = values;  qs.low = 0;  qs.high = size;  qs.depth = depth;
+        qs.[values, low, high, depth] = [values, 0, size, depth];
 } // Quicksort
 …
                         values[counter] = size - counter;                       // descending values
                 } // for
                 for ( int i = 0; i < 200; i +=1 ) {                             // random shuffle a few values
+                for ( i; 200 ) {                                                                // random shuffle a few values
                         swap( values[rand() % size], values[rand() % size] );
                 } // for

tests/concurrent/multi-monitor.cfa

-              r41efd33
+              r04e6f93
 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/signal/block.cfa

-              r41efd33
+              r04e6f93
 monitor global_data_t {
         thread_desc * last_thread;
         thread_desc * last_signaller;
+        $thread * last_thread;
+        $thread * last_signaller;
 };
 …
         if( !is_empty( cond ) ) {
                 thread_desc * next = front( cond );
+                $thread * next = front( cond );
                 if( ! signal_block( cond ) ) {

tests/errors/.expect/completeType.txt

-              r41efd33
+              r04e6f93
 completeTypeError.cfa:34:1 error: Cannot choose between 2 alternatives for expression
+errors/completeType.cfa:34:1 error: Cannot choose between 2 alternatives for expression
 Generated Cast of:
   Applying untyped:
 …
 completeTypeError.cfa:35:1 error: No reasonable alternatives for expression Applying untyped:
+errors/completeType.cfa:35:1 error: No reasonable alternatives for expression Applying untyped:
   Name: foo
 ...to:
   Name: v
 completeTypeError.cfa:36:1 error: No reasonable alternatives for expression Applying untyped:
+errors/completeType.cfa:36:1 error: No reasonable alternatives for expression Applying untyped:
   Name: baz
 ...to:
   Name: v
 completeTypeError.cfa:37:1 error: No reasonable alternatives for expression Applying untyped:
+errors/completeType.cfa:37:1 error: No reasonable alternatives for expression Applying untyped:
   Name: quux
 ...to:
   Name: v
 completeTypeError.cfa:59:1 error: No reasonable alternatives for expression Applying untyped:
+errors/completeType.cfa:59:1 error: No reasonable alternatives for expression Applying untyped:
   Name: baz
 ...to:
   Name: y
 completeTypeError.cfa:60:1 error: No reasonable alternatives for expression Applying untyped:
+errors/completeType.cfa:60:1 error: No reasonable alternatives for expression Applying untyped:
   Name: quux
 ...to:
   Name: y
 completeTypeError.cfa:72:1 error: No alternatives with satisfiable assertions for Applying untyped:
+errors/completeType.cfa:72:1 error: No alternatives with satisfiable assertions for Applying untyped:
   Name: baz
 ...to:

tests/exceptions/except-0.cfa

-              r41efd33
+              r04e6f93
 };
 void ?{}(signal_exit * this, const char * area) {
         this->area = area;
+}
 void ^?{}(signal_exit * this) {
         printf("Exiting: %s\n", this->area);
+void ?{}(signal_exit & this, const char * area) {
+        this.area = area;
+}
+void ^?{}(signal_exit & this) {
+        printf("Exiting: %s\n", this.area);
 //      sout | "Exiting:" | this->area;
+}
 …
         // Uncaught termination test.
+        /* Removed due to non-deterministic output.
         printf("Throw uncaught.\n");
         yang z;
         terminate(&z);
+}
+        */
+}

tests/exceptions/except-2.cfa

-              r41efd33
+              r04e6f93
         struct TABLE(BASE_EXCEPT) const * parent;
         size_t size;
         void (*copy)(num_error *this, num_error * other);
         void (*free)(num_error *this);
+        void (*copy)(num_error &this, num_error & other);
+        void (*free)(num_error &this);
         const char * (*msg)(num_error *this);
         int (*code)(num_error *this);
 …
         if ( ! this->msg ) {
                 static const char * base = "Num Error with code: X";
                 this->msg = malloc(22);
+                this->msg = (char *)malloc(22);
                 for (int i = 0 ; (this->msg[i] = base[i]) ; ++i);
+        }
 …
         return this->msg;
+}
 void ?{}(num_error * this, int num) {
         this->virtual_table = &INSTANCE(num_error);
         this->msg = 0;
         this->num = num;
+void ?{}(num_error & this, int num) {
+        this.virtual_table = &INSTANCE(num_error);
+        this.msg = 0;
+        this.num = num;
+}
 void ?{}(num_error * this, num_error * other) {
         this->virtual_table = other->virtual_table;
         this->msg = 0;
         this->num = other->num;
+void ?{}(num_error & this, num_error & other) {
+        this.virtual_table = other.virtual_table;
+        this.msg = 0;
+        this.num = other.num;
+}
 void ^?{}(num_error * this) {
         if( this->msg ) free( this->msg );
+void ^?{}(num_error & this) {
+        if( this.msg ) free( this.msg );
+}
 int num_error_code( num_error * this ) {

tests/nested-types.cfa

-              r41efd33
+              r04e6f93
 // Created On       : Mon Jul 9 10:20:03 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Nov  6 17:59:40 2018
 // Update Count     : 2
+// Last Modified On : Wed Feb 12 18:21:15 2020
+// Update Count     : 3
 //
 …
 //   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/test.py

r41efd33	r04e6f93
152	152	# run everything in a temp directory to make sure core file are handled properly
153	153	with tempdir():
154		# if the make command succeds continue otherwise skip to diff
	154	# if the make command succeeds continue otherwise skip to diff
155	155	if success(make_ret):
156	156	with Timed() as run_dur:

tests/userLiterals.cfa

-              r41efd33
+              r04e6f93
 // file "LICENCE" distributed with Cforall.
 //
 // user_literals.cfa --
+// userLiterals.cfa --
 //
 // Author           : Peter A. Buhr
 // Created On       : Wed Sep  6 21:40:50 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Dec  4 22:03:10 2018
 // Update Count     : 56
+// Last Modified On : Wed Feb 19 07:48:45 2020
+// Update Count     : 74
 //
 …
 int ?`__thingy_( int x ) { sout | "_thingy_" | x; return x; }
 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 ?`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 ?`_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 | w.stones; }
+ofstream & ?|?( ofstream & os, Weight w ) { return os | wd(1,1, w.stones); }
+void ?|?( ofstream & os, Weight w ) { (ofstream)(os | w); ends( os ); }
 Weight ?`st( double w ) { return (Weight){ w }; }               // backquote for user literals
 …
         sout | w;
 //      0`secs;
+`s;
 `s;
 `s;
 …
         "abc"`s;
+//      u"abc"`m;
+//      U_"abc"`h;
+//      L"abc"`_A_;
+        // FIX ME: requires char16_t, char32_t, and wchar_t be unique types
+        // u"abc"`m;
+        // U_"abc"`h;
+        // L"abc"`_A_;
         u8_"abc"`__thingy_;
 } // main
 …
 // Local Variables: //
 // tab-width: 4 //
 // compile-command: "cfa user_literals.cfa" //
+// compile-command: "cfa userLiterals.cfa" //
 // End: //

Context Navigation

Legend:

Download in other formats: