Changeset 22f94a4

.gitignore

-              r07d867b
+              r22f94a4
 # generated by configure
+aclocal.m4
+automake
 autom4te.cache
 config.h
 …
 config.log
 config.py
+configure
+libtool
 stamp-h1
-libtool
 /Makefile
+/Makefile.in
 **/Makefile
+**/Makefile.in
 /version

Jenkinsfile_disabled

-              r07d867b
+              r22f94a4
         release = Settings.RunAllTests || Settings.RunBenchmark
         build_stage('Build : configure', true) {
+                // Configure must be run inside the tree
+                dir (SrcDir) {
+                        // Generate the necessary build files
+                        sh './autogen.sh'
+                }
                 // Build outside of the src tree to ease cleaning
                 dir (BuildDir) {

benchmark/Makefile.am

r07d867b	r22f94a4
19	19
20	20	# applies to both programs
21		include $(top_srcdir)/~~src~~/cfa.make
	21	include $(top_srcdir)/tools/build/cfa.make
22	22
23	23	AM_CFLAGS = -O2 -Wall -Wextra -I$(srcdir) -lrt -pthread # -Werror

benchmark/creation/node_cor.js

-              r07d867b
+              r22f94a4
 function * coroutine() { yield }
+for ( var i = 0; i < times; i += 1 ) { // warm jit
+        cor = coroutine()
+}
 var hrstart = process.hrtime()
 for ( var i = 0; i < times; i += 1 ) {

benchmark/ctxswitch/node_cor.js

-              r07d867b
+              r22f94a4
+}
 cor = coroutine()
+for ( var i = 0; i < times; i += 1 ) { // warm git
+        cor.next();
+}
 var hrstart = process.hrtime()

benchmark/exclude

r07d867b	r22f94a4
10	10	interrupt_linux.c
11	11	exclude
	12	io
12	13	Monitor.c

benchmark/io/readv.cfa

-              r07d867b
+              r22f94a4
+#define _GNU_SOURCE
 #include <stdlib.h>
 #include <stdio.h>
 …
+}
+#include <errno.h>
 #include <unistd.h>
 #include <clock.hfa>
+#include <iofwd.hfa>
 #include <kernel.hfa>
 #include <thread.hfa>
 #include <time.hfa>
+#include <stats.hfa>
+#include "../benchcltr.hfa"
 extern bool traceHeapOn();
-extern ssize_t cfa_preadv2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags);
 int fd;
 …
 volatile size_t count = 0;
+unsigned long int buflen = 50;
+unsigned long int buflen = 512;
+bool fixed_file = false;
+cluster * the_cluster;
+thread Reader {};
+thread __attribute__((aligned(128))) Reader {};
 void ?{}( Reader & this ) {
         ((thread&)this){ "Reader Thread", *the_cluster };
+        ((thread&)this){ "Reader Thread", *the_benchmark_cluster };
+}
+struct my_processor {
+        processor p;
+};
+void ?{}( my_processor & this ) {
+        (this.p){ "I/O Processor", *the_cluster };
+int do_read(int fd, struct iovec * iov) {
+        // extern ssize_t cfa_preadv2(int, const struct iovec *, int, off_t, int, int = 0, Duration = -1`s, io_cancellation * = 0p, io_context * = 0p);
+        int sflags = 0;
+        if(fixed_file) {
+                sflags |= CFA_IO_FIXED_FD1;
+        }
+        return cfa_preadv2(fd, iov, 1, 0, 0, sflags, -1`s, 0p, 0p);
+}
 void main( Reader & ) {
+        while(!__atomic_load_n(&run, __ATOMIC_RELAXED)) yield();
+        park( __cfaabi_dbg_ctx );
+        /* paranoid */ assert( true == __atomic_load_n(&run, __ATOMIC_RELAXED) );
         char data[buflen];
+        __attribute__((aligned(512)))  char data[buflen];
         struct iovec iov = { data, buflen };
         while(__atomic_load_n(&run, __ATOMIC_RELAXED)) {
                 int r = cfa_preadv2(fd, &iov, 1, 0, 0);
                 if(r < 0) abort("%s\n", strerror(-r));
+                int r = do_read(fd, &iov);
+                if(r < 0) abort("%s\n", strerror(errno));
                 __atomic_fetch_add( &count, 1, __ATOMIC_SEQ_CST );
 …
 int main(int argc, char * argv[]) {
+        double duration   = 5.0;
+        unsigned long int nthreads = 2;
+        unsigned long int nprocs   = 1;
+        int flags = 0;
+        BENCH_DECL
+        unsigned num_io = 1;
+        io_context_params params;
+        int file_flags = 0;
+        unsigned sublen = 16;
         arg_loop:
         for(;;) {
                 static struct option options[] = {
+                        {"duration",   required_argument, 0, 'd'},
+                        {"nthreads",   required_argument, 0, 't'},
+                        {"nprocs",     required_argument, 0, 'p'},
+                        {"bufsize",    required_argument, 0, 'b'},
+                        {"userthread", no_argument      , 0, 'u'},
+                        BENCH_OPT_LONG
+                        {"bufsize",       required_argument, 0, 'b'},
+                        {"submitthread",  no_argument      , 0, 's'},
+                        {"eagersubmit",   no_argument      , 0, 'e'},
+                        {"kpollsubmit",   no_argument      , 0, 'k'},
+                        {"kpollcomplete", no_argument      , 0, 'i'},
+                        {"fixed-files",   no_argument      , 0, 'f'},
+                        {"open-direct",   no_argument      , 0, 'o'},
+                        {"submitlength",  required_argument, 0, 'l'},
                         {0, 0, 0, 0}
                 };
                 int idx = 0;
                 int opt = getopt_long(argc, argv, "d:t:p:b:u", options, &idx);
+                int opt = getopt_long(argc, argv, BENCH_OPT_SHORT "b:sekil:", options, &idx);
                 const char * arg = optarg ? optarg : "";
 …
                         case -1:
                                 break arg_loop;
+                        // Numeric Arguments
+                        case 'd':
+                                duration = strtod(arg, &end);
+                                if(*end != '\0') {
+                                        fprintf(stderr, "Duration must be a valid double, was %s\n", arg);
+                                        goto usage;
+                                }
+                                break;
+                        case 't':
+                                nthreads = strtoul(arg, &end, 10);
+                                if(*end != '\0' || nthreads < 1) {
+                                        fprintf(stderr, "Number of threads must be a positive integer, was %s\n", arg);
+                                        goto usage;
+                                }
+                                break;
+                        case 'p':
+                                nprocs = strtoul(arg, &end, 10);
+                                if(*end != '\0' || nprocs < 1) {
+                                        fprintf(stderr, "Number of processors must be a positive integer, was %s\n", arg);
+                                        goto usage;
+                                }
+                                break;
+                        BENCH_OPT_CASE
                         case 'b':
                                 buflen = strtoul(arg, &end, 10);
 …
+                                }
                                 break;
                         case 'u':
                                 flags |= CFA_CLUSTER_IO_POLLER_USER_THREAD;
+                        case 's':
+                                params.poller_submits = true;
                                 break;
+                        // Other cases
+                        case 'e':
+                                params.eager_submits = true;
+                                break;
+                        case 'k':
+                                params.poll_submit = true;
+                        case 'f':
+                                fixed_file = true;
+                                break;
+                        case 'i':
+                                params.poll_complete = true;
+                        case 'o':
+                                file_flags |= O_DIRECT;
+                                break;
+                        case 'l':
+                                sublen = strtoul(arg, &end, 10);
+                                if(*end != '\0' && sublen < 16) {
+                                        fprintf(stderr, "Submit length must be at least 16, was %s\n", arg);
+                                        goto usage;
+                                }
+                                // flags |= (sublen << CFA_CLUSTER_IO_BUFFLEN_OFFSET);
+                                break;
                         default: /* ? */
                                 fprintf(stderr, "%d\n", opt);
                         usage:
+                                fprintf(stderr, "Usage: %s : [options]\n", argv[0]);
+                                fprintf(stderr, "\n");
+                                fprintf(stderr, "  -d, --duration=DURATION  Duration of the experiment, in seconds\n");
+                                fprintf(stderr, "  -t, --nthreads=NTHREADS  Number of user threads\n");
+                                fprintf(stderr, "  -p, --nprocs=NPROCS      Number of kernel threads\n");
+                                fprintf(stderr, "  -b, --buflen=SIZE        Number of bytes to read per request\n");
+                                bench_usage( argv );
+                                fprintf( stderr, "  -b, --buflen=SIZE        Number of bytes to read per request\n" );
+                                fprintf( stderr, "  -u, --userthread         If set, cluster uses user-thread to poll I/O\n" );
+                                fprintf( stderr, "  -s, --submitthread       If set, cluster uses polling thread to submit I/O\n" );
+                                fprintf( stderr, "  -e, --eagersubmit        If set, cluster submits I/O eagerly but still aggregates submits\n" );
+                                fprintf( stderr, "  -k, --kpollsubmit        If set, cluster uses IORING_SETUP_SQPOLL\n" );
+                                fprintf( stderr, "  -i, --kpollcomplete      If set, cluster uses IORING_SETUP_IOPOLL\n" );
+                                fprintf( stderr, "  -l, --submitlength=LEN   Max number of submitions that can be submitted together\n" );
                                 exit(EXIT_FAILURE);
+                }
+        }
         fd = open(__FILE__, 0);
         if(fd < 0) {
+        int lfd = open(__FILE__, file_flags);
+        if(lfd < 0) {
                 fprintf(stderr, "Could not open source file\n");
                 exit(EXIT_FAILURE);
+        }
         printf("Running %lu threads, reading %lu bytes each, over %lu processors for %lf seconds\n", nthreads, buflen, nprocs, duration);
+        printf("Running %d threads, reading %lu bytes each, over %d processors for %f seconds\n", nthreads, buflen, nprocs, duration);
+        {
                 Time start, end;
+                cluster cl = { "IO Cluster", flags };
+                the_cluster = &cl;
+                #if !defined(__CFA_NO_STATISTICS__)
+                        print_stats_at_exit( cl );
+                #endif
+                BenchCluster cl = { num_io, params, CFA_STATS_READY_Q | CFA_STATS_IO };
+                if(fixed_file) {
+                        fd = 0;
+                        register_fixed_files( cl.self, &lfd, 1 );
+                }
+                else {
+                        fd = lfd;
+                }
+                {
                         my_processor procs[nprocs];
+                        BenchProc procs[nprocs];
+                        {
                                 Reader threads[nthreads];
                                 printf("Starting\n");
+                                start = getTime();
+                                bool is_tty = isatty(STDOUT_FILENO);
+                                start = getTimeNsec();
                                 run = true;
+                                do {
+                                        sleep(500`ms);
+                                        end = getTime();
+                                } while( (end - start) < duration`s );
+                                for(i; nthreads) {
+                                        unpark( threads[i] __cfaabi_dbg_ctx2 );
+                                }
+                                wait(duration, start, end, is_tty);
                                 run = false;
                                 end = getTime();
                                 printf("Done\n");
+                                end = getTimeNsec();
+                                printf("\nDone\n");
+                        }
+                        printf("Readers closed\n");
+                }
                 printf("Took %'ld ms\n", (end - start)`ms);
 …
+        }
         close(fd);
+        close(lfd);
+}

configure.ac

-              r07d867b
+              r22f94a4
 AC_CONFIG_AUX_DIR([automake])
 AC_CONFIG_MACRO_DIRS([automake])
-#AC_CONFIG_SRCDIR([src/main.cc])
 AC_CONFIG_HEADERS([config.h:src/config.h.in])
 AM_SILENT_RULES([yes])
 m4_include([automake/cfa.m4])
+m4_include([tools/build/cfa.m4])
 # don't use the default CFLAGS as they unconditonnaly add -O2
 …
         [  --enable-distcc     whether or not to enable distributed compilation],
         enable_distcc=$enableval, enable_distcc=no)
+AC_ARG_WITH(bwlimit,
+        [  --with-bwlimit=RATE     RATE the maximum rate at which rsync will be limited when using distributed builds],
+        [], [])
 AM_CONDITIONAL([ENABLE_DISTCC], [test x$enable_distcc = xyes])

doc/bibliography/pl.bib

-              r07d867b
+              r22f94a4
+}
 @misc{CforallBenchMarks,
+@misc{CforallConcurrentBenchmarks,
     contributer = {pabuhr@plg},
     key         = {Cforall Benchmarks},
     author      = {{\textsf{C}{$\mathbf{\forall}$} Benchmarks}},
     howpublished= {\href{https://plg.uwaterloo.ca/~cforall/benchmark.tar}{https://\-plg.uwaterloo.ca/\-$\sim$cforall/\-benchmark.tar}},
+    howpublished= {\href{https://plg.uwaterloo.ca/~cforall/doc/CforallConcurrentBenchmarks.tar}{https://\-plg.uwaterloo.ca/\-$\sim$cforall/\-doc/\-CforallConcurrentBenchmarks.tar}},
+}
 …
     author      = {Adya, Atul and Howell, Jon and Theimer, Marvin and Bolosky, William J. and Douceur, John R.},
     title       = {Cooperative Task Management Without Manual Stack Management},
     booktitle   = {Proceedings of the General Track of the Annual Conference on USENIX Annual Technical Conference},
+    booktitle   = {Proc. of the General Track USENIX Tech. Conf.},
     series      = {ATEC '02},
     year        = {2002},

doc/papers/AMA/AMA-stix/ama/WileyNJD-v2.cls

-              r07d867b
+              r22f94a4
      \@afterheading}
 \renewcommand\section{\@startsection{section}{1}{\z@}{-25pt \@plus -2pt \@minus -2pt}{12\p@}{\sectionfont}}%
 \renewcommand\subsection{\@startsection{subsection}{2}{\z@}{-22pt \@plus -2pt \@minus -2pt}{5\p@}{\subsectionfont}}%
 \renewcommand\subsubsection{\@startsection{subsubsection}{3}{\z@}{-20pt \@plus -2pt \@minus -2pt}{2\p@}{\subsubsectionfont}}%
+\renewcommand\section{\@startsection{section}{1}{\z@}{-20pt \@plus -2pt \@minus -2pt}{7\p@}{\sectionfont}}%
+\renewcommand\subsection{\@startsection{subsection}{2}{\z@}{-18pt \@plus -2pt \@minus -2pt}{5\p@}{\subsectionfont}}%
+\renewcommand\subsubsection{\@startsection{subsubsection}{3}{\z@}{-16pt \@plus -2pt \@minus -2pt}{2\p@}{\subsubsectionfont}}%
+%
 \newskip\secruleskip\secruleskip8.5\p@%

doc/papers/concurrency/Paper.tex

-              r07d867b
+              r22f94a4
 \newcommand{\CRT}{\global\columnposn=\gcolumnposn}
 % Denote newterms in particular font and index them without particular font and in lowercase, e.g., \newterm{abc}.
 % The option parameter provides an index term different from the new term, e.g., \newterm[\texttt{abc}]{abc}
+% Denote newterms in particular font and index them without particular font and in lowercase, \eg \newterm{abc}.
+% The option parameter provides an index term different from the new term, \eg \newterm[\texttt{abc}]{abc}
 % The star version does not lowercase the index information, e.g., \newterm*{IBM}.
 \newcommand{\newtermFontInline}{\emph}
 …
 {\lstset{language=python,moredelim=**[is][\protect\color{red}]{`}{`},#1}\lstset{#1}}
 {}
+\lstnewenvironment{java}[1][]
+{\lstset{language=java,moredelim=**[is][\protect\color{red}]{`}{`},#1}\lstset{#1}}
+{}
 % inline code @...@
 …
 \abstract[Summary]{
 \CFA is a polymorphic, non-object-oriented, concurrent, backwards-compatible extension of the C programming language.
+\CFA is a polymorphic, non-object-oriented, concurrent, backwards compatible extension of the C programming language.
 This paper discusses the design philosophy and implementation of its advanced control-flow and concurrent/parallel features, along with the supporting runtime written in \CFA.
 These features are created from scratch as ISO C has only low-level and/or unimplemented concurrency, so C programmers continue to rely on library approaches like pthreads.
 …
 % Library extension for executors, futures, and actors are built on these basic mechanisms.
 The runtime provides significant programmer simplification and safety by eliminating spurious wakeup and monitor barging.
 The runtime also ensures multiple monitors can be safely acquired \emph{simultaneously} (deadlock free), and this feature is fully integrated with all monitor synchronization mechanisms.
+The runtime also ensures multiple monitors can be safely acquired in a deadlock-free way, and this feature is fully integrated with all monitor synchronization mechanisms.
 All control-flow features integrate with the \CFA polymorphic type-system and exception handling, while respecting the expectations and style of C programmers.
 Experimental results show comparable performance of the new features with similar mechanisms in other concurrent programming languages.
 …
 \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.
+\CFA has object-oriented features, such as constructors, destructors, 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.},
+% 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 and there is no mechanism to designate a function parameter as a receiver, \lstinline@this@, parameter.},
 backwards-compatible extension of the C programming language.
 In many ways, \CFA is to C as Scala~\cite{Scala} is to Java, providing a \emph{research vehicle} for new typing and control-flow capabilities on top of a highly popular programming language\footnote{
 The TIOBE index~\cite{TIOBE} for December 2019 ranks the top five \emph{popular} programming languages as Java 17\%, C 16\%, Python 10\%, and \CC 6\%, \Csharp 5\% = 54\%, and over the past 30 years, C has always ranked either first or second in popularity.}
+In many ways, \CFA is to C as Scala~\cite{Scala} is to Java, providing a vehicle for new typing and control-flow capabilities on top of a highly popular programming language\footnote{
+The TIOBE index~\cite{TIOBE} for May 2020 ranks the top five \emph{popular} programming languages as C 17\%, Java 16\%, Python 9\%, \CC 6\%, and \Csharp 4\% = 52\%, 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.
+This paper discusses the design philosophy and implementation of \CFA's advanced control-flow and concurrent/parallel features, along with the supporting runtime written in \CFA.
 % 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{
+The \CFA control-flow framework extends ISO \Celeven~\cite{C11} with new call/return and concurrent/parallel control-flow.
+Call/return control-flow with argument and parameter passing appeared in the first programming languages.
+Over the past 50 years, call/return has been augmented with features like static and dynamic call, exceptions (multi-level return) and generators/coroutines (see Section~\ref{s:StatefulFunction}).
+While \CFA has mechanisms for dynamic call (algebraic effects~\cite{Zhang19}) 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.
+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 and 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}.
+If coroutines transfer to an internal event-engine for scheduling the next coroutines (as in async-await), 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).
+\Celeven and \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 and join in a function and mutex or condition locks, which is low-level and error-prone}
+Interestingly, almost a decade after the \Celeven standard, the most recent versions of gcc, clang, and msvc do not 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.
 …
 Kernel threading was chosen, largely because of its simplicity and fit with the simpler operating systems and hardware architectures at the time, which gave it a performance advantage~\cite{Drepper03}.
 Libraries like pthreads were developed for C, and the Solaris operating-system switched from user (JDK 1.1~\cite{JDK1.1}) to kernel threads.
 As a result, many current languages implementations adopt the 1:1 kernel-threading model, like Java (Scala), Objective-C~\cite{obj-c-book}, \CCeleven~\cite{C11}, C\#~\cite{Csharp} and Rust~\cite{Rust}, with a variety of presentation mechanisms.
+As a result, many languages 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}.
+The main argument for user-level threading is that it is lighter weight than kernel threading because 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.
+Finally, performant user-threading implementations, both in time and space, meet or exceed direct kernel-threading implementations, while achieving the programming advantages of high concurrency levels and safety.
 A further effort over the past two decades is the development of language memory models to deal with the conflict between language features and compiler/hardware optimizations, \eg some language features are unsafe in the presence of aggressive sequential optimizations~\cite{Buhr95a,Boehm05}.
 The consequence is that a language must provide sufficient tools to program around safety issues, as inline and library code is all sequential to the compiler.
 One solution is low-level qualifiers and functions (\eg @volatile@ and atomics) allowing \emph{programmers} to explicitly write safe (race-free~\cite{Boehm12}) programs.
 A safer solution is high-level language constructs so the \emph{compiler} knows the concurrency boundaries (where mutual exclusion and synchronization are acquired/released) and provide implicit safety at and across these boundaries.
+The consequence is that a language must provide sufficient tools to program around safety issues, as inline and library code is compiled as sequential without any explicit concurrent directive.
+One solution is low-level qualifiers and functions, \eg @volatile@ and atomics, allowing \emph{programmers} to explicitly write safe, race-free~\cite{Boehm12} programs.
+A safer solution is high-level language constructs so the \emph{compiler} knows the concurrency boundaries, \ie where mutual exclusion and synchronization are acquired and 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{
+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} or \emph{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.
+} or signals-as-hints~\cite[\S~8]{Buhr05a}, where one is a consequence of the other, \ie once there is spurious wakeup, barging follows.
 (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};
 …
 \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,
+monitor synchronization without barging, and the ability to safely acquiring multiple monitors in a deadlock-free way, while seamlessly integrating these capabilities with all monitor synchronization mechanisms,
 \item
 …
 \item
 a dynamic partitioning mechanism to segregate groups of executing user and kernel threads performing specialized work (\eg web-server or compute engine) or requiring different scheduling (\eg NUMA or real-time).
+a dynamic partitioning mechanism to segregate 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
 …
 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:MutualExclusionSynchronization} discusses the two mechanisms to restricted nondeterminism when controlling shared access to resources, called mutual exclusion, and timing relationships among threads, called synchronization.
 Section~\ref{s:Monitor} shows how both mutual exclusion and synchronization are safely embedded in the @monitor@ and @thread@ constructs.
 Section~\ref{s:CFARuntimeStructure} describes the large-scale mechanism to structure (cluster) threads and virtual processors (kernel threads).
 Section~\ref{s:Performance} uses a series of microbenchmarks to compare \CFA threading with pthreads, Java 11.0.6, Go 1.12.6, Rust 1.37.0, Python 3.7.6, Node.js 12.14.1, and \uC 7.0.0.
+Section~\ref{s:CFARuntimeStructure} describes the large-scale mechanism to structure threads and virtual processors (kernel threads).
+Section~\ref{s:Performance} uses 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 v12.18.0, and \uC 7.0.0.
 …
 \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.
+The features in a programming language should be composed of 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.
+The fundamental properties are execution state, thread, and mutual-exclusion/synchronization.
+These independent properties can be used to compose different language features, forming a compositional hierarchy, where the combination of all three is the most advanced feature, called a thread.
+While it is possible for a language to only provide threads for composing programs~\cite{Hermes90}, this unnecessarily complicates and makes inefficient solutions to certain classes of problems.
+As is shown, each of the non-rejected composed language 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 fundamental properties resulting in a complex and/or 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.
+is the state information needed by a control-flow feature to initialize and manage both compute data and execution location(s), and de-initialize.
+For example, calling a function initializes a stack frame including contained objects with constructors, manages local data in blocks and return locations during calls, and de-initializes the frame by running any object destructors and management operations.
+State is retained in fixed-sized aggregate structures (objects) and dynamic-sized stack(s), often allocated in the heap(s) managed by the runtime system.
+The lifetime of 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 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.
+is execution of code that occurs independently of other execution, where an individual thread's execution 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.
+concurrent execution becomes parallel when run on multiple processing units, \eg hyper-threading, cores, or sockets.
+A programmer needs mechanisms to create, block and unblock, and join with a thread, even if these basic mechanisms are supplied indirectly through high-level features.
+\item[\newterm{mutual-exclusion / synchronization (MES)}:]
+is the concurrency mechanism to perform an action without interruption and establish timing relationships among multiple threads.
+We contented these two properties are independent, \ie mutual exclusion cannot provide synchronization and vice versa without introducing additional threads~\cite[\S~4]{Buhr05a}.
+Limiting MES functionality 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.
+These properties are fundamental as 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 without (atomic) hardware mechanisms.
+\subsection{Structuring Execution Properties}
+Programming languages seldom present the fundamental execution properties directly to programmers.
+Instead, the properties are packaged into higher-level constructs that encapsulate details and provide safety to these low-level mechanisms.
+Interestingly, language designers often pick and choose among these execution properties proving a varying subset of constructs.
+Table~\ref{t:ExecutionPropertyComposition} shows all combinations of the three fundamental execution properties available to language designers.
+(When doing combination case-analysis, not all combinations are meaningful.)
+The combinations of state, thread, and MES compose a hierarchy of control-flow features all of which have appeared in prior programming languages, where each of these languages have found the feature useful.
+To understand the table, it is important to review the basic von Neumann execution requirement of at least one thread and 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.
+Each entry in the table, numbered \textbf{1}--\textbf{12}, is discussed with respect to how the execution properties combine to generate a high-level language construct.
 \begin{table}
 …
 \renewcommand{\arraystretch}{1.25}
 %\setlength{\tabcolsep}{5pt}
+\vspace*{-5pt}
 \begin{tabular}{c|c||l|l}
 \multicolumn{2}{c||}{execution properties} & \multicolumn{2}{c}{mutual exclusion / synchronization} \\
 …
 \hline
 \hline
 No                                      & No            & \textbf{1}\ \ \ function                              & \textbf{2}\ \ \ @monitor@ function    \\
+No                                      & No            & \textbf{1}\ \ \ @struct@                              & \textbf{2}\ \ \ @mutex@ @struct@              \\
 \hline
 Yes (stackless)         & No            & \textbf{3}\ \ \ @generator@                   & \textbf{4}\ \ \ @monitor@ @generator@ \\
+Yes (stackless)         & No            & \textbf{3}\ \ \ @generator@                   & \textbf{4}\ \ \ @mutex@ @generator@   \\
 \hline
 Yes (stackful)          & No            & \textbf{5}\ \ \ @coroutine@                   & \textbf{6}\ \ \ @monitor@ @coroutine@ \\
+Yes (stackful)          & No            & \textbf{5}\ \ \ @coroutine@                   & \textbf{6}\ \ \ @mutex@ @coroutine@   \\
 \hline
 No                                      & Yes           & \textbf{7}\ \ \ {\color{red}rejected} & \textbf{8}\ \ \ {\color{red}rejected} \\
 …
 Yes (stackless)         & Yes           & \textbf{9}\ \ \ {\color{red}rejected} & \textbf{10}\ \ \ {\color{red}rejected} \\
 \hline
 Yes (stackful)          & Yes           & \textbf{11}\ \ \ @thread@                             & \textbf{12}\ \ @monitor@ @thread@             \\
+Yes (stackful)          & Yes           & \textbf{11}\ \ \ @thread@                             & \textbf{12}\ \ @mutex@ @thread@               \\
 \end{tabular}
+\vspace*{-8pt}
 \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.
+Case 1 is a structure where access functions borrow local state (stack frame/activation) and thread from the invoker and retain this state across \emph{callees}, \ie function local-variables are retained on the borrowed stack during calls.
+Structures are a foundational mechanism for data organization, and access functions provide interface abstraction and code sharing in all programming languages.
+Case 2 is case 1 with thread safety to a structure's state where access functions provide serialization (mutual exclusion) and scheduling among calling threads (synchronization).
+A @mutex@ structure, often called a \newterm{monitor}, provides a high-level interface for race-free access of shared data in concurrent programming-languages.
+Case 3 is case 1 where the structure can implicitly retain execution state and access functions use this execution state to resume/suspend across \emph{callers}, but resume/suspend does not retain a function's local state.
+A stackless structure, often called a \newterm{generator} or \emph{iterator}, is \newterm{stackless} because it still borrows the caller's stack and thread, but the stack is used only to preserve state across its callees not callers.
+Generators provide the first step toward directly solving problems like finite-state machines that retain data and execution state between calls, whereas normal functions restart on each call.
+Case 4 is cases 2 and 3 with thread safety during execution of the generator's access functions.
+A @mutex@ generator extends generators into the concurrent domain.
+Cases 5 and 6 are like cases 3 and 4 where the structure is extended with an implicit separate stack, so only the thread is borrowed by access functions.
+A stackful generator, often called a \newterm{coroutine}, is \newterm{stackful} because resume/suspend now context switch to/from the caller's and coroutine's stack.
+A coroutine extends the state retained between calls beyond the generator's structure to arbitrary call depth in the access functions.
+Cases 7, 8, 9 and 10 are rejected because a new thread must have its own stack, where the thread begins and stack frames are stored for calls, \ie it is unrealistic for a thread to borrow a stack.
+For cases 9 and 10, the stackless frame is not growable, precluding accepting nested calls, making calls, blocking as it requires calls, or preemption as it requires pushing an interrupt frame, all of which compound to require an unknown amount of execution state.
+Hence, if this kind of uninterruptable thread exists, it must execute to completion, \ie computation only, which severely restricts runtime management.
+Cases 11 and 12 are a stackful thread with and without safe access to shared state.
+A thread is the language mechanism to start another thread of control in a program with growable execution state for call/return execution.
+In general, language constructs with more execution properties increase the cost of creation and execution along with complexity of usage.
+Given the execution-properties taxonomy, programmers now ask three basic questions: is state necessary across callers and how much, is a separate thread necessary, is thread safety necessary.
+Table~\ref{t:ExecutionPropertyComposition} then suggests the optimal language feature needed for implementing a programming problem.
+The following sections describe how \CFA fills in \emph{all} the non-rejected table entries with language features, while other programming languages may only provide a subset of the table.
 …
 \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.
+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 requires additional synchronization and possibly data transfer.
 \item
 …
 \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.
+MES must be available implicitly in language constructs, \eg Java built-in monitors, as well as explicitly for specialized requirements, \eg @java.util.concurrent@, 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
 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}.
+Otherwise, certain concurrency problems are difficult, \eg 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.
 …
 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 caller detects the action's completion through a \newterm{future} or \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.
 …
 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).
+While asynchronous calls generate new callee (server) events, we contend 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.
+The problem is when concurrent work-units need to interact and/or block as this effects the executor by stopping 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.
 …
 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.
+Selecting between stackless/stackful semantics and asymmetric/symmetric style is a tradeoff between programming requirements, performance, and design, where stackless is faster and smaller using modified call/return between closures, stackful is more general but slower and larger using context switching between distinct stacks, and asymmetric is simpler control-flow than symmetric.
+Additionally, storage management for the closure/stack must be factored into design and performance, especially in unmanaged languages without garbage collection.
 Note, creation cost (closure/stack) is amortized across usage, so activation cost (resume/suspend) is usually the dominant factor.
 …
+&
 \begin{cfa}
 void * rtn( void * arg ) { ... }
+void * `rtn`( void * arg ) { ... }
 int i = 3, rc;
 pthread_t t; $\C{// thread id}$
 …
 \hspace{3pt}
 \subfloat[C generated code for \CFA version]{\label{f:CFibonacciSim}\usebox\myboxC}
 \caption{Fibonacci (output) asymmetric generator}
+\caption{Fibonacci output asymmetric generator}
 \label{f:FibonacciAsymmetricGenerator}
 …
 \subfloat[C generated code for \CFA version]{\label{f:CFormatGenImpl}\usebox\myboxB}
 \hspace{3pt}
 \caption{Formatter (input) asymmetric generator}
+\caption{Formatter input asymmetric generator}
 \label{f:FormatterAsymmetricGenerator}
 \end{figure}
 Figure~\ref{f:FibonacciAsymmetricGenerator} shows an unbounded asymmetric generator for an infinite sequence of Fibonacci numbers written (left to right) in C, \CFA, and showing the underlying C implementation for the \CFA version.
+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.
 An additional requirement is the ability to create an arbitrary number of generators (of any kind), \ie retaining one state in global variables is insufficient;
+An additional requirement is the ability to create an arbitrary number of generators of any kind, \ie retaining one state in global variables is insufficient;
 hence, state is retained in a closure between calls.
 Figure~\ref{f:CFibonacci} shows the C approach of manually creating the closure in structure @Fib@, and multiple instances of this closure provide multiple Fibonacci generators.
 …
 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}).
+At the start of the generator main, the @static@ declaration, @states@, is initialized to the N suspend points in the generator, where operator @&&@ dereferences or 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.
+The  cost of setting @restart@ and branching via the computed @goto@ adds very little cost to the suspend and 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.
 …
 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.
+As well, C programmers are not afraid of this kind of semantic programming requirement, if it results in very small and fast generators.
 Figure~\ref{f:CFAFormatGen} shows an asymmetric \newterm{input generator}, @Fmt@, for restructuring text into groups of characters of fixed-size blocks, \ie the input on the left is reformatted into the output on the right, where newlines are ignored.
 …
 The destructor provides a newline, if formatted text ends with a full line.
 Figure~\ref{f:CFormatGenImpl} shows the C implementation of the \CFA input generator with one additional field and the computed @goto@.
 For contrast, Figure~\ref{f:PythonFormatter} shows the equivalent Python format generator with the same properties as the format generator.
+For contrast, Figure~\ref{f:PythonFormatter} shows the equivalent Python format generator with the same properties as the \CFA 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}
+An important application for the asymmetric generator is 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.
 …
 However, Adya \etal~\cite{Adya02} argue against stack ripping in Section 3.2 and suggest a hybrid approach in Section 4 using cooperatively scheduled \emph{fibers}, which is coroutining.
 As an example, the following protocol:
+Figure~\ref{f:DeviceDriverGen} shows the generator advantages in implementing a simple network device-driver with the following protocol:
 \begin{center}
 \ldots\, STX \ldots\, message \ldots\, ESC ETX \ldots\, message \ldots\, ETX 2-byte crc \ldots
 \end{center}
 is for a simple network message beginning with the control character STX, ending with an ETX, and followed by a 2-byte cyclic-redundancy check.
+where the network message begins with the control character STX, ends with an ETX, and is followed by a 2-byte cyclic-redundancy check.
 Control characters may appear in a message if preceded by an ESC.
 When a message byte arrives, it triggers an interrupt, and the operating system services the interrupt by calling the device driver with the byte read from a hardware register.
 The device driver returns a status code of its current state, and when a complete message is obtained, the operating system read the message accumulated in the supplied buffer.
 Hence, the device driver is an input/output generator, where the cost of resuming the device-driver generator is the same as call/return, so performance in an operating-system kernel is excellent.
+The device driver returns a status code of its current state, and when a complete message is obtained, the operating system reads 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 and return, so performance in an operating-system kernel is excellent.
 The key benefits of using a generator are correctness, safety, and maintenance because the execution states are transcribed directly into the programming language rather than table lookup or stack ripping.
 The conclusion is that FSMs are complex and occur in important domains, so direct generator support is important in a system programming language.
+% The conclusion is that FSMs are complex and occur in important domains, so direct generator support is important in a system programming language.
 \begin{figure}
 …
 \end{figure}
 Figure~\ref{f:CFAPingPongGen} shows a symmetric generator, where the generator resumes another generator, forming a resume/resume cycle.
+Generators can also have symmetric activation using resume/resume to create control-flow cycles among generators.
 (The trivial cycle is a generator resuming itself.)
 This control flow is similar to recursion for functions but without stack growth.
+Figure~\ref{f:PingPongFullCoroutineSteps} shows the steps for symmetric control-flow are creating, executing, and terminating the cycle.
+Figure~\ref{f:PingPongFullCoroutineSteps} shows the steps for symmetric control-flow using for the ping/pong program in Figure~\ref{f:CFAPingPongGen}.
+The program starts by creating the generators, @ping@ and @pong@, and then assigns the partners that form the cycle.
 Constructing the cycle must deal with definition-before-use to close the cycle, \ie, the first generator must know about the last generator, which is not within scope.
 (This issue occurs for any cyclic data structure.)
-The example creates the generators, @ping@/@pong@, and then assigns the partners that form the cycle.
 % (Alternatively, the constructor can assign the partners as they are declared, except the first, and the first-generator partner is set after the last generator declaration to close the cycle.)
 Once the cycle is formed, the program main resumes one of the generators, @ping@, and the generators can then traverse an arbitrary cycle using @resume@ to activate partner generator(s).
+Once the cycle is formed, the program main resumes one of the generators, @ping@, and the generators can then traverse an arbitrary number of cycles 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.
 …
 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}
+\centering
+\input{FullCoroutinePhases.pstex_t}
+\vspace*{-10pt}
+\caption{Symmetric coroutine steps: Ping / Pong}
+\label{f:PingPongFullCoroutineSteps}
+\end{figure}
 \begin{figure}
 …
 \end{figure}
-\begin{figure}
-\centering
-\input{FullCoroutinePhases.pstex_t}
-\vspace*{-10pt}
-\caption{Symmetric coroutine steps: Ping / Pong}
-\label{f:PingPongFullCoroutineSteps}
-\end{figure}
 Figure~\ref{f:CPingPongSim} shows the C implementation of the \CFA symmetric generator, where there is still only one additional field, @restart@, but @resume@ is more complex because it does a forward rather than backward jump.
 Before the jump, the parameter for the next call @partner@ is placed into the register used for the first parameter, @rdi@, and the remaining registers are reset for a return.
 …
 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.
+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.
 …
 \label{s:Coroutine}
+Stackful coroutines (Table~\ref{t:ExecutionPropertyComposition} case 5) extend generator semantics, \ie there is an implicit closure and @suspend@ may appear in a helper function called from the coroutine main.
+A coroutine is specified by replacing @generator@ with @coroutine@ for the type.
+Stackful coroutines (Table~\ref{t:ExecutionPropertyComposition} case 5) extend generator semantics with an implicit closure and @suspend@ may appear in a helper function called from the coroutine main because of the separate stack.
+Note, simulating coroutines with stacks of generators, \eg Python with @yield from@ cannot handle symmetric control-flow.
+Furthermore, all stack components must be of generators, so it is impossible to call a library function passing a generator that yields.
+Creating a generator copy of the library function maybe impossible because the library function is opaque.
+A \CFA coroutine is specified by replacing @generator@ with @coroutine@ for the type.
 Coroutine generality results in higher cost for creation, due to dynamic stack allocation, for execution, due to context switching among stacks, and for terminating, due to possible stack unwinding and dynamic stack deallocation.
 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.
+Now the coroutine type only contains communication variables between interface functions and the coroutine main.
 \begin{center}
 \begin{tabular}{@{}l|l|l|l@{}}
 …
 \begin{cfa}
 int Crc() {
+        `suspend;`
+        short int crc = byte << 8;
+        `suspend;`
+        status = (crc | byte) == sum ? MSG : ECRC;
+        `suspend;`  short int crc = byte << 8;
+        `suspend;`  status = (crc | byte) == sum ? MSG : ECRC;
         return crc;
+}
 \end{cfa}
 A call to this function is placed at the end of the driver's coroutine-main.
+A call to this function is placed at the end of the device driver's coroutine-main.
 For complex finite-state machines, refactoring is part of normal program abstraction, especially when code is used in multiple places.
 Again, this complexity is usually associated with execution state rather than data state.
 …
 \begin{comment}
 Figure~\ref{f:Coroutine3States} creates a @coroutine@ type, @`coroutine` Fib { int fn; }@, which provides communication, @fn@, for the \newterm{coroutine main}, @main@, which runs on the coroutine stack, and possibly multiple interface functions, \eg @restart@.
 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.
+Like the structure in Figure~\ref{f:ExternalState}, the coroutine type allows multiple instances, where instances of this type are passed to the overloaded coroutine main.
 The coroutine main's stack holds the state for the next generation, @f1@ and @f2@, and the code represents the three states in the Fibonacci formula via the three suspend points, to context switch back to the caller's @resume@.
 The interface function @restart@, takes a Fibonacci instance and context switches to it using @resume@;
 …
 Figure~\ref{f:ProdCons} shows the ping-pong example in Figure~\ref{f:CFAPingPongGen} extended into a producer/consumer symmetric-coroutine performing bidirectional communication.
 This example is illustrative because both producer/consumer have two interface functions with @resume@s that suspend execution in these interface (helper) functions.
+This example is illustrative because both producer and consumer have two interface functions with @resume@s that suspend execution in these interface functions.
 The program main creates the producer coroutine, passes it to the consumer coroutine in its initialization, and closes the cycle at the call to @start@ along with the number of items to be produced.
 The call to @start@ is the first @resume@ of @prod@, which remembers the program main as the starter and creates @prod@'s stack with a frame for @prod@'s coroutine main at the top, and context switches to it.
 @prod@'s coroutine main starts, creates local-state variables that are retained between coroutine activations, and executes $N$ iterations, each generating two random values, calling the consumer's @deliver@ function to transfer the values, and printing the status returned from the consumer.
 The producer call to @delivery@ transfers values into the consumer's communication variables, resumes the consumer, and returns the consumer status.
+The producer's call to @delivery@ transfers values into the consumer's communication variables, resumes the consumer, and returns the consumer status.
 Similarly on the first resume, @cons@'s stack is created and initialized, holding local-state variables retained between subsequent activations of the coroutine.
 The symmetric coroutine cycle forms when the consumer calls the producer's @payment@ function, which resumes the producer in the consumer's delivery function.
 When the producer calls @delivery@ again, it resumes the consumer in the @payment@ function.
 Both interface function than return to the their corresponding coroutine-main functions for the next cycle.
+Both interface functions then return to 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.
 …
 Terminating a coroutine cycle is more complex than a generator cycle, because it requires context switching to the program main's \emph{stack} to shutdown the program, whereas generators started by the program main run on its stack.
 Furthermore, each deallocated coroutine must execute all destructors for object allocated in the coroutine type \emph{and} allocated on the coroutine's stack at the point of suspension, which can be arbitrarily deep.
+Furthermore, each deallocated coroutine must execute all destructors for objects 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.)
 …
 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.
+The na\"{i}ve semantics for coroutine-cycle termination is to context switch to the last resumer, like executing a @suspend@ or @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).
 …
 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.
+For other scenarios, it is always possible to devise a solution with additional programming effort, such as forcing the cycle forward or backward to a safe point before starting termination.
 Note, the producer/consumer example does not illustrate the full power of the starter semantics because @cons@ always ends first.
 Assume generator @PingPong@ in Figure~\ref{f:PingPongSymmetricGenerator} is converted to a coroutine.
 Unlike generators, coroutines have a starter structure with multiple levels, where the program main starts @ping@ and @ping@ starts @pong@.
 By adjusting $N$ for either @ping@/@pong@, it is possible to have either finish first.
+By adjusting $N$ for either @ping@ or @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.
+Once back at the program main (creator), coroutines @ping@ and @pong@ are deallocated, running 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 semantics, regardless of the coroutine's state (terminated or active), like any other aggregate object.
 \subsection{Generator / Coroutine Implementation}
 A significant implementation challenge for generators/coroutines (and threads in Section~\ref{s:threads}) is adding extra fields to the custom types and related functions, \eg inserting code after/before the coroutine constructor/destructor and @main@ to create/initialize/de-initialize/destroy any extra fields, \eg stack.
 There are several solutions to these problem, which follow from the object-oriented flavour of adopting custom types.
+A significant implementation challenge for generators and coroutines (and threads in Section~\ref{s:threads}) is adding extra fields to the custom types and related functions, \eg inserting code after/before the coroutine constructor/destructor and @main@ to create/initialize/de-initialize/destroy any extra fields, \eg the coroutine stack.
+There are several solutions to this problem, which follow from the object-oriented flavour of adopting custom types.
 For object-oriented languages, inheritance is used to provide extra fields and code via explicit inheritance:
 …
 \end{cfa}
 % The problem is that the programming language and its tool chain, \eg debugger, @valgrind@, need to understand @baseCoroutine@ because it infers special property, so type @baseCoroutine@ becomes a de facto keyword and all types inheriting from it are implicitly custom types.
 The problem is that some special properties are not handled by existing language semantics, \eg the execution of constructors/destructors is in the wrong order to implicitly start threads because the thread must start \emph{after} all constructors as it relies on a completely initialized object, but the inherited constructor runs \emph{before} the derived.
+The problem is that some special properties are not handled by existing language semantics, \eg the execution of constructors and destructors is in the wrong order to implicitly start threads because the thread must start \emph{after} all constructors as it relies on a completely initialized object, but the inherited constructor runs \emph{before} the derived.
 Alternatives, such as explicitly starting threads as in Java, are repetitive and forgetting to call start is a common source of errors.
 An alternative is composition:
 …
 forall( `dtype` T | is_coroutine(T) ) void $suspend$( T & ), resume( T & );
 \end{cfa}
 Note, copying generators/coroutines/threads is undefined because muliple objects cannot execute on a shared stack and stack copying does not work in unmanaged languages (no garbage collection), like C, because the stack may contain pointers to objects within it that require updating for the copy.
 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 @main@ function has no return value or additional parameters because the coroutine type allows an arbitrary number of interface functions with corresponding arbitrary typed input/output values versus fixed ones.
+Note, copying generators, coroutines, and threads is undefined because multiple 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 or 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 read the coroutine descriptor from its handle.
+The @main@ function has no return value or additional parameters because the coroutine type allows an arbitrary number of interface functions with arbitrary typed input and output values versus fixed ones.
 The advantage of this approach is that users can easily create different types of coroutines, \eg changing the memory layout of a coroutine is trivial when implementing the @get_coroutine@ function, and possibly redefining \textsf{suspend} and @resume@.
 …
 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 handle is the @coroutine@ instance containing programmer specified type global and communication variables across interface functions.
 The coroutine descriptor contains all implicit declarations needed by the runtime, \eg @suspend@/@resume@, and can be part of the coroutine handle or separate.
 The coroutine stack can appear in a number of locations and be fixed or variable sized.
+Hence, the coroutine's stack could be a variable-length structure (VLS)\footnote{
+We are examining VLSs, where fields can be variable-sized structures or arrays.
+Once allocated, a VLS is fixed sized.}
+Hence, the coroutine's stack could be a variable-length structure (VLS)
+% \footnote{
+% We are examining VLSs, where fields can be variable-sized structures or arrays.
+% Once allocated, a VLS is fixed sized.}
 on the allocating stack, provided the allocating stack is large enough.
 For a VLS stack allocation/deallocation is an inexpensive adjustment of the stack pointer, modulo any stack constructor costs (\eg initial frame setup).
 For stack allocation in the heap, allocation/deallocation is an expensive allocation, where the heap can be a shared resource, modulo any stack constructor costs.
 It is also possible to use a split (segmented) stack calling convention, available with gcc and clang, allowing a variable-sized stack via a set of connected blocks in the heap.
 Currently, \CFA supports stack/heap allocated descriptors but only fixed-sized heap allocated stacks.
+For a VLS stack allocation and deallocation is an inexpensive adjustment of the stack pointer, modulo any stack constructor costs to initial frame setup.
+For stack allocation in the heap, allocation and 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 or 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 and 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.
 …
 The transition to concurrency, even for a single thread with multiple stacks, occurs when coroutines context switch to a \newterm{scheduling coroutine}, introducing non-determinism from the coroutine perspective~\cite[\S~3]{Buhr05a}.
 Therefore, a minimal concurrency system requires coroutines \emph{in conjunction with a nondeterministic scheduler}.
 The resulting execution system now follows a cooperative threading-model~\cite{Adya02,libdill} because context-switching points to the scheduler (blocking) are known, but the next unblocking point is unknown due to the scheduler.
+The resulting execution system now follows a cooperative threading-model~\cite{Adya02,libdill} because context-switching points to the scheduler 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.
+The reason is that the scheduler and runtime have complete knowledge about resources and how to best utilized them.
 However, the introduction of unrestricted nondeterminism results in the need for \newterm{mutual exclusion} and \newterm{synchronization}~\cite[\S~4]{Buhr05a}, which restrict nondeterminism for correctness;
 otherwise, it is impossible to write meaningful concurrent programs.
 …
 \label{s:threads}
+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@.
+Threading (Table~\ref{t:ExecutionPropertyComposition} case 11) needs the ability to start a thread and wait for its completion, where a common API is @fork@ and @join@.
 \vspace{4pt}
 \par\noindent
 …
 \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.
+\CFA has a simpler approach using a custom @thread@ type and leveraging declaration semantics, allocation and deallocation, where threads implicitly @fork@ after construction and @join@ before destruction.
 \begin{cfa}
 thread MyThread {};
 …
 } $\C{// deallocate stack-based threads, implicit joins before destruction}$
 \end{cfa}
 This semantic ensures a thread is started and stopped exactly once, eliminating some programming error, and scales to multiple threads for basic (termination) synchronization.
 For block allocation to arbitrary depth, including recursion, threads are created/destroyed in a lattice structure (tree with top and bottom).
+This semantic ensures a thread is started and stopped exactly once, eliminating some programming error, and scales to multiple threads for basic termination synchronization.
+For block allocation to arbitrary depth, including recursion, threads are created and destroyed in a lattice structure (tree with top and bottom).
 Arbitrary topologies are possible using dynamic allocation, allowing threads to outlive their declaration scope, identical to normal dynamic allocation.
 \begin{cfa}
 …
         MyThread * team = factory( 10 );
         // concurrency
         `delete( team );` $\C{// deallocate heap-based threads, implicit joins before destruction}\CRT$
+        `adelete( team );` $\C{// deallocate heap-based threads, implicit joins before destruction}\CRT$
+}
 \end{cfa}
 …
 \end{tabular}
 \end{cquote}
 Like coroutines, the @dtype@ property prevents \emph{implicit} copy operations and the @is_thread@ trait provides no \emph{explicit} copy operations, so threads must be passed by reference (pointer).
 Similarly, the function definitions ensure there is a statically typed @main@ function that is the thread starting point (first stack frame), a mechanism to get (read) the thread descriptor from its handle, and a special destructor to prevent deallocation while the thread is executing.
+Like coroutines, the @dtype@ property prevents \emph{implicit} copy operations and the @is_thread@ trait provides no \emph{explicit} copy operations, so threads must be passed by reference or pointer.
+Similarly, the function definitions ensure there is a statically typed @main@ function that is the thread starting point (first stack frame), a mechanism to read the thread descriptor from its handle, and a special destructor to prevent deallocation while the thread is executing.
 (The qualifier @mutex@ for the destructor parameter is discussed in Section~\ref{s:Monitor}.)
 The difference between the coroutine and thread is that a coroutine borrows a thread from its caller, so the first thread resuming a coroutine creates the coroutine's stack and starts running the coroutine main on the stack;
 whereas, a thread is scheduling for execution in @main@ immediately after its constructor is run.
 No return value or additional parameters are necessary for this function because the @thread@ type allows an arbitrary number of interface functions with corresponding arbitrary typed input/output values.
+No return value or additional parameters are necessary for this function because the @thread@ type allows an arbitrary number of interface functions with corresponding arbitrary typed input and output values.
 …
 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}.
 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.
+The shared data protected by mutual exclusion 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} problems.
+Without synchronization control in a critical section, an arriving thread can barge ahead of preexisting waiter threads resulting in short/long-term starvation, staleness and freshness problems, and 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}.
 …
 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.
+Hence, a programmer must learn and manipulate two sets of design and programming patterns.
 While this distinction can be hidden away in library code, effective use of the library still has to take both paradigms into account.
 In contrast, approaches based on shared-state models more closely resemble the standard call/return programming model, resulting in a single programming paradigm.
+In contrast, approaches based on shared-state models more closely resemble the standard call and 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.
 …
 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}
+`monitor` Aint { int cnt; }; $\C[4.25in]{// atomic integer counter}$
+int ++?( Aint & `mutex` this ) with( this ) { return ++cnt; } $\C{// increment}$
+int ?=?( Aint & `mutex` lhs, int rhs ) with( lhs ) { cnt = rhs; } $\C{// conversions with int, mutex optional}\CRT$
+int ?=?( int & lhs, Aint & `mutex` rhs ) with( rhs ) { lhs = cnt; }
+\end{cfa}
+The operators use the parameter-only declaration type-qualifier @mutex@ to mark which parameters require locking during function execution to protect from race conditions.
+The assignment operators provide bidirectional conversion between an atomic and normal integer without accessing field @cnt@.
+(These operations only need @mutex@, if reading/writing the implementation type is not atomic.)
+The atomic counter is used without any explicit mutual-exclusion and provides thread-safe semantics.
+\begin{cfa}
+Figure~\ref{f:AtomicCounter} compares a \CFA and Java monitor implementing an atomic counter.
+(Like other concurrent programming languages, \CFA and Java have performant specializations for the basic types using atomic instructions.)
+A \newterm{monitor} is a set of functions that ensure mutual exclusion when accessing shared state.
+(Note, in \CFA, @monitor@ is short-hand for @mutex struct@.)
+More precisely, a monitor is a programming technique that implicitly binds mutual exclusion to static function scope by call and 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 and release points eases programming, comprehension, and maintenance, at a slight cost in flexibility and efficiency.
+As for other special types, \CFA has a custom @monitor@ type.
+\begin{figure}
+\centering
+\begin{lrbox}{\myboxA}
+\begin{cfa}[aboveskip=0pt,belowskip=0pt]
+`monitor` Aint { // atomic integer counter
+        int cnt;
+};
+int ++?( Aint & `mutex` this ) with(this) { return ++cnt; }
+int ?=?( Aint & `mutex` lhs, int rhs ) with(lhs) { cnt = rhs; }
+int ?=?(int & lhs, Aint & rhs) with(rhs) { lhs = cnt; }
 int i = 0, j = 0, k = 5;
+Aint x = { 0 }, y = { 0 }, z = { 5 }; $\C{// no mutex required}$
+++x; ++y; ++z; $\C{// safe increment by multiple threads}$
+x = 2; y = i; z = k; $\C{// conversions}$
+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;
+void foo( M & mutex m ) { ... } $\C{// acquire mutual exclusion}$
+void bar( M & mutex m ) { $\C{// acquire mutual exclusion}$
+        ... `bar( m );` ... `foo( m );` ... $\C{// reacquire mutual exclusion}$
+}
+\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.
+Aint x = { 0 }, y = { 0 }, z = { 5 }; // no mutex
+++x; ++y; ++z;     // mutex
+x = 2; y = i; z = k;  // mutex
+i = x; j = y; k = z;  // no mutex
+\end{cfa}
+\end{lrbox}
+\begin{lrbox}{\myboxB}
+\begin{java}[aboveskip=0pt,belowskip=0pt]
+class Aint {
+    private int cnt;
+    public Aint( int init ) { cnt = init; }
+    `synchronized` public int inc() { return ++cnt; }
+    `synchronized` public void set( int rhs ) {cnt=rhs;}
+    public int get() { return cnt; }
+}
+int i = 0, j = 0, k = 5;
+Aint x=new Aint(0), y=new Aint(0), z=new Aint(5);
+x.inc(); y.inc(); z.inc();
+x.set( 2 ); y.set( i ); z.set( k );
+i = x.get(); j = y.get(); k = z.get();
+\end{java}
+\end{lrbox}
+\subfloat[\CFA]{\label{f:AtomicCounterCFA}\usebox\myboxA}
+\hspace{3pt}
+\vrule
+\hspace{3pt}
+\subfloat[Java]{\label{f:AtomicCounterJava}\usebox\myboxB}
+\caption{Atomic counter}
+\label{f:AtomicCounter}
+\end{figure}
+Like Java, \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.
+% \begin{cfa}
+% monitor M { ... } m;
+% void foo( M & mutex m ) { ... } $\C{// acquire mutual exclusion}$
+% void bar( M & mutex m ) { $\C{// acquire mutual exclusion}$
+%       ... `bar( m );` ... `foo( m );` ... $\C{// reacquire mutual exclusion}$
+% }
+% \end{cfa}
+\CFA monitors also ensure the monitor lock is released regardless of how an acquiring function ends, normal or exceptional, and returning a shared variable is safe via copying before the lock is released.
 Similar safety is offered by \emph{explicit} opt-in disciplines like \CC RAII versus the monitor \emph{implicit} language-enforced safety guarantee ensuring no programmer usage errors.
 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;
+However, 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}).
+Both Java and \CFA use a keyword @mutex@/\lstinline[language=java]|synchronized| to designate functions that implicitly acquire/release the monitor lock on call/return providing mutual exclusion to the stared data.
+Non-designated functions provide no mutual exclusion for read-only access or as an interface to a multi-step protocol requiring several steps of acquiring and releasing the monitor.
+Monitor objects can be passed through multiple helper functions without acquiring mutual exclusion, until a designated function associated with the object is called.
+\CFA designated functions are marked by an explicitly parameter-only pointer/reference qualifier @mutex@ (discussed further in Section\ref{s:MutexAcquisition}).
+Whereas, Java designated members are marked with \lstinline[language=java]|synchronized| that applies to the implicit reference parameter @this@.
+In the example, the increment and setter operations need mutual exclusion while the read-only getter operation can be non-mutex if reading the implementation is atomic.
 …
 \end{tabular}
 \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).
 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 @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 or pointer.
+Similarly, the function definitions ensure there is a mechanism to read the monitor descriptor from its handle, and a special destructor to prevent deallocation if a thread is using the shared data.
 The custom monitor type also inserts any locks needed to implement the mutual exclusion semantics.
+\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.
 …
 \label{s:MutexAcquisition}
+While the monitor lock provides mutual exclusion for shared data, there are implementation options for when and where the locking/unlocking occurs.
+(Much of this discussion also applies to basic locks.)
+For example, a monitor may be passed through multiple helper functions before it is necessary to acquire the monitor's mutual exclusion.
+\CFA requires programmers to identify the kind of parameter with the @mutex@ keyword and uses no keyword to mean \lstinline[morekeywords=nomutex]@nomutex@, because @mutex@ parameters are rare and no keyword is the \emph{normal} parameter semantics.
+Hence, @mutex@ parameters are documentation, at the function and its prototype, to both programmer and compiler, without other redundant keywords.
+Furthermore, \CFA relies heavily on traits as an abstraction mechanism, so the @mutex@ qualifier prevents coincidentally matching of a monitor trait with a type that is not a monitor, similar to coincidental inheritance where a shape and playing card can both be drawable.
+The next semantic decision is establishing which parameter \emph{types} may be qualified with @mutex@.
+The following has monitor parameter types that are composed of multiple objects.
+\begin{cfa}
+monitor M { ... }
+For object-oriented programming languages, the mutex property applies to one object, the implicit pointer/reference to the monitor type.
+Because \CFA uses a pointer qualifier, other possibilities exist, \eg:
+\begin{cfa}
+monitor M { ... };
 int f1( M & mutex m ); $\C{// single parameter object}$
 int f2( M * mutex m ); $\C{// single or multiple parameter object}$
 …
 int f4( stack( M * ) & mutex m ); $\C{// multiple parameters object}$
 \end{cfa}
+Function @f1@ has a single parameter object, while @f2@'s indirection could be a single or multi-element array, where static array size is often unknown in C.
+Function @f3@ has a multiple object matrix, and @f4@ a multiple object data structure.
+While shown shortly, multiple object acquisition is possible, but the number of objects must be statically known.
+Therefore, \CFA only acquires one monitor per parameter with exactly one level of indirection, and exclude pointer types to unknown sized arrays.
+For object-oriented monitors, \eg Java, calling a mutex member \emph{implicitly} acquires mutual exclusion of the receiver object, @`rec`.foo(...)@.
+\CFA has no receiver, and hence, the explicit @mutex@ qualifier is used to specify which objects acquire mutual exclusion.
+A positive consequence of this design decision is the ability to support multi-monitor functions,\footnote{
+Function @f1@ has a single object parameter, while functions @f2@ to @f4@ can be a single or multi-element parameter with statically unknown size.
+Because of the statically unknown size, \CFA only supports a single reference @mutex@ parameter, @f1@.
+The \CFA @mutex@ qualifier does allow the ability to support multi-monitor functions,\footnote{
 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}.
+where the number of acquisitions is statically known, called \newterm{bulk acquire}.
 \CFA guarantees bulk acquisition order is consistent across calls to @mutex@ functions using the same monitors as arguments, so acquiring multiple monitors in a bulk acquire is safe from deadlock.
 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.
 …
 % There are many aspects of scheduling in a concurrency system, all related to resource utilization by waiting threads, \ie which thread gets the resource next.
 % Different forms of scheduling include access to processors by threads (see Section~\ref{s:RuntimeStructureCluster}), another is access to a shared resource by a lock or monitor.
 This section discusses scheduling for waiting threads eligible for monitor entry, \ie which user thread gets the shared resource next. (See Section~\ref{s:RuntimeStructureCluster} for scheduling kernel threads on virtual processors.)
+This section discusses scheduling for waiting threads eligible for monitor entry~\cite{Buhr95b}, \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} (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.
+\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 versa, 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 are 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}.
 …
 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).
+Monitor mutual-exclusion means signalling cannot have the signaller and signalled thread in the monitor simultaneously, so only the signaller or signallee can proceed and the other waits on an implicit urgent list~\cite[p.~551]{Hoare74}.
+Figure~\ref{f:MonitorScheduling} shows internal and 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 the urgent list (see Figure~\ref{f:MonitorScheduling}) 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.
+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.)
 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.
+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 consuming 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.
+In \CFA, a condition queue can be created and stored independently.
 \begin{figure}
 …
 \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.
+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)~\cite[p.~551]{Hoare74}.
+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++}.
+For external scheduling in Figure~\ref{f:BBExt}, the internal scheduling is replaced, 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.
 …
 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@.
+Hence, this mechanism is done in terms of control flow, next call, versus in terms of data, channels, as in Go and 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.
+Figure~\ref{f:ReadersWriterLock} shows internal and external scheduling for a readers/writer lock with no barging and threads are serviced in FIFO order to eliminate staleness and 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@.
 …
 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.
+Note, 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.
+Given external and internal scheduling, what guidelines can a programmer use to select between them?
+In general, external scheduling is easier to understand and code because only the next logical action (mutex function(s)) is stated, and the monitor implicitly handles all the details.
+Therefore, there are no condition variables, and hence, no wait and signal, which reduces coding complexity and synchronization errors.
+If external scheduling is simpler than internal, why not use it all the time?
+Unfortunately, external scheduling cannot be used if: scheduling depends on parameter value(s) or scheduling must block across an unknown series of calls on a condition variable, \ie internal scheduling.
+For example, the dating service cannot be written using external scheduling.
+First, scheduling requires knowledge of calling parameters to make matching decisions and parameters of calling threads are unavailable within the monitor.
+Specifically, a thread within the monitor cannot examine the @ccode@ of threads waiting on the calling queue to determine if there is a matching partner.
+(Similarly, if the bounded buffer or readers/writer are restructured with a single interface function with a parameter denoting producer/consumer or reader/write, they cannot be solved with external scheduling.)
+Second, a scheduling decision may be delayed across an unknown number of calls when there is no immediate match so the thread in the monitor must block on a condition.
+Specifically, if a thread determines there is no opposite calling thread with the same @ccode@, it must wait an unknown period until a matching thread arrives.
+For complex synchronization, both external and internal scheduling can be used to take advantage of best of properties of each.
+Finally, both internal and external scheduling extend to multiple monitors in a natural way.
 \begin{cquote}
 \begin{tabular}{@{}l@{\hspace{2\parindentlnth}}l@{}}
 …
 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.
+@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 prototype must be accessible.
 % When an overloaded function appears in an @waitfor@ statement, calls to any function with that name are accepted.
 % The rationale is that members with the same name should perform a similar function, and therefore, all should be eligible to accept a call.
+% The rationale is that functions with the same name should perform a similar actions, and therefore, all should be eligible to accept a call.
 Overloaded functions can be disambiguated using a cast
 \begin{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 ) {
 …
 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.
 For a @waitfor@ clause to be executed, its @when@ must be true and an outstanding call to its corresponding member(s) must exist.
+For a @waitfor@ clause to be executed, its @when@ must be true and an outstanding call to its corresponding function(s) must exist.
 The \emph{conditional-expression} of a @when@ may call a function, but the function must not block or context switch.
 If there are multiple acceptable mutex calls, selection occurs top-to-bottom (prioritized) among the @waitfor@ clauses, whereas some programming languages with similar mechanisms accept nondeterministically for this case, \eg Go \lstinline[morekeywords=select]@select@.
 If some accept guards are true and there are no outstanding calls to these members, the acceptor is blocked until a call to one of these members is made.
+If there are multiple acceptable mutex calls, selection is prioritized top-to-bottom among the @waitfor@ clauses, whereas some programming languages with similar mechanisms accept nondeterministically for this case, \eg Go \lstinline[morekeywords=select]@select@.
+If some accept guards are true and there are no outstanding calls to these functions, the acceptor is blocked until a call to one of these functions is made.
 If there is a @timeout@ clause, it provides an upper bound on waiting.
 If all the accept guards are false, the statement does nothing, unless there is a terminating @else@ clause with a true guard, which is executed instead.
 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{cfa}
 `when` ( $\emph{conditional-expression}$ )      $\C{// optional guard}$
         waitfor( $\emph{mutex-member-name}$ ) $\emph{statement}$ $\C{// action after call}$
+        waitfor( $\emph{mutex-function-name}$ ) $\emph{statement}$ $\C{// action after call}$
 `or` `when` ( $\emph{conditional-expression}$ ) $\C{// any number of functions}$
         waitfor( $\emph{mutex-member-name}$ ) $\emph{statement}$
+        waitfor( $\emph{mutex-function-name}$ ) $\emph{statement}$
 `or`    ...
 `when` ( $\emph{conditional-expression}$ ) $\C{// optional guard}$
 …
 The left example only accepts @mem1@ if @C1@ is true or only @mem2@ if @C2@ is true.
 The right example accepts either @mem1@ or @mem2@ if @C1@ and @C2@ are true.
+Hence, the @waitfor@ has parallel semantics, accepting any true @when@ clause.
 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@.
 …
 One scheduling solution is for the signaller S to keep ownership of all locks until the last lock is ready to be transferred, because this semantics fits most closely to the behaviour of single-monitor scheduling.
 However, this solution is inefficient if W2 waited first and can be immediate passed @m2@ when released, while S retains @m1@ until completion of the outer mutex statement.
+However, this solution is inefficient if W2 waited first and immediate passed @m2@ when released, while S retains @m1@ until completion of the outer mutex statement.
 If W1 waited first, the signaller must retain @m1@ amd @m2@ until completion of the outer mutex statement and then pass both to W1.
 % Furthermore, there is an execution sequence where the signaller always finds waiter W2, and hence, waiter W1 starves.
 To support this efficient semantics (and prevent barging), the implementation maintains a list of monitors acquired for each blocked thread.
 When a signaller exits or waits in a monitor function/statement, the front waiter on urgent is unblocked if all its monitors are released.
 Implementing a fast subset check for the necessary released monitors is important and discussed in the following sections.
+To support these 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 mutex function or statement, the front waiter on urgent is unblocked if all its monitors are released.
+Implementing a fast subset check for the necessarily 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.
 …
 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.
+Knowing all members at compilation, even separate compilation, allows uniquely numbered them so the accept-statement implementation can use a fast and compact bit mask with $O(1)$ compare.
 \begin{figure}
 …
 Hence, function pointers are used to identify the functions listed in the @waitfor@ statement, stored in a variable-sized array.
 Then, the same implementation approach used for the urgent stack (see Section~\ref{s:Scheduling}) is used for the calling queue.
 Each caller has a list of monitors acquired, and the @waitfor@ statement performs a (short) linear search matching functions in the @waitfor@ list with called functions, and then verifying the associated mutex locks can be transfers.
+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 transferred.
 …
 struct Msg { int i, j; };
 monitor thread GoRtn { int i;  float f;  Msg m; };
+mutex 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; }
 …
 void ^?{}( GoRtn & mutex ) {}
 void main( GoRtn & gortn ) with( gortn ) {  // thread starts
+void main( GoRtn & mutex gortn ) with(gortn) { // thread starts
         for () {
 …
 \begin{cfa}
 monitor thread DatingService {
+mutex thread DatingService {
         condition Girls[CompCodes], Boys[CompCodes];
         int girlPhoneNo, boyPhoneNo, ccode;
 …
 % \label{f:pingpong}
 % \end{figure}
 Note, the ping/pong threads are globally declared, @pi@/@po@, and hence, start (and possibly complete) before the program main starts.
+Note, the ping/pong threads are globally declared, @pi@/@po@, and hence, start and possibly complete before the program main starts.
 \end{comment}
 \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.
+\subsection{\texorpdfstring{\protect\lstinline@mutex@ Generators / Coroutines / Threads}{monitor Generators / Coroutines / Threads}}
+\CFA generators, coroutines, and threads can also be @mutex@ (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:
+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 ) }
 …
 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.
+% The thread main function is by default @mutex@, so the @mutex@ qualifier for the thread parameter is optional.
+% The reason is that the thread logically starts instantaneously in the thread main acquiring its mutual exclusion, so it starts before any calls to prepare for synchronizing these calls.
+The \CFA program @main@ uses the call/return paradigm to directly communicate with the @GoRtn main@, whereas Go switches to the unbuffered 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.
+The difference between call and channel send occurs for buffered channels making the send asynchronous.
+In \CFA, asynchronous call and multiple buffers are provided using an administrator and worker threads~\cite{Gentleman81} and/or futures (not discussed).
 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.
+When converting a monitor to a thread (server), the coding pattern is to move as much code as possible from the accepted functions into the thread main so it does as 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 complex servers, \eg 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 to build the \CFA runtime, but we always advocate using high-level mechanisms whenever possible.
+Some of these low-level mechanisms are used to build the \CFA runtime, but we always advocate using high-level mechanisms whenever possible.
 …
 In contrast to direct threading is indirect \newterm{thread pools}, \eg Java @executor@, where small jobs (work units) are inserted into a work pool for execution.
 If the jobs are dependent, \ie interact, there is an implicit/explicit dependency graph that ties them together.
+If the jobs are dependent, \ie interact, there is an implicit dependency graph that ties them together.
 While removing direct concurrency, and hence the amount of context switching, thread pools significantly limit the interaction that can occur among jobs.
 Indeed, jobs should not block because that also blocks the underlying thread, which effectively means the CPU utilization, and therefore throughput, suffers.
 …
 \label{s:RuntimeStructureProcessor}
 A virtual processor is implemented by a kernel thread (\eg UNIX process), which are scheduled for execution on a hardware processor by the underlying operating system.
+A virtual processor is implemented by a kernel thread, \eg UNIX process, which are scheduled for execution on a hardware processor by the underlying operating system.
 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.
 …
 \label{s:Implementation}
 A primary implementation challenge is avoiding contention from dynamically allocating memory because of bulk acquire, \eg the internal-scheduling design is (almost) free of allocations.
+A primary implementation challenge is avoiding contention from dynamically allocating memory because of bulk acquire, \eg the internal-scheduling design is almost free of allocations.
 All blocking operations are made by parking threads onto queues, therefore all queues are designed with intrusive nodes, where each node has preallocated link fields for chaining.
 Furthermore, several bulk-acquire operations need a variable amount of memory.
 …
 There are two versions of the \CFA runtime kernel: debug and non-debug.
 The debugging version has many runtime checks and internal assertions, \eg stack (non-writable) guard page, and checks for stack overflow whenever context switches occur among coroutines and threads, which catches most stack overflows.
+The debugging version has many runtime checks and internal assertions, \eg stack non-writable guard page, and checks for stack overflow whenever context switches occur among coroutines and threads, which catches most stack overflows.
 After a program is debugged, the non-debugging version can be used to significantly decrease space and increase performance.
 …
 To test the performance of the \CFA runtime, a series of microbenchmarks are used to compare \CFA with pthreads, Java 11.0.6, Go 1.12.6, Rust 1.37.0, Python 3.7.6, Node.js 12.14.1, and \uC 7.0.0.
 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.
+For comparison, the package must be multi-processor (M:N), which excludes libdil and libmil~\cite{libdill} (M:1)), and use a shared-memory programming model, \eg not message passing.
 The benchmark computer is an AMD Opteron\texttrademark\ 6380 NUMA 64-core, 8 socket, 2.5 GHz processor, running Ubuntu 16.04.6 LTS, and pthreads/\CFA/\uC are compiled with gcc 9.2.1.
 …
 The 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}.
+All omitted tests for other languages are functionally identical to the \CFA tests and available online~\cite{CforallConcurrentBenchmarks}.
 % 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 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} 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.
+% cp -p benchmark.tar /u/cforall/public_html/doc/concurrent_benchmark.tar
+\paragraph{Creation}
+Creation is measured by creating and 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}[aboveskip=0pt,belowskip=0pt]
+@coroutine@ C {};
+void main( C & ) { for () { @suspend;@ } }
+int main() { // coroutine test
+        C c;
+        BENCH( for ( N ) { @resume( c );@ } )
+\begin{cfa}
+@coroutine@ MyCoroutine {};
+void ?{}( MyCoroutine & this ) {
+#ifdef EAGER
+        resume( this );
+#endif
+}
+void main( MyCoroutine & ) {}
+int main() {
+        BENCH( for ( N ) { @MyCoroutine c;@ } )
         sout | result;
+}
+int main() { // thread test
+        BENCH( for ( N ) { @yield();@ } )
+        sout | result;
+}
+\end{cfa}
+\captionof{figure}{\CFA context-switch benchmark}
+\label{f:ctx-switch}
+\end{cfa}
+\captionof{figure}{\CFA creation benchmark}
+\label{f:creation}
 \columnbreak
 \vspace*{-16pt}
+\captionof{table}{Context switch comparison (nanoseconds)}
+\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           & 0.0   \\
+\CFA generator          & 1.8           & 2.0           & 0.3   \\
+\CFA coroutine          & 32.5          & 32.9          & 0.8   \\
+\CFA thread                     & 93.8          & 93.6          & 2.2   \\
+\uC coroutine           & 50.3          & 50.3          & 0.2   \\
+\uC thread                      & 97.3          & 97.4          & 1.0   \\
+Python generator        & 40.9          & 41.3          & 1.5   \\
+Node.js generator       & 32.6          & 32.2          & 1.0   \\
+Node.js await           & 1852.2        & 1854.7        & 16.4  \\
+Goroutine thread        & 143.0         & 143.3         & 1.1   \\
+Rust thread                     & 332.0         & 331.4         & 2.4   \\
+Java thread                     & 405.0         & 415.0         & 17.6  \\
+Pthreads thread         & 334.3         & 335.2         & 3.9
+\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               & 33.4          & 33.5          & 0.3           \\
+Goroutine thread                & 751.0         & 750.5         & 3.1           \\
+Rust tokio thread               & 1860.0        & 1881.1        & 37.6          \\
+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}
+\vspace*{-10pt}
 \paragraph{Internal Scheduling}
 …
 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.
+Java scheduling is significantly greater because the benchmark explicitly creates multiple threads in order to prevent the JIT from making the program sequential, \ie removing all locking.
 \begin{multicols}{2}
 …
+}
 \end{cfa}
+\vspace*{-8pt}
 \captionof{figure}{\CFA Internal-scheduling benchmark}
 \label{f:schedint}
 …
 \paragraph{Mutual-Exclusion}
 Uncontented mutual exclusion, which frequently occurs, is measured by entering/leaving a critical section.
 For monitors, entering and leaving a monitor function is measured, otherwise the language-appropriate mutex-lock is measured.
+Uncontented mutual exclusion, which frequently occurs, is measured by entering and leaving a critical section.
+For monitors, entering and leaving a mutex 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}.
 …
 \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.
+\paragraph{Context Switching}
+In procedural programming, the cost of a function call is important as modularization (refactoring) increases.
+(In many cases, a compiler inlines function calls to increase the size and number of basic blocks for optimizing.)
+Similarly, when modularization extends to coroutines and 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} shows the \CFA code for a coroutine and 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}
 \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;@ } )
+\begin{cfa}[aboveskip=0pt,belowskip=0pt]
+@coroutine@ C {};
+void main( C & ) { for () { @suspend;@ } }
+int main() { // coroutine test
+        C c;
+        BENCH( for ( N ) { @resume( c );@ } )
         sout | result;
+}
+\end{cfa}
+\captionof{figure}{\CFA creation benchmark}
+\label{f:creation}
+int main() { // thread test
+        BENCH( for ( N ) { @yield();@ } )
+        sout | result;
+}
+\end{cfa}
+\captionof{figure}{\CFA context-switch benchmark}
+\label{f:ctx-switch}
 \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
+\captionof{table}{Context switch comparison (nanoseconds)}
+\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           & 0.0   \\
+\CFA generator          & 1.8           & 2.0           & 0.3   \\
+\CFA coroutine          & 32.5          & 32.9          & 0.8   \\
+\CFA thread                     & 93.8          & 93.6          & 2.2   \\
+\uC coroutine           & 50.3          & 50.3          & 0.2   \\
+\uC thread                      & 97.3          & 97.4          & 1.0   \\
+Python generator        & 40.9          & 41.3          & 1.5   \\
+Node.js await           & 1852.2        & 1854.7        & 16.4  \\
+Node.js generator       & 33.3          & 33.4          & 0.3   \\
+Goroutine thread        & 143.0         & 143.3         & 1.1   \\
+Rust async await        & 32.0          & 32.0          & 0.0   \\
+Rust tokio thread       & 143.0         & 143.0         & 1.7   \\
+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}
 …
 \subsection{Discussion}
 Languages using 1:1 threading based on pthreads can at best meet or exceed (due to language overhead) the pthread results.
+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.
 …
 \section{Conclusion}
+\section{Conclusions and Future Work}
 Advanced control-flow will always be difficult, especially when there is temporal ordering and nondeterminism.
 …
 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@.
+\CFA high-level race-free monitors and threads, when used with mutex access function, provide the core mechanisms for mutual exclusion and synchronization, without having to resort to magic qualifiers like @volatile@ or @atomic@.
 Extending these mechanisms to handle high-level deadlock-free bulk acquire across both mutual exclusion and synchronization is a unique contribution.
 The \CFA runtime provides concurrency based on a preemptive M:N user-level threading-system, executing in clusters, which encapsulate scheduling of work on multiple kernel threads providing parallelism.
 The M:N model is judged to be efficient and provide greater flexibility than a 1:1 threading model.
 These concepts and the \CFA runtime-system are written in the \CFA language, extensively leveraging the \CFA type-system, which demonstrates the expressiveness of the \CFA language.
 Performance comparisons with other concurrent systems/languages show the \CFA approach is competitive across all basic operations, which translates directly into good performance in well-written applications with advanced control-flow.
+Performance comparisons with other concurrent systems and 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.
-\section{Future Work}
 While control flow in \CFA has a strong start, development is still underway to complete a number of missing features.
+\paragraph{Flexible Scheduling}
+\label{futur:sched}
+\medskip
+\textbf{Flexible Scheduling:}
 An important part of concurrency is scheduling.
 Different scheduling algorithms can affect performance (both in terms of average and variation).
+Different scheduling algorithms can affect performance, both in terms of average and variation.
 However, no single scheduler is optimal for all workloads and therefore there is value in being able to change the scheduler for given programs.
 One solution is to offer various tuning options, allowing the scheduler to be adjusted to the requirements of the workload.
 …
 Currently, the \CFA pluggable scheduler is too simple to handle complex scheduling, \eg quality of service and real-time, where the scheduler must interact with mutex objects to deal with issues like priority inversion~\cite{Buhr00b}.
+\paragraph{Non-Blocking I/O}
+\label{futur:nbio}
+\smallskip
+\textbf{Non-Blocking I/O:}
 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.
 …
 A non-blocking I/O library is currently under development for \CFA.
+\paragraph{Other Concurrency Tools}
+\label{futur:tools}
+\smallskip
+\textbf{Other Concurrency Tools:}
 While monitors offer flexible and powerful concurrency for \CFA, other concurrency tools are also necessary for a complete multi-paradigm concurrency package.
 Examples of such tools can include futures and promises~\cite{promises}, executors and actors.
 …
 As well, new \CFA extensions should make it possible to create a uniform interface for virtually all mutual exclusion, including monitors and low-level locks.
+\paragraph{Implicit Threading}
+\label{futur:implcit}
+Basic concurrent (embarrassingly parallel) applications can benefit greatly from implicit concurrency, where sequential programs are converted to concurrent, possibly with some help from pragmas to guide the conversion.
+\smallskip
+\textbf{Implicit Threading:}
+Basic \emph{embarrassingly parallel} applications can benefit greatly from implicit concurrency, where sequential programs are converted to concurrent, with some help from pragmas to guide the conversion.
 This type of concurrency can be achieved both at the language level and at the library level.
 The canonical example of implicit concurrency is concurrent nested @for@ loops, which are amenable to divide and conquer algorithms~\cite{uC++book}.
 The \CFA language features should make it possible to develop a reasonable number of implicit concurrency mechanism to solve basic HPC data-concurrency problems.
+The \CFA language features should make it possible to develop a reasonable number of implicit concurrency mechanisms to solve basic HPC data-concurrency problems.
 However, implicit concurrency is a restrictive solution with significant limitations, so it can never replace explicit concurrent programming.

doc/papers/concurrency/figures/RunTimeStructure.fig

-              r07d867b
+              r22f94a4
 -2
 2
 3855 2775 4155 2925
 3 0 1 0 0 0 0 0 0.000 1 0.0000 3930 2850 30 30 3930 2850 3960 2880
 3 0 1 0 0 0 0 0 0.000 1 0.0000 4035 2850 30 30 4035 2850 4065 2880
+3255 2475 3555 2625
+3 0 1 0 0 0 0 0 0.000 1 0.0000 3330 2550 30 30 3330 2550 3360 2580
+3 0 1 0 0 0 0 0 0.000 1 0.0000 3435 2550 30 30 3435 2550 3465 2580
 -6
 4755 3525 5055 3675
 3 0 1 0 0 0 0 0 0.000 1 0.0000 4830 3600 30 30 4830 3600 4860 3630
 3 0 1 0 0 0 0 0 0.000 1 0.0000 4935 3600 30 30 4935 3600 4965 3630
+4155 3225 4455 3375
+3 0 1 0 0 0 0 0 0.000 1 0.0000 4230 3300 30 30 4230 3300 4260 3330
+3 0 1 0 0 0 0 0 0.000 1 0.0000 4335 3300 30 30 4335 3300 4365 3330
 -6
 4650 2775 4950 2925
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4725 2850 15 15 4725 2850 4740 2865
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4800 2850 15 15 4800 2850 4815 2865
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4875 2850 15 15 4875 2850 4890 2865
+4050 2475 4350 2625
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4125 2550 15 15 4125 2550 4140 2565
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4200 2550 15 15 4200 2550 4215 2565
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4275 2550 15 15 4275 2550 4290 2565
 -6
 3225 2400 3525 2550
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3300 2475 15 15 3300 2475 3315 2490
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3375 2475 15 15 3375 2475 3390 2490
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3450 2475 15 15 3450 2475 3465 2490
+2625 2100 2925 2250
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 2700 2175 15 15 2700 2175 2715 2190
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 2775 2175 15 15 2775 2175 2790 2190
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 2850 2175 15 15 2850 2175 2865 2190
 -6
 5475 3450 5625 3750
 3 0 1 -1 -1 0 0 20 0.000 1 4.7120 5550 3525 15 15 5550 3525 5535 3540
 3 0 1 -1 -1 0 0 20 0.000 1 4.7120 5550 3600 15 15 5550 3600 5535 3615
 3 0 1 -1 -1 0 0 20 0.000 1 4.7120 5550 3675 15 15 5550 3675 5535 3690
+4875 3150 5025 3450
+3 0 1 -1 -1 0 0 20 0.000 1 4.7120 4950 3225 15 15 4950 3225 4935 3240
+3 0 1 -1 -1 0 0 20 0.000 1 4.7120 4950 3300 15 15 4950 3300 4935 3315
+3 0 1 -1 -1 0 0 20 0.000 1 4.7120 4950 3375 15 15 4950 3375 4935 3390
 -6
 4275 3525 4575 3675
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4350 3600 15 15 4350 3600 4365 3615
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4425 3600 15 15 4425 3600 4440 3615
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4500 3600 15 15 4500 3600 4515 3615
+3675 3225 3975 3375
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3750 3300 15 15 3750 3300 3765 3315
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3825 3300 15 15 3825 3300 3840 3315
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3900 3300 15 15 3900 3300 3915 3315
 -6
 3225 4125 4650 4425
 4350 4200 4650 4350
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4425 4275 15 15 4425 4275 4440 4290
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4500 4275 15 15 4500 4275 4515 4290
 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 4575 4275 15 15 4575 4275 4590 4290
+2625 3825 4050 4125
+3750 3900 4050 4050
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3825 3975 15 15 3825 3975 3840 3990
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3900 3975 15 15 3900 3975 3915 3990
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3975 3975 15 15 3975 3975 3990 3990
 -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
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 2850 3975 225 150 2850 3975 3075 4125
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3450 3975 225 150 3450 3975 3675 4125
 -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
+6075 3825 6900 4125
+6600 3900 6900 4050
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 6675 3975 15 15 6675 3975 6690 3990
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 6750 3975 15 15 6750 3975 6765 3990
+3 0 1 -1 -1 0 0 20 0.000 1 0.0000 6825 3975 15 15 6825 3975 6840 3990
 -6
 1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 6900 4275 225 150 6900 4275 7125 4425
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 6300 3975 225 150 6300 3975 6525 4125
 -6
 6675 3525 8025 3975
+6075 3225 7425 3675
 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
+3450 6375 3450
 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
+3450 6750 3450
 2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
 3975 7800 3525 7350 3525 7350 3975 7800 3975
+3675 7200 3225 6750 3225 6750 3675 7200 3675
 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
+3450 7425 3450
 -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 5550 3225 150 150 5550 3225 5700 3225
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 5550 3975 150 150 5550 3975 5700 3975
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3525 2850 150 150 3525 2850 3675 2850
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4200 2475 150 150 4200 2475 4350 2475
 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 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 3750 2475 150 150 3750 2475 3900 2625
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4875 3600 150 150 4875 3600 5025 3750
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3975 2850 150 150 3975 2850 4125 2850
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 7200 2775 150 150 7200 2775 7350 2775
 3 0 1 0 0 0 0 0 0.000 1 0.0000 2250 4830 30 30 2250 4830 2280 4860
 3 0 1 0 0 0 0 0 0.000 1 0.0000 7200 2775 30 30 7200 2775 7230 2805
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3525 3600 150 150 3525 3600 3675 3600
 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3875 4800 100 100 3875 4800 3975 4800
 1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4650 4800 150 75 4650 4800 4800 4875
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4950 2325 150 150 4950 2325 5100 2325
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4950 2925 150 150 4950 2925 5100 2925
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4950 3675 150 150 4950 3675 5100 3675
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 2925 2550 150 150 2925 2550 3075 2550
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3600 2175 150 150 3600 2175 3750 2175
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3825 2550 150 150 3825 2550 3975 2550
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4050 2175 150 150 4050 2175 4200 2175
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3375 3300 150 150 3375 3300 3525 3300
+3 0 1 0 0 0 0 0 0.000 1 0.0000 2925 3300 30 30 2925 3300 2955 3330
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3150 2175 150 150 3150 2175 3300 2325
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4275 3300 150 150 4275 3300 4425 3450
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3375 2550 150 150 3375 2550 3525 2550
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 6600 2475 150 150 6600 2475 6750 2475
+3 0 1 0 0 0 0 0 0.000 1 0.0000 1650 4530 30 30 1650 4530 1680 4560
+3 0 1 0 0 0 0 0 0.000 1 0.0000 6600 2475 30 30 6600 2475 6630 2505
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 2925 3300 150 150 2925 3300 3075 3300
+3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 3275 4500 100 100 3275 4500 3375 4500
+1 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4050 4500 150 75 4050 4500 4200 4575
 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
+3900 1800 3450 1350 3450 1350 3900 1800 3900
 2 1 1 -1 -1 0 0 -1 4.000 0 0 0 0 0 5
 4500 6300 1800 3000 1800 3000 4500 6300 4500
+4200 5700 1500 2400 1500 2400 4200 5700 4200
 2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
 2850 5775 2400 5325 2400 5325 2850 5775 2850
+2550 5175 2100 4725 2100 4725 2550 5175 2550
 2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
 4200 5775 3750 5325 3750 5325 4200 5775 4200
+3900 5175 3450 4725 3450 4725 3900 5175 3900
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3975 5325 3975
+3675 4725 3675
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3225 5325 3225
+2925 4725 2925
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 2625 5325 2625
+2325 4725 2325
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3975 5925 3975
+3675 5325 3675
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3225 5925 3225
+2925 5325 2925
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 2625 5925 2625
+2325 5325 2325
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 0 0 2
 3975 5175 2625
+3675 4575 2325
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3975 5925 2025
+3675 5325 1725
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3750 6225 3750
+3450 5625 3450
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 2625 3225 2625
+2325 2625 2325
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 3
 1 1.00 45.00 90.00
 2025 4200 2025 4200 2250
+1725 3600 1725 3600 1950
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 0 0 2
 2625 3225 3600
+2325 2625 3300
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3600 3375 3600
+3300 2775 3300
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3600 3825 3600
+3300 3225 3300
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3600 4275 3600
+3300 3675 3300
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3600 4725 3600
+3300 4125 3300
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 2
 1 1.00 45.00 90.00
 3600 5175 3600
+3300 4575 3300
 2 0 1 -1 -1 0 0 -1 0.000 0 0 0 0 0 5
 3450 5775 3000 5325 3000 5325 3450 5775 3450
+3150 5175 2700 4725 2700 4725 3150 5175 3150
 2 1 1 -1 -1 0 0 -1 4.000 0 0 0 0 0 5
 4500 8100 1800 6600 1800 6600 4500 8100 4500
+4200 7500 1500 6000 1500 6000 4200 7500 4200
 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
+2475 6225 2475
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 0 0 2
 2775 6825 3750
+2475 6225 3450
 1 0 1 -1 -1 0 0 -1 0.000 0 0 -1 1 0 4
 1 1.00 45.00 90.00
 3750 7875 2325 7200 2325 7200 2550
+3450 7275 2025 6600 2025 6600 2250
 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
+4650 5250 4425 5025 4425 5025 4650 5250 4650
 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 2 150 870 4200 3975 Ready Tasks\001
 1 -1 0 0 0 10 0.0000 2 135 1095 7350 1725 Other Cluster(s)\001
 1 -1 0 0 0 10 0.0000 2 105 840 4650 1725 User Cluster\001
 1 -1 0 0 0 10 0.0000 2 150 615 2175 3675 Manager\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
+4650 6150 4650 6150 4425 6375 4425 6375 4650
+1 -1 0 0 0 10 0.0000 2 105 720 4950 4125 Processors\001
+1 -1 0 0 0 10 0.0000 2 120 1005 3600 2925 Blocked Tasks\001
+1 -1 0 0 0 10 0.0000 2 150 870 3600 3675 Ready Tasks\001
+1 -1 0 0 0 10 0.0000 2 135 1095 6750 1425 Other Cluster(s)\001
+1 -1 0 0 0 10 0.0000 2 105 840 4050 1425 User Cluster\001
+1 -1 0 0 0 10 0.0000 2 150 615 1575 3375 Manager\001
+1 -1 0 0 0 10 0.0000 2 105 990 1575 3225 Discrete-event\001
+1 -1 0 0 0 10 0.0000 2 135 795 1575 4050 preemption\001
+0 -1 0 0 0 10 0.0000 2 150 1365 1725 4575 generator/coroutine\001
+0 -1 0 0 0 10 0.0000 2 120 270 3450 4575 task\001
+0 -1 0 0 0 10 0.0000 2 105 450 6450 4575 cluster\001
+0 -1 0 0 0 10 0.0000 2 105 660 5325 4575 processor\001
+0 -1 0 0 0 10 0.0000 2 105 555 4275 4575 monitor\001

doc/papers/concurrency/mail2

-              r07d867b
+              r22f94a4
 Software: Practice and Experience Editorial Office
+Date: Sat, 18 Apr 2020 10:42:13 +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.R1
+-Apr-2020
+Dear Dr Buhr,
+Many thanks for submitting SPE-19-0219.R1 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.
+I believe that we are making progress here towards a paper that can be published in Software: Practice and Experience.  However the referees still have significant concerns about the paper. The journal's focus is on practice and experience, and one of the the reviewers' concerns remains that your submission should focus the narrative more on the perspective of the programmer than the language designer. I agree that this would strengthen your submission, and I ask you to address this as well as the referees' other comments.
+A revised version of your manuscript that takes into account the comments of the referee(s) will be reconsidered for publication.
+Please note that submitting a revision of your manuscript does not guarantee eventual acceptance, and that your revision may 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 a short 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 would like help with English language editing, or other article preparation support, Wiley Editing Services offers expert help with English Language Editing, as well as translation, manuscript formatting, and figure formatting at www.wileyauthors.com/eeo/preparation. You can also check out our resources for Preparing Your Article for general guidance about writing and preparing your manuscript at www.wileyauthors.com/eeo/prepresources.
+Once again, thank you for submitting your manuscript to Software: Practice and Experience and I look forward to receiving your revision.
+Sincerely,
+Richard
+Prof. Richard Jones
+Software: Practice and Experience
+R.E.Jones@kent.ac.uk
+Referee(s)' Comments to Author:
+Reviewing: 1
+Comments to the Author
+(A relatively short second review)
+I thank the authors for their revisions and comprehensive response to
+reviewers' comments --- many of my comments have been successfully
+addressed by the revisions.  Here I'll structure my comments around
+the main salient points in that response which I consider would
+benefit from further explanation.
+>  Table 1 is moved to the start and explained in detail.
+I consider this change makes a significant improvement to the paper,
+laying out the landscape of language features at the start, and thus
+addresses my main concerns about the paper.
+I still have a couple of issues --- perhaps the largest is that it's
+still not clear at this point in the paper what some of these options
+are, or crucially how they would be used. I don't know if it's
+possbile to give high-level examples or use cases to be clear about
+these up front - or if that would duplicate too much information from
+later in the paper - either way expanding out the discussion - even if
+just two a couple of sentences for each row - would help me more.  The
+point is not just to define these categories but to ensure the
+readers' understanding of these definitons agrees with that used in
+the paper.
+in a little more detail:
+ * 1st para section 2 begs the question: why not support each
+   dimension independently, and let the programmer or library designer
+   combiine features?
+ * "execution state" seems a relatively low-level description here.
+  I don't think of e.g. the lambda calculus that way. Perhaps it's as
+  good a term as any.
+ * Why must there "be language mechanisms to create, block/unblock,
+   and join with a thread"?  There aren't in Smalltalk (although there
+   are in the runtime).  Especially given in Cforall those mechanisms
+   are *implicit* on thread creation and destruction?
+ * "Case 1 is a function that borrows storage for its state (stack
+   frame/activation) and a thread from its invoker"
+   this much makes perfect sense to me, but I don't understand how a
+   non-stateful, non-theaded function can then retain
+   "this state across callees, ie, function local-variables are
+   retained on the stack across calls."
+   how can it retain function-local values *across calls* when it
+   doesn't have any functional-local state?
+   I'm not sure if I see two separate cases here - rougly equivalent
+   to C functions without static storage, and then C functions *with*
+   static storage. I assumed that was the distinction between cases 1
+   & 3; but perhpas the actual distinction is that 3 has a
+   suspend/resume point, and so the "state" in figure 1 is this
+   component of execution state (viz figs 1 & 2), not the state
+   representing the cross-call variables?
+>    but such evaluation isn't appropriate for garbage-collected or JITTed
+   languages like Java or Go.
+For JITTed languages in particular, reporting peak performance needs
+to "warm up" the JIT with a number of iterators before beginning
+measurement. Actually for JIT's its even worse: see Edd Barrett et al
+OOPSLA 2017.
+minor issues:
+ * footnote A - I've looked at various other papers & the website to
+   try to understand how "object-oriented" Cforall is - I'm still not
+   sure.  This footnote says Cforall has "virtuals" - presumably
+   virtual functions, i.e. dynamic dispatch - and inheritance: that
+   really is OO as far as I (and most OO people) are concerned.  For
+   example Haskell doesn't have inheritance, so it's not OO; while
+   CLOS (the Common Lisp *Object* System) or things like Cecil and
+   Dylan are considered OO even though they have "multiple function
+   parameters as receivers", lack "lexical binding between a structure
+   and set of functions", and don't have explicit receiver invocation
+   syntax.  Python has receiver syntax, but unlike Java or Smalltalk
+   or C++, method declarations still need to have an explicit "self"
+   receiver parameter.  Seems to me that Go, for example, is
+   more-or-less OO with interfaces, methods, and dynamic dispatch (yes
+   also and an explicit receiver syntax but that's not
+   determiniative); while Rust lacks dynamic dispatch built-in.  C is
+   not OO as a language, but as you say given it supports function
+   pointers with structures, it does support an OO programm style.
+   This is why I again recommend just not buying into this fight: not
+   making any claims about whether Cforall is OO or is not - because
+   as I see it, the rest of the paper doesn't depend on whether
+   Cforall is OO or not.  That said: this is just a recommendation,
+   and I won't quibble over this any further.
+ * is a "monitor function" the same as a "mutex function"?
+   if so the paper should pick one term; if not, make the distinction clear.
+ * "As stated on line 1 because state declarations from the generator
+    type can be moved out of the coroutine type into the coroutine main"
+    OK sure, but again: *why* would a programmer want to do that?
+    (Other than, I guess, to show the difference between coroutines &
+    generators?)  Perhaps another way to put this is that the first
+    para of 3.2 gives the disadvantages of coroutines vs-a-vs
+    generators, briefly describes the extended semantics, but never
+    actualy says why a programmer may want those extended semantics,
+    or how they would benefit.  I don't mean to belabour the point,
+    but (generalist?) readers like me would generally benefit from
+    those kinds of discussions about each feature throughout the
+    paper: why might a programmer want to use them?
+> p17 if the multiple-monitor entry procedure really is novel, write a paper
+> about that, and only about that.
+> We do not believe this is a practical suggestion.
+ * I'm honestly not trying to be snide here: I'm not an expert on
+   monitor or concurrent implementations. Brinch Hansen's original
+   monitors were single acquire; this draft does not cite any other
+   previous work that I could see. I'm not suggesting that the brief
+   mention of this mechanism necessarily be removed from this paper,
+   but if this is novel (and a clear advance over a classical OO
+   monitor a-la Java which only acquires the distinguished reciever)
+   then that would be worth another paper in itself.
+> * conclusion should conclude the paper, not the related.
+> We do not understand this comment.if ithis
+My typo: the paper's conclusion should come at the end, after the
+future work section.
+To encourage accountability, I'm signing my reviews in 2020.
+For the record, I am James Noble, kjx@ecs.vuw.ac.nz.
+Reviewing: 2
+Comments to the Author
+I thank the authors for their detailed response. To respond to a couple of points raised  in response to my review (number 2):
+- on the Boehm paper and whether code is "all sequential to the compiler": I now understand the authors' position better and suspect we are in violent agreement, except for whether it's appropriate to use the rather breezy phrase "all sequential to the compiler". It would be straightforward to clarify that code not using the atomics features is optimized *as if* it were sequential, i.e. on the assumption of a lack of data races.
+- on the distinction between "mutual exclusion" and "synchronization": the added citation does help, in that it makes a coherent case for the definition the authors prefer. However, the text could usefully clarify that this is a matter of definition not of fact, given especially that in my assessment the authors' preferred definition is not the most common one. (Although the mention of Hoare's apparent use of this definition is one data point, countervailing ones are found in many contemporaneous or later papers, e.g. Habermann's 1972 "Synchronization of Communicating Processes" (CACM 15(3)), Reed & Kanodia's 1979 "Synchronization with eventcounts and sequencers" (CACM (22(2)) and so on.)
+I am glad to see that the authors have taken on board most of the straightforward improvements I suggested.
+However, a recurring problem of unclear writing still remains through many parts of the paper, including much of sections 2, 3 and 6. To highlight a couple of problem patches (by no means exhaustive):
+- section 2 (an expanded version of what was previously section 5.9) lacks examples and is generally obscure and allusory ("the most advanced feature" -- name it! "in triplets" -- there is only one triplet!; what are "execution locations"? "initialize" and "de-initialize" what? "borrowed from the invoker" is a concept in need of explaining or at least a fully explained example -- in what sense does a plain function borrow" its stack frame? "computation only" as opposed to what? in 2.2, in what way is a "request" fundamental to "synchronization"? and the "implicitly" versus "explicitly" point needs stating as elsewhere, with a concrete example e.g. Java built-in mutexes versus java.util.concurrent).
+- section 6: 6.2 omits the most important facts in preference for otherwise inscrutable detail: "identify the kind of parameter" (first say *that there are* kinds of parameter, and what "kinds" means!); "mutex parameters are documentation" is misleading (they are also semantically significant!) and fails to say *what* they mean; the most important thing is surely that 'mutex' is a language feature for performing lock/unlock operations at function entry/exit. So say it! The meanings of examples f3 and f4 remain unclear. Meanwhile in 6.3, "urgent" is not introduced (we are supposed to infer its meaning from Figure 12, but that Figure is incomprehensible to me), and we are told of "external scheduling"'s long history in Ada but not clearly what it actually means; 6.4's description of "waitfor" tells us it is different from an if-else chain but tries to use two *different* inputs to tell us that the behavior is different; tell us an instance where *the same* values of C1 and C2 give different behavior (I even wrote out a truth table and still don't see the semantic difference)
+The authors frequently use bracketed phrases, and sometimes slashes "/", in ways that are confusing and/or detrimental to readability. Page 13 line 2's "forward (backward)" is one particularly egregious example. In general I would recommend the the authors try to limit their use of parentheses and slashes as a means of forcing a clearer wording to emerge. Also, the use of "eg." is often cursory and does not explain the examples given, which are frequently a one- or two-word phrase of unclear referent.
+Considering the revision more broadly, none of the more extensive or creative rewrites I suggested in my previous review have been attempted, nor any equivalent efforts to improve its readability. The hoisting of the former section 5.9 is a good idea, but the newly added material accompanying it (around Table 1) suffers fresh deficiencies in clarity. Overall the paper is longer than before, even though (as my previous review stated), I believe a shorter paper is required in order to serve the likely purpose of publication. (Indeed, the authors' letter implies that a key goal of publication is to build community and gain external users.)
+Given this trajectory, I no longer see a path to an acceptable revision of the present submission. Instead I suggest the authors consider splitting the paper in two: one half about coroutines and stack management, the other about mutexes, monitors and the runtime. (A briefer presentation of the runtime may be helpful in the first paper also, and a brief recap of the generator and coroutine support is obviously needed in the second too.) Both of these new papers would need to be written with a strong emphasis on clarity, paying great care to issues of structure, wording, choices of example, and restraint (saying what's important, not everything that could be said). I am confident the authors could benefit from getting early feedback from others at their institution. For the performance experiments, of course these do not split evenly -- most (but not all) belong in the second of these two hypothetical papers. But the first of them would still have plenty of meat to it; for me, a clear and thorough study of the design space around coroutines is the most interesting and tantalizing prospect.
+I do not buy the authors' defense of the limited practical experience or "non-micro" benchmarking presented. Yes, gaining external users is hard and I am sympathetic on that point. But building something at least *somewhat* substantial with your own system should be within reach, and without it the "practice and experience" aspects of the work have not been explored. Clearly C\/ is the product of a lot of work over an extended period, so it is a surprise that no such experience is readily available for inclusion.
+Some smaller points:
+It does not seem right to state that a stack is essential to Von Neumann architectures -- since the earliest Von Neumann machines (and indeed early Fortran) did not use one.
+To elaborate on something another reviewer commented on: it is a surprise to find a "Future work" section *after* the "Conclusion" section. A "Conclusions and future work" section often works well.
+Reviewing: 3
+Comments to the Author
+This is the second round of reviewing.
+As in the first review, I found that the paper (and Cforall) contains
+a lot of really interesting ideas, but it remains really difficult to
+have a good sense of which idea I should use and when. This applies in
+different ways to different features from the language:
+* coroutines/generators/threads: here there is
+  some discussion, but it can be improved.
+* interal/external scheduling: I didn't find any direct comparison
+  between these features, except by way of example.
+I requested similar things in my previous review and I see that
+content was added in response to those requests. Unfortunately, I'm
+not sure that I can say it improved the paper's overall read. I think
+in some sense the additions were "too much" -- I would have preferred
+something more like a table or a few paragraphs highlighting the key
+reasons one would pick one construct or the other.
+In general, I do wonder if the paper is just trying to do too much.
+The discussion of clusters and pre-emption in particular feels quite
+rushed.
+## Summary
+I make a number of suggestions below but the two most important
+I think are:
+* Recommend to shorten the comparison on coroutine/generator/threads
+  in Section 2 to a paragraph with a few examples, or possibly a table
+  explaining the trade-offs between the constructs
+* Recommend to clarify the relationship between internal/external
+  scheduling -- is one more general but more error-prone or low-level?
+## Coroutines/generators/threads
+There is obviously a lot of overlap between these features, and in
+particular between coroutines and generators. As noted in the previous
+review, many languages have chosen to offer *only* generators, and to
+build coroutines by stacks of generators invoking one another.
+I believe the newly introduced Section 2 of the paper is trying to
+motivate why each of these constructs exist, but I did not find it
+effective. It was dense and difficult to understand. I think the
+problem is that Section 2 seems to be trying to derive "from first
+principles" why each construct exists, but I think that a more "top
+down" approach would be easier to understand.
+In fact, the end of Section 2.1 (on page 5) contains a particular
+paragraph that embodies this "top down" approach. It starts,
+"programmers can now answer three basic questions", and thus gives
+some practical advice for which construct you should use and when. I
+think giving some examples of specific applications that this
+paragraph, combined with some examples of cases where each construct
+was needed, would be a better approach.
+I don't think this compariosn needs to be very long. It seems clear
+enough that one would
+* prefer generators for simple computations that yield up many values,
+* prefer coroutines for more complex processes that have significant
+  internal structure,
+* prefer threads for cases where parallel execution is desired or
+  needed.
+I did appreciate the comparison in Section 2.3 between async-await in
+JS/Java and generators/coroutines. I agree with its premise that those
+mechanisms are a poor replacement for generators (and, indeed, JS has
+a distinct generator mechanism, for example, in part for this reason).
+I believe I may have asked for this in a previous review, but having
+read it, I wonder if it is really necessary, since those mechanisms
+are so different in purpose.
+## Internal vs external scheduling
+I find the motivation for supporting both internal and external
+scheduling to be fairly implicit. After several reads through the
+section, I came to the conclusion that internal scheduling is more
+expressive than external scheduling, but sometimes less convenient or
+clear. Is this correct? If not, it'd be useful to clarify where
+external scheduling is more expressive.
+The same is true, I think, of the `signal_block` function, which I
+have not encountered before; it seems like its behavior can be modeled
+with multiple condition variables, but that's clearly more complex.
+One question I had about `signal_block`: what happens if one signals
+but no other thread is waiting? Does it block until some other thread
+waits? Or is that user error?
+I would find it very interesting to try and capture some of the
+properties that make internal vs external scheduling the better
+choice.
+For example, it seems to me that external scheduling works well if
+there are only a few "key" operations, but that internal scheduling
+might be better otherwise, simply because it would be useful to have
+the ability to name a signal that can be referenced by many
+methods. Consider the bounded buffer from Figure 13: if it had
+multiple methods for removing elements, and not just `remove`, then
+the `waitfor(remove)` call in `insert` might not be sufficient.
+## Comparison of external scheduling to messaging
+I did enjoy the section comparing external scheduling to Go's
+messaging mechanism, which I believe is a new addition.
+I believe that one difference between the Go program and the Cforall
+equivalent is that the Goroutine has an associated queue, so that
+multiple messages could be enqueued, whereas the Cforall equivalent is
+effectively a "bounded buffer" of length 1. Is that correct? I think
+this should be stated explicitly. (Presumably, one could modify the
+Cforall program to include an explicit vector of queued messages if
+desired, but you would also be reimplementing the channel
+abstraction.)
+Also, in Figure 20, I believe that there is a missing `mutex` keyword.
+The fiugre states:
+```
+void main(GoRtn & gortn) with(gortn) {
+```
+but I think it should probably be as follows:
+```
+void main(GoRtn & mutex gortn) with(gortn) {
+```
+Unless there is some implicit `mutex` associated with being a main
+function for a `monitor thread`.
+## Atomic operations and race freedom
+I was glad to see that the paper acknowledged that Cforall still had
+low-level atomic operations, even if their use is discouraged in favor
+of higher-level alternatives.
+However, I still feel that the conclusion overstates the value of the
+contribution here when it says that "Cforall high-level race-free
+monitors and threads provide the core mechanisms for mutual exclusion
+and synchronization, without the need for volatile and atomics". I
+feel confident that Java programmers, for example, would be advised to
+stick with synchronized methods whenever possible, and it seems to me
+that they offer similar advantages -- but they sometimes wind up using
+volatiles for performance reasons.
+I was also confused by the term "race-free" in that sentence. In
+particular, I don't think that Cforall has any mechanisms for
+preventing *data races*, and it clearly doesn't prevent "race
+conditions" (which would bar all sorts of useful programs). I suppose
+that "race free" here might be referring to the improvements such as
+removing barging behavior.
+## Performance comparisons
+In my previous review, I requested comparisons against Rust and
+node.js, and I see that the new version of the paper includes both,
+which is a good addition.
+One note on the Rust results: I believe that the results are comparing
+against the threads found in Rust's standard library, which are
+essentially a shallow wrapper around pthreads, and hence the
+performance is quite close to pthread performance (as one would
+expect). It would perhaps be more interesting to see a comparison
+built using [tokio] or [async-std], two of the more prominent
+user-space threading libraries that build on Rust's async-await
+feature (which operates quite differently than Javascript's
+async-await, in that it doesn't cause every aync function call to
+schedule a distinct task).
+[tokio]: https://tokio.rs/
+[async-std]: https://async.rs/
+That said, I am satisfied with the performance results as they are in
+the current revision.
+## Minor notes and typos
+Several figures used the `with` keyword. I deduced that `with(foo)`
+permits one to write `bar` instead of `foo.bar`. It seems worth
+introducing. Apologies if this is stated in the paper, if so I missed
+it.
+On page 20, section 6.3, "external scheduling and vice versus" should be
+"external scheduling and vice versa".
+On page 5, section 2.3, the paper states "we content" but it should be
+"we contend".
+Reviewing: Editor
+A few small comments in addition to those of the referees.
+Page 1. I don't believe that it s fair to imply that Scala is  "research vehicle" as it is used by major players, Twitter being the most prominent example.
+Page 15. Must Cforall threads start after construction (e.g. see your example on page 15, line 21)? I can think of examples where it is not desirable that threads start immediately after construction, e.g. a game with N players, each of whom is expensive to create, but all of whom should be started at the same time.
+Page 18, line 17: is using
+Date: Tue, 16 Jun 2020 13:45:03 +0000
+From: Aaron Thomas <onbehalfof@manuscriptcentral.com>
+Reply-To: speoffice@wiley.com
+To: tdelisle@uwaterloo.ca, pabuhr@uwaterloo.ca
+Subject: SPE-19-0219.R2 successfully submitted
+-Jun-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.R2.  Please mention this number in all future correspondence regarding this submission.
+You can view the status of your manuscript at any time by checking your Author Center after logging into https://mc.manuscriptcentral.com/spe.  If you have difficulty using this site, please click the 'Get Help Now' link at the top right corner of the site.
+Thank you for submitting your manuscript to Software: Practice and Experience.
+Sincerely,
+Software: Practice and Experience Editorial Office

doc/proposals/vtable.md

-              r07d867b
+              r22f94a4
 default is provided or not, the second syntax can be used to pick a
 parameter on instantiation.
+### Extension: Object Access
+This requires that the resolution scope (see below) is at the type level or
+has explicate points with names. These are the tables and table names used
+here.
+The system already knows where to find the virtual table and the object. If
+the tables have particular identities, or on the user side names, then it is
+meaningful to check if a binding virtual table is the same* as another. The
+main use of this is virtual table declarations also give the type they bind
+and if a binding table matches a known table then the underlyind object in the
+trait object must be of that type.
+* By identity, by value would work and in some senses be more flexiable. But
+  it would be slower and refering to further away functions would be harder.
+This gives one of the main new features of the hierarchical use of virtual
+tables (see below); the ability to recover the underlying object. Or a pointer
+of the approprate type it which both reflects the implementation and gives a
+convenent way to encode the boolean/conditional aspect of the operation which
+is that a different virtual table might be in use.
+There are two general ways to reperent this; a cast or a field access. The
+cast is traditional and would definitely fit if a single pointer repersents
+a trait object with the virtual table as part of the object. However for a
+double pointer field access might be more approprate. By this system though
+it is not the type that is used as the identifier but the virtual table. If
+there is one table per type than it becomes equivilant again. Otherwise the
+table has to be used as the identifier and the type is just a result of that
+which seems important for syntax.
 Hierarchy
 …
 be used in only some of the declarations.
     trait combiner fee = (summation_instance, sum);
+    trait combiner fee = {summation_instance, sum};
     trait combiner foe = summation_instance;

doc/theses/thierry_delisle_PhD/code/relaxed_list.cpp

-              r07d867b
+              r22f94a4
+#include "relaxed_list.hpp"
+#if !defined(LIST_VARIANT_HPP)
+#define LIST_VARIANT_HPP "relaxed_list.hpp"
+#endif
+#include LIST_VARIANT_HPP
+#if !defined(LIST_VARIANT)
+#error not variant selected
+#endif
 #include <array>
 …
 template<>
 thread_local relaxed_list<Node>::TLS relaxed_list<Node>::tls = {};
+thread_local LIST_VARIANT<Node>::TLS LIST_VARIANT<Node>::tls = {};
 template<>
 relaxed_list<Node> * relaxed_list<Node>::head = nullptr;
+std::atomic_uint32_t LIST_VARIANT<Node>::ticket = { 0 };
 #ifndef NO_STATS
 template<>
 relaxed_list<Node>::GlobalStats relaxed_list<Node>::global_stats = {};
+LIST_VARIANT<Node>::GlobalStats LIST_VARIANT<Node>::global_stats = {};
 #endif
 …
         size_t valmax = 0;
         size_t valmin = 100000000ul;
+        struct {
+                size_t val = 0;
+                size_t cnt = 0;
+        } comp;
+        struct {
+                size_t val = 0;
+                size_t cnt = 0;
+        } subm;
 };
 …
         std::atomic_size_t valmax = { 0 };
         std::atomic_size_t valmin = { 100000000ul };
+        struct {
+                std::atomic_size_t val = { 0 };
+                std::atomic_size_t cnt = { 0 };
+        } comp;
+        struct {
+                std::atomic_size_t val = { 0 };
+                std::atomic_size_t cnt = { 0 };
+        } subm;
 };
 …
         global.crc_out += local.crc_out;
+        global.comp.val += local.comp.val;
+        global.comp.cnt += local.comp.cnt;
+        global.subm.val += local.subm.val;
+        global.subm.cnt += local.subm.cnt;
         atomic_max(global.valmax, local.valmax);
         atomic_min(global.valmin, local.valmin);
         relaxed_list<Node>::stats_tls_tally();
+        LIST_VARIANT<Node>::stats_tls_tally();
+}
 …
         auto before = Clock::now();
         barrier.wait(0);
+        bool is_tty = isatty(STDOUT_FILENO);
         while(true) {
 …
                         break;
+                }
+                std::cout << "\r" << std::setprecision(4) << durr.count();
+                std::cout.flush();
+                if(is_tty) {
+                        std::cout << "\r" << std::setprecision(4) << durr.count();
+                        std::cout.flush();
+                }
+        }
 …
         auto dur_nano = duration_cast<std::nano>(1.0);
+        if(global.valmax != 0) {
+                std::cout << "Max runs      : " << global.valmax << "\n";
+                std::cout << "Min runs      : " << global.valmin << "\n";
+        }
+        if(global.comp.cnt != 0) {
+                std::cout << "Submit count  : " << global.subm.cnt << "\n";
+                std::cout << "Submit average: " << ((double(global.subm.val)) / global.subm.cnt) << "\n";
+                std::cout << "Complete count: " << global.comp.cnt << "\n";
+                std::cout << "Complete avg  : " << ((double(global.comp.val)) / global.comp.cnt) << "\n";
+        }
         std::cout << "Duration      : " << duration << "s\n";
         std::cout << "ns/Op         : " << ( dur_nano / ops_thread )<< "\n";
 …
         std::cout << "Ops/sec       : " << ops_sec << "\n";
         std::cout << "Total ops     : " << ops << "(" << global.in << "i, " << global.out << "o, " << global.empty << "e)\n";
-        if(global.valmax != 0) {
-                std::cout << "Max runs      : " << global.valmax << "\n";
-                std::cout << "Min runs      : " << global.valmin << "\n";
+        }
         #ifndef NO_STATS
                 relaxed_list<Node>::stats_print(std::cout);
+                LIST_VARIANT<Node>::stats_print(std::cout);
         #endif
+}
 …
         unsigned nslots,
         local_stat_t & local,
         relaxed_list<Node> & list
+        LIST_VARIANT<Node> & list
 ) {
         while(__builtin_expect(!done.load(std::memory_order_relaxed), true)) {
 …
         std::cout << "Initializing ";
         size_t npushed = 0;
         relaxed_list<Node> list = { nthread * nqueues };
+        LIST_VARIANT<Node> list = { nthread, nqueues };
+        {
                 Node** all_nodes[nthread];
 …
         unsigned nnodes,
         local_stat_t & local,
         relaxed_list<Node> & list
+        LIST_VARIANT<Node> & list
 ) {
         Node * nodes[nnodes];
 …
         std::cout << "Initializing ";
         // List being tested
         relaxed_list<Node> list = { nthread * nqueues };
+        LIST_VARIANT<Node> list = { nthread, nqueues };
+        {
                 enable_stats = true;
 …
                 enable_stats = false;
+        }
+        print_stats(duration, nthread, global);
+}
+// ================================================================================================
+struct __attribute__((aligned(64))) Slot {
+        Node * volatile node;
+};
+__attribute__((noinline)) void runProducer_body(
+        std::atomic<bool>& done,
+        Random & rand,
+        Slot * slots,
+        int nslots,
+        local_stat_t & local,
+        LIST_VARIANT<Node> & list
+) {
+        while(__builtin_expect(!done.load(std::memory_order_relaxed), true)) {
+                Node * node = list.pop();
+                if(!node) {
+                        local.empty ++;
+                        continue;
+                }
+                local.crc_out += node->value;
+                local.out++;
+                if(node->id == 0) {
+                        unsigned cnt = 0;
+                        for(int i = 0; i < nslots; i++) {
+                                Node * found = __atomic_exchange_n( &slots[i].node, nullptr, __ATOMIC_SEQ_CST );
+                                if( found ) {
+                                        local.crc_in += found->value;
+                                        local.in++;
+                                        cnt++;
+                                        list.push( found );
+                                }
+                        }
+                        local.crc_in += node->value;
+                        local.in++;
+                        list.push( node );
+                        local.comp.cnt++;
+                        local.comp.val += cnt;
+                }
+                else {
+                        unsigned len = 0;
+                        while(true) {
+                                auto off = rand.next();
+                                for(int i = 0; i < nslots; i++) {
+                                        Node * expected = nullptr;
+                                        int idx = (i + off) % nslots;
+                                        Slot & slot = slots[ idx ];
+                                        if(
+                                                slot.node == nullptr &&
+                                                __atomic_compare_exchange_n( &slot.node, &expected, node, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST )
+                                        ) {
+                                                local.subm.cnt++;
+                                                local.subm.val += len;
+                                                goto LOOP;
+                                        }
+                                        assert( expected != node );
+                                        len++;
+                                }
+                        }
+                }
+                LOOP:;
+        }
+}
+void runProducer(unsigned nthread, unsigned nqueues, double duration, unsigned nnodes) {
+        std::cout << "Producer Benchmark" << std::endl;
+        // Barrier for synchronization
+        barrier_t barrier(nthread + 1);
+        // Data to check everything is OK
+        global_stat_t global;
+        // Flag to signal termination
+        std::atomic_bool done  = { false };
+        std::cout << "Initializing ";
+        int nslots = nnodes * 4;
+        Slot * slots = new Slot[nslots];
+        std::cout << nnodes << " nodes (" << nslots << " slots)" << std::endl;
+        // List being tested
+        LIST_VARIANT<Node> list = { nthread, nqueues };
+        {
+                Random rand(rdtscl());
+                for(unsigned i = 0; i < nnodes; i++) {
+                        Node * node = new Node(rand.next() % 100);
+                        node->id = i;
+                        global.crc_in += node->value;
+                        list.push(node);
+                }
+                for(int i = 0; i < nslots; i++) {
+                        slots[i].node = nullptr;
+                }
+        }
+        {
+                enable_stats = true;
+                std::thread * threads[nthread];
+                unsigned i = 1;
+                for(auto & t : threads) {
+                        t = new std::thread([&done, &list, &barrier, &global, slots, nslots](unsigned tid) {
+                                Random rand(tid + rdtscl());
+                                local_stat_t local;
+                                barrier.wait(tid);
+                                // EXPERIMENT START
+                                runProducer_body(done, rand, slots, nslots, local, list);
+                                // EXPERIMENT END
+                                barrier.wait(tid);
+                                tally_stats(global, local);
+                        }, i++);
+                }
+                waitfor(duration, barrier, done);
+                for(auto t : threads) {
+                        t->join();
+                        delete t;
+                }
+                enable_stats = false;
+        }
+        {
+                while(Node * node = list.pop()) {
+                        global.crc_out += node->value;
+                        delete node;
+                }
+                for(int i = 0; i < nslots; i++) {
+                        delete slots[i].node;
+                }
+                delete [] slots;
+        }
 …
         unsigned nnodes,
         local_stat_t & local,
         relaxed_list<Node> & list
+        LIST_VARIANT<Node> & list
 ) {
         Node * nodes[nnodes];
 …
         // List being tested
         relaxed_list<Node> list = { nthread * nqueues };
+        LIST_VARIANT<Node> list = { nthread, nqueues };
+        {
                 enable_stats = true;
 …
         print_stats(duration, nthread, global);
         save_fairness(data_out.get(), 100, nthread, width, length, output);
+        // save_fairness(data_out.get(), 100, nthread, width, length, output);
+}
 …
                 Churn,
                 PingPong,
+                Producer,
                 Fairness,
                 NONE
 …
                                 case PingPong:
                                         nnodes = 1;
-                                        nslots = 1;
                                         switch(argc - optind) {
                                         case 0: break;
 …
                                                 break;
                                         default:
+                                                std::cerr << "'PingPong' benchmark doesn't accept more than 2 extra arguments" << std::endl;
+                                                std::cerr << "'PingPong' benchmark doesn't accept more than 1 extra arguments" << std::endl;
+                                                goto usage;
+                                        }
+                                        break;
+                                case Producer:
+                                        nnodes = 32;
+                                        switch(argc - optind) {
+                                        case 0: break;
+                                        case 1:
+                                                try {
+                                                        arg = optarg = argv[optind];
+                                                        nnodes = stoul(optarg, &len);
+                                                        if(len != arg.size()) { throw std::invalid_argument(""); }
+                                                } catch(std::invalid_argument &) {
+                                                        std::cerr << "Number of nodes must be a positive integer, was " << arg << std::endl;
+                                                        goto usage;
+                                                }
+                                                break;
+                                        default:
+                                                std::cerr << "'Producer' benchmark doesn't accept more than 1 extra arguments" << std::endl;
                                                 goto usage;
+                                        }
 …
                                         break;
+                                }
+                                if(iequals(arg, "producer")) {
+                                        benchmark = Producer;
+                                        break;
+                                }
                                 if(iequals(arg, "fairness")) {
                                         benchmark = Fairness;
 …
                                 std::cerr << "Usage: " << argv[0] << ": [options] -b churn [NNODES] [NSLOTS = NNODES]" << std::endl;
                                 std::cerr << "  or:  " << argv[0] << ": [options] -b pingpong [NNODES]" << std::endl;
+                                std::cerr << "  or:  " << argv[0] << ": [options] -b producer [NNODES]" << std::endl;
                                 std::cerr << std::endl;
                                 std::cerr << "  -d, --duration=DURATION  Duration of the experiment, in seconds" << std::endl;
 …
         std::cout << "Running " << nthreads << " threads (" << (nthreads * nqueues) << " queues) for " << duration << " seconds" << std::endl;
+        std::cout << "Relaxed list variant: " << LIST_VARIANT<Node>::name() << std::endl;
         switch(benchmark) {
                 case Churn:
 …
                 case PingPong:
                         runPingPong(nthreads, nqueues, duration, nnodes);
+                        break;
+                case Producer:
+                        runProducer(nthreads, nqueues, duration, nnodes);
                         break;
                 case Fairness:
 …
+}
 void save_fairness(const int data[], int factor, unsigned nthreads, size_t columns, size_t rows, const std::string & output) {
         std::ofstream os(output);
         os << "<html>\n";
         os << "<head>\n";
         os << "<style>\n";
         os << "</style>\n";
         os << "</head>\n";
         os << "<body>\n";
         os << "<table style=\"width=100%\">\n";
         size_t idx = 0;
         for(size_t r = 0ul; r < rows; r++) {
                 os << "<tr>\n";
                 for(size_t c = 0ul; c < columns; c++) {
                         os << "<td class=\"custom custom" << data[idx] << "\"></td>\n";
                         idx++;
+                }
                 os << "</tr>\n";
+        }
         os << "</table>\n";
         os << "</body>\n";
         os << "</html>\n";
         os << std::endl;
+}
 #include <png.h>
 #include <setjmp.h>
+// void save_fairness(const int data[], int factor, unsigned nthreads, size_t columns, size_t rows, const std::string & output) {
+//      std::ofstream os(output);
+//      os << "<html>\n";
+//      os << "<head>\n";
+//      os << "<style>\n";
+//      os << "</style>\n";
+//      os << "</head>\n";
+//      os << "<body>\n";
+//      os << "<table style=\"width=100%\">\n";
+//      size_t idx = 0;
+//      for(size_t r = 0ul; r < rows; r++) {
+//              os << "<tr>\n";
+//              for(size_t c = 0ul; c < columns; c++) {
+//                      os << "<td class=\"custom custom" << data[idx] << "\"></td>\n";
+//                      idx++;
+//              }
+//              os << "</tr>\n";
+//      }
+//      os << "</table>\n";
+//      os << "</body>\n";
+//      os << "</html>\n";
+//      os << std::endl;
+// }
+// #include <png.h>
+// #include <setjmp.h>
 /*

doc/theses/thierry_delisle_PhD/code/relaxed_list.hpp

-              r07d867b
+              r22f94a4
 #pragma once
+#define LIST_VARIANT relaxed_list
+#define VANILLA 0
+#define SNZI 1
+#define BITMASK 2
+#define DISCOVER 3
+#define SNZM 4
+#define BIAS 5
+#define BACK 6
+#define BACKBIAS 7
+#ifndef VARIANT
+#define VARIANT VANILLA
+#endif
 #ifndef NO_STATS
 …
 #endif
+#include <cmath>
+#include <functional>
 #include <memory>
 #include <mutex>
+#include <thread>
 #include <type_traits>
 #include "assert.hpp"
 #include "utils.hpp"
+#include "links.hpp"
+#include "snzi.hpp"
+#include "snzi-packed.hpp"
+#include "snzm.hpp"
 using namespace std;
-struct spinlock_t {
-        std::atomic_bool ll = { false };
-        inline void lock() {
-                while( __builtin_expect(ll.exchange(true),false) ) {
-                        while(ll.load(std::memory_order_relaxed))
-                                asm volatile("pause");
+                }
+        }
-        inline bool try_lock() {
-                return false == ll.exchange(true);
+        }
-        inline void unlock() {
-                ll.store(false, std::memory_order_release);
+        }
-        inline explicit operator bool() {
-                return ll.load(std::memory_order_relaxed);
+        }
-};
-static inline bool bts(std::atomic_size_t & target, size_t bit ) {
-        //*
-        int result = 0;
-        asm volatile(
-                "LOCK btsq %[bit], %[target]\n\t"
-                :"=@ccc" (result)
-                : [target] "m" (target), [bit] "r" (bit)
-        );
-        return result != 0;
-        /*/
-        size_t mask = 1ul << bit;
-        size_t ret = target.fetch_or(mask, std::memory_order_relaxed);
-        return (ret & mask) != 0;
-        //*/
+}
-static inline bool btr(std::atomic_size_t & target, size_t bit ) {
-        //*
-        int result = 0;
-        asm volatile(
-                "LOCK btrq %[bit], %[target]\n\t"
-                :"=@ccc" (result)
-                : [target] "m" (target), [bit] "r" (bit)
-        );
-        return result != 0;
-        /*/
-        size_t mask = 1ul << bit;
-        size_t ret = target.fetch_and(~mask, std::memory_order_relaxed);
-        return (ret & mask) != 0;
-        //*/
+}
-extern bool enable_stats;
 struct pick_stat {
 …
                 size_t attempt = 0;
                 size_t success = 0;
+                size_t local = 0;
         } push;
         struct {
 …
                 size_t success = 0;
                 size_t mask_attempt = 0;
+                size_t mask_reset = 0;
+                size_t local = 0;
         } pop;
 };
 …
 template<typename node_t>
-struct _LinksFields_t {
-        node_t * prev = nullptr;
-        node_t * next = nullptr;
-        unsigned long long ts = 0;
-};
-template<typename node_t>
 class __attribute__((aligned(128))) relaxed_list {
         static_assert(std::is_same<decltype(node_t::_links), _LinksFields_t<node_t>>::value, "Node must have a links field");
 public:
+        relaxed_list(unsigned numLists)
+                : lists(new intrusive_queue_t[numLists])
+                , numLists(numLists)
+        static const char * name() {
+                const char * names[] = {
+                        "RELAXED: VANILLA",
+                        "RELAXED: SNZI",
+                        "RELAXED: BITMASK",
+                        "RELAXED: SNZI + DISCOVERED MASK",
+                        "RELAXED: SNZI + MASK",
+                        "RELAXED: SNZI + LOCAL BIAS",
+                        "RELAXED: SNZI + REVERSE RNG",
+                        "RELAXED: SNZI + LOCAL BIAS + REVERSE RNG"
+                };
+                return names[VARIANT];
+        }
+        relaxed_list(unsigned numThreads, unsigned numQueues)
+                : numLists(numThreads * numQueues)
+                , lists(new intrusive_queue_t<node_t>[numLists])
+                #if VARIANT == SNZI || VARIANT == BACK
+                        , snzi( std::log2( numLists / (2 * numQueues) ), 2 )
+                #elif VARIANT == BIAS || VARIANT == BACKBIAS
+                        #ifdef SNZI_PACKED
+                                , snzi( std::ceil( std::log2(numLists) ) )
+                        #else
+                                , snzi( std::log2( numLists / (2 * numQueues) ), 2 )
+                        #endif
+                #elif VARIANT == SNZM || VARIANT == DISCOVER
+                        , snzm( numLists )
+                #endif
+        {
                 assertf(7 * 8 * 8 >= numLists, "List currently only supports 448 sublists");
-                // assert(sizeof(*this) == 128);
                 std::cout << "Constructing Relaxed List with " << numLists << std::endl;
-                #ifndef NO_STATS
-                        if(head) this->next = head;
-                        head = this;
-                #endif
+        }
 …
                 while(true) {
                         // Pick a random list
                         unsigned i = tls.rng.next() % numLists;
+                        unsigned i = idx_from_r(tls.rng1.next(), VARIANT == BIAS || VARIANT == BACKBIAS);
                         #ifndef NO_STATS
 …
                         if( !lists[i].lock.try_lock() ) continue;
+                        __attribute__((unused)) int num = numNonEmpty;
+                        #if VARIANT == VANILLA || VARIANT == BITMASK
+                                __attribute__((unused)) int num = numNonEmpty;
+                        #endif
                         // Actually push it
                         if(lists[i].push(node)) {
+                                numNonEmpty++;
+                                size_t qword = i >> 6ull;
+                                size_t bit   = i & 63ull;
+                                assertf((list_mask[qword] & (1ul << bit)) == 0, "Before set %zu:%zu (%u), %zx & %zx", qword, bit, i, list_mask[qword].load(), (1ul << bit));
+                                __attribute__((unused)) bool ret = bts(list_mask[qword], bit);
+                                assert(!ret);
+                                assertf((list_mask[qword] & (1ul << bit)) != 0, "After set %zu:%zu (%u), %zx & %zx", qword, bit, i, list_mask[qword].load(), (1ul << bit));
+                        }
+                        assert(numNonEmpty <= (int)numLists);
+                                #if VARIANT == DISCOVER
+                                        size_t qword = i >> 6ull;
+                                        size_t bit   = i & 63ull;
+                                        assert(qword == 0);
+                                        bts(tls.mask, bit);
+                                        snzm.arrive(i);
+                                #elif VARIANT == SNZI || VARIANT == BIAS
+                                        snzi.arrive(i);
+                                #elif VARIANT == BACK || VARIANT == BACKBIAS
+                                        snzi.arrive(i);
+                                        tls.rng2.set_raw_state( tls.rng1.get_raw_state());
+                                #elif VARIANT == SNZM
+                                        snzm.arrive(i);
+                                #elif VARIANT == BITMASK
+                                        numNonEmpty++;
+                                        size_t qword = i >> 6ull;
+                                        size_t bit   = i & 63ull;
+                                        assertf((list_mask[qword] & (1ul << bit)) == 0, "Before set %zu:%zu (%u), %zx & %zx", qword, bit, i, list_mask[qword].load(), (1ul << bit));
+                                        __attribute__((unused)) bool ret = bts(list_mask[qword], bit);
+                                        assert(!ret);
+                                        assertf((list_mask[qword] & (1ul << bit)) != 0, "After set %zu:%zu (%u), %zx & %zx", qword, bit, i, list_mask[qword].load(), (1ul << bit));
+                                #else
+                                        numNonEmpty++;
+                                #endif
+                        }
+                        #if VARIANT == VANILLA || VARIANT == BITMASK
+                                assert(numNonEmpty <= (int)numLists);
+                        #endif
                         // Unlock and return
 …
                         #ifndef NO_STATS
                                 tls.pick.push.success++;
+                                tls.empty.push.value += num;
+                                tls.empty.push.count += 1;
+                                #if VARIANT == VANILLA || VARIANT == BITMASK
+                                        tls.empty.push.value += num;
+                                        tls.empty.push.count += 1;
+                                #endif
                         #endif
                         return;
 …
         __attribute__((noinline, hot)) node_t * pop() {
+                #if !defined(NO_BITMASK)
+                        // for(int r = 0; r < 10 && numNonEmpty != 0; r++) {
+                        //      // Pick two lists at random
+                        //      unsigned i = tls.rng.next() % numLists;
+                        //      unsigned j = tls.rng.next() % numLists;
+                        //      if(auto node = try_pop(i, j)) return node;
+                        // }
+                #if VARIANT == DISCOVER
+                        assert(numLists <= 64);
+                        while(snzm.query()) {
+                                tls.pick.pop.mask_attempt++;
+                                unsigned i, j;
+                                {
+                                        // Pick first list totally randomly
+                                        i = tls.rng1.next() % numLists;
+                                        // Pick the other according to the bitmask
+                                        unsigned r = tls.rng1.next();
+                                        size_t mask = tls.mask.load(std::memory_order_relaxed);
+                                        if(mask == 0) {
+                                                tls.pick.pop.mask_reset++;
+                                                mask = (1U << numLists) - 1;
+                                                tls.mask.store(mask, std::memory_order_relaxed);
+                                        }
+                                        unsigned b = rand_bit(r, mask);
+                                        assertf(b < 64, "%zu %u", mask, b);
+                                        j = b;
+                                        assert(j < numLists);
+                                }
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                #elif VARIANT == SNZI
+                        while(snzi.query()) {
+                                // Pick two lists at random
+                                int i = tls.rng1.next() % numLists;
+                                int j = tls.rng1.next() % numLists;
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                #elif VARIANT == BACK
+                        while(snzi.query()) {
+                                // Pick two lists at random
+                                int i = tls.rng2.prev() % numLists;
+                                int j = tls.rng2.prev() % numLists;
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                #elif VARIANT == BACKBIAS
+                        while(snzi.query()) {
+                                // Pick two lists at random
+                                int i = idx_from_r(tls.rng2.prev(), true);
+                                int j = idx_from_r(tls.rng2.prev(), true);
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                #elif VARIANT == BIAS
+                        while(snzi.query()) {
+                                // Pick two lists at random
+                                unsigned ri = tls.rng1.next();
+                                unsigned i;
+                                unsigned j = tls.rng1.next();
+                                if(0 == (ri & 0xF)) {
+                                        i = (ri >> 4) % numLists;
+                                } else {
+                                        i = tls.my_queue + ((ri >> 4) % 4);
+                                        j = tls.my_queue + ((j >> 4) % 4);
+                                        tls.pick.pop.local++;
+                                }
+                                i %= numLists;
+                                j %= numLists;
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                #elif VARIANT == SNZM
+                        //*
+                        while(snzm.query()) {
+                                tls.pick.pop.mask_attempt++;
+                                unsigned i, j;
+                                {
+                                        // Pick two random number
+                                        unsigned ri = tls.rng1.next();
+                                        unsigned rj = tls.rng1.next();
+                                        // Pick two nodes from it
+                                        unsigned wdxi = ri & snzm.mask;
+                                        // unsigned wdxj = rj & snzm.mask;
+                                        // Get the masks from the nodes
+                                        // size_t maski = snzm.masks(wdxi);
+                                        size_t maskj = snzm.masks(wdxj);
+                                        if(maski == 0 && maskj == 0) continue;
+                                        #if defined(__BMI2__)
+                                                uint64_t idxsi = _pext_u64(snzm.indexes, maski);
+                                                // uint64_t idxsj = _pext_u64(snzm.indexes, maskj);
+                                                auto pi = __builtin_popcountll(maski);
+                                                // auto pj = __builtin_popcountll(maskj);
+                                                ri = pi ? ri & ((pi >> 3) - 1) : 0;
+                                                rj = pj ? rj & ((pj >> 3) - 1) : 0;
+                                                unsigned bi = (idxsi >> (ri << 3)) & 0xff;
+                                                unsigned bj = (idxsj >> (rj << 3)) & 0xff;
+                                        #else
+                                                unsigned bi = rand_bit(ri >> snzm.depth, maski);
+                                                unsigned bj = rand_bit(rj >> snzm.depth, maskj);
+                                        #endif
+                                        i = (bi << snzm.depth) | wdxi;
+                                        j = (bj << snzm.depth) | wdxj;
+                                        /* paranoid */ assertf(i < numLists, "%u %u", bj, wdxi);
+                                        /* paranoid */ assertf(j < numLists, "%u %u", bj, wdxj);
+                                }
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                        /*/
+                        while(snzm.query()) {
+                                // Pick two lists at random
+                                int i = tls.rng1.next() % numLists;
+                                int j = tls.rng1.next() % numLists;
+                                if(auto node = try_pop(i, j)) return node;
+                        }
+                        //*/
+                #elif VARIANT == BITMASK
                         int nnempty;
                         while(0 != (nnempty = numNonEmpty)) {
                                 tls.pick.pop.mask_attempt++;
                                 unsigned i, j;
-                                // if( numLists < 4 || (numLists / nnempty) < 4 ) {
-                                //      // Pick two lists at random
-                                //      i = tls.rng.next() % numLists;
-                                //      j = tls.rng.next() % numLists;
-                                // } else
+                                {
-                                        #ifndef NO_STATS
-                                                // tls.pick.push.mask_attempt++;
-                                        #endif
                                         // Pick two lists at random
                                         unsigned num = ((numLists - 1) >> 6) + 1;
                                         unsigned ri = tls.rng.next();
                                         unsigned rj = tls.rng.next();
+                                        unsigned ri = tls.rng1.next();
+                                        unsigned rj = tls.rng1.next();
                                         unsigned wdxi = (ri >> 6u) % num;
 …
                         while(numNonEmpty != 0) {
                                 // Pick two lists at random
                                 int i = tls.rng.next() % numLists;
                                 int j = tls.rng.next() % numLists;
+                                int i = tls.rng1.next() % numLists;
+                                int j = tls.rng1.next() % numLists;
                                 if(auto node = try_pop(i, j)) return node;
 …
                 #ifndef NO_STATS
                         tls.pick.pop.attempt++;
+                #endif
+                #if VARIANT == DISCOVER
+                        if(lists[i].ts() > 0) bts(tls.mask, i); else btr(tls.mask, i);
+                        if(lists[j].ts() > 0) bts(tls.mask, j); else btr(tls.mask, j);
                 #endif
 …
                 if( !list.lock.try_lock() ) return nullptr;
+                __attribute__((unused)) int num = numNonEmpty;
+                #if VARIANT == VANILLA || VARIANT == BITMASK
+                        __attribute__((unused)) int num = numNonEmpty;
+                #endif
                 // If list is empty, unlock and retry
 …
                 if(emptied) {
+                        numNonEmpty--;
+                        size_t qword = w >> 6ull;
+                        size_t bit   = w & 63ull;
+                        assert((list_mask[qword] & (1ul << bit)) != 0);
+                        __attribute__((unused)) bool ret = btr(list_mask[qword], bit);
+                        assert(ret);
+                        assert((list_mask[qword] & (1ul << bit)) == 0);
+                        #if VARIANT == DISCOVER
+                                size_t qword = w >> 6ull;
+                                size_t bit   = w & 63ull;
+                                assert(qword == 0);
+                                __attribute__((unused)) bool ret = btr(tls.mask, bit);
+                                snzm.depart(w);
+                        #elif VARIANT == SNZI || VARIANT == BIAS || VARIANT == BACK || VARIANT == BACKBIAS
+                                snzi.depart(w);
+                        #elif VARIANT == SNZM
+                                snzm.depart(w);
+                        #elif VARIANT == BITMASK
+                                numNonEmpty--;
+                                size_t qword = w >> 6ull;
+                                size_t bit   = w & 63ull;
+                                assert((list_mask[qword] & (1ul << bit)) != 0);
+                                __attribute__((unused)) bool ret = btr(list_mask[qword], bit);
+                                assert(ret);
+                                assert((list_mask[qword] & (1ul << bit)) == 0);
+                        #else
+                                numNonEmpty--;
+                        #endif
+                }
                 // Unlock and return
                 list.lock.unlock();
+                assert(numNonEmpty >= 0);
+                #if VARIANT == VANILLA || VARIANT == BITMASK
+                        assert(numNonEmpty >= 0);
+                #endif
                 #ifndef NO_STATS
                         tls.pick.pop.success++;
+                        tls.empty.pop.value += num;
+                        tls.empty.pop.count += 1;
+                        #if VARIANT == VANILLA || VARIANT == BITMASK
+                                tls.empty.pop.value += num;
+                                tls.empty.pop.count += 1;
+                        #endif
                 #endif
                 return node;
+        }
+private:
+        class __attribute__((aligned(128))) intrusive_queue_t {
+        public:
+                typedef spinlock_t lock_t;
+                friend class relaxed_list<node_t>;
+                struct stat {
+                        ssize_t diff = 0;
+                        size_t  push = 0;
+                        size_t  pop  = 0;
+                        // size_t value = 0;
+                        // size_t count = 0;
+                };
+        private:
+                struct sentinel_t {
+                        _LinksFields_t<node_t> _links;
+                };
+                lock_t lock;
+                sentinel_t before;
+                sentinel_t after;
+                #ifndef NO_STATS
+                        stat s;
+                #endif
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Winvalid-offsetof"
+                static constexpr auto fields_offset = offsetof( node_t, _links );
+#pragma GCC diagnostic pop
+        public:
+                intrusive_queue_t()
+                        : before{{ nullptr, tail() }}
+                        , after {{ head(), nullptr }}
+                {
+                        /* paranoid */ assert((reinterpret_cast<uintptr_t>( head() ) + fields_offset) == reinterpret_cast<uintptr_t>(&before));
+                        /* paranoid */ assert((reinterpret_cast<uintptr_t>( tail() ) + fields_offset) == reinterpret_cast<uintptr_t>(&after ));
+                        /* paranoid */ assert(head()->_links.prev == nullptr);
+                        /* paranoid */ assert(head()->_links.next == tail() );
+                        /* paranoid */ assert(tail()->_links.next == nullptr);
+                        /* paranoid */ assert(tail()->_links.prev == head() );
+                        /* paranoid */ assert(sizeof(*this) == 128);
+                        /* paranoid */ assert((intptr_t(this) % 128) == 0);
+        inline unsigned idx_from_r(unsigned r, bool bias) {
+                unsigned i;
+                if(bias) {
+                        if(0 == (r & 0x3F)) {
+                                i = r >> 6;
+                        } else {
+                                i = tls.my_queue + ((r >> 6) % 4);
+                                tls.pick.push.local++;
+                        }
+                } else {
+                        i = r;
+                }
+                ~intrusive_queue_t() = default;
+                inline node_t * head() const {
+                        node_t * rhead = reinterpret_cast<node_t *>(
+                                reinterpret_cast<uintptr_t>( &before ) - fields_offset
+                        );
+                        assert(rhead);
+                        return rhead;
+                }
+                inline node_t * tail() const {
+                        node_t * rtail = reinterpret_cast<node_t *>(
+                                reinterpret_cast<uintptr_t>( &after ) - fields_offset
+                        );
+                        assert(rtail);
+                        return rtail;
+                }
+                inline bool push(node_t * node) {
+                        assert(lock);
+                        assert(node->_links.ts != 0);
+                        node_t * tail = this->tail();
+                        node_t * prev = tail->_links.prev;
+                        // assertf(node->_links.ts >= prev->_links.ts,
+                        //      "New node has smaller timestamp: %llu < %llu", node->_links.ts, prev->_links.ts);
+                        node->_links.next = tail;
+                        node->_links.prev = prev;
+                        prev->_links.next = node;
+                        tail->_links.prev = node;
+                        #ifndef NO_STATS
+                                if(enable_stats) {
+                                        s.diff++;
+                                        s.push++;
+                                }
+                        #endif
+                        if(before._links.ts == 0l) {
+                                before._links.ts = node->_links.ts;
+                                assert(node->_links.prev == this->head());
+                                return true;
+                        }
+                        return false;
+                }
+                inline std::pair<node_t *, bool> pop() {
+                        assert(lock);
+                        node_t * head = this->head();
+                        node_t * tail = this->tail();
+                        node_t * node = head->_links.next;
+                        node_t * next = node->_links.next;
+                        if(node == tail) return {nullptr, false};
+                        head->_links.next = next;
+                        next->_links.prev = head;
+                        #ifndef NO_STATS
+                                if(enable_stats) {
+                                        s.diff--;
+                                        s.pop ++;
+                                }
+                        #endif
+                        if(next == tail) {
+                                before._links.ts = 0l;
+                                return {node, true};
+                        }
+                        else {
+                                assert(next->_links.ts != 0);
+                                before._links.ts = next->_links.ts;
+                                assert(before._links.ts != 0);
+                                return {node, false};
+                        }
+                }
+                long long ts() const {
+                        return before._links.ts;
+                }
+        };
+                return i % numLists;
+        }
 public:
         static __attribute__((aligned(128))) thread_local struct TLS {
+                Random     rng = { int(rdtscl()) };
+                Random     rng1 = { unsigned(std::hash<std::thread::id>{}(std::this_thread::get_id()) ^ rdtscl()) };
+                Random     rng2 = { unsigned(std::hash<std::thread::id>{}(std::this_thread::get_id()) ^ rdtscl()) };
+                unsigned   my_queue = (ticket++) * 4;
                 pick_stat  pick;
                 empty_stat empty;
+                __attribute__((aligned(64))) std::atomic_size_t mask = { 0 };
         } tls;
+private:
+        const unsigned numLists;
+        __attribute__((aligned(64))) std::unique_ptr<intrusive_queue_t<node_t> []> lists;
+private:
+        #if VARIANT == SNZI || VARIANT == BACK
+                snzi_t snzi;
+        #elif VARIANT == BIAS || VARIANT == BACKBIAS
+                #ifdef SNZI_PACKED
+                        snzip_t snzi;
+                #else
+                        snzi_t snzi;
+                #endif
+        #elif VARIANT == SNZM || VARIANT == DISCOVER
+                snzm_t snzm;
+        #else
+                std::atomic_int numNonEmpty  = { 0 };  // number of non-empty lists
+        #endif
+        #if VARIANT == BITMASK
+                std::atomic_size_t list_mask[7] = { {0}, {0}, {0}, {0}, {0}, {0}, {0} }; // which queues are empty
+        #endif
 public:
+        std::atomic_int numNonEmpty  = { 0 };  // number of non-empty lists
+        std::atomic_size_t list_mask[7] = { {0}, {0}, {0}, {0}, {0}, {0}, {0} }; // which queues are empty
+private:
+        __attribute__((aligned(64))) std::unique_ptr<intrusive_queue_t []> lists;
+        const unsigned numLists;
+public:
+        static const constexpr size_t sizeof_queue = sizeof(intrusive_queue_t);
+        static const constexpr size_t sizeof_queue = sizeof(intrusive_queue_t<node_t>);
+        static std::atomic_uint32_t ticket;
 #ifndef NO_STATS
-        static void stats_print(std::ostream & os) {
-                auto it = head;
-                while(it) {
-                        it->stats_print_local(os);
-                        it = it->next;
+                }
+        }
         static void stats_tls_tally() {
                 global_stats.pick.push.attempt += tls.pick.push.attempt;
                 global_stats.pick.push.success += tls.pick.push.success;
+                global_stats.pick.push.local += tls.pick.push.local;
                 global_stats.pick.pop .attempt += tls.pick.pop.attempt;
                 global_stats.pick.pop .success += tls.pick.pop.success;
                 global_stats.pick.pop .mask_attempt += tls.pick.pop.mask_attempt;
+                global_stats.pick.pop .mask_reset += tls.pick.pop.mask_reset;
+                global_stats.pick.pop .local += tls.pick.pop.local;
                 global_stats.qstat.push.value += tls.empty.push.value;
 …
                                 std::atomic_size_t attempt = { 0 };
                                 std::atomic_size_t success = { 0 };
+                                std::atomic_size_t local = { 0 };
                         } push;
                         struct {
 …
                                 std::atomic_size_t success = { 0 };
                                 std::atomic_size_t mask_attempt = { 0 };
+                                std::atomic_size_t mask_reset = { 0 };
+                                std::atomic_size_t local = { 0 };
                         } pop;
                 } pick;
 …
         } global_stats;
+        // Link list of all lists for stats
+        __attribute__((aligned(64))) relaxed_list<node_t> * next = nullptr;
+        static relaxed_list<node_t> * head;
+        void stats_print_local(std::ostream & os ) {
+public:
+        static void stats_print(std::ostream & os ) {
                 std::cout << "----- Relaxed List Stats -----" << std::endl;
+                {
-                        ssize_t diff = 0;
-                        size_t  num  = 0;
-                        ssize_t max  = 0;
-                        for(size_t i = 0; i < numLists; i++) {
-                                const auto & list = lists[i];
-                                diff+= list.s.diff;
-                                num ++;
-                                max  = std::abs(max) > std::abs(list.s.diff) ? max : list.s.diff;
-                                os << "Local Q ops   : " << (list.s.push + list.s.pop) << "(" << list.s.push << "i, " << list.s.pop << "o)\n";
+                        }
-                        os << "Difference   : " << ssize_t(double(diff) / num  ) << " avg\t" << max << "max" << std::endl;
+                }
                 const auto & global = global_stats;
 …
                 double pop_sur  = (100.0 * double(global.pick.pop .success) / global.pick.pop .attempt);
                 double mpop_sur = (100.0 * double(global.pick.pop .success) / global.pick.pop .mask_attempt);
+                os << "Push   Pick % : " << push_sur << "(" << global.pick.push.success << " / " << global.pick.push.attempt << ")\n";
+                os << "Pop    Pick % : " << pop_sur  << "(" << global.pick.pop .success << " / " << global.pick.pop .attempt << ")\n";
+                os << "TryPop Pick % : " << mpop_sur << "(" << global.pick.pop .success << " / " << global.pick.pop .mask_attempt << ")\n";
+                double rpop_sur = (100.0 * double(global.pick.pop .success) / global.pick.pop .mask_reset);
+                double push_len = double(global.pick.push.attempt     ) / global.pick.push.success;
+                double pop_len  = double(global.pick.pop .attempt     ) / global.pick.pop .success;
+                double mpop_len = double(global.pick.pop .mask_attempt) / global.pick.pop .success;
+                double rpop_len = double(global.pick.pop .mask_reset  ) / global.pick.pop .success;
+                os << "Push   Pick   : " << push_sur << " %, len " << push_len << " (" << global.pick.push.attempt      << " / " << global.pick.push.success << ")\n";
+                os << "Pop    Pick   : " << pop_sur  << " %, len " << pop_len  << " (" << global.pick.pop .attempt      << " / " << global.pick.pop .success << ")\n";
+                os << "TryPop Pick   : " << mpop_sur << " %, len " << mpop_len << " (" << global.pick.pop .mask_attempt << " / " << global.pick.pop .success << ")\n";
+                os << "Pop M Reset   : " << rpop_sur << " %, len " << rpop_len << " (" << global.pick.pop .mask_reset   << " / " << global.pick.pop .success << ")\n";
                 double avgQ_push = double(global.qstat.push.value) / global.qstat.push.count;
 …
                 os << "Pop    Avg Qs : " << avgQ_pop  << " (" << global.qstat.pop .count << "ops)\n";
                 os << "Global Avg Qs : " << avgQ      << " (" << (global.qstat.push.count + global.qstat.pop .count) << "ops)\n";
+                os << "Local Push    : " << global.pick.push.local << "\n";
+                os << "Local Pop     : " << global.pick.pop .local << "\n";
+        }
 #endif

doc/theses/thierry_delisle_PhD/code/utils.hpp

-              r07d867b
+              r22f94a4
 };
+// class Random {
+// private:
+//      unsigned int seed;
+// public:
+//      Random(int seed) {
+//              this->seed = seed;
+//      }
+//      /** returns pseudorandom x satisfying 0 <= x < n. **/
+//      unsigned int next() {
+//              seed ^= seed << 6;
+//              seed ^= seed >> 21;
+//              seed ^= seed << 7;
+//              return seed;
+//      }
+// };
+constexpr uint64_t extendedEuclidY(uint64_t a, uint64_t b);
+constexpr uint64_t extendedEuclidX(uint64_t a, uint64_t b){
+    return (b==0) ? 1 : extendedEuclidY(b, a - b * (a / b));
+}
+constexpr uint64_t extendedEuclidY(uint64_t a, uint64_t b){
+    return (b==0) ? 0 : extendedEuclidX(b, a - b * (a / b)) - (a / b) * extendedEuclidY(b, a - b * (a / b));
+}
 class Random {
 private:
+        unsigned int seed;
+        uint64_t x;
+        static constexpr const uint64_t M  = 1ul << 48ul;
+        static constexpr const uint64_t A  = 25214903917;
+        static constexpr const uint64_t C  = 11;
+        static constexpr const uint64_t D  = 16;
 public:
+        Random(int seed) {
+                this->seed = seed;
+        static constexpr const uint64_t m  = M;
+        static constexpr const uint64_t a  = A;
+        static constexpr const uint64_t c  = C;
+        static constexpr const uint64_t d  = D;
+        static constexpr const uint64_t ai = extendedEuclidX(A, M);
+public:
+        Random(unsigned int seed) {
+                this->x = seed * a;
+        }
         /** returns pseudorandom x satisfying 0 <= x < n. **/
         unsigned int next() {
+                seed ^= seed << 6;
+                seed ^= seed >> 21;
+                seed ^= seed << 7;
+                return seed;
+        }
+                //nextx = (a * x + c) % m;
+                x = (A * x + C) & (M - 1);
+                return x >> D;
+        }
+        unsigned int prev() {
+                //prevx = (ainverse * (x - c)) mod m
+                unsigned int r = x >> D;
+                x = ai * (x - C) & (M - 1);
+                return r;
+        }
+        void set_raw_state(uint64_t _x) {
+                this->x = _x;
+        }
+        uint64_t get_raw_state() {
+                return this->x;
+        }
 };
 …
+}
+static inline unsigned rand_bit(unsigned rnum, size_t mask) __attribute__((artificial));
 static inline unsigned rand_bit(unsigned rnum, size_t mask) {
         unsigned bit = mask ? rnum % __builtin_popcountl(mask) : 0;
 …
 #endif
+}
+struct spinlock_t {
+        std::atomic_bool ll = { false };
+        inline void lock() {
+                while( __builtin_expect(ll.exchange(true),false) ) {
+                        while(ll.load(std::memory_order_relaxed))
+                                asm volatile("pause");
+                }
+        }
+        inline bool try_lock() {
+                return false == ll.exchange(true);
+        }
+        inline void unlock() {
+                ll.store(false, std::memory_order_release);
+        }
+        inline explicit operator bool() {
+                return ll.load(std::memory_order_relaxed);
+        }
+};
+static inline bool bts(std::atomic_size_t & target, size_t bit ) {
+        //*
+        int result = 0;
+        asm volatile(
+                "LOCK btsq %[bit], %[target]\n\t"
+                :"=@ccc" (result)
+                : [target] "m" (target), [bit] "r" (bit)
+        );
+        return result != 0;
+        /*/
+        size_t mask = 1ul << bit;
+        size_t ret = target.fetch_or(mask, std::memory_order_relaxed);
+        return (ret & mask) != 0;
+        //*/
+}
+static inline bool btr(std::atomic_size_t & target, size_t bit ) {
+        //*
+        int result = 0;
+        asm volatile(
+                "LOCK btrq %[bit], %[target]\n\t"
+                :"=@ccc" (result)
+                : [target] "m" (target), [bit] "r" (bit)
+        );
+        return result != 0;
+        /*/
+        size_t mask = 1ul << bit;
+        size_t ret = target.fetch_and(~mask, std::memory_order_relaxed);
+        return (ret & mask) != 0;
+        //*/
+}

driver/cc1.cc

-              r07d867b
+              r22f94a4
 // Created On       : Fri Aug 26 14:23:51 2005
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Oct 20 08:14:33 2019
 // Update Count     : 385
+// Last Modified On : Sat May 30 18:09:05 2020
+// Update Count     : 404
 //
 …
 static string o_file;
 static string bprefix;
+static string lang;                                                                             // -x flag
 …
                         if ( prefix( val, "-compiler=" ) ) {
                                 compiler_path = val.substr( 10 );
+                        } else if ( prefix( val, "-x=" ) ) {
+                                lang = val.substr( 3 );
                         } // if
                 } // if
 …
                         } else if ( prefix( val, "-B=" ) ) {            // location of cfa-cpp
                                 bprefix = val.substr( 3 );
+                        } else if ( prefix( val, "-x=" ) ) {            // ignore
                         } else {                                                                        // normal flag for cfa-cpp
                                 args[nargs++] = ( *new string( arg.substr( arg.find_first_of( "=" ) + 1 ) ) ).c_str();
 …
                 args[0] = compiler_path.c_str();
+                suffix( cpp_in, args, nargs );                                  // check suffix
+                if ( lang.size() == 0 ) {
+                        suffix( cpp_in, args, nargs );                          // check suffix
+                } else {
+                        args[nargs++] = "-x";
+                        args[nargs++] = ( *new string( lang.c_str() ) ).c_str();
+                } // if
                 args[nargs++] = cpp_in;
                 if ( o_flag ) {                                                                 // location for output
 …
                 args[0] = compiler_path.c_str();
+                suffix( cpp_in, args, nargs );                                  // check suffix
+                if ( lang.size() == 0 ) {
+                        suffix( cpp_in, args, nargs );                          // check suffix
+                } else {
+                        args[nargs++] = "-x";
+                        args[nargs++] = ( *new string( lang.c_str() ) ).c_str();
+                } // if
                 args[nargs++] = cpp_in;                                                 // input to cpp
                 args[nargs] = nullptr;                                                  // terminate argument list

driver/cfa.cc

-              r07d867b
+              r22f94a4
 // Created On       : Tue Aug 20 13:44:49 2002
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Jan 31 16:48:03 2020
 // Update Count     : 421
+// Last Modified On : Sat May 30 18:28:23 2020
+// Update Count     : 433
 //
 …
                                 args[nargs++] = argv[i];                                // pass argument along
                                 if ( arg == "-o" ) o_file = i;                  // remember file
+                        } else if ( arg == "-XCFA" ) {                          // CFA pass through
+                                i += 1;
+                                if ( i == argc ) continue;                              // next argument available ?
+                                Putenv( argv, argv[i] );
+                                // CFA specific arguments
+                        } else if ( strncmp(arg.c_str(), "-XCFA", 5) == 0 ) {                           // CFA pass through
+                                if(arg.size() == 5) {
+                                        i += 1;
+                                        if ( i == argc ) continue;                              // next argument available ?
+                                        Putenv( argv, argv[i] );
+                                        // CFA specific arguments
+                                }
+                                else if(arg[5] == ',') {
+                                        Putenv( argv, argv[i] + 6 );
+                                        // CFA specific arguments
+                                }
+                                else {
+                                        args[nargs++] = argv[i];
+                                }
                         } else if ( arg == "-CFA" ) {
 …
                                         lang = arg.substr( 2 );
                                 } // if
+                                x_flag = lang != "none";
+                                if ( x_flag ) {
+                                        cerr << argv[0] << " warning, only one -x flag per compile, ignoring subsequent flag." << endl;
+                                } else {
+                                        x_flag = true;
+                                        Putenv( argv, string( "-x=" ) + lang );
+                                } // if
                         } else if ( prefix( arg, "-std=" ) || prefix( arg, "--std=" ) ) {
                                 std_flag = true;                                                // -std=XX provided
 …
         #ifdef __x86_64__
         args[nargs++] = "-mcx16";                                                       // allow double-wide CAA
+        args[nargs++] = "-mcx16";                                                       // allow double-wide CAS
         #endif // __x86_64__
 …
                 args[nargs++] = "-Wl,--pop-state";
                 args[nargs++] = "-pthread";
+                #ifdef __x86_64__
+                args[nargs++] = "-latomic";                                             // allow double-wide CAS
+                #endif // __x86_64__
                 args[nargs++] = "-ldl";
                 args[nargs++] = "-lrt";

examples/io/cat.c

-              r07d867b
+              r22f94a4
+/*
+This is a simple "cat" example that uses io_uring in IORING_SETUP_IOPOLL mode.
+It demonstrates the bare minimum needed to use io_uring in polling mode.
+It uses liburing for simplicity.
+*/
+#ifndef _GNU_SOURCE
+#define _GNU_SOURCE
+#endif
 #include <fcntl.h>
 #include <liburing.h>
 #include <stdio.h>
+#include <string.h>
 #include <unistd.h>
 struct io_uring ring;
+__attribute__((aligned(1024))) char data[1024];
 int main(int argc,  char * argv[]) {
 …
+      }
       int fd = open(argv[1], 0);
+      int fd = open(argv[1], O_DIRECT);
       if(fd < 0) {
             printf("Could not open file %s.\n", argv[1]);
 …
       /* prep the array */
+      char data[512];
+      struct iovec iov = { data, 512 };
+      struct iovec iov = { data, 1024 };
       /* init liburing */
       io_uring_queue_init(256, &ring, 0);
+      io_uring_queue_init(256, &ring, IORING_SETUP_IOPOLL);
       /* declare required structs */
 …
       sqe = io_uring_get_sqe(&ring);
       io_uring_prep_readv(sqe, fd, &iov, 1, 0);
+      // io_uring_prep_read(sqe, fd, data, 1024, 0);
       sqe->user_data = data;
+      sqe->user_data = (uint64_t)(uintptr_t)data;
       /* tell the kernel we have an sqe ready for consumption */
 …
       /* wait for the sqe to complete */
       int ret = io_uring_wait_cqe(&ring, &cqe);
       /* read and process cqe event */
       if(ret == 0) {
             char * out = cqe->user_data;
+            char * out = (char *)(uintptr_t)cqe->user_data;
             signed int len = cqe->res;
             io_uring_cqe_seen(&ring, cqe);
 …
             if(len > 0) {
                   printf("%.*s", len, out);
+            }
+            else if( len < 0 ) {
+                  fprintf(stderr, "readv/read returned error : %s\n", strerror(-len));
+            }
+      }

examples/io/filereader.c

-              r07d867b
+              r22f94a4
+/*
+This is a file reading example that users io_uring in non-blocking mode.
+It demonstrates the bare minimum needed to use io_uring.
+It also optionally pre-registers the file descriptors (and a pipe, just to show it works).
+It uses liburing for simplicity.
+*/
 #include <errno.h>
 #include <fcntl.h>
 …
 int main(int argc, char * argv[]) {
+        if ( argc != 3 ) {
+                printf( "usage\n" );
+                exit( EXIT_FAILURE );
+        }
+        if(argc != 3) {
+            printf("usage:   %s FILE TIMES - read FILE from disk TIMES times\n", argv[0]);
+        if(argc != 3 && argc != 4) {
+            printf("usage:   %s FILE TIMES [fixed] - read FILE from disk TIMES times\n", argv[0]);
             return EXIT_FAILURE;
+      }
+        bool fixed = false;
+        if(argc == 4) {
+                fixed = 0 == strcmp(argv[3], "fixed");
+        }
       int times = atoi( argv[2] );
 …
+      }
+        int rfd = fd;
         /* prep the array */
       char data[100];
 …
       io_uring_queue_init(256, &ring, 0);
+        int pipefds[2];
+        if(fixed) {
+                int ret = pipe(pipefds);
+                if( ret < 0 ) {
+                        printf("Pipe Error : %s\n", strerror( errno ));
+                        return EXIT_FAILURE;
+                }
+                rfd = 0;
+                int fds[] = {
+                        fd, pipefds[0], pipefds[1]
+                };
+                int cnt = sizeof(fds) / sizeof(fds[0]);
+                printf("Registering %d files as fixed\n", cnt);
+                ret = io_uring_register_files(&ring, fds, cnt);
+                if( ret < 0 ) {
+                        printf("Register Error : %s\n", strerror( -ret ));
+                        return EXIT_FAILURE;
+                }
+        }
       /* declare required structs */
         printf("Reading %s(%d) %d times\n", argv[1], fd, times);
 …
                 /* get an sqe and fill in a READV operation */
               struct io_uring_sqe * sqe = io_uring_get_sqe(&ring);
+                io_uring_prep_readv(sqe, fd, &iov, 1, 0);
+                io_uring_prep_readv(sqe, rfd, &iov, 1, 0);
+                if(fixed) {
+                        sqe->flags = IOSQE_FIXED_FILE;
+                }
                 /* tell the kernel we have an sqe ready for consumption */
 …
       close(fd);
+        if(fixed) {
+                close(pipefds[0]);
+                close(pipefds[1]);
+        }
+}

examples/io/simple/server.c

-              r07d867b
+              r22f94a4
+/*
+This is a simple server that users io_uring in blocking mode.
+It demonstrates the bare minimum needed to use io_uring.
+It uses liburing for simplicity.
+*/
 #include <assert.h>
 #include <stdio.h>
 …
 char data[256];
 struct iovec iov = { data, 256 };
 struct msghdr msg = { "", 0, &iov, 1, NULL, 0, 0 };
+struct msghdr msg = { (void *)"", 0, &iov, 1, NULL, 0, 0 };
 static void async_read(int sock) {
         /* get an sqe and fill in a READ operation */

libcfa/configure.ac

-              r07d867b
+              r22f94a4
 AM_SILENT_RULES([yes])
 m4_include([../automake/cfa.m4])
+m4_include([../tools/build/cfa.m4])
 AM_INIT_AUTOMAKE([subdir-objects])
 …
         [  --enable-distcc     whether or not to enable distributed compilation],
         enable_distcc=$enableval, enable_distcc=no)
+AC_ARG_WITH(bwlimit,
+        [  --with-bwlimit=RATE     RATE the maximum rate at which rsync will be limited when using distributed builds],
+        DIST_BWLIMIT=$withval, DIST_BWLIMIT=0)
 echo -n "checking for distributated build... "
 …
 AC_SUBST(CFADIR_HASH)
 AC_SUBST(CFA_VERSION)
+AC_SUBST(DIST_BWLIMIT)
 #==============================================================================
 …
 AM_CONDITIONAL([BUILDLIB], [test "x${CONFIG_BUILDLIB}" = "xyes"])
+AM_T='$(T)'
+AC_SUBST(AM_T)
 #==============================================================================
 #Trasforming cc1 will break compilation
 …
 AC_PROG_MAKE_SET
+AC_CHECK_HEADERS([linux/io_uring.h])
+#io_uring 5.4 and earlier uses defines
+#io_uring 5.5 uses enum values
+#io_uring 5.6 and later uses probes
+AH_TEMPLATE([CFA_HAVE_LINUX_IO_URING_H],[Defined if io_uring support is present when compiling libcfathread.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_NOP],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_NOP.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_READV],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_READV.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_WRITEV],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_WRITEV.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_FSYNC],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_FSYNC.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_READ_FIXED],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_READ_FIXED.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_WRITE_FIXED],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_WRITE_FIXED.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_POLL_ADD],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_POLL_ADD.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_POLL_REMOVE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_POLL_REMOVE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_SYNC_FILE_RANGE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_SYNC_FILE_RANGE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_SENDMSG],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_SENDMSG.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_RECVMSG],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_RECVMSG.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_TIMEOUT],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_TIMEOUT.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_TIMEOUT_REMOVE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_TIMEOUT_REMOVE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_ACCEPT],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_ACCEPT.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_ASYNC_CANCEL],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_ASYNC_CANCEL.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_LINK_TIMEOUT],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_LINK_TIMEOUT.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_CONNECT],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_CONNECT.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_FALLOCATE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_FALLOCATE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_OPENAT],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_OPENAT.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_CLOSE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_CLOSE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_FILES_UPDATE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_FILES_UPDATE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_STATX],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_STATX.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_READ],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_READ.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_WRITE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_WRITE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_FADVISE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_FADVISE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_MADVISE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_MADVISE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_SEND],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_SEND.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_RECV],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_RECV.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_OPENAT2],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_OPENAT2.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_EPOLL_CTL],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_EPOLL_CTL.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_SPLICE],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_SPLICE.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_PROVIDE_BUFFERS],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_PROVIDE_BUFFERS.])
+AH_TEMPLATE([CFA_HAVE_IORING_OP_REMOVE_BUFFER],[Defined if io_uring support is present when compiling libcfathread and supports the operation IORING_OP_REMOVE_BUFFER.])
+AH_TEMPLATE([CFA_HAVE_IOSQE_FIXED_FILE],[Defined if io_uring support is present when compiling libcfathread and supports the flag FIXED_FILE.])
+AH_TEMPLATE([CFA_HAVE_IOSQE_IO_DRAIN],[Defined if io_uring support is present when compiling libcfathread and supports the flag IO_DRAIN.])
+AH_TEMPLATE([CFA_HAVE_IOSQE_ASYNC],[Defined if io_uring support is present when compiling libcfathread and supports the flag ASYNC.])
+AH_TEMPLATE([CFA_HAVE_IOSQE_IO_LINK],[Defined if io_uring support is present when compiling libcfathread and supports the flag IO_LINK.])
+AH_TEMPLATE([CFA_HAVE_IOSQE_IO_HARDLINK],[Defined if io_uring support is present when compiling libcfathread and supports the flag IO_HARDLINK.])
+AH_TEMPLATE([CFA_HAVE_SPLICE_F_FD_IN_FIXED],[Defined if io_uring support is present when compiling libcfathread and supports the flag SPLICE_F_FD_IN_FIXED.])
+AH_TEMPLATE([CFA_HAVE_IORING_SETUP_ATTACH_WQ],[Defined if io_uring support is present when compiling libcfathread and supports the flag IORING_SETUP_ATTACH_WQ.])
+AH_TEMPLATE([HAVE_PREADV2],[Defined if preadv2 support is present when compiling libcfathread.])
+AH_TEMPLATE([HAVE_PWRITEV2],[Defined if pwritev2 support is present when compiling libcfathread.])
+AH_TEMPLATE([__CFA_NO_STATISTICS__],[Defined if libcfathread was compiled without support for statistics.])
+define(ioring_ops, [IORING_OP_NOP,IORING_OP_READV,IORING_OP_WRITEV,IORING_OP_FSYNC,IORING_OP_READ_FIXED,IORING_OP_WRITE_FIXED,IORING_OP_POLL_ADD,IORING_OP_POLL_REMOVE,IORING_OP_SYNC_FILE_RANGE,IORING_OP_SENDMSG,IORING_OP_RECVMSG,IORING_OP_TIMEOUT,IORING_OP_TIMEOUT_REMOVE,IORING_OP_ACCEPT,IORING_OP_ASYNC_CANCEL,IORING_OP_LINK_TIMEOUT,IORING_OP_CONNECT,IORING_OP_FALLOCATE,IORING_OP_OPENAT,IORING_OP_CLOSE,IORING_OP_FILES_UPDATE,IORING_OP_STATX,IORING_OP_READ,IORING_OP_WRITE,IORING_OP_FADVISE,IORING_OP_MADVISE,IORING_OP_SEND,IORING_OP_RECV,IORING_OP_OPENAT2,IORING_OP_EPOLL_CTL,IORING_OP_SPLICE,IORING_OP_PROVIDE_BUFFERS,IORING_OP_REMOVE_BUFFER])
+define(ioring_flags, [IOSQE_FIXED_FILE,IOSQE_IO_DRAIN,IOSQE_ASYNC,IOSQE_IO_LINK,IOSQE_IO_HARDLINK,SPLICE_F_FD_IN_FIXED,IORING_SETUP_ATTACH_WQ])
+define(ioring_from_decls, [
+        m4_foreach([op], [ioring_ops], [
+                AC_CHECK_DECL(op, [AC_DEFINE([CFA_HAVE_]op)], [], [[#include <linux/io_uring.h>]])
+        ])
+])
+AC_CHECK_HEADERS([linux/io_uring.h], [
+        AC_DEFINE(CFA_HAVE_LINUX_IO_URING_H)
+        AC_CHECK_HEADER([liburing.h], [
+                AC_CHECK_LIB([uring], [io_uring_get_probe], [
+                        m4_foreach([op], [ioring_ops], [
+                                AC_CHECK_DECL(op, [
+                                        AC_RUN_IFELSE([
+                                                AC_LANG_PROGRAM(
+                                                        [[#include <liburing.h>]],
+                                                        [[int main() {]]
+                                                        [[      struct io_uring_probe *probe = io_uring_get_probe();]]
+                                                        [[      if(io_uring_opcode_supported(probe, ]]op[[))]]
+                                                        [[              return 0;]]
+                                                        [[      else]]
+                                                        [[              return 1;]]
+                                                        [[}]]
+                                                )
+                                        ],[
+                                                AC_DEFINE([CFA_HAVE_]op)
+                                        ],[
+                                                AC_MSG_FAILURE([Check support for] op [ with liburing failed])
+                                        ])
+                                ], [], [[#include <linux/io_uring.h>]])
+                        ])
+                ], [
+                        ioring_from_decls
+                ])
+        ], [
+                ioring_from_decls
+        ])
+        # check support for various io_uring flags
+        m4_foreach([op], [ioring_flags], [
+                AC_CHECK_DECL(op, [AC_DEFINE([CFA_HAVE_]op)], [], [[#include <linux/io_uring.h>]])
+        ])
+])
 AC_CHECK_FUNCS([preadv2 pwritev2])

libcfa/prelude/Makefile.am

r07d867b	r22f94a4
72	72	if ENABLE_DISTCC
73	73	distribution: @LOCAL_CFACC@ @LOCAL_CC1@ @CFACPP@ gcc-builtins.cf builtins.cf extras.cf prelude.cfa bootloader.c $(srcdir)/../../tools/build/push2dist.sh
74		${AM_V_GEN}$(srcdir)/../../tools/build/push2dist.sh @CFADIR_HASH@
	74	${AM_V_GEN}$(srcdir)/../../tools/build/push2dist.sh @CFADIR_HASH@ @DIST_BWLIMIT@
75	75	@echo "Dummy file to track distribution to remote hosts" > ${@}
76	76

libcfa/prelude/builtins.c

-              r07d867b
+              r22f94a4
 // Created On       : Fri Jul 21 16:21:03 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Nov 21 16:31:39 2019
 // Update Count     : 101
+// Last Modified On : Mon Jul 13 21:10:02 2020
+// Update Count     : 109
 //
 …
 static inline forall( dtype DT ) DT * intptr( uintptr_t addr ) { return (DT *)addr; }
+#if defined(__SIZEOF_INT128__)
+// constructor for 128-bit numbers (all constants are unsigned as +/- are operators)
+static inline void ?{}( unsigned int128 & this, unsigned long int h, unsigned long int l ) {
+        this = (unsigned int128)h << 64 | (unsigned int128)l;
+} // ?{}
+#endif // __SIZEOF_INT128__
 // exponentiation operator implementation

libcfa/prelude/defines.hfa.in

-              r07d867b
+              r22f94a4
+//
+// Cforall Version 1.0.0 Copyright (C) 2016 University of Waterloo
+//
+// The contents of this file are covered under the licence agreement in the
+// file "LICENCE" distributed with Cforall.
+//
+// defines.hfa.in --
+//
+// Author           : Thierry Delisle
+// Created On       : Thu Apr 30 15:23:00 2020
+// Last Modified By :
+// Last Modified On :
+// Update Count     :
+//
+/* prelude/defines.hfa.in.  Generated from configure.ac by autoheader.  */
+/* Location of cfa command. */
+#undef CFA_BINDIR
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_ACCEPT. */
+#undef CFA_HAVE_IORING_OP_ACCEPT
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_ASYNC_CANCEL. */
+#undef CFA_HAVE_IORING_OP_ASYNC_CANCEL
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_CLOSE. */
+#undef CFA_HAVE_IORING_OP_CLOSE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_CONNECT. */
+#undef CFA_HAVE_IORING_OP_CONNECT
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_EPOLL_CTL. */
+#undef CFA_HAVE_IORING_OP_EPOLL_CTL
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_FADVISE. */
+#undef CFA_HAVE_IORING_OP_FADVISE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_FALLOCATE. */
+#undef CFA_HAVE_IORING_OP_FALLOCATE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_FILES_UPDATE. */
+#undef CFA_HAVE_IORING_OP_FILES_UPDATE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_FSYNC. */
+#undef CFA_HAVE_IORING_OP_FSYNC
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_LINK_TIMEOUT. */
+#undef CFA_HAVE_IORING_OP_LINK_TIMEOUT
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_MADVISE. */
+#undef CFA_HAVE_IORING_OP_MADVISE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_NOP. */
+#undef CFA_HAVE_IORING_OP_NOP
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_OPENAT. */
+#undef CFA_HAVE_IORING_OP_OPENAT
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_OPENAT2. */
+#undef CFA_HAVE_IORING_OP_OPENAT2
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_POLL_ADD. */
+#undef CFA_HAVE_IORING_OP_POLL_ADD
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_POLL_REMOVE. */
+#undef CFA_HAVE_IORING_OP_POLL_REMOVE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_PROVIDE_BUFFERS. */
+#undef CFA_HAVE_IORING_OP_PROVIDE_BUFFERS
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_READ. */
+#undef CFA_HAVE_IORING_OP_READ
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_READV. */
+#undef CFA_HAVE_IORING_OP_READV
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_READ_FIXED. */
+#undef CFA_HAVE_IORING_OP_READ_FIXED
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_RECV. */
+#undef CFA_HAVE_IORING_OP_RECV
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_RECVMSG. */
+#undef CFA_HAVE_IORING_OP_RECVMSG
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_REMOVE_BUFFER. */
+#undef CFA_HAVE_IORING_OP_REMOVE_BUFFER
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_SEND. */
+#undef CFA_HAVE_IORING_OP_SEND
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_SENDMSG. */
+#undef CFA_HAVE_IORING_OP_SENDMSG
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_SPLICE. */
+#undef CFA_HAVE_IORING_OP_SPLICE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_STATX. */
+#undef CFA_HAVE_IORING_OP_STATX
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_SYNC_FILE_RANGE. */
+#undef CFA_HAVE_IORING_OP_SYNC_FILE_RANGE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_TIMEOUT. */
+#undef CFA_HAVE_IORING_OP_TIMEOUT
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_TIMEOUT_REMOVE. */
+#undef CFA_HAVE_IORING_OP_TIMEOUT_REMOVE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_WRITE. */
+#undef CFA_HAVE_IORING_OP_WRITE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_WRITEV. */
+#undef CFA_HAVE_IORING_OP_WRITEV
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the operation IORING_OP_WRITE_FIXED. */
+#undef CFA_HAVE_IORING_OP_WRITE_FIXED
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the flag IORING_SETUP_ATTACH_WQ. */
+#undef CFA_HAVE_IORING_SETUP_ATTACH_WQ
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the flag ASYNC. */
+#undef CFA_HAVE_IOSQE_ASYNC
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the flag FIXED_FILE. */
+#undef CFA_HAVE_IOSQE_FIXED_FILE
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the flag IO_DRAIN. */
+#undef CFA_HAVE_IOSQE_IO_DRAIN
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the flag IO_HARDLINK. */
+#undef CFA_HAVE_IOSQE_IO_HARDLINK
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the flag IO_LINK. */
+#undef CFA_HAVE_IOSQE_IO_LINK
+/* Defined if io_uring support is present when compiling libcfathread. */
+#undef CFA_HAVE_LINUX_IO_URING_H
+/* Defined if io_uring support is present when compiling libcfathread and
+   supports the flag SPLICE_F_FD_IN_FIXED. */
+#undef CFA_HAVE_SPLICE_F_FD_IN_FIXED
+/* Location of include files. */
+#undef CFA_INCDIR
+/* Location of cc1 and cfa-cpp commands. */
+#undef CFA_LIBDIR
+/* Location of cfa install. */
+#undef CFA_PREFIX
+/* Define to 1 if you have the <dlfcn.h> header file. */
+#undef HAVE_DLFCN_H
+/* Define to 1 if you have the <inttypes.h> header file. */
+#undef HAVE_INTTYPES_H
+/* Define to 1 if you have the <linux/io_uring.h> header file. */
 #undef HAVE_LINUX_IO_URING_H
+/* Define to 1 if you have the <memory.h> header file. */
+#undef HAVE_MEMORY_H
+/* Define to 1 if you have the `preadv2' function. */
 #undef HAVE_PREADV2
+/* Define to 1 if you have the `pwritev2' function. */
 #undef HAVE_PWRITEV2
+/* Define to 1 if you have the <stdint.h> header file. */
+#undef HAVE_STDINT_H
+/* Define to 1 if you have the <stdlib.h> header file. */
+#undef HAVE_STDLIB_H
+/* Define to 1 if you have the <strings.h> header file. */
+#undef HAVE_STRINGS_H
+/* Define to 1 if you have the <string.h> header file. */
+#undef HAVE_STRING_H
+/* Define to 1 if you have the <sys/stat.h> header file. */
+#undef HAVE_SYS_STAT_H
+/* Define to 1 if you have the <sys/types.h> header file. */
+#undef HAVE_SYS_TYPES_H
+/* Define to 1 if you have the <unistd.h> header file. */
+#undef HAVE_UNISTD_H
+/* Define to the sub-directory where libtool stores uninstalled libraries. */
+#undef LT_OBJDIR
+/* Name of package */
+#undef PACKAGE
+/* Define to the address where bug reports for this package should be sent. */
+#undef PACKAGE_BUGREPORT
+/* Define to the full name of this package. */
+#undef PACKAGE_NAME
+/* Define to the full name and version of this package. */
+#undef PACKAGE_STRING
+/* Define to the one symbol short name of this package. */
+#undef PACKAGE_TARNAME
+/* Define to the home page for this package. */
+#undef PACKAGE_URL
+/* Define to the version of this package. */
+#undef PACKAGE_VERSION
+/* Define to 1 if you have the ANSI C header files. */
+#undef STDC_HEADERS
+/* Version number of package */
+#undef VERSION
+/* Defined if libcfathread was compiled without support for statistics. */
 #undef __CFA_NO_STATISTICS__

libcfa/src/Makefile.am

-              r07d867b
+              r22f94a4
 ## Created On       : Sun May 31 08:54:01 2015
 ## Last Modified By : Peter A. Buhr
 ## Last Modified On : Mon Mar 16 18:07:59 2020
 ## Update Count     : 242
+## Last Modified On : Mon Jun  1 13:35:33 2020
+## Update Count     : 248
 ###############################################################################
 …
 ACLOCAL_AMFLAGS  = -I automake
 include $(srcdir)/../../src/cfa.make
+include $(top_srcdir)/../tools/build/cfa.make
 libdir = ${CFA_LIBDIR}
 …
 #----------------------------------------------------------------------------------------------------------------
 if BUILDLIB
 headers_nosrc = bitmanip.hfa math.hfa gmp.hfa time_t.hfa bits/align.hfa bits/containers.hfa bits/defs.hfa bits/debug.hfa bits/locks.hfa containers/list.hfa
 headers = fstream.hfa iostream.hfa iterator.hfa limits.hfa rational.hfa time.hfa stdlib.hfa common.hfa \
           containers/maybe.hfa containers/pair.hfa containers/result.hfa containers/vector.hfa
+headers_nosrc = bitmanip.hfa exception.hfa math.hfa gmp.hfa time_t.hfa clock.hfa \
+                bits/align.hfa bits/containers.hfa bits/defs.hfa bits/debug.hfa bits/locks.hfa \
+                containers/list.hfa containers/stackLockFree.hfa concurrency/iofwd.hfa
+libsrc = startup.cfa interpose.cfa bits/debug.cfa assert.cfa exception.c virtual.c heap.cfa ${headers:.hfa=.cfa}
+headers = common.hfa fstream.hfa heap.hfa iostream.hfa iterator.hfa limits.hfa rational.hfa \
+                time.hfa stdlib.hfa memory.hfa parseargs.hfa \
+                containers/maybe.hfa containers/pair.hfa containers/result.hfa containers/vector.hfa
+libsrc = startup.cfa interpose.cfa bits/debug.cfa assert.cfa exception.c virtual.c ${headers:.hfa=.cfa}
 # not all platforms support concurrency, add option do disable it
+thread_headers_nosrc = concurrency/invoke.h
+thread_headers = concurrency/coroutine.hfa concurrency/thread.hfa concurrency/kernel.hfa concurrency/monitor.hfa concurrency/mutex.hfa
+thread_libsrc = concurrency/CtxSwitch-@ARCHITECTURE@.S concurrency/alarm.cfa concurrency/invoke.c concurrency/io.cfa concurrency/preemption.cfa ${thread_headers:.hfa=.cfa}
+thread_headers_nosrc = bits/random.hfa concurrency/invoke.h concurrency/kernel/fwd.hfa
+thread_headers = concurrency/coroutine.hfa concurrency/thread.hfa concurrency/kernel.hfa \
+                concurrency/monitor.hfa concurrency/mutex.hfa
+thread_libsrc = concurrency/CtxSwitch-@ARCHITECTURE@.S concurrency/alarm.cfa \
+                concurrency/invoke.c concurrency/io.cfa concurrency/iocall.cfa \
+                concurrency/io/setup.cfa \
+                concurrency/kernel/startup.cfa concurrency/preemption.cfa \
+                concurrency/ready_queue.cfa concurrency/stats.cfa \
+                ${thread_headers:.hfa=.cfa}
 else
 headers =
 …
 prelude.o : prelude.cfa extras.cf gcc-builtins.cf builtins.cf @LOCAL_CFACC@ @CFACPP@
         ${AM_V_GEN}$(CFACOMPILE) -quiet -XCFA -l ${<} -c -o ${@}
+        ${AM_V_GEN}$(CFACOMPILE) -quiet -XCFA,-l ${<} -c -o ${@}
 prelude.lo: prelude.cfa extras.cf gcc-builtins.cf builtins.cf @LOCAL_CFACC@ @CFACPP@
         ${AM_V_GEN}$(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) $(LIBTOOLFLAGS) --mode=compile \
         $(CFACOMPILE) -quiet -XCFA -l ${<} -c -o ${@}
+        $(CFACOMPILE) -quiet -XCFA,-l ${<} -c -o ${@}
 #----------------------------------------------------------------------------------------------------------------

libcfa/src/bitmanip.hfa

-              r07d867b
+              r22f94a4
 // Created On       : Sat Mar 14 18:12:27 2020
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Apr 19 22:29:58 2020
 // Update Count     : 121
+// Last Modified On : Mon Aug 10 09:21:02 2020
+// Update Count     : 139
 //
 …
 // Bits are numbered 1-N.
 //#include <assert.h>
+#include <assert.h>
 #define __bitsizeof( n ) (sizeof(n) * __CHAR_BIT__)
 …
         // Count all 0 bits.
         unsigned int all0s( unsigned char n ) { return __bitsizeof(n) - __builtin_popcount( n ); }
         unsigned int all0s( unsigned short int n ) { return __bitsizeof(n) - __builtin_popcount( n ); }
         unsigned int all0s( unsigned int n ) { return __bitsizeof(n) - __builtin_popcount( n ); }
         unsigned int all0s( unsigned long int n ) { return __bitsizeof(n) - __builtin_popcountl( n ); }
         unsigned int all0s( unsigned long long int n ) { return __bitsizeof(n) - __builtin_popcountll( n ); }
+        unsigned int all0s( unsigned char n ) { return __builtin_popcount( (typeof(n))~n ); }
+        unsigned int all0s( unsigned short int n ) { return __builtin_popcount( (typeof(n))~n ); }
+        unsigned int all0s( unsigned int n ) { return __builtin_popcount( ~n ); }
+        unsigned int all0s( unsigned long int n ) { return __builtin_popcountl( ~n ); }
+        unsigned int all0s( unsigned long long int n ) { return __builtin_popcountll( ~n ); }
         // Find least significiant zero bit. (ffs)
 …
         // Returns n aligned at the floor of align, clear bits above or equal to align, giving n % align.
         signed char floor2( signed char n, char align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         unsigned char floor2( unsigned char n, unsigned char align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         short int floor2( short int n, short int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         unsigned short int floor2( unsigned short int n, unsigned short int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         int floor2( int n, int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         unsigned int floor2( unsigned int n, unsigned int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         long int floor2( long int n, long int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         unsigned long int floor2( unsigned long int n, unsigned long int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         long long int floor2( long long int n, long long int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
         unsigned long long int floor2( unsigned long long int n, unsigned long long int align ) { /*assert( is_pow2( align ) );*/ return n & -align; }
+        signed char floor2( signed char n, signed char align ) { verify( is_pow2( align ) ); return n & -align; }
+        unsigned char floor2( unsigned char n, unsigned char align ) { verify( is_pow2( align ) ); return n & -align; }
+        short int floor2( short int n, short int align ) { verify( is_pow2( align ) ); return n & -align; }
+        unsigned short int floor2( unsigned short int n, unsigned short int align ) { verify( is_pow2( align ) ); return n & -align; }
+        int floor2( int n, int align ) { verify( is_pow2( align ) ); return n & -align; }
+        unsigned int floor2( unsigned int n, unsigned int align ) { verify( is_pow2( align ) ); return n & -align; }
+        long int floor2( long int n, long int align ) { verify( is_pow2( align ) ); return n & -align; }
+        unsigned long int floor2( unsigned long int n, unsigned long int align ) { verify( is_pow2( align ) ); return n & -align; }
+        long long int floor2( long long int n, long long int align ) { verify( is_pow2( align ) ); return n & -align; }
+        unsigned long long int floor2( unsigned long long int n, unsigned long long int align ) { verify( is_pow2( align ) ); return n & -align; }
         // forall( otype T | { T ?&?( T, T ); T -?( T ); } )
         // T floor2( T n, T align ) { /* assert( is_pow2( align ) ); */ return n & -align; }
+        // T floor2( T n, T align ) { verify( is_pow2( align ) ); return n & -align; }
         signed char floor( signed char n, char align ) { return n / align * align; }
+        signed char floor( signed char n, signed char align ) { return n / align * align; }
         unsigned char floor( unsigned char n, unsigned char align ) { return n / align * align; }
         short int floor( short int n, short int align ) { return n / align * align; }
 …
         // Returns n aligned at the ceiling of align, negate, round down, negate is the same as round up.
         signed char ceiling2( signed char n, char align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         unsigned char ceiling2( unsigned char n, unsigned char align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         short int ceiling2( short int n, short int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         unsigned short int ceiling2( unsigned short int n, unsigned short int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         int ceiling2( int n, int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         unsigned int ceiling2( unsigned int n, unsigned int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         long int ceiling2( long int n, long int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         unsigned long int ceiling2( unsigned long int n, unsigned long int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         long long int ceiling2( long long int n, long long int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
         unsigned long long int ceiling2( unsigned long long int n, unsigned long long int align ) { /*assert( is_pow2( align ) );*/ return -floor2( -n, align ); }
+        signed char ceiling2( signed char n, signed char align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        unsigned char ceiling2( unsigned char n, unsigned char align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        short int ceiling2( short int n, short int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        unsigned short int ceiling2( unsigned short int n, unsigned short int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        int ceiling2( int n, int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        unsigned int ceiling2( unsigned int n, unsigned int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        long int ceiling2( long int n, long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        unsigned long int ceiling2( unsigned long int n, unsigned long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        long long int ceiling2( long long int n, long long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        unsigned long long int ceiling2( unsigned long long int n, unsigned long long int align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
         // forall( otype T | { T floor2( T, T ); T -?( T ); } )
         // T ceiling2( T n, T align ) { /* assert( is_pow2( align ) ); */ return -floor2( -n, align ); }
+        // T ceiling2( T n, T align ) { verify( is_pow2( align ) ); return -floor2( -n, align ); }
+        signed char ceiling( signed char n, char align ) { return (n + (align - 1)) / align; }
+        unsigned char ceiling( unsigned char n, unsigned char align ) { return (n + (align - 1)) / align; }
+        short int ceiling( short int n, short int align ) { return (n + (align - 1)) / align; }
+        unsigned short int ceiling( unsigned short int n, unsigned short int align ) { return (n + (align - 1)) / align; }
+        int ceiling( int n, int align ) { return (n + (align - 1)) / align; }
+        unsigned int ceiling( unsigned int n, unsigned int align ) { return (n + (align - 1)) / align; }
+        long int ceiling( long int n, long int align ) { return (n + (align - 1)) / align; }
+        unsigned long int ceiling( unsigned long int n, unsigned long int align ) { return (n + (align - 1)) / align; }
+        long long int ceiling( long long int n, long long int align ) { return (n + (align - 1)) / align; }
+        unsigned long long int ceiling( unsigned long long int n, unsigned long long int align ) { return (n + (align - 1)) / align; }
+        signed char ceiling_div( signed char n, char align ) { return (n + (align - 1)) / align; }
+        unsigned char ceiling_div( unsigned char n, unsigned char align ) { return (n + (align - 1)) / align; }
+        short int ceiling_div( short int n, short int align ) { return (n + (align - 1)) / align; }
+        unsigned short int ceiling_div( unsigned short int n, unsigned short int align ) { return (n + (align - 1)) / align; }
+        int ceiling_div( int n, int align ) { return (n + (align - 1)) / align; }
+        unsigned int ceiling_div( unsigned int n, unsigned int align ) { return (n + (align - 1)) / align; }
+        long int ceiling_div( long int n, long int align ) { return (n + (align - 1)) / align; }
+        unsigned long int ceiling_div( unsigned long int n, unsigned long int align ) { return (n + (align - 1)) / align; }
+        long long int ceiling_div( long long int n, long long int align ) { return (n + (align - 1)) / align; }
+        unsigned long long int ceiling_div( unsigned long long int n, unsigned long long int align ) { return (n + (align - 1)) / align; }
+        // forall( otype T | { T ?+?( T, T ); T ?-?( T, T ); T ?%?( T, T ); } )
+        // T ceiling_div( T n, T align ) { verify( is_pow2( align ) );return (n + (align - 1)) / align; }
+        // gcc notices the div/mod pair and saves both so only one div.
+        signed char ceiling( signed char n, signed char align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        unsigned char ceiling( unsigned char n, unsigned char align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        short int ceiling( short int n, short int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        unsigned short int ceiling( unsigned short int n, unsigned short int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        int ceiling( int n, int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        unsigned int ceiling( unsigned int n, unsigned int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        long int ceiling( long int n, long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        unsigned long int ceiling( unsigned long int n, unsigned long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0) , align); }
+        long long int ceiling( long long int n, long long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
+        unsigned long long int ceiling( unsigned long long int n, unsigned long long int align ) { return floor( n + (n % align != 0 ? align - 1 : 0), align ); }
         // forall( otype T | { void ?{}( T &, one_t ); T ?+?( T, T ); T ?-?( T, T ); T ?/?( T, T ); } )
         // T ceiling( T n, T align ) { return (n + (align - (T){1})) / align; }
+        // T ceiling( T n, T align ) { return return floor( n + (n % align != 0 ? align - 1 : 0), align ); *}
 } // distribution

libcfa/src/bits/containers.hfa

r07d867b	r22f94a4
194	194
195	195	int ?!=?( const __queue(T) & this, __attribute__((unused)) zero_t zero ) {
196		return this.head != 0;
	196	return this.head != 1p;
197	197	}
198	198	}

libcfa/src/bits/debug.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Thu Mar 30 12:30:01 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 13:03:16 2020
 // Update Count     : 11
+// Last Modified On : Wed Jun 17 11:07:13 2020
+// Update Count     : 12
 //
-extern "C" {
 #include <stdio.h>
 #include <stdlib.h>
 …
 #include <stdarg.h>
 #include <unistd.h>
+}
 enum { buffer_size = 4096 };

libcfa/src/bits/debug.hfa

-              r07d867b
+              r22f94a4
 #pragma once
+#include <assert.h>
 #ifdef __CFA_DEBUG__
 …
                 || defined(__CFA_DEBUG_PRINT_IO__) || defined(__CFA_DEBUG_PRINT_IO_CORE__) \
                 || defined(__CFA_DEBUG_PRINT_MONITOR__) || defined(__CFA_DEBUG_PRINT_PREEMPTION__) \
+                || defined(__CFA_DEBUG_PRINT_RUNTIME_CORE__) || defined(__CFA_DEBUG_PRINT_EXCEPTION__)
+                || defined(__CFA_DEBUG_PRINT_RUNTIME_CORE__) || defined(__CFA_DEBUG_PRINT_EXCEPTION__) \
+                || defined(__CFA_DEBUG_PRINT_READY_QUEUE__)
         #include <stdio.h>
         #include <unistd.h>

libcfa/src/bits/defs.hfa

r07d867b	r22f94a4
16	16	#pragma once
17	17
18		~~#include <stdbool.h>~~
19		~~#include <stddef.h>~~
20	18	#include <stdint.h>
21	19

libcfa/src/bits/locks.hfa

-              r07d867b
+              r22f94a4
                 pthread_mutex_init(&lock, &mattr);
                 pthread_cond_init (&cond, 0p);
+                pthread_cond_init (&cond, (const pthread_condattr_t *)0p);  // workaround trac#208: cast should not be required
                 val = 0;
+        }
 …
         #undef CHECKED
+        struct $thread;
+        extern void park( __cfaabi_dbg_ctx_param );
+        extern void unpark( struct $thread * this __cfaabi_dbg_ctx_param2 );
+        static inline struct $thread * active_thread ();
+        // Semaphore which only supports a single thread
+        struct single_sem {
+                struct $thread * volatile ptr;
+        };
+        static inline {
+                void  ?{}(single_sem & this) {
+                        this.ptr = 0p;
+                }
+                void ^?{}(single_sem & this) {}
+                bool wait(single_sem & this) {
+                        for() {
+                                struct $thread * expected = this.ptr;
+                                if(expected == 1p) {
+                                        if(__atomic_compare_exchange_n(&this.ptr, &expected, 0p, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)) {
+                                                return false;
+                                        }
+                                }
+                                else {
+                                        /* paranoid */ verify( expected == 0p );
+                                        if(__atomic_compare_exchange_n(&this.ptr, &expected, active_thread(), false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)) {
+                                                park( __cfaabi_dbg_ctx );
+                                                return true;
+                                        }
+                                }
+                        }
+                }
+                bool post(single_sem & this) {
+                        for() {
+                                struct $thread * expected = this.ptr;
+                                if(expected == 1p) return false;
+                                if(expected == 0p) {
+                                        if(__atomic_compare_exchange_n(&this.ptr, &expected, 1p, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)) {
+                                                return false;
+                                        }
+                                }
+                                else {
+                                        if(__atomic_compare_exchange_n(&this.ptr, &expected, 0p, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)) {
+                                                unpark( expected __cfaabi_dbg_ctx2 );
+                                                return true;
+                                        }
+                                }
+                        }
+                }
+        }
 #endif

libcfa/src/bits/signal.hfa

-              r07d867b
+              r22f94a4
 #include "bits/defs.hfa"
-extern "C" {
 #include <errno.h>
 #define __USE_GNU
 …
 #include <stdlib.h>
 #include <string.h>
+}
 // Short hands for signal context information

libcfa/src/concurrency/alarm.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Fri Jun 2 11:31:25 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Jan  5 08:41:36 2020
 // Update Count     : 69
+// Last Modified On : Wed Jun 17 16:11:35 2020
+// Update Count     : 75
 //
 #define __cforall_thread__
-extern "C" {
 #include <errno.h>
 #include <stdio.h>
+#include <unistd.h>
 #include <string.h>
-#include <unistd.h>
 #include <sys/time.h>
+}
 #include "alarm.hfa"
 #include "kernel_private.hfa"
+#include "kernel/fwd.hfa"
 #include "preemption.hfa"
 …
+}
 void ?{}( alarm_node_t & this, processor   * proc, Time alarm, Duration period ) with( this ) {
+void ?{}( alarm_node_t & this, processor * proc, Time alarm, Duration period ) with( this ) {
         this.proc = proc;
         this.alarm = alarm;

libcfa/src/concurrency/alarm.hfa

r07d867b	r22f94a4
23	23	#include "time.hfa"
24	24
25		#include ~~<containers/list.hfa>~~
	25	#include "containers/list.hfa"
26	26
27	27	struct $thread;

libcfa/src/concurrency/coroutine.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Mon Nov 28 12:27:26 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 12:29:25 2020
 // Update Count     : 16
+// Last Modified On : Tue May 26 22:06:09 2020
+// Update Count     : 21
 //
 …
 #include "coroutine.hfa"
-extern "C" {
 #include <stddef.h>
 #include <malloc.h>
 …
 #include <string.h>
 #include <unistd.h>
+// use this define to make unwind.h play nice, definetely a hack
+#include <sys/mman.h>                                                                   // mprotect
+extern "C" {
+// use this define to make unwind.h play nice, definitely a hack
 #define HIDE_EXPORTS
 #include <unwind.h>
 #undef HIDE_EXPORTS
-#include <sys/mman.h>
+}

libcfa/src/concurrency/invoke.c

-              r07d867b
+              r22f94a4
 #elif defined( __ARM_ARCH )
+#error ARM needs to be upgrade to use to parameters like X86/X64 (A.K.A. : I broke this and do not know how to fix it)
+#error ARM needs to be upgrade to use two parameters like X86/X64 (A.K.A. : I broke this and do not know how to fix it)
+        // More details about the error:
+        // To avoid the thunk problem, I changed the invoke routine to pass the main explicitly
+        // instead of relying on an assertion. This effectively hoists any required thunk one level
+        // which was enough to get to global scope in most cases.
+        // This means that __cfactx_invoke_... now takes two parameters and the FakeStack needs
+        // to be adjusted as a consequence of that.
+        // I don't know how to do that for ARM, hence the #error
         struct FakeStack {
                 float fpRegs[16];                       // floating point registers

libcfa/src/concurrency/invoke.h

-              r07d867b
+              r22f94a4
 #include "bits/defs.hfa"
 #include "bits/locks.hfa"
+#include "kernel/fwd.hfa"
 #ifdef __cforall
 …
 #ifndef _INVOKE_H_
 #define _INVOKE_H_
-#ifdef __ARM_ARCH
-        // function prototypes are only really used by these macros on ARM
-        void disable_global_interrupts();
-        void enable_global_interrupts();
-        #define TL_GET( member ) ( { __typeof__( kernelTLS.member ) target; \
-                disable_global_interrupts(); \
-                target = kernelTLS.member; \
-                enable_global_interrupts(); \
-                target; } )
-        #define TL_SET( member, value ) disable_global_interrupts(); \
-                kernelTLS.member = value; \
-                enable_global_interrupts();
-#else
-        #define TL_GET( member ) kernelTLS.member
-        #define TL_SET( member, value ) kernelTLS.member = value;
-#endif
-        #ifdef __cforall
-        extern "Cforall" {
-                extern __attribute__((aligned(128))) thread_local struct KernelThreadData {
-                        struct $thread    * volatile this_thread;
-                        struct processor      * volatile this_processor;
-                        struct {
-                                volatile unsigned short disable_count;
-                                volatile bool enabled;
-                                volatile bool in_progress;
-                        } preemption_state;
-                        uint32_t rand_seed;
-                } kernelTLS __attribute__ ((tls_model ( "initial-exec" )));
+        }
-        #endif
         struct __stack_context_t {
 …
         };
+        enum coroutine_state { Halted, Start, Primed, Blocked, Ready, Active, Rerun };
+        enum __Preemption_Reason { __NO_PREEMPTION, __ALARM_PREEMPTION, __POLL_PREEMPTION, __MANUAL_PREEMPTION };
+        enum __Coroutine_State { Halted, Start, Primed, Blocked, Ready, Active };
         struct $coroutine {
 …
                 // current execution status for coroutine
                 enum coroutine_state state;
+                enum __Coroutine_State state;
                 // first coroutine to resume this one
 …
         };
+        // Link lists fields
+        // instrusive link field for threads
+        struct __thread_desc_link {
+                struct $thread * next;
+                struct $thread * prev;
+                volatile unsigned long long ts;
+                int preferred;
+        };
         struct $thread {
                 // Core threading fields
 …
                 // current execution status for coroutine
+                volatile int state;
+                enum __Preemption_Reason preempted;
+                volatile int ticket;
+                enum __Coroutine_State state:8;
+                enum __Preemption_Reason preempted:8;
                 //SKULLDUGGERY errno is not save in the thread data structure because returnToKernel appears to be the only function to require saving and restoring it
+                // pointer to the cluster on which the thread is running
+                struct cluster * curr_cluster;
+                // Link lists fields
+                // instrusive link field for threads
+                struct __thread_desc_link link;
                 // coroutine body used to store context
 …
                 struct $monitor *  self_mon_p;
-                // pointer to the cluster on which the thread is running
-                struct cluster * curr_cluster;
                 // monitors currently held by this thread
                 struct __monitor_group_t monitors;
-                // Link lists fields
-                // instrusive link field for threads
-                struct $thread * next;
                 struct {
 …
                         // previous function to park/unpark the thread
                         const char * park_caller;
+                        enum coroutine_state park_result;
+                        int park_result;
+                        enum __Coroutine_State park_state;
                         bool park_stale;
                         const char * unpark_caller;
+                        enum coroutine_state unpark_result;
+                        int unpark_result;
+                        enum __Coroutine_State unpark_state;
                         bool unpark_stale;
                 #endif
 …
         #ifdef __cforall
         extern "Cforall" {
                 static inline $thread *& get_next( $thread & this ) __attribute__((const)) {
                         return this.next;
+                        return this.link.next;
+                }

libcfa/src/concurrency/io.cfa

-              r07d867b
+              r22f94a4
 //
+// #define __CFA_DEBUG_PRINT_IO__
+// #define __CFA_DEBUG_PRINT_IO_CORE__
+#include "kernel.hfa"
+#if !defined(HAVE_LINUX_IO_URING_H)
+        void __kernel_io_startup( cluster &, int, bool ) {
+                // Nothing to do without io_uring
+        }
+        void __kernel_io_finish_start( cluster & ) {
+                // Nothing to do without io_uring
+        }
+        void __kernel_io_prepare_stop( cluster & ) {
+                // Nothing to do without io_uring
+        }
+        void __kernel_io_shutdown( cluster &, bool ) {
+                // Nothing to do without io_uring
+        }
+#else
+#define __cforall_thread__
+#if defined(__CFA_DEBUG__)
+        // #define __CFA_DEBUG_PRINT_IO__
+        // #define __CFA_DEBUG_PRINT_IO_CORE__
+#endif
+#if defined(CFA_HAVE_LINUX_IO_URING_H)
+        #define _GNU_SOURCE         /* See feature_test_macros(7) */
+        #include <errno.h>
+        #include <signal.h>
+        #include <stdint.h>
+        #include <string.h>
+        #include <unistd.h>
         extern "C" {
+                #define _GNU_SOURCE         /* See feature_test_macros(7) */
+                #include <errno.h>
+                #include <stdint.h>
+                #include <string.h>
+                #include <unistd.h>
+                #include <sys/mman.h>
+                #include <sys/epoll.h>
                 #include <sys/syscall.h>
 …
+        }
+        #include "bits/signal.hfa"
+        #include "kernel_private.hfa"
+        #include "thread.hfa"
+        uint32_t entries_per_cluster() {
+                return 256;
+        }
+        static void * __io_poller_slow( void * arg );
+        // Weirdly, some systems that do support io_uring don't actually define these
+        #ifdef __alpha__
+                /*
+                * alpha is the only exception, all other architectures
+                * have common numbers for new system calls.
+                */
+                #ifndef __NR_io_uring_setup
+                        #define __NR_io_uring_setup           535
+                #endif
+                #ifndef __NR_io_uring_enter
+                        #define __NR_io_uring_enter           536
+                #endif
+                #ifndef __NR_io_uring_register
+                        #define __NR_io_uring_register        537
+                #endif
+        #else /* !__alpha__ */
+                #ifndef __NR_io_uring_setup
+                        #define __NR_io_uring_setup           425
+                #endif
+                #ifndef __NR_io_uring_enter
+                        #define __NR_io_uring_enter           426
+                #endif
+                #ifndef __NR_io_uring_register
+                        #define __NR_io_uring_register        427
+                #endif
+        #endif
+        // Fast poller user-thread
+        // Not using the "thread" keyword because we want to control
+        // more carefully when to start/stop it
+        struct __io_poller_fast {
+                struct __io_data * ring;
+                bool waiting;
+                $thread thrd;
+        };
+        void ?{}( __io_poller_fast & this, struct cluster & cltr ) {
+                this.ring = cltr.io;
+                this.waiting = true;
+                (this.thrd){ "Fast I/O Poller", cltr };
+        }
+        void ^?{}( __io_poller_fast & mutex this );
+        void main( __io_poller_fast & this );
+        static inline $thread * get_thread( __io_poller_fast & this ) { return &this.thrd; }
+        void ^?{}( __io_poller_fast & mutex this ) {}
+        struct __submition_data {
+                // Head and tail of the ring (associated with array)
+                volatile uint32_t * head;
+                volatile uint32_t * tail;
+                // The actual kernel ring which uses head/tail
+                // indexes into the sqes arrays
+                uint32_t * array;
+                // number of entries and mask to go with it
+                const uint32_t * num;
+                const uint32_t * mask;
+                // Submission flags (Not sure what for)
+                uint32_t * flags;
+                // number of sqes not submitted (whatever that means)
+                uint32_t * dropped;
+                // Like head/tail but not seen by the kernel
+                volatile uint32_t alloc;
+                volatile uint32_t ready;
+                __spinlock_t lock;
+                // A buffer of sqes (not the actual ring)
+                struct io_uring_sqe * sqes;
+                // The location and size of the mmaped area
+                void * ring_ptr;
+                size_t ring_sz;
+                // Statistics
+                #if !defined(__CFA_NO_STATISTICS__)
+                        struct {
+                                struct {
+                                        volatile unsigned long long int val;
+                                        volatile unsigned long long int cnt;
+                                        volatile unsigned long long int block;
+                                } submit_avg;
+                        } stats;
+                #endif
+        };
+        struct __completion_data {
+                // Head and tail of the ring
+                volatile uint32_t * head;
+                volatile uint32_t * tail;
+                // number of entries and mask to go with it
+                const uint32_t * mask;
+                const uint32_t * num;
+                // number of cqes not submitted (whatever that means)
+                uint32_t * overflow;
+                // the kernel ring
+                struct io_uring_cqe * cqes;
+                // The location and size of the mmaped area
+                void * ring_ptr;
+                size_t ring_sz;
+                // Statistics
+                #if !defined(__CFA_NO_STATISTICS__)
+                        struct {
+                                struct {
+                                        unsigned long long int val;
+                                        unsigned long long int slow_cnt;
+                                        unsigned long long int fast_cnt;
+                                } completed_avg;
+                        } stats;
+                #endif
+        };
+        struct __io_data {
+                struct __submition_data submit_q;
+                struct __completion_data completion_q;
+                uint32_t ring_flags;
+                int cltr_flags;
+                int fd;
+                semaphore submit;
+                volatile bool done;
+                struct {
+                        struct {
+                                void * stack;
+                                pthread_t kthrd;
+                        } slow;
+                        __io_poller_fast fast;
+                        __bin_sem_t sem;
+                } poller;
+        };
+//=============================================================================================
+// I/O Startup / Shutdown logic
+//=============================================================================================
+        void __kernel_io_startup( cluster & this, int io_flags, bool main_cluster ) {
+                this.io = malloc();
+                // Step 1 : call to setup
+                struct io_uring_params params;
+                memset(&params, 0, sizeof(params));
+                uint32_t nentries = entries_per_cluster();
+                int fd = syscall(__NR_io_uring_setup, nentries, &params );
+                if(fd < 0) {
+                        abort("KERNEL ERROR: IO_URING SETUP - %s\n", strerror(errno));
+                }
+                // Step 2 : mmap result
+                memset( this.io, 0, sizeof(struct __io_data) );
+                struct __submition_data  & sq = this.io->submit_q;
+                struct __completion_data & cq = this.io->completion_q;
+                // calculate the right ring size
+                sq.ring_sz = params.sq_off.array + (params.sq_entries * sizeof(unsigned)           );
+                cq.ring_sz = params.cq_off.cqes  + (params.cq_entries * sizeof(struct io_uring_cqe));
+                // Requires features
+                #if defined(IORING_FEAT_SINGLE_MMAP)
+                        // adjust the size according to the parameters
+                        if ((params.features & IORING_FEAT_SINGLE_MMAP) != 0) {
+                                cq->ring_sz = sq->ring_sz = max(cq->ring_sz, sq->ring_sz);
+                        }
+                #endif
+                // mmap the Submit Queue into existence
+                sq.ring_ptr = mmap(0, sq.ring_sz, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_POPULATE, fd, IORING_OFF_SQ_RING);
+                if (sq.ring_ptr == (void*)MAP_FAILED) {
+                        abort("KERNEL ERROR: IO_URING MMAP1 - %s\n", strerror(errno));
+                }
+                // Requires features
+                #if defined(IORING_FEAT_SINGLE_MMAP)
+                        // mmap the Completion Queue into existence (may or may not be needed)
+                        if ((params.features & IORING_FEAT_SINGLE_MMAP) != 0) {
+                                cq->ring_ptr = sq->ring_ptr;
+                        }
+                        else
+                #endif
+                {
+                        // We need multiple call to MMAP
+                        cq.ring_ptr = mmap(0, cq.ring_sz, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_POPULATE, fd, IORING_OFF_CQ_RING);
+                        if (cq.ring_ptr == (void*)MAP_FAILED) {
+                                munmap(sq.ring_ptr, sq.ring_sz);
+                                abort("KERNEL ERROR: IO_URING MMAP2 - %s\n", strerror(errno));
+                        }
+                }
+                // mmap the submit queue entries
+                size_t size = params.sq_entries * sizeof(struct io_uring_sqe);
+                sq.sqes = (struct io_uring_sqe *)mmap(0, size, PROT_READ | PROT_WRITE, MAP_SHARED | MAP_POPULATE, fd, IORING_OFF_SQES);
+                if (sq.sqes == (struct io_uring_sqe *)MAP_FAILED) {
+                        munmap(sq.ring_ptr, sq.ring_sz);
+                        if (cq.ring_ptr != sq.ring_ptr) munmap(cq.ring_ptr, cq.ring_sz);
+                        abort("KERNEL ERROR: IO_URING MMAP3 - %s\n", strerror(errno));
+                }
+                // Get the pointers from the kernel to fill the structure
+                // submit queue
+                sq.head    = (volatile uint32_t *)(((intptr_t)sq.ring_ptr) + params.sq_off.head);
+                sq.tail    = (volatile uint32_t *)(((intptr_t)sq.ring_ptr) + params.sq_off.tail);
+                sq.mask    = (   const uint32_t *)(((intptr_t)sq.ring_ptr) + params.sq_off.ring_mask);
+                sq.num     = (   const uint32_t *)(((intptr_t)sq.ring_ptr) + params.sq_off.ring_entries);
+                sq.flags   = (         uint32_t *)(((intptr_t)sq.ring_ptr) + params.sq_off.flags);
+                sq.dropped = (         uint32_t *)(((intptr_t)sq.ring_ptr) + params.sq_off.dropped);
+                sq.array   = (         uint32_t *)(((intptr_t)sq.ring_ptr) + params.sq_off.array);
+                sq.alloc = *sq.tail;
+                sq.ready = *sq.tail;
+                // completion queue
+                cq.head     = (volatile uint32_t *)(((intptr_t)cq.ring_ptr) + params.cq_off.head);
+                cq.tail     = (volatile uint32_t *)(((intptr_t)cq.ring_ptr) + params.cq_off.tail);
+                cq.mask     = (   const uint32_t *)(((intptr_t)cq.ring_ptr) + params.cq_off.ring_mask);
+                cq.num      = (   const uint32_t *)(((intptr_t)cq.ring_ptr) + params.cq_off.ring_entries);
+                cq.overflow = (         uint32_t *)(((intptr_t)cq.ring_ptr) + params.cq_off.overflow);
+                cq.cqes   = (struct io_uring_cqe *)(((intptr_t)cq.ring_ptr) + params.cq_off.cqes);
+                // some paranoid checks
+                /* paranoid */ verifyf( (*cq.mask) == ((*cq.num) - 1ul32), "IO_URING Expected mask to be %u (%u entries), was %u", (*cq.num) - 1ul32, *cq.num, *cq.mask  );
+                /* paranoid */ verifyf( (*cq.num)  >= nentries, "IO_URING Expected %u entries, got %u", nentries, *cq.num );
+                /* paranoid */ verifyf( (*cq.head) == 0, "IO_URING Expected head to be 0, got %u", *cq.head );
+                /* paranoid */ verifyf( (*cq.tail) == 0, "IO_URING Expected tail to be 0, got %u", *cq.tail );
+                /* paranoid */ verifyf( (*sq.mask) == ((*sq.num) - 1ul32), "IO_URING Expected mask to be %u (%u entries), was %u", (*sq.num) - 1ul32, *sq.num, *sq.mask );
+                /* paranoid */ verifyf( (*sq.num) >= nentries, "IO_URING Expected %u entries, got %u", nentries, *sq.num );
+                /* paranoid */ verifyf( (*sq.head) == 0, "IO_URING Expected head to be 0, got %u", *sq.head );
+                /* paranoid */ verifyf( (*sq.tail) == 0, "IO_URING Expected tail to be 0, got %u", *sq.tail );
+                // Update the global ring info
+                this.io->ring_flags = params.flags;
+                this.io->cltr_flags = io_flags;
+                this.io->fd         = fd;
+                this.io->done       = false;
+                (this.io->submit){ min(*sq.num, *cq.num) };
+                // Initialize statistics
+                #if !defined(__CFA_NO_STATISTICS__)
+                        this.io->submit_q.stats.submit_avg.val   = 0;
+                        this.io->submit_q.stats.submit_avg.cnt   = 0;
+                        this.io->submit_q.stats.submit_avg.block = 0;
+                        this.io->completion_q.stats.completed_avg.val = 0;
+                        this.io->completion_q.stats.completed_avg.slow_cnt = 0;
+                        this.io->completion_q.stats.completed_avg.fast_cnt = 0;
+                #endif
+                if(!main_cluster) {
+                        __kernel_io_finish_start( this );
+                }
+        }
+        void __kernel_io_finish_start( cluster & this ) {
+                if( this.io->cltr_flags & CFA_CLUSTER_IO_POLLER_USER_THREAD ) {
+                        __cfadbg_print_safe(io_core, "Kernel I/O : Creating fast poller for cluter %p\n", &this);
+                        (this.io->poller.fast){ this };
+                        __thrd_start( this.io->poller.fast, main );
+                }
+                // Create the poller thread
+                __cfadbg_print_safe(io_core, "Kernel I/O : Creating slow poller for cluter %p\n", &this);
+                this.io->poller.slow.stack = __create_pthread( &this.io->poller.slow.kthrd, __io_poller_slow, &this );
+        }
+        void __kernel_io_prepare_stop( cluster & this ) {
+                __cfadbg_print_safe(io_core, "Kernel I/O : Stopping pollers for cluster\n", &this);
+                // Notify the poller thread of the shutdown
+                __atomic_store_n(&this.io->done, true, __ATOMIC_SEQ_CST);
+                // Stop the IO Poller
+                sigval val = { 1 };
+                pthread_sigqueue( this.io->poller.slow.kthrd, SIGUSR1, val );
+                post( this.io->poller.sem );
+                // Wait for the poller thread to finish
+                pthread_join( this.io->poller.slow.kthrd, 0p );
+                free( this.io->poller.slow.stack );
+                __cfadbg_print_safe(io_core, "Kernel I/O : Slow poller stopped for cluster\n", &this);
+                if( this.io->cltr_flags & CFA_CLUSTER_IO_POLLER_USER_THREAD ) {
+                        with( this.io->poller.fast ) {
+                                /* paranoid */ verify( waiting ); // The thread shouldn't be in a system call
+                                /* paranoid */ verify( this.procs.head == 0p || &this == mainCluster );
+                                /* paranoid */ verify( this.idles.head == 0p || &this == mainCluster );
+                                // We need to adjust the clean-up based on where the thread is
+                                if( thrd.preempted != __NO_PREEMPTION ) {
+                                        // This is the tricky case
+                                        // The thread was preempted and now it is on the ready queue
+                                        /* paranoid */ verify( thrd.state == Active );           // The thread better be in this state
+                                        /* paranoid */ verify( thrd.next == 1p );                // The thread should be the last on the list
+                                        /* paranoid */ verify( this.ready_queue.head == &thrd ); // The thread should be the only thing on the list
+                                        // Remove the thread from the ready queue of this cluster
+                                        this.ready_queue.head = 1p;
+                                        thrd.next = 0p;
+                                        // Fixup the thread state
+                                        thrd.state = Blocked;
+                                        thrd.preempted = __NO_PREEMPTION;
+                                        // Pretend like the thread was blocked all along
+                                }
+                                // !!! This is not an else if !!!
+                                if( thrd.state == Blocked ) {
+                                        // This is the "easy case"
+                                        // The thread is parked and can easily be moved to active cluster
+                                        verify( thrd.curr_cluster != active_cluster() || thrd.curr_cluster == mainCluster );
+                                        thrd.curr_cluster = active_cluster();
+                        // unpark the fast io_poller
+                                        unpark( &thrd __cfaabi_dbg_ctx2 );
+                                }
+                                else {
+                                        // The thread is in a weird state
+                                        // I don't know what to do here
+                                        abort("Fast poller thread is in unexpected state, cannot clean-up correctly\n");
+                                }
+                        }
+                        ^(this.io->poller.fast){};
+                        __cfadbg_print_safe(io_core, "Kernel I/O : Fast poller stopped for cluster\n", &this);
+                }
+        }
+        void __kernel_io_shutdown( cluster & this, bool main_cluster ) {
+                if(!main_cluster) {
+                        __kernel_io_prepare_stop( this );
+                }
+                // print statistics
+                #if !defined(__CFA_NO_STATISTICS__)
+                        if(this.print_stats) {
+                                with(this.io->submit_q.stats, this.io->completion_q.stats) {
+                                        __cfaabi_bits_print_safe( STDERR_FILENO,
+                                                "----- I/O uRing Stats -----\n"
+                                                "- total submit calls  : %'15llu\n"
+                                                "- avg submit          : %'18.2lf\n"
+                                                "- pre-submit block %%  : %'18.2lf\n"
+                                                "- total wait calls    : %'15llu   (%'llu slow, %'llu fast)\n"
+                                                "- avg completion/wait : %'18.2lf\n",
+                                                submit_avg.cnt,
+                                                ((double)submit_avg.val) / submit_avg.cnt,
+                                                (100.0 * submit_avg.block) / submit_avg.cnt,
+                                                completed_avg.slow_cnt + completed_avg.fast_cnt,
+                                                completed_avg.slow_cnt,  completed_avg.fast_cnt,
+                                                ((double)completed_avg.val) / (completed_avg.slow_cnt + completed_avg.fast_cnt)
+                                        );
+                                }
+                        }
+                #endif
+                // Shutdown the io rings
+                struct __submition_data  & sq = this.io->submit_q;
+                struct __completion_data & cq = this.io->completion_q;
+                // unmap the submit queue entries
+                munmap(sq.sqes, (*sq.num) * sizeof(struct io_uring_sqe));
+                // unmap the Submit Queue ring
+                munmap(sq.ring_ptr, sq.ring_sz);
+                // unmap the Completion Queue ring, if it is different
+                if (cq.ring_ptr != sq.ring_ptr) {
+                        munmap(cq.ring_ptr, cq.ring_sz);
+                }
+                // close the file descriptor
+                close(this.io->fd);
+                free( this.io );
+        #include "stats.hfa"
+        #include "kernel.hfa"
+        #include "kernel/fwd.hfa"
+        #include "io/types.hfa"
+//=============================================================================================
+// I/O Syscall
+//=============================================================================================
+        static int __io_uring_enter( struct __io_data & ring, unsigned to_submit, bool get ) {
+                bool need_sys_to_submit = false;
+                bool need_sys_to_complete = false;
+                unsigned flags = 0;
+                TO_SUBMIT:
+                if( to_submit > 0 ) {
+                        if( !(ring.ring_flags & IORING_SETUP_SQPOLL) ) {
+                                need_sys_to_submit = true;
+                                break TO_SUBMIT;
+                        }
+                        if( (*ring.submit_q.flags) & IORING_SQ_NEED_WAKEUP ) {
+                                need_sys_to_submit = true;
+                                flags |= IORING_ENTER_SQ_WAKEUP;
+                        }
+                }
+                if( get && !(ring.ring_flags & IORING_SETUP_SQPOLL) ) {
+                        flags |= IORING_ENTER_GETEVENTS;
+                        if( (ring.ring_flags & IORING_SETUP_IOPOLL) ) {
+                                need_sys_to_complete = true;
+                        }
+                }
+                int ret = 0;
+                if( need_sys_to_submit || need_sys_to_complete ) {
+                        ret = syscall( __NR_io_uring_enter, ring.fd, to_submit, 0, flags, 0p, _NSIG / 8);
+                        if( ret < 0 ) {
+                                switch((int)errno) {
+                                case EAGAIN:
+                                case EINTR:
+                                        ret = -1;
+                                        break;
+                                default:
+                                        abort( "KERNEL ERROR: IO_URING SYSCALL - (%d) %s\n", (int)errno, strerror(errno) );
+                                }
+                        }
+                }
+                // Memory barrier
+                __atomic_thread_fence( __ATOMIC_SEQ_CST );
+                return ret;
+        }
 …
 // I/O Polling
 //=============================================================================================
+        struct io_user_data {
+                int32_t result;
+                $thread * thrd;
+        };
+        static unsigned __collect_submitions( struct __io_data & ring );
+        static uint32_t __release_consumed_submission( struct __io_data & ring );
+        static inline void process(struct io_uring_cqe & cqe ) {
+                struct __io_user_data_t * data = (struct __io_user_data_t *)(uintptr_t)cqe.user_data;
+                __cfadbg_print_safe( io, "Kernel I/O : Syscall completed : cqe %p, result %d for %p\n", data, cqe.res, data->thrd );
+                data->result = cqe.res;
+                unpark( data->thrd __cfaabi_dbg_ctx2 );
+        }
         // Process a single completion message from the io_uring
         // This is NOT thread-safe
+        static int __drain_io( struct __io_data & ring, sigset_t * mask, int waitcnt, bool in_kernel ) {
+                int ret = syscall( __NR_io_uring_enter, ring.fd, 0, waitcnt, IORING_ENTER_GETEVENTS, mask, _NSIG / 8);
+        static [int, bool] __drain_io( & struct __io_data ring ) {
+                /* paranoid */ verify( !kernelTLS.preemption_state.enabled );
+                unsigned to_submit = 0;
+                if( ring.poller_submits ) {
+                        // If the poller thread also submits, then we need to aggregate the submissions which are ready
+                        to_submit = __collect_submitions( ring );
+                }
+                int ret = __io_uring_enter(ring, to_submit, true);
                 if( ret < 0 ) {
+                        switch((int)errno) {
+                        case EAGAIN:
+                        case EINTR:
+                                return -EAGAIN;
+                        default:
+                                abort( "KERNEL ERROR: IO_URING WAIT - %s\n", strerror(errno) );
+                        }
+                }
+                        return [0, true];
+                }
+                // update statistics
+                if (to_submit > 0) {
+                        __STATS__( true,
+                                if( to_submit > 0 ) {
+                                        io.submit_q.submit_avg.rdy += to_submit;
+                                        io.submit_q.submit_avg.csm += ret;
+                                        io.submit_q.submit_avg.cnt += 1;
+                                }
+                        )
+                }
+                // Release the consumed SQEs
+                __release_consumed_submission( ring );
                 // Drain the queue
                 unsigned head = *ring.completion_q.head;
+                unsigned tail = __atomic_load_n(ring.completion_q.tail, __ATOMIC_ACQUIRE);
+                unsigned tail = *ring.completion_q.tail;
+                const uint32_t mask = *ring.completion_q.mask;
                 // Nothing was new return 0
                 if (head == tail) {
                         return 0;
+                        return [0, to_submit > 0];
+                }
                 uint32_t count = tail - head;
+                /* paranoid */ verify( count != 0 );
                 for(i; count) {
                         unsigned idx = (head + i) & (*ring.completion_q.mask);
+                        unsigned idx = (head + i) & mask;
                         struct io_uring_cqe & cqe = ring.completion_q.cqes[idx];
                         /* paranoid */ verify(&cqe);
+                        struct io_user_data * data = (struct io_user_data *)cqe.user_data;
+                        __cfadbg_print_safe( io, "Kernel I/O : Performed reading io cqe %p, result %d for %p\n", data, cqe.res, data->thrd );
+                        data->result = cqe.res;
+                        if(!in_kernel) { unpark( data->thrd __cfaabi_dbg_ctx2 ); }
+                        else         { __unpark( data->thrd __cfaabi_dbg_ctx2 ); }
+                }
+                // Allow new submissions to happen
+                V(ring.submit, count);
+                        process( cqe );
+                }
                 // Mark to the kernel that the cqe has been seen
                 // Ensure that the kernel only sees the new value of the head index after the CQEs have been read.
+                __atomic_thread_fence( __ATOMIC_SEQ_CST );
                 __atomic_fetch_add( ring.completion_q.head, count, __ATOMIC_RELAXED );
+                return count;
+        }
+        static void * __io_poller_slow( void * arg ) {
+                cluster * cltr = (cluster *)arg;
+                struct __io_data & ring = *cltr->io;
+                sigset_t mask;
+                sigfillset(&mask);
+                if ( pthread_sigmask( SIG_BLOCK, &mask, 0p ) == -1 ) {
+                        abort( "KERNEL ERROR: IO_URING - pthread_sigmask" );
+                }
+                sigdelset( &mask, SIGUSR1 );
+                verify( (*ring.submit_q.head) == (*ring.submit_q.tail) );
+                verify( (*ring.completion_q.head) == (*ring.completion_q.tail) );
+                __cfadbg_print_safe(io_core, "Kernel I/O : Slow poller for ring %p ready\n", &ring);
+                if( ring.cltr_flags & CFA_CLUSTER_IO_POLLER_USER_THREAD ) {
+                        while(!__atomic_load_n(&ring.done, __ATOMIC_SEQ_CST)) {
+                                // In the user-thread approach drain and if anything was drained,
+                                // batton pass to the user-thread
+                                int count = __drain_io( ring, &mask, 1, true );
+                return [count, count > 0 || to_submit > 0];
+        }
+        void main( $io_ctx_thread & this ) {
+                epoll_event ev;
+                __ioctx_register( this, ev );
+                __cfadbg_print_safe(io_core, "Kernel I/O : IO poller %p for ring %p ready\n", &this, &this.ring);
+                int reset = 0;
+                // Then loop until we need to start
+                while(!__atomic_load_n(&this.done, __ATOMIC_SEQ_CST)) {
+                        // Drain the io
+                        int count;
+                        bool again;
+                        disable_interrupts();
+                                [count, again] = __drain_io( *this.ring );
+                                if(!again) reset++;
                                 // Update statistics
+                                #if !defined(__CFA_NO_STATISTICS__)
+                                        ring.completion_q.stats.completed_avg.val += count;
+                                        ring.completion_q.stats.completed_avg.slow_cnt += 1;
+                                #endif
+                                if(count > 0) {
+                                        __cfadbg_print_safe(io_core, "Kernel I/O : Moving to ring %p to fast poller\n", &ring);
+                                        __unpark( &ring.poller.fast.thrd __cfaabi_dbg_ctx2 );
+                                        wait( ring.poller.sem );
+                                }
+                        }
+                }
+                else {
+                        while(!__atomic_load_n(&ring.done, __ATOMIC_SEQ_CST)) {
+                                //In the naive approach, just poll the io completion queue directly
+                                int count = __drain_io( ring, &mask, 1, true );
+                                // Update statistics
+                                #if !defined(__CFA_NO_STATISTICS__)
+                                        ring.completion_q.stats.completed_avg.val += count;
+                                        ring.completion_q.stats.completed_avg.slow_cnt += 1;
+                                #endif
+                        }
+                }
+                __cfadbg_print_safe(io_core, "Kernel I/O : Slow poller for ring %p stopping\n", &ring);
+                return 0p;
+        }
+        void main( __io_poller_fast & this ) {
+                verify( this.ring->cltr_flags & CFA_CLUSTER_IO_POLLER_USER_THREAD );
+                // Start parked
+                park( __cfaabi_dbg_ctx );
+                __cfadbg_print_safe(io_core, "Kernel I/O : Fast poller for ring %p ready\n", &this.ring);
+                int reset = 0;
+                // Then loop until we need to start
+                while(!__atomic_load_n(&this.ring->done, __ATOMIC_SEQ_CST)) {
+                        // Drain the io
+                        this.waiting = false;
+                        int count = __drain_io( *this.ring, 0p, 0, false );
+                        reset += count > 0 ? 1 : 0;
+                        // Update statistics
+                        #if !defined(__CFA_NO_STATISTICS__)
+                                this.ring->completion_q.stats.completed_avg.val += count;
+                                this.ring->completion_q.stats.completed_avg.fast_cnt += 1;
+                        #endif
+                        this.waiting = true;
+                                __STATS__( true,
+                                        io.complete_q.completed_avg.val += count;
+                                        io.complete_q.completed_avg.fast_cnt += 1;
+                                )
+                        enable_interrupts( __cfaabi_dbg_ctx );
+                        // If we got something, just yield and check again
                         if(reset < 5) {
-                                // If we got something, just yield and check again
                                 yield();
+                        }
+                        // We didn't get anything baton pass to the slow poller
                         else {
+                                // We didn't get anything baton pass to the slow poller
+                                __cfadbg_print_safe(io_core, "Kernel I/O : Moving to ring %p to slow poller\n", &this.ring);
+                                post( this.ring->poller.sem );
+                                park( __cfaabi_dbg_ctx );
+                                __cfadbg_print_safe(io_core, "Kernel I/O : Parking io poller %p\n", &this.self);
                                 reset = 0;
+                                // block this thread
+                                __ioctx_prepare_block( this, ev );
+                                wait( this.sem );
+                        }
+                }
 …
 // Submition steps :
+// 1 - We need to make sure we don't overflow any of the buffer, P(ring.submit) to make sure
+//     entries are available. The semaphore make sure that there is no more operations in
+//     progress then the number of entries in the buffer. This probably limits concurrency
+//     more than necessary since submitted but not completed operations don't need any
+//     entries in user space. However, I don't know what happens if we overflow the buffers
+//     because too many requests completed at once. This is a safe approach in all cases.
+//     Furthermore, with hundreds of entries, this may be okay.
+//
+// 2 - Allocate a queue entry. The ring already has memory for all entries but only the ones
+// 1 - Allocate a queue entry. The ring already has memory for all entries but only the ones
 //     listed in sq.array are visible by the kernel. For those not listed, the kernel does not
 //     offer any assurance that an entry is not being filled by multiple flags. Therefore, we
 //     need to write an allocator that allows allocating concurrently.
 //
 // 3 - Actually fill the submit entry, this is the only simple and straightforward step.
 //
 // 4 - Append the entry index to the array and adjust the tail accordingly. This operation
+// 2 - Actually fill the submit entry, this is the only simple and straightforward step.
+//
+// 3 - Append the entry index to the array and adjust the tail accordingly. This operation
 //     needs to arrive to two concensus at the same time:
 //     A - The order in which entries are listed in the array: no two threads must pick the
 …
 //
+        static inline [* struct io_uring_sqe, uint32_t] __submit_alloc( struct __io_data & ring ) {
+                // Wait for a spot to be available
+                __attribute__((unused)) bool blocked = P(ring.submit);
+                #if !defined(__CFA_NO_STATISTICS__)
+                        __atomic_fetch_add( &ring.submit_q.stats.submit_avg.block, blocked ? 1ul64 : 0ul64, __ATOMIC_RELAXED );
+                #endif
+                // Allocate the sqe
+                uint32_t idx = __atomic_fetch_add(&ring.submit_q.alloc, 1ul32, __ATOMIC_SEQ_CST);
+                // Validate that we didn't overflow anything
+                // Check that nothing overflowed
+                /* paranoid */ verify( true );
+                // Check that it goes head -> tail -> alloc and never head -> alloc -> tail
+                /* paranoid */ verify( true );
+                // Return the sqe
+                return [&ring.submit_q.sqes[ idx & (*ring.submit_q.mask)], idx];
+        }
+        static inline void __submit( struct __io_data & ring, uint32_t idx ) {
+                // get mutual exclusion
+                lock(ring.submit_q.lock __cfaabi_dbg_ctx2);
+                // Append to the list of ready entries
+                uint32_t * tail = ring.submit_q.tail;
+        [* struct io_uring_sqe, uint32_t] __submit_alloc( struct __io_data & ring, uint64_t data ) {
+                /* paranoid */ verify( data != 0 );
+                // Prepare the data we need
+                __attribute((unused)) int len   = 0;
+                __attribute((unused)) int block = 0;
+                uint32_t cnt = *ring.submit_q.num;
+                uint32_t mask = *ring.submit_q.mask;
+                disable_interrupts();
+                        uint32_t off = __tls_rand();
+                enable_interrupts( __cfaabi_dbg_ctx );
+                // Loop around looking for an available spot
+                for() {
+                        // Look through the list starting at some offset
+                        for(i; cnt) {
+                                uint64_t expected = 0;
+                                uint32_t idx = (i + off) & mask;
+                                struct io_uring_sqe * sqe = &ring.submit_q.sqes[idx];
+                                volatile uint64_t * udata = (volatile uint64_t *)&sqe->user_data;
+                                if( *udata == expected &&
+                                        __atomic_compare_exchange_n( udata, &expected, data, true, __ATOMIC_SEQ_CST, __ATOMIC_RELAXED ) )
+                                {
+                                        // update statistics
+                                        __STATS__( false,
+                                                io.submit_q.alloc_avg.val   += len;
+                                                io.submit_q.alloc_avg.block += block;
+                                                io.submit_q.alloc_avg.cnt   += 1;
+                                        )
+                                        // Success return the data
+                                        return [sqe, idx];
+                                }
+                                verify(expected != data);
+                                len ++;
+                        }
+                        block++;
+                        yield();
+                }
+        }
+        static inline uint32_t __submit_to_ready_array( struct __io_data & ring, uint32_t idx, const uint32_t mask ) {
+                /* paranoid */ verify( idx <= mask   );
+                /* paranoid */ verify( idx != -1ul32 );
+                // We need to find a spot in the ready array
+                __attribute((unused)) int len   = 0;
+                __attribute((unused)) int block = 0;
+                uint32_t ready_mask = ring.submit_q.ready_cnt - 1;
+                disable_interrupts();
+                        uint32_t off = __tls_rand();
+                enable_interrupts( __cfaabi_dbg_ctx );
+                uint32_t picked;
+                LOOKING: for() {
+                        for(i; ring.submit_q.ready_cnt) {
+                                picked = (i + off) & ready_mask;
+                                uint32_t expected = -1ul32;
+                                if( __atomic_compare_exchange_n( &ring.submit_q.ready[picked], &expected, idx, true, __ATOMIC_SEQ_CST, __ATOMIC_RELAXED ) ) {
+                                        break LOOKING;
+                                }
+                                verify(expected != idx);
+                                len ++;
+                        }
+                        block++;
+                        if( try_lock(ring.submit_q.lock __cfaabi_dbg_ctx2) ) {
+                                __release_consumed_submission( ring );
+                                unlock( ring.submit_q.lock );
+                        }
+                        else {
+                                yield();
+                        }
+                }
+                // update statistics
+                __STATS__( false,
+                        io.submit_q.look_avg.val   += len;
+                        io.submit_q.look_avg.block += block;
+                        io.submit_q.look_avg.cnt   += 1;
+                )
+                return picked;
+        }
+        void __submit( struct io_context * ctx, uint32_t idx ) __attribute__((nonnull (1))) {
+                __io_data & ring = *ctx->thrd.ring;
+                // Get now the data we definetely need
+                volatile uint32_t * const tail = ring.submit_q.tail;
+                const uint32_t mask  = *ring.submit_q.mask;
+                // There are 2 submission schemes, check which one we are using
+                if( ring.poller_submits ) {
+                        // If the poller thread submits, then we just need to add this to the ready array
+                        __submit_to_ready_array( ring, idx, mask );
+                        post( ctx->thrd.sem );
+                        __cfadbg_print_safe( io, "Kernel I/O : Added %u to ready for %p\n", idx, active_thread() );
+                }
+                else if( ring.eager_submits ) {
+                        uint32_t picked = __submit_to_ready_array( ring, idx, mask );
+                        for() {
+                                yield();
+                                // If some one else collected our index, we are done
+                                #warning ABA problem
+                                if( ring.submit_q.ready[picked] != idx ) {
+                                        __STATS__( false,
+                                                io.submit_q.helped += 1;
+                                        )
+                                        return;
+                                }
+                                if( try_lock(ring.submit_q.lock __cfaabi_dbg_ctx2) ) {
+                                        __STATS__( false,
+                                                io.submit_q.leader += 1;
+                                        )
+                                        break;
+                                }
+                                __STATS__( false,
+                                        io.submit_q.busy += 1;
+                                )
+                        }
+                        // We got the lock
+                        unsigned to_submit = __collect_submitions( ring );
+                        int ret = __io_uring_enter( ring, to_submit, false );
+                        if( ret < 0 ) {
+                                unlock(ring.submit_q.lock);
+                                return;
+                        }
+                        /* paranoid */ verify( ret > 0 || to_submit == 0 || (ring.ring_flags & IORING_SETUP_SQPOLL) );
+                        // Release the consumed SQEs
+                        __release_consumed_submission( ring );
+                        // update statistics
+                        __STATS__( true,
+                                io.submit_q.submit_avg.rdy += to_submit;
+                                io.submit_q.submit_avg.csm += ret;
+                                io.submit_q.submit_avg.cnt += 1;
+                        )
+                        unlock(ring.submit_q.lock);
+                }
+                else {
+                        // get mutual exclusion
+                        lock(ring.submit_q.lock __cfaabi_dbg_ctx2);
+                        /* paranoid */ verifyf( ring.submit_q.sqes[ idx ].user_data != 0,
+                        /* paranoid */  "index %u already reclaimed\n"
+                        /* paranoid */  "head %u, prev %u, tail %u\n"
+                        /* paranoid */  "[-0: %u,-1: %u,-2: %u,-3: %u]\n",
+                        /* paranoid */  idx,
+                        /* paranoid */  *ring.submit_q.head, ring.submit_q.prev_head, *tail
+                        /* paranoid */  ,ring.submit_q.array[ ((*ring.submit_q.head) - 0) & (*ring.submit_q.mask) ]
+                        /* paranoid */  ,ring.submit_q.array[ ((*ring.submit_q.head) - 1) & (*ring.submit_q.mask) ]
+                        /* paranoid */  ,ring.submit_q.array[ ((*ring.submit_q.head) - 2) & (*ring.submit_q.mask) ]
+                        /* paranoid */  ,ring.submit_q.array[ ((*ring.submit_q.head) - 3) & (*ring.submit_q.mask) ]
+                        /* paranoid */ );
+                        // Append to the list of ready entries
+                        /* paranoid */ verify( idx <= mask );
+                        ring.submit_q.array[ (*tail) & mask ] = idx;
+                        __atomic_fetch_add(tail, 1ul32, __ATOMIC_SEQ_CST);
+                        // Submit however, many entries need to be submitted
+                        int ret = __io_uring_enter( ring, 1, false );
+                        if( ret < 0 ) {
+                                switch((int)errno) {
+                                default:
+                                        abort( "KERNEL ERROR: IO_URING SUBMIT - %s\n", strerror(errno) );
+                                }
+                        }
+                        // update statistics
+                        __STATS__( false,
+                                io.submit_q.submit_avg.csm += 1;
+                                io.submit_q.submit_avg.cnt += 1;
+                        )
+                        // Release the consumed SQEs
+                        __release_consumed_submission( ring );
+                        unlock(ring.submit_q.lock);
+                        __cfadbg_print_safe( io, "Kernel I/O : Performed io_submit for %p, returned %d\n", active_thread(), ret );
+                }
+        }
+        static unsigned __collect_submitions( struct __io_data & ring ) {
+                /* paranoid */ verify( ring.submit_q.ready != 0p );
+                /* paranoid */ verify( ring.submit_q.ready_cnt > 0 );
+                unsigned to_submit = 0;
+                uint32_t tail = *ring.submit_q.tail;
                 const uint32_t mask = *ring.submit_q.mask;
+                ring.submit_q.array[ (*tail) & mask ] = idx & mask;
+                __atomic_fetch_add(tail, 1ul32, __ATOMIC_SEQ_CST);
+                // Submit however, many entries need to be submitted
+                int ret = syscall( __NR_io_uring_enter, ring.fd, 1, 0, 0, 0p, 0);
+                if( ret < 0 ) {
+                        switch((int)errno) {
+                        default:
+                                abort( "KERNEL ERROR: IO_URING SUBMIT - %s\n", strerror(errno) );
+                        }
+                }
+                // update statistics
+                #if !defined(__CFA_NO_STATISTICS__)
+                        ring.submit_q.stats.submit_avg.val += 1;
+                        ring.submit_q.stats.submit_avg.cnt += 1;
+                #endif
+                unlock(ring.submit_q.lock);
+                // Make sure that idx was submitted
+                // Be careful to not get false positive if we cycled the entire list or that someone else submitted for us
+                __cfadbg_print_safe( io, "Kernel I/O : Performed io_submit for %p, returned %d\n", active_thread(), ret );
+        }
+        static inline void ?{}(struct io_uring_sqe & this, uint8_t opcode, int fd) {
+                this.opcode = opcode;
+                #if !defined(IOSQE_ASYNC)
+                        this.flags = 0;
+                #else
+                        this.flags = IOSQE_ASYNC;
+                #endif
+                this.ioprio = 0;
+                this.fd = fd;
+                this.off = 0;
+                this.addr = 0;
+                this.len = 0;
+                this.rw_flags = 0;
+                this.__pad2[0] = this.__pad2[1] = this.__pad2[2] = 0;
+        }
+        static inline void ?{}(struct io_uring_sqe & this, uint8_t opcode, int fd, void * addr, uint32_t len, uint64_t off ) {
+                (this){ opcode, fd };
+                this.off = off;
+                this.addr = (uint64_t)addr;
+                this.len = len;
+        }
+//=============================================================================================
+// I/O Interface
+//=============================================================================================
+        #define __submit_prelude \
+                struct __io_data & ring = *active_cluster()->io; \
+                struct io_uring_sqe * sqe; \
+                uint32_t idx; \
+                [sqe, idx] = __submit_alloc( ring );
+        #define __submit_wait \
+                io_user_data data = { 0, active_thread() }; \
+                /*__cfaabi_bits_print_safe( STDERR_FILENO, "Preparing user data %p for %p\n", &data, data.thrd );*/ \
+                sqe->user_data = (uint64_t)&data; \
+                __submit( ring, idx ); \
+                park( __cfaabi_dbg_ctx ); \
+                return data.result;
+                // Go through the list of ready submissions
+                for( i; ring.submit_q.ready_cnt ) {
+                        // replace any submission with the sentinel, to consume it.
+                        uint32_t idx = __atomic_exchange_n( &ring.submit_q.ready[i], -1ul32, __ATOMIC_RELAXED);
+                        // If it was already the sentinel, then we are done
+                        if( idx == -1ul32 ) continue;
+                        // If we got a real submission, append it to the list
+                        ring.submit_q.array[ (tail + to_submit) & mask ] = idx & mask;
+                        to_submit++;
+                }
+                // Increment the tail based on how many we are ready to submit
+                __atomic_fetch_add(ring.submit_q.tail, to_submit, __ATOMIC_SEQ_CST);
+                return to_submit;
+        }
+        static uint32_t __release_consumed_submission( struct __io_data & ring ) {
+                const uint32_t smask = *ring.submit_q.mask;
+                if( !try_lock(ring.submit_q.release_lock __cfaabi_dbg_ctx2) ) return 0;
+                uint32_t chead = *ring.submit_q.head;
+                uint32_t phead = ring.submit_q.prev_head;
+                ring.submit_q.prev_head = chead;
+                unlock(ring.submit_q.release_lock);
+                uint32_t count = chead - phead;
+                for( i; count ) {
+                        uint32_t idx = ring.submit_q.array[ (phead + i) & smask ];
+                        ring.submit_q.sqes[ idx ].user_data = 0;
+                }
+                return count;
+        }
 #endif
-// Some forward declarations
-extern "C" {
-        #include <unistd.h>
-        #include <sys/types.h>
-        #include <sys/socket.h>
-        #include <sys/syscall.h>
-#if defined(HAVE_PREADV2)
-        struct iovec;
-        extern ssize_t preadv2 (int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags);
-#endif
-#if defined(HAVE_PWRITEV2)
-        struct iovec;
-        extern ssize_t pwritev2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags);
-#endif
-        extern int fsync(int fd);
-        extern int sync_file_range(int fd, int64_t offset, int64_t nbytes, unsigned int flags);
-        struct msghdr;
-        struct sockaddr;
-        extern ssize_t sendmsg(int sockfd, const struct msghdr *msg, int flags);
-        extern ssize_t recvmsg(int sockfd, struct msghdr *msg, int flags);
-        extern ssize_t send(int sockfd, const void *buf, size_t len, int flags);
-        extern ssize_t recv(int sockfd, void *buf, size_t len, int flags);
-        extern int accept4(int sockfd, struct sockaddr *addr, socklen_t *addrlen, int flags);
-        extern int connect(int sockfd, const struct sockaddr *addr, socklen_t addrlen);
-        extern int fallocate(int fd, int mode, uint64_t offset, uint64_t len);
-        extern int posix_fadvise(int fd, uint64_t offset, uint64_t len, int advice);
-        extern int madvise(void *addr, size_t length, int advice);
-        extern int openat(int dirfd, const char *pathname, int flags, mode_t mode);
-        extern int close(int fd);
-        extern ssize_t read (int fd, void *buf, size_t count);
+}
-//-----------------------------------------------------------------------------
-// Asynchronous operations
-#if defined(HAVE_PREADV2)
-        ssize_t cfa_preadv2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags) {
-                #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_READV)
-                        return preadv2(fd, iov, iovcnt, offset, flags);
-                #else
-                        __submit_prelude
-                        (*sqe){ IORING_OP_READV, fd, iov, iovcnt, offset };
-                        __submit_wait
-                #endif
+        }
-#endif
-#if defined(HAVE_PWRITEV2)
-        ssize_t cfa_pwritev2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags) {
-                #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_WRITEV)
-                        return pwritev2(fd, iov, iovcnt, offset, flags);
-                #else
-                        __submit_prelude
-                        (*sqe){ IORING_OP_WRITEV, fd, iov, iovcnt, offset };
-                        __submit_wait
-                #endif
+        }
-#endif
-int cfa_fsync(int fd) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_FSYNC)
-                return fsync(fd);
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_FSYNC, fd };
-                __submit_wait
-        #endif
+}
-int cfa_sync_file_range(int fd, int64_t offset, int64_t nbytes, unsigned int flags) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_SYNC_FILE_RANGE)
-                return sync_file_range(fd, offset, nbytes, flags);
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_SYNC_FILE_RANGE, fd };
-                sqe->off = offset;
-                sqe->len = nbytes;
-                sqe->sync_range_flags = flags;
-                __submit_wait
-        #endif
+}
-ssize_t cfa_sendmsg(int sockfd, const struct msghdr *msg, int flags) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_SENDMSG)
-                return sendmsg(sockfd, msg, flags);
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_SENDMSG, sockfd, msg, 1, 0 };
-                sqe->msg_flags = flags;
-                __submit_wait
-        #endif
+}
-ssize_t cfa_recvmsg(int sockfd, struct msghdr *msg, int flags) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_RECVMSG)
-                return recvmsg(sockfd, msg, flags);
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_RECVMSG, sockfd, msg, 1, 0 };
-                sqe->msg_flags = flags;
-                __submit_wait
-        #endif
+}
-ssize_t cfa_send(int sockfd, const void *buf, size_t len, int flags) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_SEND)
-                return send( sockfd, buf, len, flags );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_SEND, sockfd };
-                sqe->addr = (uint64_t)buf;
-                sqe->len = len;
-                sqe->msg_flags = flags;
-                __submit_wait
-        #endif
+}
-ssize_t cfa_recv(int sockfd, void *buf, size_t len, int flags) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_RECV)
-                return recv( sockfd, buf, len, flags );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_RECV, sockfd };
-                sqe->addr = (uint64_t)buf;
-                sqe->len = len;
-                sqe->msg_flags = flags;
-                __submit_wait
-        #endif
+}
-int cfa_accept4(int sockfd, struct sockaddr *addr, socklen_t *addrlen, int flags) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_ACCEPT)
-                return accept4( sockfd, addr, addrlen, flags );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_ACCEPT, sockfd };
-                sqe->addr = addr;
-                sqe->addr2 = addrlen;
-                sqe->accept_flags = flags;
-                __submit_wait
-        #endif
+}
-int cfa_connect(int sockfd, const struct sockaddr *addr, socklen_t addrlen) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_CONNECT)
-                return connect( sockfd, addr, addrlen );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_CONNECT, sockfd };
-                sqe->addr = (uint64_t)addr;
-                sqe->off = addrlen;
-                __submit_wait
-        #endif
+}
-int cfa_fallocate(int fd, int mode, uint64_t offset, uint64_t len) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_FALLOCATE)
-                return fallocate( fd, mode, offset, len );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_FALLOCATE, fd };
-                sqe->off = offset;
-                sqe->len = length;
-                sqe->mode = mode;
-                __submit_wait
-        #endif
+}
-int cfa_fadvise(int fd, uint64_t offset, uint64_t len, int advice) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_FADVISE)
-                return posix_fadvise( fd, offset, len, advice );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_FADVISE, fd };
-                sqe->off = (uint64_t)offset;
-                sqe->len = length;
-                sqe->fadvise_advice = advice;
-                __submit_wait
-        #endif
+}
-int cfa_madvise(void *addr, size_t length, int advice) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_MADVISE)
-                return madvise( addr, length, advice );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_MADVISE, 0 };
-                sqe->addr = (uint64_t)addr;
-                sqe->len = length;
-                sqe->fadvise_advice = advice;
-                __submit_wait
-        #endif
+}
-int cfa_openat(int dirfd, const char *pathname, int flags, mode_t mode) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_OPENAT)
-                return openat( dirfd, pathname, flags, mode );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_OPENAT, dirfd };
-                sqe->addr = (uint64_t)pathname;
-                sqe->open_flags = flags;
-                sqe->mode = mode;
-                __submit_wait
-        #endif
+}
-int cfa_close(int fd) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_CLOSE)
-                return close( fd );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_CLOSE, fd };
-                __submit_wait
-        #endif
+}
-ssize_t cfa_read(int fd, void *buf, size_t count) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_READ)
-                return read( fd, buf, count );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_READ, fd, buf, count, 0 };
-                __submit_wait
-        #endif
+}
-ssize_t cfa_write(int fd, void *buf, size_t count) {
-        #if !defined(HAVE_LINUX_IO_URING_H) || !defined(IORING_OP_WRITE)
-                return read( fd, buf, count );
-        #else
-                __submit_prelude
-                (*sqe){ IORING_OP_WRITE, fd, buf, count, 0 };
-                __submit_wait
-        #endif
+}
-//-----------------------------------------------------------------------------
-// Check if a function is asynchronous
-// Macro magic to reduce the size of the following switch case
-#define IS_DEFINED_APPLY(f, ...) f(__VA_ARGS__)
-#define IS_DEFINED_SECOND(first, second, ...) second
-#define IS_DEFINED_TEST(expansion) _CFA_IO_FEATURE_##expansion
-#define IS_DEFINED(macro) IS_DEFINED_APPLY( IS_DEFINED_SECOND,IS_DEFINED_TEST(macro) false, true)
-bool has_user_level_blocking( fptr_t func ) {
-        #if defined(HAVE_LINUX_IO_URING_H)
-                #if defined(HAVE_PREADV2)
-                        if( /*func == (fptr_t)preadv2 || */
-                                func == (fptr_t)cfa_preadv2 )
-                                #define _CFA_IO_FEATURE_IORING_OP_READV ,
-                                return IS_DEFINED(IORING_OP_READV);
-                #endif
-                #if defined(HAVE_PWRITEV2)
-                        if( /*func == (fptr_t)pwritev2 || */
-                                func == (fptr_t)cfa_pwritev2 )
-                                #define _CFA_IO_FEATURE_IORING_OP_WRITEV ,
-                                return IS_DEFINED(IORING_OP_WRITEV);
-                #endif
-                if( /*func == (fptr_t)fsync || */
-                        func == (fptr_t)cfa_fsync )
-                        #define _CFA_IO_FEATURE_IORING_OP_FSYNC ,
-                        return IS_DEFINED(IORING_OP_FSYNC);
-                if( /*func == (fptr_t)ync_file_range || */
-                        func == (fptr_t)cfa_sync_file_range )
-                        #define _CFA_IO_FEATURE_IORING_OP_SYNC_FILE_RANGE ,
-                        return IS_DEFINED(IORING_OP_SYNC_FILE_RANGE);
-                if( /*func == (fptr_t)sendmsg || */
-                        func == (fptr_t)cfa_sendmsg )
-                        #define _CFA_IO_FEATURE_IORING_OP_SENDMSG ,
-                        return IS_DEFINED(IORING_OP_SENDMSG);
-                if( /*func == (fptr_t)recvmsg || */
-                        func == (fptr_t)cfa_recvmsg )
-                        #define _CFA_IO_FEATURE_IORING_OP_RECVMSG ,
-                        return IS_DEFINED(IORING_OP_RECVMSG);
-                if( /*func == (fptr_t)send || */
-                        func == (fptr_t)cfa_send )
-                        #define _CFA_IO_FEATURE_IORING_OP_SEND ,
-                        return IS_DEFINED(IORING_OP_SEND);
-                if( /*func == (fptr_t)recv || */
-                        func == (fptr_t)cfa_recv )
-                        #define _CFA_IO_FEATURE_IORING_OP_RECV ,
-                        return IS_DEFINED(IORING_OP_RECV);
-                if( /*func == (fptr_t)accept4 || */
-                        func == (fptr_t)cfa_accept4 )
-                        #define _CFA_IO_FEATURE_IORING_OP_ACCEPT ,
-                        return IS_DEFINED(IORING_OP_ACCEPT);
-                if( /*func == (fptr_t)connect || */
-                        func == (fptr_t)cfa_connect )
-                        #define _CFA_IO_FEATURE_IORING_OP_CONNECT ,
-                        return IS_DEFINED(IORING_OP_CONNECT);
-                if( /*func == (fptr_t)fallocate || */
-                        func == (fptr_t)cfa_fallocate )
-                        #define _CFA_IO_FEATURE_IORING_OP_FALLOCATE ,
-                        return IS_DEFINED(IORING_OP_FALLOCATE);
-                if( /*func == (fptr_t)posix_fadvise || */
-                        func == (fptr_t)cfa_fadvise )
-                        #define _CFA_IO_FEATURE_IORING_OP_FADVISE ,
-                        return IS_DEFINED(IORING_OP_FADVISE);
-                if( /*func == (fptr_t)madvise || */
-                        func == (fptr_t)cfa_madvise )
-                        #define _CFA_IO_FEATURE_IORING_OP_MADVISE ,
-                        return IS_DEFINED(IORING_OP_MADVISE);
-                if( /*func == (fptr_t)openat || */
-                        func == (fptr_t)cfa_openat )
-                        #define _CFA_IO_FEATURE_IORING_OP_OPENAT ,
-                        return IS_DEFINED(IORING_OP_OPENAT);
-                if( /*func == (fptr_t)close || */
-                        func == (fptr_t)cfa_close )
-                        #define _CFA_IO_FEATURE_IORING_OP_CLOSE ,
-                        return IS_DEFINED(IORING_OP_CLOSE);
-                if( /*func == (fptr_t)read || */
-                        func == (fptr_t)cfa_read )
-                        #define _CFA_IO_FEATURE_IORING_OP_READ ,
-                        return IS_DEFINED(IORING_OP_READ);
-                if( /*func == (fptr_t)write || */
-                        func == (fptr_t)cfa_write )
-                        #define _CFA_IO_FEATURE_IORING_OP_WRITE ,
-                        return IS_DEFINED(IORING_OP_WRITE);
-        #endif
-        return false;
+}

libcfa/src/concurrency/iofwd.hfa

-              r07d867b
+              r22f94a4
 #pragma once
+ssize_t cfa_preadv2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags);
+ssize_t cfa_pwritev2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags);
+int cfa_fsync(int fd);
+int cfa_sync_file_range(int fd, int64_t offset, int64_t nbytes, unsigned int flags);
+ssize_t cfa_sendmsg(int sockfd, const struct msghdr *msg, int flags);
+ssize_t cfa_recvmsg(int sockfd, struct msghdr *msg, int flags);
+ssize_t cfa_send(int sockfd, const void *buf, size_t len, int flags);
+ssize_t cfa_recv(int sockfd, void *buf, size_t len, int flags);
+int cfa_accept4(int sockfd, struct sockaddr *addr, socklen_t *addrlen, int flags);
+int cfa_connect(int sockfd, const struct sockaddr *addr, socklen_t addrlen);
+int cfa_fallocate(int fd, int mode, uint64_t offset, uint64_t len);
+int cfa_fadvise(int fd, uint64_t offset, uint64_t len, int advice);
+int cfa_madvise(void *addr, size_t length, int advice);
+int cfa_openat(int dirfd, const char *pathname, int flags, mode_t mode);
+int cfa_close(int fd);
+int cfa_statx(int dirfd, const char *pathname, int flags, unsigned int mask, struct statx *statxbuf);
+ssize_t cfa_read(int fd, void *buf, size_t count);
+ssize_t cfa_write(int fd, void *buf, size_t count)
+#include <unistd.h>
+extern "C" {
+        #include <sys/types.h>
+        #if CFA_HAVE_LINUX_IO_URING_H
+                #include <linux/io_uring.h>
+        #endif
+}
+#include "bits/defs.hfa"
+#include "time.hfa"
+#if defined(CFA_HAVE_IOSQE_FIXED_FILE)
+        #define CFA_IO_FIXED_FD1 IOSQE_FIXED_FILE
+#endif
+#if defined(CFA_HAVE_SPLICE_F_FD_IN_FIXED)
+        #define CFA_IO_FIXED_FD2 SPLICE_F_FD_IN_FIXED
+#endif
+#if defined(CFA_HAVE_IOSQE_IO_DRAIN)
+        #define CFA_IO_DRAIN IOSQE_IO_DRAIN
+#endif
+#if defined(CFA_HAVE_IOSQE_ASYNC)
+        #define CFA_IO_ASYNC IOSQE_ASYNC
+#endif
+struct cluster;
+struct io_context;
+struct io_cancellation;
+struct iovec;
+struct msghdr;
+struct sockaddr;
+struct statx;
+extern ssize_t cfa_preadv2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_pwritev2(int fd, const struct iovec *iov, int iovcnt, off_t offset, int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_fsync(int fd, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_sync_file_range(int fd, int64_t offset, int64_t nbytes, unsigned int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_sendmsg(int sockfd, const struct msghdr *msg, int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_recvmsg(int sockfd, struct msghdr *msg, int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_send(int sockfd, const void *buf, size_t len, int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_recv(int sockfd, void *buf, size_t len, int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_accept4(int sockfd, struct sockaddr *addr, socklen_t *addrlen, int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_connect(int sockfd, const struct sockaddr *addr, socklen_t addrlen, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_fallocate(int fd, int mode, uint64_t offset, uint64_t len, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_fadvise(int fd, uint64_t offset, uint64_t len, int advice, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_madvise(void *addr, size_t length, int advice, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_openat(int dirfd, const char *pathname, int flags, mode_t mode, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_close(int fd, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern int cfa_statx(int dirfd, const char *pathname, int flags, unsigned int mask, struct statx *statxbuf, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_read(int fd, void *buf, size_t count, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_write(int fd, void *buf, size_t count, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_splice(int fd_in, loff_t *off_in, int fd_out, loff_t *off_out, size_t len, unsigned int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
+extern ssize_t cfa_tee(int fd_in, int fd_out, size_t len, unsigned int flags, int submit_flags = 0, Duration timeout = -1`s, io_cancellation * cancellation = 0p, io_context * context = 0p);
 //-----------------------------------------------------------------------------
 // Check if a function is blocks a only the user thread
 bool has_user_level_blocking( fptr_t func );
+//-----------------------------------------------------------------------------
+void register_fixed_files( io_context & ctx , int * files, unsigned count );
+void register_fixed_files( cluster    & cltr, int * files, unsigned count );

libcfa/src/concurrency/kernel.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Tue Jan 17 12:27:26 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 13:03:15 2020
 // Update Count     : 58
+// Last Modified On : Thu Jul  9 06:22:54 2020
+// Update Count     : 66
 //
 …
 //C Includes
-#include <stddef.h>
 #include <errno.h>
-#include <string.h>
-extern "C" {
 #include <stdio.h>
-#include <fenv.h>
-#include <sys/resource.h>
 #include <signal.h>
 #include <unistd.h>
-#include <limits.h>                                                                             // PTHREAD_STACK_MIN
-#include <sys/mman.h>                                                                   // mprotect
+}
 //CFA Includes
-#include "time.hfa"
 #include "kernel_private.hfa"
 #include "preemption.hfa"
-#include "startup.hfa"
 //Private includes
 …
 // Some assembly required
 #if defined( __i386 )
-        #define CtxGet( ctx )        \
-                __asm__ volatile (     \
-                        "movl %%esp,%0\n"\
-                        "movl %%ebp,%1\n"\
-                        : "=rm" (ctx.SP),\
-                                "=rm" (ctx.FP) \
+                )
         // mxcr : SSE Status and Control bits (control bits are preserved across function calls)
         // fcw  : X87 FPU control word (preserved across function calls)
 …
 #elif defined( __x86_64 )
-        #define CtxGet( ctx )        \
-                __asm__ volatile (     \
-                        "movq %%rsp,%0\n"\
-                        "movq %%rbp,%1\n"\
-                        : "=rm" (ctx.SP),\
-                                "=rm" (ctx.FP) \
+                )
         #define __x87_store         \
                 uint32_t __mxcr;      \
 …
 #elif defined( __ARM_ARCH )
-#define CtxGet( ctx ) __asm__ ( \
-                "mov %0,%%sp\n"   \
-                "mov %1,%%r11\n"   \
-        : "=rm" (ctx.SP), "=rm" (ctx.FP) )
 #else
         #error unknown hardware architecture
 #endif
+//-----------------------------------------------------------------------------
+//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 ) ));
+extern $thread * mainThread;
+extern processor * mainProcessor;
 //-----------------------------------------------------------------------------
 // Kernel Scheduling logic
 static $thread * __next_thread(cluster * this);
+static $thread * __next_thread_slow(cluster * this);
 static void __run_thread(processor * this, $thread * dst);
+static $thread * __halt(processor * this);
+static bool __wake_one(cluster * cltr, bool was_empty);
+static bool __wake_proc(processor *);
+//-----------------------------------------------------------------------------
+// Kernel storage
+KERNEL_STORAGE(cluster,         mainCluster);
+KERNEL_STORAGE(processor,       mainProcessor);
+KERNEL_STORAGE($thread, mainThread);
+KERNEL_STORAGE(__stack_t,       mainThreadCtx);
+cluster     * mainCluster;
+processor   * mainProcessor;
+$thread * mainThread;
+extern "C" {
+        struct { __dllist_t(cluster) list; __spinlock_t lock; } __cfa_dbg_global_clusters;
+}
+size_t __page_size = 0;
+//-----------------------------------------------------------------------------
+// Global state
+thread_local struct KernelThreadData kernelTLS __attribute__ ((tls_model ( "initial-exec" ))) = {
+        NULL,                                                                                           // cannot use 0p
+        NULL,
+        { 1, false, false },
+u //this should be seeded better but due to a bug calling rdtsc doesn't work
+};
+//-----------------------------------------------------------------------------
+// Struct to steal stack
+struct current_stack_info_t {
+        __stack_t * storage;                                                            // pointer to stack object
+        void * base;                                                                            // base of stack
+        void * limit;                                                                           // stack grows towards stack limit
+        void * context;                                                                         // address of cfa_context_t
+};
+void ?{}( current_stack_info_t & this ) {
+        __stack_context_t ctx;
+        CtxGet( ctx );
+        this.base = ctx.FP;
+        rlimit r;
+        getrlimit( RLIMIT_STACK, &r);
+        size_t size = r.rlim_cur;
+        this.limit = (void *)(((intptr_t)this.base) - size);
+        this.context = &storage_mainThreadCtx;
+}
+//-----------------------------------------------------------------------------
+// Main thread construction
+void ?{}( $coroutine & this, current_stack_info_t * info) with( this ) {
+        stack.storage = info->storage;
+        with(*stack.storage) {
+                limit     = info->limit;
+                base      = info->base;
+        }
+        __attribute__((may_alias)) intptr_t * istorage = (intptr_t*) &stack.storage;
+        *istorage |= 0x1;
+        name = "Main Thread";
+        state = Start;
+        starter = 0p;
+        last = 0p;
+        cancellation = 0p;
+}
+void ?{}( $thread & this, current_stack_info_t * info) with( this ) {
+        state = Start;
+        self_cor{ info };
+        curr_cor = &self_cor;
+        curr_cluster = mainCluster;
+        self_mon.owner = &this;
+        self_mon.recursion = 1;
+        self_mon_p = &self_mon;
+        next = 0p;
+        node.next = 0p;
+        node.prev = 0p;
+        doregister(curr_cluster, this);
+        monitors{ &self_mon_p, 1, (fptr_t)0 };
+}
+//-----------------------------------------------------------------------------
+// Processor coroutine
+void ?{}(processorCtx_t & this) {
+}
+// Construct the processor context of non-main processors
+static void ?{}(processorCtx_t & this, processor * proc, current_stack_info_t * info) {
+        (this.__cor){ info };
+        this.proc = proc;
+}
+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;
+        pending_preemption = false;
+        runner.proc = &this;
+        idle{};
+        __cfadbg_print_safe(runtime_core, "Kernel : Starting core %p\n", &this);
+        this.stack = __create_pthread( &this.kernel_thread, __invoke_processor, (void *)&this );
+        __cfadbg_print_safe(runtime_core, "Kernel : core %p created\n", &this);
+}
+void ^?{}(processor & this) with( this ){
+        if( ! __atomic_load_n(&do_terminate, __ATOMIC_ACQUIRE) ) {
+                __cfadbg_print_safe(runtime_core, "Kernel : core %p signaling termination\n", &this);
+                __atomic_store_n(&do_terminate, true, __ATOMIC_RELAXED);
+                __wake_proc( &this );
+                P( terminated );
+                verify( kernelTLS.this_processor != &this);
+        }
+        int err = pthread_join( kernel_thread, 0p );
+        if( err != 0 ) abort("KERNEL ERROR: joining processor %p caused error %s\n", &this, strerror(err));
+        free( this.stack );
+}
+void ?{}(cluster & this, const char name[], Duration preemption_rate, int io_flags) with( this ) {
+        this.name = name;
+        this.preemption_rate = preemption_rate;
+        ready_queue{};
+        ready_queue_lock{};
+        #if !defined(__CFA_NO_STATISTICS__)
+                print_stats = false;
+        #endif
+        procs{ __get };
+        idles{ __get };
+        threads{ __get };
+        __kernel_io_startup( this, io_flags, &this == mainCluster );
+        doregister(this);
+}
+void ^?{}(cluster & this) {
+        __kernel_io_shutdown( this, &this == mainCluster );
+        unregister(this);
+}
+static void __wake_one(struct __processor_id_t * id, cluster * cltr);
+static void push  (__cluster_idles & idles, processor & proc);
+static void remove(__cluster_idles & idles, processor & proc);
+static [unsigned idle, unsigned total, * processor] query( & __cluster_idles idles );
 //=============================================================================================
 …
         __cfadbg_print_safe(runtime_core, "Kernel : core %p starting\n", this);
+        doregister(this->cltr, this);
+        #if !defined(__CFA_NO_STATISTICS__)
+                if( this->print_halts ) {
+                        __cfaabi_bits_print_safe( STDOUT_FILENO, "Processor : %d - %s (%p)\n", this->id, this->name, (void*)this);
+                }
+        #endif
+        {
 …
                 $thread * readyThread = 0p;
+                for( unsigned int spin_count = 0; ! __atomic_load_n(&this->do_terminate, __ATOMIC_SEQ_CST); spin_count++ ) {
+                MAIN_LOOP:
+                for() {
                         // Try to get the next thread
                         readyThread = __next_thread( this->cltr );
+                        // If no ready thread
+                        if( readyThread == 0p ) {
+                                // Block until a thread is ready
+                                readyThread = __halt(this);
+                        if( !readyThread ) {
+                                readyThread = __next_thread_slow( this->cltr );
+                        }
+                        // Check if we actually found a thread
+                        if( readyThread ) {
+                                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                                /* paranoid */ verifyf( readyThread->state == Ready || readyThread->preempted != __NO_PREEMPTION, "state : %d, preempted %d\n", readyThread->state, readyThread->preempted);
+                                /* paranoid */ verifyf( readyThread->next == 0p, "Expected null got %p", readyThread->next );
+                                // We found a thread run it
+                                __run_thread(this, readyThread);
+                                /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                        HALT:
+                        if( !readyThread ) {
+                                // Don't block if we are done
+                                if( __atomic_load_n(&this->do_terminate, __ATOMIC_SEQ_CST) ) break MAIN_LOOP;
+                                #if !defined(__CFA_NO_STATISTICS__)
+                                        __tls_stats()->ready.sleep.halts++;
+                                #endif
+                                // Push self to idle stack
+                                push(this->cltr->idles, * this);
+                                // Confirm the ready-queue is empty
+                                readyThread = __next_thread_slow( this->cltr );
+                                if( readyThread ) {
+                                        // A thread was found, cancel the halt
+                                        remove(this->cltr->idles, * this);
+                                        #if !defined(__CFA_NO_STATISTICS__)
+                                                __tls_stats()->ready.sleep.cancels++;
+                                        #endif
+                                        // continue the mai loop
+                                        break HALT;
+                                }
+                                #if !defined(__CFA_NO_STATISTICS__)
+                                        if(this->print_halts) {
+                                                __cfaabi_bits_print_safe( STDOUT_FILENO, "PH:%d - %lld 0\n", this->id, rdtscl());
+                                        }
+                                #endif
+                                wait( this->idle );
+                                #if !defined(__CFA_NO_STATISTICS__)
+                                        if(this->print_halts) {
+                                                __cfaabi_bits_print_safe( STDOUT_FILENO, "PH:%d - %lld 1\n", this->id, rdtscl());
+                                        }
+                                #endif
+                                // We were woken up, remove self from idle
+                                remove(this->cltr->idles, * this);
+                                // DON'T just proceed, start looking again
+                                continue MAIN_LOOP;
+                        }
+                        /* paranoid */ verify( readyThread );
+                        // We found a thread run it
+                        __run_thread(this, readyThread);
+                        // Are we done?
+                        if( __atomic_load_n(&this->do_terminate, __ATOMIC_SEQ_CST) ) break MAIN_LOOP;
+                }
 …
+        }
-        unregister(this->cltr, this);
         V( this->terminated );
+        if(this == mainProcessor) {
+                // HACK : the coroutine context switch expects this_thread to be set
+                // and it make sense for it to be set in all other cases except here
+                // fake it
+                kernelTLS.this_thread = mainThread;
+        }
         __cfadbg_print_safe(runtime_core, "Kernel : core %p terminated\n", this);
-        // HACK : the coroutine context switch expects this_thread to be set
-        // and it make sense for it to be set in all other cases except here
-        // fake it
-        if( this == mainProcessor ) kernelTLS.this_thread = mainThread;
+}
 …
 // from the processor coroutine to the target thread
 static void __run_thread(processor * this, $thread * thrd_dst) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ verifyf( thrd_dst->state == Ready || thrd_dst->preempted != __NO_PREEMPTION, "state : %d, preempted %d\n", thrd_dst->state, thrd_dst->preempted);
+        /* paranoid */ verifyf( thrd_dst->link.next == 0p, "Expected null got %p", thrd_dst->link.next );
+        __builtin_prefetch( thrd_dst->context.SP );
         $coroutine * proc_cor = get_coroutine(this->runner);
 …
         // Actually run the thread
         RUNNING:  while(true) {
+                if(unlikely(thrd_dst->preempted)) {
+                        thrd_dst->preempted = __NO_PREEMPTION;
+                        verify(thrd_dst->state == Active  || thrd_dst->state == Rerun);
+                } else {
+                        verify(thrd_dst->state == Blocked || thrd_dst->state == Ready); // Ready means scheduled normally, blocked means rerun
+                        thrd_dst->state = Active;
+                }
+                thrd_dst->preempted = __NO_PREEMPTION;
+                thrd_dst->state = Active;
                 __cfaabi_dbg_debug_do(
 …
                 /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+                /* paranoid */ verify( kernelTLS.this_thread == thrd_dst );
                 /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) < ((uintptr_t)__get_stack(thrd_dst->curr_cor)->base ) || thrd_dst->curr_cor == proc_cor, "ERROR : Destination $thread %p has been corrupted.\n StackPointer too small.\n", thrd_dst ); // add escape condition if we are setting up the processor
                 /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) > ((uintptr_t)__get_stack(thrd_dst->curr_cor)->limit) || thrd_dst->curr_cor == proc_cor, "ERROR : Destination $thread %p has been corrupted.\n StackPointer too large.\n", thrd_dst ); // add escape condition if we are setting up the processor
 …
                 /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) > ((uintptr_t)__get_stack(thrd_dst->curr_cor)->limit), "ERROR : Destination $thread %p has been corrupted.\n StackPointer too large.\n", thrd_dst );
                 /* paranoid */ verifyf( ((uintptr_t)thrd_dst->context.SP) < ((uintptr_t)__get_stack(thrd_dst->curr_cor)->base ), "ERROR : Destination $thread %p has been corrupted.\n StackPointer too small.\n", thrd_dst );
+                /* paranoid */ verify( kernelTLS.this_thread == thrd_dst );
                 /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
 …
                 if(unlikely(thrd_dst->preempted != __NO_PREEMPTION)) {
                         // The thread was preempted, reschedule it and reset the flag
                         __schedule_thread( thrd_dst );
+                        __schedule_thread( (__processor_id_t*)this, thrd_dst );
                         break RUNNING;
+                }
+                if(unlikely(thrd_dst->state == Halted)) {
+                        // The thread has halted, it should never be scheduled/run again
+                        // We may need to wake someone up here since
+                        unpark( this->destroyer __cfaabi_dbg_ctx2 );
+                        this->destroyer = 0p;
+                        break RUNNING;
+                }
+                /* paranoid */ verify( thrd_dst->state == Active );
+                thrd_dst->state = Blocked;
                 // set state of processor coroutine to active and the thread to inactive
+                static_assert(sizeof(thrd_dst->state) == sizeof(int));
+                enum coroutine_state old_state = __atomic_exchange_n(&thrd_dst->state, Blocked, __ATOMIC_SEQ_CST);
+                __cfaabi_dbg_debug_do( thrd_dst->park_result = old_state; )
+                switch(old_state) {
+                        case Halted:
+                                // The thread has halted, it should never be scheduled/run again, leave it back to Halted and move on
+                                thrd_dst->state = Halted;
+                                // We may need to wake someone up here since
+                                unpark( this->destroyer __cfaabi_dbg_ctx2 );
+                                this->destroyer = 0p;
+                                break RUNNING;
+                        case Active:
+                int old_ticket = __atomic_fetch_sub(&thrd_dst->ticket, 1, __ATOMIC_SEQ_CST);
+                __cfaabi_dbg_debug_do( thrd_dst->park_result = old_ticket; )
+                switch(old_ticket) {
+                        case 1:
                                 // This is case 1, the regular case, nothing more is needed
                                 break RUNNING;
                         case Rerun:
+                        case 2:
                                 // This is case 2, the racy case, someone tried to run this thread before it finished blocking
                                 // In this case, just run it again.
 …
                         default:
                                 // This makes no sense, something is wrong abort
                                 abort("Finished running a thread that was Blocked/Start/Primed %d\n", old_state);
+                                abort();
+                }
+        }
 …
         proc_cor->state = Active;
         kernelTLS.this_thread = 0p;
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+}
 …
         $coroutine * proc_cor = get_coroutine(kernelTLS.this_processor->runner);
         $thread * thrd_src = kernelTLS.this_thread;
+        #if !defined(__CFA_NO_STATISTICS__)
+                struct processor * last_proc = kernelTLS.this_processor;
+        #endif
         // Run the thread on this processor
 …
+        }
+        #if !defined(__CFA_NO_STATISTICS__)
+                if(last_proc != kernelTLS.this_processor) {
+                        __tls_stats()->ready.threads.migration++;
+                }
+        #endif
         /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
         /* paranoid */ verifyf( ((uintptr_t)thrd_src->context.SP) < ((uintptr_t)__get_stack(thrd_src->curr_cor)->base ), "ERROR : Returning $thread %p has been corrupted.\n StackPointer too small.\n", thrd_src );
         /* paranoid */ verifyf( ((uintptr_t)thrd_src->context.SP) > ((uintptr_t)__get_stack(thrd_src->curr_cor)->limit), "ERROR : Returning $thread %p has been corrupted.\n StackPointer too large.\n", thrd_src );
+}
-// KERNEL_ONLY
-// Context invoker for processors
-// This is the entry point for processors (kernel threads)
-// It effectively constructs a coroutine by stealing the pthread stack
-static void * __invoke_processor(void * arg) {
-        processor * proc = (processor *) arg;
-        kernelTLS.this_processor = proc;
-        kernelTLS.this_thread    = 0p;
-        kernelTLS.preemption_state.[enabled, disable_count] = [false, 1];
-        // SKULLDUGGERY: We want to create a context for the processor coroutine
-        // which is needed for the 2-step context switch. However, there is no reason
-        // to waste the perfectly valid stack create by pthread.
-        current_stack_info_t info;
-        __stack_t ctx;
-        info.storage = &ctx;
-        (proc->runner){ proc, &info };
-        __cfaabi_dbg_print_safe("Coroutine : created stack %p\n", get_coroutine(proc->runner)->stack.storage);
-        //Set global state
-        kernelTLS.this_thread = 0p;
-        //We now have a proper context from which to schedule threads
-        __cfadbg_print_safe(runtime_core, "Kernel : core %p created (%p, %p)\n", proc, &proc->runner, &ctx);
-        // SKULLDUGGERY: Since the coroutine doesn't have its own stack, we can't
-        // resume it to start it like it normally would, it will just context switch
-        // back to here. Instead directly call the main since we already are on the
-        // appropriate stack.
-        get_coroutine(proc->runner)->state = Active;
-        main( proc->runner );
-        get_coroutine(proc->runner)->state = Halted;
-        // Main routine of the core returned, the core is now fully terminated
-        __cfadbg_print_safe(runtime_core, "Kernel : core %p main ended (%p)\n", proc, &proc->runner);
-        return 0p;
+}
-static void Abort( int ret, const char func[] ) {
-        if ( ret ) {                                                                            // pthread routines return errno values
-                abort( "%s : internal error, error(%d) %s.", func, ret, strerror( ret ) );
-        } // if
-} // Abort
-void * __create_pthread( pthread_t * pthread, void * (*start)(void *), void * arg ) {
-        pthread_attr_t attr;
-        Abort( pthread_attr_init( &attr ), "pthread_attr_init" ); // initialize attribute
-        size_t stacksize;
-        // default stack size, normally defined by shell limit
-        Abort( pthread_attr_getstacksize( &attr, &stacksize ), "pthread_attr_getstacksize" );
-        assert( stacksize >= PTHREAD_STACK_MIN );
-        void * stack;
-        __cfaabi_dbg_debug_do(
-                stack = memalign( __page_size, stacksize + __page_size );
-                // pthread has no mechanism to create the guard page in user supplied stack.
-                if ( mprotect( stack, __page_size, PROT_NONE ) == -1 ) {
-                        abort( "mprotect : internal error, mprotect failure, error(%d) %s.", errno, strerror( errno ) );
-                } // if
-        );
-        __cfaabi_dbg_no_debug_do(
-                stack = malloc( stacksize );
-        );
-        Abort( pthread_attr_setstack( &attr, stack, stacksize ), "pthread_attr_setstack" );
-        Abort( pthread_create( pthread, &attr, start, arg ), "pthread_create" );
-        return stack;
+}
-// KERNEL_ONLY
-static void __kernel_first_resume( processor * this ) {
-        $thread * src = mainThread;
-        $coroutine * dst = get_coroutine(this->runner);
-        verify( ! kernelTLS.preemption_state.enabled );
-        kernelTLS.this_thread->curr_cor = dst;
-        __stack_prepare( &dst->stack, 65000 );
-        __cfactx_start(main, dst, this->runner, __cfactx_invoke_coroutine);
-        verify( ! kernelTLS.preemption_state.enabled );
-        dst->last = &src->self_cor;
-        dst->starter = dst->starter ? dst->starter : &src->self_cor;
-        // make sure the current state is still correct
-        /* paranoid */ verify(src->state == Ready);
-        // context switch to specified coroutine
-        verify( dst->context.SP );
-        __cfactx_switch( &src->context, &dst->context );
-        // when __cfactx_switch returns we are back in the src coroutine
-        mainThread->curr_cor = &mainThread->self_cor;
-        // make sure the current state has been update
-        /* paranoid */ verify(src->state == Active);
-        verify( ! kernelTLS.preemption_state.enabled );
+}
-// KERNEL_ONLY
-static void __kernel_last_resume( processor * this ) {
-        $coroutine * src = &mainThread->self_cor;
-        $coroutine * dst = get_coroutine(this->runner);
-        verify( ! kernelTLS.preemption_state.enabled );
-        verify( dst->starter == src );
-        verify( dst->context.SP );
-        // SKULLDUGGERY in debug the processors check that the
-        // stack is still within the limit of the stack limits after running a thread.
-        // that check doesn't make sense if we context switch to the processor using the
-        // coroutine semantics. Since this is a special case, use the current context
-        // info to populate these fields.
-        __cfaabi_dbg_debug_do(
-                __stack_context_t ctx;
-                CtxGet( ctx );
-                mainThread->context.SP = ctx.SP;
-                mainThread->context.FP = ctx.FP;
+        )
-        // context switch to the processor
-        __cfactx_switch( &src->context, &dst->context );
+}
 …
 // Scheduler routines
 // KERNEL ONLY
+void __schedule_thread( $thread * thrd ) with( *thrd->curr_cluster ) {
+void __schedule_thread( struct __processor_id_t * id, $thread * thrd ) {
+        /* paranoid */ verify( thrd );
+        /* paranoid */ verify( thrd->state != Halted );
         /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
         /* paranoid */ #if defined( __CFA_WITH_VERIFY__ )
         /* paranoid */ if( thrd->state == Blocked || thrd->state == Start ) assertf( thrd->preempted == __NO_PREEMPTION,
                           "Error inactive thread marked as preempted, state %d, preemption %d\n", thrd->state, thrd->preempted );
         /* paranoid */ if( thrd->preempted != __NO_PREEMPTION ) assertf(thrd->state == Active || thrd->state == Rerun,
                           "Error preempted thread marked as not currently running, state %d, preemption %d\n", thrd->state, thrd->preempted );
+        /* paranoid */  if( thrd->state == Blocked || thrd->state == Start ) assertf( thrd->preempted == __NO_PREEMPTION,
+                                        "Error inactive thread marked as preempted, state %d, preemption %d\n", thrd->state, thrd->preempted );
+        /* paranoid */  if( thrd->preempted != __NO_PREEMPTION ) assertf(thrd->state == Active,
+                                        "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 );
+        /* paranoid */ verifyf( thrd->link.next == 0p, "Expected null got %p", thrd->link.next );
         if (thrd->preempted == __NO_PREEMPTION) thrd->state = Ready;
+        lock  ( ready_queue_lock __cfaabi_dbg_ctx2 );
+        bool was_empty = !(ready_queue != 0);
+        append( ready_queue, thrd );
+        unlock( ready_queue_lock );
+        __wake_one(thrd->curr_cluster, was_empty);
+        ready_schedule_lock  ( id );
+                push( thrd->curr_cluster, thrd );
+                __wake_one(id, thrd->curr_cluster);
+        ready_schedule_unlock( id );
         /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
 …
 // KERNEL ONLY
+static $thread * __next_thread(cluster * this) with( *this ) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        lock( ready_queue_lock __cfaabi_dbg_ctx2 );
+        $thread * head = pop_head( ready_queue );
+        unlock( ready_queue_lock );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        return head;
+static inline $thread * __next_thread(cluster * this) with( *this ) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        ready_schedule_lock  ( (__processor_id_t*)kernelTLS.this_processor );
+                $thread * thrd = pop( this );
+        ready_schedule_unlock( (__processor_id_t*)kernelTLS.this_processor );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        return thrd;
+}
+// KERNEL ONLY
+static inline $thread * __next_thread_slow(cluster * this) with( *this ) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        ready_schedule_lock  ( (__processor_id_t*)kernelTLS.this_processor );
+                $thread * thrd = pop_slow( this );
+        ready_schedule_unlock( (__processor_id_t*)kernelTLS.this_processor );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        return thrd;
+}
 // KERNEL ONLY unpark with out disabling interrupts
+void __unpark( $thread * thrd __cfaabi_dbg_ctx_param2 ) {
+        static_assert(sizeof(thrd->state) == sizeof(int));
+void __unpark(  struct __processor_id_t * id, $thread * thrd __cfaabi_dbg_ctx_param2 ) {
         // record activity
         __cfaabi_dbg_record_thrd( *thrd, false, caller );
         enum coroutine_state old_state = __atomic_exchange_n(&thrd->state, Rerun, __ATOMIC_SEQ_CST);
         __cfaabi_dbg_debug_do( thrd->unpark_result = old_state; )
         switch(old_state) {
                 case Active:
+        int old_ticket = __atomic_fetch_add(&thrd->ticket, 1, __ATOMIC_SEQ_CST);
+        __cfaabi_dbg_debug_do( thrd->unpark_result = old_ticket; thrd->unpark_state = thrd->state; )
+        switch(old_ticket) {
+                case 1:
                         // Wake won the race, the thread will reschedule/rerun itself
                         break;
                 case Blocked:
+                case 0:
                         /* paranoid */ verify( ! thrd->preempted != __NO_PREEMPTION );
+                        /* paranoid */ verify( thrd->state == Blocked );
                         // Wake lost the race,
+                        thrd->state = Blocked;
+                        __schedule_thread( thrd );
+                        __schedule_thread( id, 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
 …
         disable_interrupts();
         __unpark( thrd __cfaabi_dbg_ctx_fwd2 );
+        __unpark( (__processor_id_t*)kernelTLS.this_processor, thrd __cfaabi_dbg_ctx_fwd2 );
         enable_interrupts( __cfaabi_dbg_ctx );
+}
 …
         $thread * thrd = kernelTLS.this_thread;
         /* paranoid */ verify(thrd->state == Active || thrd->state == Rerun);
+        /* paranoid */ verify(thrd->state == Active);
         // SKULLDUGGERY: It is possible that we are preempting this thread just before
 …
         // If that is the case, abandon the preemption.
         bool preempted = false;
         if(thrd->next == 0p) {
+        if(thrd->link.next == 0p) {
                 preempted = true;
                 thrd->preempted = reason;
 …
 //=============================================================================================
-// Kernel Setup logic
-//=============================================================================================
-//-----------------------------------------------------------------------------
-// Kernel boot procedures
-static void __kernel_startup(void) {
-        verify( ! kernelTLS.preemption_state.enabled );
-        __cfadbg_print_safe(runtime_core, "Kernel : Starting\n");
-        __page_size = sysconf( _SC_PAGESIZE );
-        __cfa_dbg_global_clusters.list{ __get };
-        __cfa_dbg_global_clusters.lock{};
-        // Initialize the main cluster
-        mainCluster = (cluster *)&storage_mainCluster;
-        (*mainCluster){"Main Cluster"};
-        __cfadbg_print_safe(runtime_core, "Kernel : Main cluster ready\n");
-        // Start by initializing the main thread
-        // SKULLDUGGERY: the mainThread steals the process main thread
-        // which will then be scheduled by the mainProcessor normally
-        mainThread = ($thread *)&storage_mainThread;
-        current_stack_info_t info;
-        info.storage = (__stack_t*)&storage_mainThreadCtx;
-        (*mainThread){ &info };
-        __cfadbg_print_safe(runtime_core, "Kernel : Main thread ready\n");
-        // Construct the processor context of the main processor
-        void ?{}(processorCtx_t & this, processor * proc) {
-                (this.__cor){ "Processor" };
-                this.__cor.starter = 0p;
-                this.proc = proc;
+        }
-        void ?{}(processor & this) with( this ) {
-                name = "Main Processor";
-                cltr = mainCluster;
-                terminated{ 0 };
-                do_terminate = false;
-                preemption_alarm = 0p;
-                pending_preemption = false;
-                kernel_thread = pthread_self();
-                runner{ &this };
-                __cfadbg_print_safe(runtime_core, "Kernel : constructed main processor context %p\n", &runner);
+        }
-        // Initialize the main processor and the main processor ctx
-        // (the coroutine that contains the processing control flow)
-        mainProcessor = (processor *)&storage_mainProcessor;
-        (*mainProcessor){};
-        //initialize the global state variables
-        kernelTLS.this_processor = mainProcessor;
-        kernelTLS.this_thread    = mainThread;
-        // Enable preemption
-        kernel_start_preemption();
-        // Add the main thread to the ready queue
-        // once resume is called on mainProcessor->runner the mainThread needs to be scheduled like any normal thread
-        __schedule_thread(mainThread);
-        // SKULLDUGGERY: Force a context switch to the main processor to set the main thread's context to the current UNIX
-        // context. Hence, the main thread does not begin through __cfactx_invoke_thread, like all other threads. The trick here is that
-        // mainThread is on the ready queue when this call is made.
-        __kernel_first_resume( kernelTLS.this_processor );
-        // THE SYSTEM IS NOW COMPLETELY RUNNING
-        // Now that the system is up, finish creating systems that need threading
-        __kernel_io_finish_start( *mainCluster );
-        __cfadbg_print_safe(runtime_core, "Kernel : Started\n--------------------------------------------------\n\n");
-        verify( ! kernelTLS.preemption_state.enabled );
-        enable_interrupts( __cfaabi_dbg_ctx );
-        verify( TL_GET( preemption_state.enabled ) );
+}
-static void __kernel_shutdown(void) {
-        //Before we start shutting things down, wait for systems that need threading to shutdown
-        __kernel_io_prepare_stop( *mainCluster );
-        /* paranoid */ verify( TL_GET( preemption_state.enabled ) );
-        disable_interrupts();
-        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
-        __cfadbg_print_safe(runtime_core, "\n--------------------------------------------------\nKernel : Shutting down\n");
-        // SKULLDUGGERY: Notify the mainProcessor it needs to terminates.
-        // When its coroutine terminates, it return control to the mainThread
-        // which is currently here
-        __atomic_store_n(&mainProcessor->do_terminate, true, __ATOMIC_RELEASE);
-        __kernel_last_resume( kernelTLS.this_processor );
-        mainThread->self_cor.state = Halted;
-        // THE SYSTEM IS NOW COMPLETELY STOPPED
-        // Disable preemption
-        kernel_stop_preemption();
-        // Destroy the main processor and its context in reverse order of construction
-        // These were manually constructed so we need manually destroy them
-        void ^?{}(processor & this) with( this ){
-                /* paranoid */ verify( this.do_terminate == true );
+        }
-        ^(*mainProcessor){};
-        // Final step, destroy the main thread since it is no longer needed
-        // Since we provided a stack to this taxk it will not destroy anything
-        /* paranoid */ verify(mainThread->self_cor.stack.storage == (__stack_t*)(((uintptr_t)&storage_mainThreadCtx)| 0x1));
-        ^(*mainThread){};
-        ^(*mainCluster){};
-        ^(__cfa_dbg_global_clusters.list){};
-        ^(__cfa_dbg_global_clusters.lock){};
-        __cfadbg_print_safe(runtime_core, "Kernel : Shutdown complete\n");
+}
-//=============================================================================================
 // Kernel Idle Sleep
 //=============================================================================================
-static $thread * __halt(processor * this) with( *this ) {
-        if( do_terminate ) return 0p;
-        // First, lock the cluster idle
-        lock( cltr->idle_lock __cfaabi_dbg_ctx2 );
-        // Check if we can find a thread
-        if( $thread * found = __next_thread( cltr ) ) {
-                unlock( cltr->idle_lock );
-                return found;
+        }
-        // Move this processor from the active list to the idle list
-        move_to_front(cltr->procs, cltr->idles, *this);
-        // Unlock the idle lock so we don't go to sleep with a lock
-        unlock    (cltr->idle_lock);
-        // We are ready to sleep
-        __cfadbg_print_safe(runtime_core, "Kernel : Processor %p ready to sleep\n", this);
-        wait( idle );
-        // We have woken up
-        __cfadbg_print_safe(runtime_core, "Kernel : Processor %p woke up and ready to run\n", this);
-        // Get ourself off the idle list
-        with( *cltr ) {
-                lock  (idle_lock __cfaabi_dbg_ctx2);
-                move_to_front(idles, procs, *this);
-                unlock(idle_lock);
+        }
-        // Don't check the ready queue again, we may not be in a position to run a thread
-        return 0p;
+}
 // Wake a thread from the front if there are any
+static bool __wake_one(cluster * this, __attribute__((unused)) bool force) {
+        // if we don't want to force check if we know it's false
+        // if( !this->idles.head && !force ) return false;
+        // First, lock the cluster idle
+        lock( this->idle_lock __cfaabi_dbg_ctx2 );
+        // Check if there is someone to wake up
+        if( !this->idles.head ) {
+                // Nope unlock and return false
+                unlock( this->idle_lock );
+                return false;
+        }
+        // Wake them up
+        __cfadbg_print_safe(runtime_core, "Kernel : waking Processor %p\n", this->idles.head);
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        post( this->idles.head->idle );
+        // Unlock and return true
+        unlock( this->idle_lock );
+        return true;
+static void __wake_one(struct __processor_id_t * id, cluster * this) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        /* paranoid */ verify( ready_schedule_islocked( id ) );
+        // Check if there is a sleeping processor
+        processor * p;
+        unsigned idle;
+        unsigned total;
+        [idle, total, p] = query(this->idles);
+        // If no one is sleeping, we are done
+        if( idle == 0 ) return;
+        // We found a processor, wake it up
+        post( p->idle );
+        #if !defined(__CFA_NO_STATISTICS__)
+                __tls_stats()->ready.sleep.wakes++;
+        #endif
+        /* paranoid */ verify( ready_schedule_islocked( id ) );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        return;
+}
 // Unconditionnaly wake a thread
 static bool __wake_proc(processor * this) {
+void __wake_proc(processor * this) {
         __cfadbg_print_safe(runtime_core, "Kernel : waking Processor %p\n", this);
 …
                 bool ret = post( this->idle );
         enable_interrupts( __cfaabi_dbg_ctx );
+        return ret;
+}
+static void push  (__cluster_idles & this, processor & proc) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        lock( this );
+                this.idle++;
+                /* paranoid */ verify( this.idle <= this.total );
+                insert_first(this.list, proc);
+        unlock( this );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+}
+static void remove(__cluster_idles & this, processor & proc) {
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+        lock( this );
+                this.idle--;
+                /* paranoid */ verify( this.idle >= 0 );
+                remove(proc);
+        unlock( this );
+        /* paranoid */ verify( ! kernelTLS.preemption_state.enabled );
+}
+static [unsigned idle, unsigned total, * processor] query( & __cluster_idles this ) {
+        for() {
+                uint64_t l = __atomic_load_n(&this.lock, __ATOMIC_SEQ_CST);
+                if( 1 == (l % 2) ) { Pause(); continue; }
+                unsigned idle    = this.idle;
+                unsigned total   = this.total;
+                processor * proc = &this.list`first;
+                if(l != __atomic_load_n(&this.lock, __ATOMIC_SEQ_CST)) { Pause(); continue; }
+                return [idle, total, proc];
+        }
+}
 …
 void kernel_abort_msg( void * kernel_data, char * abort_text, int abort_text_size ) {
         $thread * thrd = kernel_data;
+        $thread * thrd = ( $thread * ) kernel_data;
         if(thrd) {
 …
 //-----------------------------------------------------------------------------
-// Global Queues
-void doregister( cluster     & cltr ) {
-        lock      ( __cfa_dbg_global_clusters.lock __cfaabi_dbg_ctx2);
-        push_front( __cfa_dbg_global_clusters.list, cltr );
-        unlock    ( __cfa_dbg_global_clusters.lock );
+}
-void unregister( cluster     & cltr ) {
-        lock  ( __cfa_dbg_global_clusters.lock __cfaabi_dbg_ctx2);
-        remove( __cfa_dbg_global_clusters.list, cltr );
-        unlock( __cfa_dbg_global_clusters.lock );
+}
-void doregister( cluster * cltr, $thread & thrd ) {
-        lock      (cltr->thread_list_lock __cfaabi_dbg_ctx2);
-        cltr->nthreads += 1;
-        push_front(cltr->threads, thrd);
-        unlock    (cltr->thread_list_lock);
+}
-void unregister( cluster * cltr, $thread & thrd ) {
-        lock  (cltr->thread_list_lock __cfaabi_dbg_ctx2);
-        remove(cltr->threads, thrd );
-        cltr->nthreads -= 1;
-        unlock(cltr->thread_list_lock);
+}
-void doregister( cluster * cltr, processor * proc ) {
-        lock      (cltr->idle_lock __cfaabi_dbg_ctx2);
-        cltr->nprocessors += 1;
-        push_front(cltr->procs, *proc);
-        unlock    (cltr->idle_lock);
+}
-void unregister( cluster * cltr, processor * proc ) {
-        lock  (cltr->idle_lock __cfaabi_dbg_ctx2);
-        remove(cltr->procs, *proc );
-        cltr->nprocessors -= 1;
-        unlock(cltr->idle_lock);
+}
-//-----------------------------------------------------------------------------
 // Debug
 __cfaabi_dbg_debug_do(
 …
         return true;
+}
+//-----------------------------------------------------------------------------
+// Statistics
+#if !defined(__CFA_NO_STATISTICS__)
+        void print_halts( processor & this ) {
+                this.print_halts = true;
+        }
+#endif
 // Local Variables: //
 // mode: c //

libcfa/src/concurrency/kernel.hfa

-              r07d867b
+              r22f94a4
 #pragma once
-#include <stdbool.h>
-#include <stdint.h>
 #include "invoke.h"
 #include "time_t.hfa"
 #include "coroutine.hfa"
+#include "containers/list.hfa"
 extern "C" {
+#include <pthread.h>
+#include <semaphore.h>
+#include <bits/pthreadtypes.h>
+}
 …
 extern struct cluster * mainCluster;
+// Processor
+// Processor id, required for scheduling threads
+struct __processor_id_t {
+        unsigned id;
+        #if !defined(__CFA_NO_STATISTICS__)
+                struct __stats_t * stats;
+        #endif
+};
 coroutine processorCtx_t {
         struct processor * proc;
 …
 // Wrapper around kernel threads
 struct processor {
+struct __attribute__((aligned(128))) processor {
         // Main state
+        inline __processor_id_t;
+        // Cluster from which to get threads
+        struct cluster * cltr;
+        // Set to true to notify the processor should terminate
+        volatile bool do_terminate;
         // Coroutine ctx who does keeps the state of the processor
         struct processorCtx_t runner;
-        // Cluster from which to get threads
-        struct cluster * cltr;
         // Name of the processor
 …
         __bin_sem_t idle;
-        // Termination
-        // Set to true to notify the processor should terminate
-        volatile bool do_terminate;
         // Termination synchronisation (user semaphore)
         semaphore terminated;
 …
         // Link lists fields
+        struct __dbg_node_proc {
+                struct processor * next;
+                struct processor * prev;
+        } node;
+        DLISTED_MGD_IMPL_IN(processor)
+        #if !defined(__CFA_NO_STATISTICS__)
+                int print_stats;
+                bool print_halts;
+        #endif
 #ifdef __CFA_DEBUG__
 …
 static inline void  ?{}(processor & this, const char name[]) { this{name, *mainCluster }; }
+static inline [processor *&, processor *& ] __get( processor & this ) __attribute__((const)) { return this.node.[next, prev]; }
+DLISTED_MGD_IMPL_OUT(processor)
 //-----------------------------------------------------------------------------
 …
 struct __io_data;
+#define CFA_CLUSTER_IO_POLLER_USER_THREAD 1 << 0
+// #define CFA_CLUSTER_IO_POLLER_KERNEL_SIDE 1 << 1
+// IO poller user-thread
+// Not using the "thread" keyword because we want to control
+// more carefully when to start/stop it
+struct $io_ctx_thread {
+        struct __io_data * ring;
+        single_sem sem;
+        volatile bool done;
+        $thread self;
+};
+struct io_context {
+        $io_ctx_thread thrd;
+};
+struct io_context_params {
+        int num_entries;
+        int num_ready;
+        int submit_aff;
+        bool eager_submits:1;
+        bool poller_submits:1;
+        bool poll_submit:1;
+        bool poll_complete:1;
+};
+void  ?{}(io_context_params & this);
+void  ?{}(io_context & this, struct cluster & cl);
+void  ?{}(io_context & this, struct cluster & cl, const io_context_params & params);
+void ^?{}(io_context & this);
+struct io_cancellation {
+        uint32_t target;
+};
+static inline void  ?{}(io_cancellation & this) { this.target = -1u; }
+static inline void ^?{}(io_cancellation & this) {}
+bool cancel(io_cancellation & this);
+//-----------------------------------------------------------------------------
+// Cluster Tools
+// Intrusives lanes which are used by the relaxed ready queue
+struct __attribute__((aligned(128))) __intrusive_lane_t;
+void  ?{}(__intrusive_lane_t & this);
+void ^?{}(__intrusive_lane_t & this);
+// Counter used for wether or not the lanes are all empty
+struct __attribute__((aligned(128))) __snzi_node_t;
+struct __snzi_t {
+        unsigned mask;
+        int root;
+        __snzi_node_t * nodes;
+};
+void  ?{}( __snzi_t & this, unsigned depth );
+void ^?{}( __snzi_t & this );
+//TODO adjust cache size to ARCHITECTURE
+// Structure holding the relaxed ready queue
+struct __ready_queue_t {
+        // Data tracking how many/which lanes are used
+        // Aligned to 128 for cache locality
+        __snzi_t snzi;
+        // Data tracking the actual lanes
+        // On a seperate cacheline from the used struct since
+        // used can change on each push/pop but this data
+        // only changes on shrink/grow
+        struct {
+                // Arary of lanes
+                __intrusive_lane_t * volatile data;
+                // Number of lanes (empty or not)
+                volatile size_t count;
+        } lanes;
+};
+void  ?{}(__ready_queue_t & this);
+void ^?{}(__ready_queue_t & this);
+// Idle Sleep
+struct __cluster_idles {
+        // Spin lock protecting the queue
+        volatile uint64_t lock;
+        // Total number of processors
+        unsigned total;
+        // Total number of idle processors
+        unsigned idle;
+        // List of idle processors
+        dlist(processor, processor) list;
+};
 //-----------------------------------------------------------------------------
 // Cluster
+struct cluster {
+        // Ready queue locks
+        __spinlock_t ready_queue_lock;
+struct __attribute__((aligned(128))) cluster {
         // Ready queue for threads
         __queue_t($thread) ready_queue;
+        __ready_queue_t ready_queue;
         // Name of the cluster
 …
         Duration preemption_rate;
+        // List of processors
+        __spinlock_t idle_lock;
+        __dllist_t(struct processor) procs;
+        __dllist_t(struct processor) idles;
+        unsigned int nprocessors;
+        // List of idle processors
+        __cluster_idles idles;
         // List of threads
 …
         } node;
+        struct __io_data * io;
+        struct {
+                io_context * ctxs;
+                unsigned cnt;
+        } io;
         #if !defined(__CFA_NO_STATISTICS__)
+                bool print_stats;
+                struct __stats_t * stats;
+                int print_stats;
         #endif
 };
 extern Duration default_preemption();
 void ?{} (cluster & this, const char name[], Duration preemption_rate, int flags);
+void ?{} (cluster & this, const char name[], Duration preemption_rate, unsigned num_io, const io_context_params & io_params);
 void ^?{}(cluster & this);
+static inline void ?{} (cluster & this)                                      { this{"Anonymous Cluster", default_preemption(), 0}; }
+static inline void ?{} (cluster & this, Duration preemption_rate)            { this{"Anonymous Cluster", preemption_rate, 0}; }
+static inline void ?{} (cluster & this, const char name[])                   { this{name, default_preemption(), 0}; }
+static inline void ?{} (cluster & this, int flags)                           { this{"Anonymous Cluster", default_preemption(), flags}; }
+static inline void ?{} (cluster & this, Duration preemption_rate, int flags) { this{"Anonymous Cluster", preemption_rate, flags}; }
+static inline void ?{} (cluster & this, const char name[], int flags)        { this{name, default_preemption(), flags}; }
+static inline void ?{} (cluster & this)                                            { io_context_params default_params;    this{"Anonymous Cluster", default_preemption(), 1, default_params}; }
+static inline void ?{} (cluster & this, Duration preemption_rate)                  { io_context_params default_params;    this{"Anonymous Cluster", preemption_rate, 1, default_params}; }
+static inline void ?{} (cluster & this, const char name[])                         { io_context_params default_params;    this{name, default_preemption(), 1, default_params}; }
+static inline void ?{} (cluster & this, unsigned num_io)                           { io_context_params default_params;    this{"Anonymous Cluster", default_preemption(), num_io, default_params}; }
+static inline void ?{} (cluster & this, Duration preemption_rate, unsigned num_io) { io_context_params default_params;    this{"Anonymous Cluster", preemption_rate, num_io, default_params}; }
+static inline void ?{} (cluster & this, const char name[], unsigned num_io)        { io_context_params default_params;    this{name, default_preemption(), num_io, default_params}; }
+static inline void ?{} (cluster & this, const io_context_params & io_params)                                            { this{"Anonymous Cluster", default_preemption(), 1, io_params}; }
+static inline void ?{} (cluster & this, Duration preemption_rate, const io_context_params & io_params)                  { this{"Anonymous Cluster", preemption_rate, 1, io_params}; }
+static inline void ?{} (cluster & this, const char name[], const io_context_params & io_params)                         { this{name, default_preemption(), 1, io_params}; }
+static inline void ?{} (cluster & this, unsigned num_io, const io_context_params & io_params)                           { this{"Anonymous Cluster", default_preemption(), num_io, io_params}; }
+static inline void ?{} (cluster & this, Duration preemption_rate, unsigned num_io, const io_context_params & io_params) { this{"Anonymous Cluster", preemption_rate, num_io, io_params}; }
+static inline void ?{} (cluster & this, const char name[], unsigned num_io, const io_context_params & io_params)        { this{name, default_preemption(), num_io, io_params}; }
 static inline [cluster *&, cluster *& ] __get( cluster & this ) __attribute__((const)) { return this.node.[next, prev]; }
 …
 #if !defined(__CFA_NO_STATISTICS__)
         static inline void print_stats_at_exit( cluster & this ) {
                 this.print_stats = true;
+        static inline void print_stats_at_exit( cluster & this, int flags ) {
+                this.print_stats |= flags;
+        }
+        static inline void print_stats_at_exit( processor & this, int flags ) {
+                this.print_stats |= flags;
+        }
+        void print_halts( processor & this );
 #endif

libcfa/src/concurrency/kernel_private.hfa

-              r07d867b
+              r22f94a4
 #include "alarm.hfa"
+#include "stats.hfa"
 //-----------------------------------------------------------------------------
 // Scheduler
+struct __attribute__((aligned(128))) __scheduler_lock_id_t;
 extern "C" {
 …
+}
+void __schedule_thread( $thread * ) __attribute__((nonnull (1)));
+void __schedule_thread( struct __processor_id_t *, $thread * )
+#if defined(NDEBUG) || (!defined(__CFA_DEBUG__) && !defined(__CFA_VERIFY__))
+        __attribute__((nonnull (2)))
+#endif
+;
 //Block current thread and release/wake-up the following resources
 …
-struct event_kernel_t {
-        alarm_list_t alarms;
-        __spinlock_t lock;
-};
-extern event_kernel_t * event_kernel;
-struct __cfa_kernel_preemption_state_t {
-        bool enabled;
-        bool in_progress;
-        unsigned short disable_count;
-};
-extern volatile thread_local __cfa_kernel_preemption_state_t preemption_state __attribute__ ((tls_model ( "initial-exec" )));
 extern cluster * mainCluster;
 …
 // KERNEL ONLY unpark with out disabling interrupts
+void __unpark( $thread * thrd __cfaabi_dbg_ctx_param2 );
+//-----------------------------------------------------------------------------
+// I/O
+void __kernel_io_startup     ( cluster &, int, bool );
+void __kernel_io_finish_start( cluster & );
+void __kernel_io_prepare_stop( cluster & );
+void __kernel_io_shutdown    ( cluster &, bool );
+void __unpark( struct __processor_id_t *, $thread * thrd __cfaabi_dbg_ctx_param2 );
+static inline bool __post(single_sem & this, struct __processor_id_t * id) {
+        for() {
+                struct $thread * expected = this.ptr;
+                if(expected == 1p) return false;
+                if(expected == 0p) {
+                        if(__atomic_compare_exchange_n(&this.ptr, &expected, 1p, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)) {
+                                return false;
+                        }
+                }
+                else {
+                        if(__atomic_compare_exchange_n(&this.ptr, &expected, 0p, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)) {
+                                __unpark( id, expected __cfaabi_dbg_ctx2 );
+                                return true;
+                        }
+                }
+        }
+}
 //-----------------------------------------------------------------------------
 // Utils
-#define KERNEL_STORAGE(T,X) static char storage_##X[sizeof(T)]
-static inline uint32_t __tls_rand() {
-        kernelTLS.rand_seed ^= kernelTLS.rand_seed << 6;
-        kernelTLS.rand_seed ^= kernelTLS.rand_seed >> 21;
-        kernelTLS.rand_seed ^= kernelTLS.rand_seed << 7;
-        return kernelTLS.rand_seed;
+}
-void doregister( struct cluster & cltr );
-void unregister( struct cluster & cltr );
 void doregister( struct cluster * cltr, struct $thread & thrd );
 void unregister( struct cluster * cltr, struct $thread & thrd );
+void doregister( struct cluster * cltr, struct processor * proc );
+void unregister( struct cluster * cltr, struct processor * proc );
+//-----------------------------------------------------------------------------
+// I/O
+void ^?{}(io_context & this, bool );
+//=======================================================================
+// Cluster lock API
+//=======================================================================
+// Cells use by the reader writer lock
+// while not generic it only relies on a opaque pointer
+struct __attribute__((aligned(128))) __scheduler_lock_id_t {
+        // Spin lock used as the underlying lock
+        volatile bool lock;
+        // Handle pointing to the proc owning this cell
+        // Used for allocating cells and debugging
+        __processor_id_t * volatile handle;
+        #ifdef __CFA_WITH_VERIFY__
+                // Debug, check if this is owned for reading
+                bool owned;
+        #endif
+};
+static_assert( sizeof(struct __scheduler_lock_id_t) <= __alignof(struct __scheduler_lock_id_t));
+// Lock-Free registering/unregistering of threads
+// Register a processor to a given cluster and get its unique id in return
+unsigned doregister( struct __processor_id_t * proc );
+// Unregister a processor from a given cluster using its id, getting back the original pointer
+void     unregister( struct __processor_id_t * proc );
+//-----------------------------------------------------------------------
+// Cluster idle lock/unlock
+static inline void lock(__cluster_idles & this) {
+        for() {
+                uint64_t l = this.lock;
+                if(
+                        (0 == (l % 2))
+                        && __atomic_compare_exchange_n(&this.lock, &l, l + 1, false, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST)
+                ) return;
+                Pause();
+        }
+}
+static inline void unlock(__cluster_idles & this) {
+        /* paranoid */ verify( 1 == (this.lock % 2) );
+        __atomic_fetch_add( &this.lock, 1, __ATOMIC_SEQ_CST );
+}
+//=======================================================================
+// Reader-writer lock implementation
+// Concurrent with doregister/unregister,
+//    i.e., threads can be added at any point during or between the entry/exit
+//-----------------------------------------------------------------------
+// simple spinlock underlying the RWLock
+// Blocking acquire
+static inline void __atomic_acquire(volatile bool * ll) {
+        while( __builtin_expect(__atomic_exchange_n(ll, (bool)true, __ATOMIC_SEQ_CST), false) ) {
+                while(__atomic_load_n(ll, (int)__ATOMIC_RELAXED))
+                        asm volatile("pause");
+        }
+        /* paranoid */ verify(*ll);
+}
+// Non-Blocking acquire
+static inline bool __atomic_try_acquire(volatile bool * ll) {
+        return !__atomic_exchange_n(ll, (bool)true, __ATOMIC_SEQ_CST);
+}
+// Release
+static inline void __atomic_unlock(volatile bool * ll) {
+        /* paranoid */ verify(*ll);
+        __atomic_store_n(ll, (bool)false, __ATOMIC_RELEASE);
+}
+//-----------------------------------------------------------------------
+// Reader-Writer lock protecting the ready-queues
+// while this lock is mostly generic some aspects
+// have been hard-coded to for the ready-queue for
+// simplicity and performance
+struct __scheduler_RWLock_t {
+        // total cachelines allocated
+        unsigned int max;
+        // cachelines currently in use
+        volatile unsigned int alloc;
+        // cachelines ready to itereate over
+        // (!= to alloc when thread is in second half of doregister)
+        volatile unsigned int ready;
+        // writer lock
+        volatile bool lock;
+        // data pointer
+        __scheduler_lock_id_t * data;
+};
+void  ?{}(__scheduler_RWLock_t & this);
+void ^?{}(__scheduler_RWLock_t & this);
+extern __scheduler_RWLock_t * __scheduler_lock;
+//-----------------------------------------------------------------------
+// Reader side : acquire when using the ready queue to schedule but not
+//  creating/destroying queues
+static inline void ready_schedule_lock( struct __processor_id_t * proc) with(*__scheduler_lock) {
+        unsigned iproc = proc->id;
+        /*paranoid*/ verify(data[iproc].handle == proc);
+        /*paranoid*/ verify(iproc < ready);
+        // Step 1 : make sure no writer are in the middle of the critical section
+        while(__atomic_load_n(&lock, (int)__ATOMIC_RELAXED))
+                asm volatile("pause");
+        // Fence needed because we don't want to start trying to acquire the lock
+        // before we read a false.
+        // Not needed on x86
+        // std::atomic_thread_fence(std::memory_order_seq_cst);
+        // Step 2 : acquire our local lock
+        __atomic_acquire( &data[iproc].lock );
+        /*paranoid*/ verify(data[iproc].lock);
+        #ifdef __CFA_WITH_VERIFY__
+                // Debug, check if this is owned for reading
+                data[iproc].owned = true;
+        #endif
+}
+static inline void ready_schedule_unlock( struct __processor_id_t * proc) with(*__scheduler_lock) {
+        unsigned iproc = proc->id;
+        /*paranoid*/ verify(data[iproc].handle == proc);
+        /*paranoid*/ verify(iproc < ready);
+        /*paranoid*/ verify(data[iproc].lock);
+        /*paranoid*/ verify(data[iproc].owned);
+        #ifdef __CFA_WITH_VERIFY__
+                // Debug, check if this is owned for reading
+                data[iproc].owned = false;
+        #endif
+        __atomic_unlock(&data[iproc].lock);
+}
+#ifdef __CFA_WITH_VERIFY__
+        static inline bool ready_schedule_islocked( struct __processor_id_t * proc) {
+                return __scheduler_lock->data[proc->id].owned;
+        }
+        static inline bool ready_mutate_islocked() {
+                return __scheduler_lock->lock;
+        }
+#endif
+//-----------------------------------------------------------------------
+// Writer side : acquire when changing the ready queue, e.g. adding more
+//  queues or removing them.
+uint_fast32_t ready_mutate_lock( void );
+void ready_mutate_unlock( uint_fast32_t /* value returned by lock */ );
+//=======================================================================
+// Ready-Queue API
+//-----------------------------------------------------------------------
+// pop thread from the ready queue of a cluster
+// returns 0p if empty
+__attribute__((hot)) bool query(struct cluster * cltr);
+//-----------------------------------------------------------------------
+// push thread onto a ready queue for a cluster
+// returns true if the list was previously empty, false otherwise
+__attribute__((hot)) bool push(struct cluster * cltr, struct $thread * thrd);
+//-----------------------------------------------------------------------
+// pop thread from the ready queue of a cluster
+// returns 0p if empty
+// May return 0p spuriously
+__attribute__((hot)) struct $thread * pop(struct cluster * cltr);
+//-----------------------------------------------------------------------
+// pop thread from the ready queue of a cluster
+// returns 0p if empty
+// guaranteed to find any threads added before this call
+__attribute__((hot)) struct $thread * pop_slow(struct cluster * cltr);
+//-----------------------------------------------------------------------
+// remove thread from the ready queue of a cluster
+// returns bool if it wasn't found
+bool remove_head(struct cluster * cltr, struct $thread * thrd);
+//-----------------------------------------------------------------------
+// Increase the width of the ready queue (number of lanes) by 4
+void ready_queue_grow  (struct cluster * cltr, int target);
+//-----------------------------------------------------------------------
+// Decrease the width of the ready queue (number of lanes) by 4
+void ready_queue_shrink(struct cluster * cltr, int target);
 // Local Variables: //

libcfa/src/concurrency/monitor.cfa

-              r07d867b
+              r22f94a4
                 // Some one else has the monitor, wait in line for it
                 /* paranoid */ verify( thrd->next == 0p );
+                /* paranoid */ verify( thrd->link.next == 0p );
                 append( this->entry_queue, thrd );
                 /* paranoid */ verify( thrd->next == 1p );
+                /* paranoid */ verify( thrd->link.next == 1p );
                 unlock( this->lock );
 …
                 // Some one else has the monitor, wait in line for it
                 /* paranoid */ verify( thrd->next == 0p );
+                /* paranoid */ verify( thrd->link.next == 0p );
                 append( this->entry_queue, thrd );
                 /* paranoid */ verify( thrd->next == 1p );
+                /* paranoid */ verify( thrd->link.next == 1p );
                 unlock( this->lock );
 …
         $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 );
+        /* paranoid */ verify( !new_owner || new_owner->link.next == 0p );
         __set_owner( this, new_owner );
 …
+        }
         __cfaabi_dbg_print_safe( "Kernel :  Runing %i (%p)\n", ready2run, ready2run ? node->waiting_thread : 0p );
+        __cfaabi_dbg_print_safe( "Kernel :  Runing %i (%p)\n", ready2run, ready2run ? (thread*)node->waiting_thread : (thread*)0p );
         return ready2run ? node->waiting_thread : 0p;
+}
 …
         // For each thread in the entry-queue
         for(    $thread ** thrd_it = &entry_queue.head;
                 *thrd_it != 1p;
                 thrd_it = &(*thrd_it)->next
+                (*thrd_it) != 1p;
+                thrd_it = &(*thrd_it)->link.next
         ) {
                 // For each acceptable check if it matches

libcfa/src/concurrency/mutex.cfa

r07d867b	r22f94a4
30	30	this.lock{};
31	31	this.blocked_threads{};
	32	this.is_locked = false;
32	33	}
33	34

libcfa/src/concurrency/preemption.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Mon Jun 5 14:20:42 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Dec  5 16:34:05 2019
 // Update Count     : 43
+// Last Modified On : Wed Jun 17 11:36:25 2020
+// Update Count     : 46
 //
 …
 #include <assert.h>
-extern "C" {
 #include <errno.h>
 #include <stdio.h>
 …
 #include <unistd.h>
 #include <limits.h>                                                                             // PTHREAD_STACK_MIN
+}
 #include "bits/signal.hfa"
+#include "kernel_private.hfa"
 #if !defined(__CFA_DEFAULT_PREEMPTION__)
 …
 // FwdDeclarations : timeout handlers
 static void preempt( processor   * this );
 static void timeout( $thread * this );
+static void timeout( struct __processor_id_t * id, $thread * this );
 // FwdDeclarations : Signal handlers
 …
 // Tick one frame of the Discrete Event Simulation for alarms
 static void tick_preemption() {
+static void tick_preemption( struct __processor_id_t * id ) {
         alarm_node_t * node = 0p;                                                       // Used in the while loop but cannot be declared in the while condition
         alarm_list_t * alarms = &event_kernel->alarms;          // Local copy for ease of reading
 …
+                }
                 else {
                         timeout( node->thrd );
+                        timeout( id, node->thrd );
+                }
 …
         // If there are still alarms pending, reset the timer
         if( & (*alarms)`first ) {
                 __cfaabi_dbg_print_buffer_decl( " KERNEL: @%ju(%ju) resetting alarm to %ju.\n", currtime.tv, __kernel_get_time().tv, (alarms->head->alarm - currtime).tv);
+                __cfadbg_print_buffer_decl(preemption, " KERNEL: @%ju(%ju) resetting alarm to %ju.\n", currtime.tv, __kernel_get_time().tv, (alarms->head->alarm - currtime).tv);
                 Duration delta = (*alarms)`first.alarm - currtime;
                 Duration capped = max(delta, 50`us);
 …
         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
+                /* paranoid */ verify( proc );
                 with( kernelTLS.preemption_state ){
 …
 // reserved for future use
+static void timeout( $thread * this ) {
+        __unpark( this __cfaabi_dbg_ctx2 );
+static void timeout( struct __processor_id_t * id, $thread * this ) {
+        #if !defined( __CFA_NO_STATISTICS__ )
+                kernelTLS.this_stats = this->curr_cluster->stats;
+        #endif
+        __unpark( id, this __cfaabi_dbg_ctx2 );
+}
 …
 // Startup routine to activate preemption
 // Called from kernel_startup
 void kernel_start_preemption() {
+void __kernel_alarm_startup() {
         __cfaabi_dbg_print_safe( "Kernel : Starting preemption\n" );
 …
 // Shutdown routine to deactivate preemption
 // Called from kernel_shutdown
 void kernel_stop_preemption() {
+void __kernel_alarm_shutdown() {
         __cfaabi_dbg_print_safe( "Kernel : Preemption stopping\n" );
 …
 // Waits on SIGALRM and send SIGUSR1 to whom ever needs it
 static void * alarm_loop( __attribute__((unused)) void * args ) {
+        __processor_id_t id;
+        id.id = doregister(&id);
         // Block sigalrms to control when they arrive
         sigset_t mask;
 …
                         // __cfaabi_dbg_print_safe( "Kernel : Preemption thread tick\n" );
                         lock( event_kernel->lock __cfaabi_dbg_ctx2 );
                         tick_preemption();
+                        tick_preemption( &id );
                         unlock( event_kernel->lock );
                         break;
 …
 EXIT:
         __cfaabi_dbg_print_safe( "Kernel : Preemption thread stopping\n" );
+        unregister(&id);
         return 0p;
+}
 …
         sigset_t oldset;
         int ret;
         ret = pthread_sigmask(0, 0p, &oldset);
+        ret = pthread_sigmask(0, ( const sigset_t * ) 0p, &oldset);  // workaround trac#208: cast should be unnecessary
         if(ret != 0) { abort("ERROR sigprocmask returned %d", ret); }

libcfa/src/concurrency/preemption.hfa

-              r07d867b
+              r22f94a4
 #pragma once
+#include "bits/locks.hfa"
 #include "alarm.hfa"
-#include "kernel_private.hfa"
+void kernel_start_preemption();
+void kernel_stop_preemption();
+struct event_kernel_t {
+        alarm_list_t alarms;
+        __spinlock_t lock;
+};
+extern event_kernel_t * event_kernel;
 void update_preemption( processor * this, Duration duration );

libcfa/src/concurrency/thread.cfa

-              r07d867b
+              r22f94a4
         context{ 0p, 0p };
         self_cor{ name, storage, storageSize };
+        ticket = 1;
         state = Start;
         preempted = __NO_PREEMPTION;
 …
         self_mon_p = &self_mon;
         curr_cluster = &cl;
+        next = 0p;
+        link.next = 0p;
+        link.prev = 0p;
+        link.preferred = -1;
         node.next = 0p;
 …
         verify( this_thrd->context.SP );
         __schedule_thread(this_thrd);
+        __schedule_thread( (__processor_id_t *)kernelTLS.this_processor, this_thrd);
         enable_interrupts( __cfaabi_dbg_ctx );
+}

libcfa/src/concurrency/thread.hfa

-              r07d867b
+              r22f94a4
 //-----------------------------------------------------------------------------
-// Thread getters
-static inline struct $thread * active_thread () { return TL_GET( this_thread ); }
-//-----------------------------------------------------------------------------
 // Scheduler API
 …
 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();
+        }
+}
 //----------
 // sleep: force thread to block and be rescheduled after Duration duration

libcfa/src/containers/list.hfa

-              r07d867b
+              r22f94a4
 //
+#pragma once
 #include <assert.h>
 …
+\
 static inline NODE& $tempcv_e2n(ELEM &node) { \
         return node; \
+        return ( NODE & ) node; \
 } \
+\
 …
         Telem& $tempcv_n2e(Tnode &);
         Tnode& $tempcv_e2n(Telem &);
+        Telem& ?`next(Tnode &);
+        Telem& ?`prev(Tnode &);
 };
 …
                 $next_link(singleton_to_insert) = $next_link(list_pos);
                 if ($next_link(list_pos).is_terminator) {
                         dlist(Tnode, Telem) *list = $next_link(list_pos).terminator;
+                        dlist(Tnode, Telem) *list = ( dlist(Tnode, Telem) * ) $next_link(list_pos).terminator;
                         $dlinks(Telem) *list_links = & list->$links;
                         $mgd_link(Telem) *list_last = & list_links->prev;
 …
                 $prev_link(singleton_to_insert) = $prev_link(list_pos);
                 if ($prev_link(list_pos).is_terminator) {
                         dlist(Tnode, Telem) *list = $prev_link(list_pos).terminator;
+                        dlist(Tnode, Telem) *list = ( dlist(Tnode, Telem) * ) $prev_link(list_pos).terminator;
                         $dlinks(Telem) *list_links = & list->$links;
                         $mgd_link(Telem) *list_first = & list_links->next;
 …
                 if ( $prev_link(list_pos).is_terminator ) {
                         dlist(Tnode, Telem) * tgt_before = $prev_link(list_pos).terminator;
+                        dlist(Tnode, Telem) * tgt_before = ( dlist(Tnode, Telem) * ) $prev_link(list_pos).terminator;
                         $dlinks(Telem) * links_before = & tgt_before->$links;
                         &incoming_from_prev = & links_before->next;
 …
                 if ( $next_link(list_pos).is_terminator ) {
                         dlist(Tnode, Telem) * tgt_after = $next_link(list_pos).terminator;
+                        dlist(Tnode, Telem) * tgt_after = ( dlist(Tnode, Telem) * ) $next_link(list_pos).terminator;
                         $dlinks(Telem) * links_after = & tgt_after->$links;
                         &incoming_from_next = & links_after->prev;
 …
                 $prev_link(list_pos) = (Telem*) 0p;
+        }
+        static inline bool ?`is_empty(dlist(Tnode, Telem) &list) {
+                assert( &list != 0p );
+                $dlinks(Telem) *listLinks = & list.$links;
+                if (listLinks->next.is_terminator) {
+                        assert(listLinks->prev.is_terminator);
+                        assert(listLinks->next.terminator);
+                        assert(listLinks->prev.terminator);
+                        return true;
+                } else {
+                        assert(!listLinks->prev.is_terminator);
+                        assert(listLinks->next.elem);
+                        assert(listLinks->prev.elem);
+                        return false;
+                }
+        }
+        static inline Telem & pop_first(dlist(Tnode, Telem) &list) {
+                assert( &list != 0p );
+                assert( !list`is_empty );
+                $dlinks(Telem) *listLinks = & list.$links;
+                Telem & first = *listLinks->next.elem;
+                Tnode & list_pos_first  = $tempcv_e2n( first );
+                remove(list_pos_first);
+                return first;
+        }
+        static inline Telem & pop_last(dlist(Tnode, Telem) &list) {
+                assert( &list != 0p );
+                assert( !list`is_empty );
+                $dlinks(Telem) *listLinks = & list.$links;
+                Telem & last = *listLinks->prev.elem;
+                Tnode & list_pos_last  = $tempcv_e2n( last );
+                remove(list_pos_last);
+                return last;
+        }
+}

libcfa/src/containers/vector.hfa

-              r07d867b
+              r22f94a4
 // Created On       : Tue Jul  5 18:00:07 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Jul 22 10:01:18 2017
 // Update Count     : 3
+// Last Modified On : Wed Jun 17 11:02:46 2020
+// Update Count     : 4
 //
 #pragma once
-extern "C" {
 #include <stdbool.h>
+}
 //------------------------------------------------------------------------------

libcfa/src/exception.c

-              r07d867b
+              r22f94a4
 // Created On       : Mon Jun 26 15:13:00 2017
 // Last Modified By : Andrew Beach
 // Last Modified On : Tue Apr 14 12:01:00 2020
 // Update Count     : 18
+// Last Modified On : Thr May 21 12:18:00 2020
+// Update Count     : 20
 //
 …
+}
 void __cfaehm_throw_resume(exception_t * except) {
+void __cfaehm_throw_resume(exception_t * except, void (*defaultHandler)(exception_t *)) {
         struct exception_context_t * context = this_exception_context();
         __cfadbg_print_safe(exception, "Throwing resumption exception\n");
+        __attribute__((cleanup(reset_top_resume)))
+        struct __cfaehm_try_resume_node * original_head = context->top_resume;
+        struct __cfaehm_try_resume_node * current = context->top_resume;
+        for ( ; current ; current = current->next) {
+                context->top_resume = current->next;
+                if (current->handler(except)) {
+                        return;
+        {
+                __attribute__((cleanup(reset_top_resume)))
+                struct __cfaehm_try_resume_node * original_head = context->top_resume;
+                struct __cfaehm_try_resume_node * current = context->top_resume;
+                for ( ; current ; current = current->next) {
+                        context->top_resume = current->next;
+                        if (current->handler(except)) {
+                                return;
+                        }
+                }
+        }
+        } // End the search and return to the top of the stack.
+        // No handler found, fall back to the default operation.
         __cfadbg_print_safe(exception, "Unhandled exception\n");
+        // Fall back to termination:
+        __cfaehm_throw_terminate(except);
+        // TODO: Default handler for resumption.
+        defaultHandler(except);
+}
 …
+// TERMINATION ===============================================================
+// MEMORY MANAGEMENT (still for integers)
+// May have to move to cfa for constructors and destructors (references).
+// MEMORY MANAGEMENT =========================================================
 // How to clean up an exception in various situations.
 …
+}
+// If this isn't a rethrow (*except==0), delete the provided exception.
+void __cfaehm_cleanup_terminate( void * except ) {
+        if ( *(void**)except ) __cfaehm_delete_exception( *(exception_t **)except );
+}
+// CANCELLATION ==============================================================
 // Function needed by force unwind
 …
+}
+// Cancel the current stack, prefroming approprate clean-up and messaging.
+void __cfaehm_cancel_stack( exception_t * exception ) {
+        // TODO: Detect current stack and pick a particular stop-function.
+        _Unwind_Reason_Code ret;
+        ret = _Unwind_ForcedUnwind( &this_exception_storage, _Stop_Fn, (void*)0x22 );
+        printf("UNWIND ERROR %d after force unwind\n", ret);
+        abort();
+}
+// TERMINATION ===============================================================
+// If this isn't a rethrow (*except==0), delete the provided exception.
+void __cfaehm_cleanup_terminate( void * except ) {
+        if ( *(void**)except ) __cfaehm_delete_exception( *(exception_t **)except );
+}
+static void __cfaehm_cleanup_default( exception_t ** except ) {
+        __cfaehm_delete_exception( *except );
+        *except = NULL;
+}
 // The exception that is being thrown must already be stored.
+static __attribute__((noreturn)) void __cfaehm_begin_unwind(void) {
+        if ( ! this_exception_context()->current_exception ) {
+static void __cfaehm_begin_unwind(void(*defaultHandler)(exception_t *)) {
+        struct exception_context_t * context = this_exception_context();
+        struct _Unwind_Exception * storage = &this_exception_storage;
+        if ( NULL == context->current_exception ) {
                 printf("UNWIND ERROR missing exception in begin unwind\n");
                 abort();
 …
         // Call stdlibc to raise the exception
+        _Unwind_Reason_Code ret = _Unwind_RaiseException( &this_exception_storage );
+        __cfadbg_print_safe(exception, "Begin unwinding (storage &p, context %p)\n", storage, context);
+        _Unwind_Reason_Code ret = _Unwind_RaiseException( storage );
         // If we reach here it means something happened. For resumption to work we need to find a way
 …
         // the whole stack.
+        if ( ret == _URC_END_OF_STACK ) {
+                // No proper handler was found. This can be handled in many ways, C++ calls std::terminate.
+                // Here we force unwind the stack, basically raising a cancellation.
+                printf("Uncaught exception %p\n", &this_exception_storage);
+                ret = _Unwind_ForcedUnwind( &this_exception_storage, _Stop_Fn, (void*)0x22 );
+                printf("UNWIND ERROR %d after force unwind\n", ret);
+        // We did not simply reach the end of the stack without finding a handler. This is an error.
+        if ( ret != _URC_END_OF_STACK ) {
+                printf("UNWIND ERROR %d after raise exception\n", ret);
                 abort();
+        }
+        // We did not simply reach the end of the stack without finding a handler. This is an error.
+        printf("UNWIND ERROR %d after raise exception\n", ret);
+        // No handler found, go to the default operation.
+        __cfadbg_print_safe(exception, "Uncaught exception %p\n", storage);
+        __attribute__((cleanup(__cfaehm_cleanup_default)))
+        exception_t * exception = context->current_exception;
+        defaultHandler( exception );
+}
+void __cfaehm_throw_terminate( exception_t * val, void (*defaultHandler)(exception_t *) ) {
+        __cfadbg_print_safe(exception, "Throwing termination exception\n");
+        __cfaehm_allocate_exception( val );
+        __cfaehm_begin_unwind( defaultHandler );
+}
+static __attribute__((noreturn)) void __cfaehm_rethrow_adapter( exception_t * except ) {
+        // TODO: Print some error message.
+        (void)except;
         abort();
+}
-void __cfaehm_throw_terminate( exception_t * val ) {
-        __cfadbg_print_safe(exception, "Throwing termination exception\n");
-        __cfaehm_allocate_exception( val );
-        __cfaehm_begin_unwind();
+}
 …
         __cfadbg_print_safe(exception, "Rethrowing termination exception\n");
+        __cfaehm_begin_unwind();
+        __cfaehm_begin_unwind( __cfaehm_rethrow_adapter );
+        abort();
+}

libcfa/src/exception.h

-              r07d867b
+              r22f94a4
 // Created On       : Mon Jun 26 15:11:00 2017
 // Last Modified By : Andrew Beach
 // Last Modified On : Fri Mar 27 10:16:00 2020
 // Update Count     : 9
+// Last Modified On : Tue May 19 14:17:00 2020
+// Update Count     : 10
 //
 …
+void __cfaehm_cancel_stack(exception_t * except) __attribute__((noreturn));
 // Used in throw statement translation.
 void __cfaehm_throw_terminate(exception_t * except) __attribute__((noreturn));
+void __cfaehm_throw_terminate(exception_t * except, void (*)(exception_t *));
 void __cfaehm_rethrow_terminate() __attribute__((noreturn));
 void __cfaehm_throw_resume(exception_t * except);
+void __cfaehm_throw_resume(exception_t * except, void (*)(exception_t *));
 // Function catches termination exceptions.
 …
 #ifdef __cforall
+}
+// Not all the built-ins can be expressed in C. These can't be
+// implemented in the .c file either so they all have to be inline.
+trait is_exception(dtype T) {
+        /* The first field must be a pointer to a virtual table.
+         * That virtual table must be a decendent of the base exception virtual tab$
+         */
+        void mark_exception(T *);
+        // This is never used and should be a no-op.
+};
+trait is_termination_exception(dtype T | is_exception(T)) {
+        void defaultTerminationHandler(T &);
+};
+trait is_resumption_exception(dtype T | is_exception(T)) {
+        void defaultResumptionHandler(T &);
+};
+forall(dtype T | is_termination_exception(T))
+static inline void $throw(T & except) {
+        __cfaehm_throw_terminate(
+                (exception_t *)&except,
+                (void(*)(exception_t *))defaultTerminationHandler
+        );
+}
+forall(dtype T | is_resumption_exception(T))
+static inline void $throwResume(T & except) {
+        __cfaehm_throw_resume(
+                (exception_t *)&except,
+                (void(*)(exception_t *))defaultResumptionHandler
+        );
+}
+forall(dtype T | is_exception(T))
+static inline void cancel_stack(T & except) __attribute__((noreturn)) {
+        __cfaehm_cancel_stack( (exception_t *)&except );
+}
+forall(dtype T | is_exception(T))
+static inline void defaultTerminationHandler(T & except) {
+        return cancel_stack( except );
+}
+forall(dtype T | is_exception(T))
+static inline void defaultResumptionHandler(T & except) {
+        throw except;
+}
 #endif

libcfa/src/exception.hfa

-              r07d867b
+              r22f94a4
 // Created On       : Thu Apr  7 10:25:00 2020
 // Last Modified By : Andrew Beach
 // Last Modified On : Thu Apr  7 10:25:00 2020
 // Update Count     : 0
+// Last Modified On : Tue Aug  4 16:22:00 2020
+// Update Count     : 3
 //
 …
 // -----------------------------------------------------------------------------------------------
+// All internals helper macros begin with an underscore.
+#define _CLOSE(...) __VA_ARGS__ }
+#define _GLUE2(left, right) left##right
+#define _GLUE3(left, middle, right) left##middle##right
+#define _EXC_DISPATCH(to, ...) to(__VA_ARGS__,__cfaehm_base_exception_t,)
+// FWD_TRIVIAL_EXCEPTION(exception_name);
+// TRIVIAL_EXCEPTION_DECLARATION(exception_name);
 // Declare a trivial exception, one that adds no fields or features.
 // This will make the exception visible and may go in a .hfa or .cfa file.
+#define FWD_TRIVIAL_EXCEPTION(...) _EXC_DISPATCH(_FWD_TRIVIAL_EXCEPTION, __VA_ARGS__)
+// INST_TRIVIAL_EXCEPTION(exception_name);
+#define TRIVIAL_EXCEPTION_DECLARATION(...) \
+        _EXC_DISPATCH(_TRIVIAL_EXCEPTION_DECLARATION, __VA_ARGS__)
+// TRIVIAL_EXCEPTION_INSTANCE(exception_name);
 // Create the trival exception. This must be used exactly once and should be used in a .cfa file,
 // as it creates the unique instance of the virtual table.
+#define INST_TRIVIAL_EXCEPTION(...) _EXC_DISPATCH(_INST_TRIVIAL_EXCEPTION, __VA_ARGS__)
+#define TRIVIAL_EXCEPTION_INSTANCE(...) _EXC_DISPATCH(_TRIVIAL_EXCEPTION_INSTANCE, __VA_ARGS__)
 // TRIVIAL_EXCEPTION(exception_name[, parent_name]);
 // Does both of the above, a short hand if the exception is only used in one .cfa file.
 …
 // base exception. This feature may be removed or changed.
 #define TRIVIAL_EXCEPTION(...) \
+        _EXC_DISPATCH(_FWD_TRIVIAL_EXCEPTION, __VA_ARGS__); \
+        _EXC_DISPATCH(_INST_TRIVIAL_EXCEPTION, __VA_ARGS__)
+#define _FWD_TRIVIAL_EXCEPTION(exception_name, parent_name, ...) \
+        _EXC_DISPATCH(_TRIVIAL_EXCEPTION_DECLARATION, __VA_ARGS__); \
+        _EXC_DISPATCH(_TRIVIAL_EXCEPTION_INSTANCE, __VA_ARGS__)
+// FORALL_TRIVIAL_EXCEPTION(exception_name, (assertions...), (parameters...));
+// Forward declare a polymorphic but otherwise trivial exception type. You must provide the entire
+// assertion list (exactly what would go in the forall clause) and parameters list (only the
+// parameter names from the assertion list, same order and comma seperated). This should be
+// visible where ever use the exception. This just generates the polymorphic framework, see
+// POLY_VTABLE_DECLARATION to allow instantiations.
+#define FORALL_TRIVIAL_EXCEPTION(exception_name, assertions, parameters) \
+        _FORALL_TRIVIAL_EXCEPTION(exception_name, __cfaehm_base_exception_t, assertions, parameters, )
+// FORALL_TRIVIAL_INSTANCE(exception_name, (assertions...), (parameters...))
+// Create the forall trivial exception. The assertion list and parameters must match.
+// There must be exactly one use of this in a program for each exception type. This just
+// generates the polymorphic framework, see POLY_VTABLE_INSTANCE to allow instantiations.
+#define FORALL_TRIVIAL_INSTANCE(exception_name, assertions, parameters) \
+        _FORALL_CTOR0_INSTANCE(exception_name, assertions, parameters)
+// DATA_EXCEPTION(exception_name)(fields...);
+// Forward declare an exception that adds fields but no features. The added fields go in the
+// second argument list. The virtual table instance must be provided later (see VTABLE_INSTANCE).
+#define DATA_EXCEPTION(...) _EXC_DISPATCH(_DATA_EXCEPTION, __VA_ARGS__)
+// FORALL_DATA_EXCEPTION(exception_name, (assertions...), (parameters...))(fields...);
+// Define a polymorphic exception that adds fields but no additional features. The assertion list
+// and matching parameters must match. Then you can give the list of fields. This should be
+// visible where ever you use the exception. This just generates the polymorphic framework, see
+// POLY_VTABLE_DECLARATION to allow instantiations.
+#define FORALL_DATA_EXCEPTION(exception_name, assertions, parameters) \
+        _FORALL_DATA_EXCEPTION(exception_name, __cfaehm_base_exception_t, assertions, parameters, )
+// FORALL_DATA_INSTANCE(exception_name, (assertions...), (parameters...))
+// Create a polymorphic data exception. The assertion list and parameters must match. This should
+// appear once in each program. This just generates the polymorphic framework, see
+// POLY_VTABLE_INSTANCE to allow instantiations.
+#define FORALL_DATA_INSTANCE(exception_name, assertions, parameters) \
+        _FORALL_CTOR0_INSTANCE(exception_name, assertions, parameters)
+// VTABLE_DECLARATION(exception_name)([new_features...]);
+// Declare a virtual table type for an exception with exception_name. You may also add features
+// (fields on the virtual table) by including them in the second list.
+#define VTABLE_DECLARATION(...) _EXC_DISPATCH(_VTABLE_DECLARATION, __VA_ARGS__)
+// VTABLE_INSTANCE(exception_name)(msg [, others...]);
+// Create the instance of the virtual table. There must be exactly one instance of a virtual table
+// for each exception type. This fills in most of the fields of the virtual table (uses ?=? and
+// ^?{}) but you must provide the message function and any other fields added in the declaration.
+#define VTABLE_INSTANCE(...) _EXC_DISPATCH(_VTABLE_INSTANCE, __VA_ARGS__)
+// FORALL_VTABLE_DECLARATION(exception_name, (assertions...), (parameters...))([new_features...]);
+// Declare a polymorphic virtual table type for an exception with exception_name, the given
+// assertions and parameters. You may also add features (fields on the virtual table). This just
+// generates the polymorphic framework, see POLY_VTABLE_DECLARATION to allow instantiations.
+#define FORALL_VTABLE_DECLARATION(exception_name, assertions, parameters) \
+        _FORALL_VTABLE_DECLARATION(exception_name, __cfaehm_base_exception_t, assertions, parameters, )
+// POLY_VTABLE_DECLARATION(exception_name, types...);
+// Declares that an instantiation for this exception exists for the given types. This should be
+// visible anywhere you use the instantiation of the exception is used.
+#define POLY_VTABLE_DECLARATION(exception_name, ...) \
+        void mark_exception(exception_name(__VA_ARGS__) *); \
+        extern VTABLE_TYPE(exception_name)(__VA_ARGS__) VTABLE_NAME(exception_name)
+// POLY_VTABLE_INSTANCE(exception_name, types...)(msg [, others...]);
+// Creates an instantiation for the given exception for the given types. This should occur only
+// once in the entire program. You must fill in all features, message and any others given in the
+// initial declaration.
+#define POLY_VTABLE_INSTANCE(exception_name, ...) \
+        _POLY_VTABLE_INSTANCE(exception_name, __cfaehm_base_exception_t, __VA_ARGS__)
+// VTABLE_TYPE(exception_name) | VTABLE_NAME(exception_name)
+// Get the name of the vtable type or the name of the vtable instance for an exception type.
+#define VTABLE_TYPE(exception_name) struct _GLUE2(exception_name,_vtable)
+#define VTABLE_NAME(exception_name) _GLUE3(_,exception_name,_vtable_instance)
+// VTABLE_FIELD(exception_name);
+// FORALL_VTABLE_FIELD(exception_name, (parameters-or-types));
+// The declaration of the virtual table field. Should be the first declaration in a virtual type.
+#define VTABLE_FIELD(exception_name) VTABLE_TYPE(exception_name) const * virtual_table
+#define FORALL_VTABLE_FIELD(exception_name, parameters) \
+        VTABLE_TYPE(exception_name) parameters const * virtual_table
+// VTABLE_INIT(object_reference, exception_name);
+// Sets a virtual table field on an object to the virtual table instance for the type.
+#define VTABLE_INIT(this, exception_name) (this).virtual_table = &VTABLE_NAME(exception_name)
+// VTABLE_ASSERTION(exception_name, (parameters...))
+// The assertion that there is an instantiation of the vtable for the exception and types.
+#define VTABLE_ASSERTION(exception_name, parameters) \
+        { VTABLE_TYPE(exception_name) parameters VTABLE_NAME(exception_name); }
+// All internal helper macros begin with an underscore.
+#define _CLOSE(...) __VA_ARGS__ }
+#define _GLUE2(left, right) left##right
+#define _GLUE3(left, middle, right) left##middle##right
+#define _EXC_DISPATCH(to, ...) to(__VA_ARGS__,__cfaehm_base_exception_t,)
+#define _UNPACK(...) __VA_ARGS__
+#define _TRIVIAL_EXCEPTION_DECLARATION(exception_name, parent_name, ...) \
         _VTABLE_DECLARATION(exception_name, parent_name)(); \
         struct exception_name { \
 …
         void ?{}(exception_name & this); \
         const char * _GLUE2(exception_name,_msg)(exception_name * this)
+#define _INST_TRIVIAL_EXCEPTION(exception_name, parent_name, ...) \
+#define _TRIVIAL_EXCEPTION_INSTANCE(exception_name, parent_name, ...) \
         void ?{}(exception_name & this) { \
                 VTABLE_INIT(this, exception_name); \
 …
         _VTABLE_INSTANCE(exception_name, parent_name,)(_GLUE2(exception_name,_msg))
+// DATA_EXCEPTION(exception_name)(fields...);
+// Forward declare an exception that adds fields but no features. The added fields go in the
+// second argument list. The virtual table instance must be provided later (see VTABLE_INSTANCE).
+#define DATA_EXCEPTION(...) _EXC_DISPATCH(_DATA_EXCEPTION, __VA_ARGS__)
+#define _FORALL_TRIVIAL_EXCEPTION(exception_name, parent_name, assertions, \
+                parameters, parent_parameters) \
+        _FORALL_VTABLE_DECLARATION(exception_name, parent_name, assertions, \
+                parameters, parent_parameters)(); \
+        forall assertions struct exception_name { \
+                FORALL_VTABLE_FIELD(exception_name, parameters); \
+        }; \
+        _FORALL_CTOR0_DECLARATION(exception_name, assertions, parameters)
+#define _FORALL_CTOR0_DECLARATION(exception_name, assertions, parameters) \
+        forall(_UNPACK assertions | VTABLE_ASSERTION(exception_name, parameters) ) \
+        void ?{}(exception_name parameters & this)
+#define _FORALL_CTOR0_INSTANCE(exception_name, assertions, parameters) \
+        _FORALL_CTOR0_DECLARATION(exception_name, assertions, parameters) { \
+                VTABLE_INIT(this, exception_name); \
+        }
 #define _DATA_EXCEPTION(exception_name, parent_name, ...) \
         _VTABLE_DECLARATION(exception_name, parent_name)(); \
+        struct exception_name { VTABLE_FIELD(exception_name); _CLOSE
+// VTABLE_DECLARATION(exception_name)([new_features...]);
+// Declare a virtual table type for an exception with exception_name. You may also add features
+// (fields on the virtual table) by including them in the second list.
+#define VTABLE_DECLARATION(...) _EXC_DISPATCH(_VTABLE_DECLARATION, __VA_ARGS__)
+        struct exception_name { \
+                VTABLE_FIELD(exception_name); \
+                _CLOSE
+#define _FORALL_DATA_EXCEPTION(exception_name, parent_name, \
+                assertions, parameters, parent_parameters) \
+        _FORALL_VTABLE_DECLARATION(exception_name, parent_name, \
+                assertions, parameters, parent_parameters)(); \
+        _FORALL_CTOR0_DECLARATION(exception_name, assertions, parameters); \
+        forall assertions struct exception_name { \
+                FORALL_VTABLE_FIELD(exception_name, parameters); \
+                _CLOSE
 #define _VTABLE_DECLARATION(exception_name, parent_name, ...) \
         struct exception_name; \
+        void mark_exception(exception_name *); \
         VTABLE_TYPE(exception_name); \
         extern VTABLE_TYPE(exception_name) VTABLE_NAME(exception_name); \
 …
                 _CLOSE
-// VTABLE_INSTANCE(exception_name)(msg [, others...]);
-// Create the instance of the virtual table. There must be exactly one instance of a virtual table
-// for each exception type. This fills in most of the fields of the virtual table (uses ?=? and
-// ^?{}) but you must provide the message function and any other fields added in the declaration.
-#define VTABLE_INSTANCE(...) _EXC_DISPATCH(_VTABLE_INSTANCE, __VA_ARGS__)
 #define _VTABLE_INSTANCE(exception_name, parent_name, ...) \
+        void mark_exception(exception_name *) {} \
         void _GLUE2(exception_name,_copy)(exception_name * this, exception_name * other) { \
                 *this = *other; \
 …
                 _CLOSE
+// VTABLE_TYPE(exception_name) | VTABLE_NAME(exception_name)
+// Get the name of the vtable type or the name of the vtable instance for an exception type.
+#define VTABLE_TYPE(exception_name) struct _GLUE2(exception_name,_vtable)
+#define VTABLE_NAME(exception_name) _GLUE3(_,exception_name,_vtable_instance)
+// VTABLE_FIELD(exception_name);
+// The declaration of the virtual table field. Should be the first declaration in a virtual type.
+#define VTABLE_FIELD(exception_name) VTABLE_TYPE(exception_name) const * virtual_table
+// VTABLE_INIT(object_reference, exception_name);
+// Sets a virtual table field on an object to the virtual table instance for the type.
+#define VTABLE_INIT(this, exception_name) (this).virtual_table = &VTABLE_NAME(exception_name)
+#define _FORALL_VTABLE_DECLARATION(exception_name, parent_name, assertions, \
+                parameters, parent_parameters) \
+        forall assertions struct exception_name; \
+        forall assertions VTABLE_TYPE(exception_name) { \
+                VTABLE_TYPE(parent_name) parent_parameters const * parent; \
+                size_t size; \
+                void (*copy)(exception_name parameters * this, exception_name parameters * other); \
+                void (*free)(exception_name parameters & this); \
+                const char * (*msg)(exception_name parameters * this); \
+                _CLOSE
+#define _POLY_VTABLE_INSTANCE(exception_name, parent_name, ...) \
+        void mark_exception(exception_name(__VA_ARGS__) *) {} \
+        void _GLUE2(exception_name,_copy)( \
+                        exception_name(__VA_ARGS__) * this, exception_name(__VA_ARGS__) * other) { \
+                *this = *other; \
+        } \
+        VTABLE_TYPE(exception_name)(__VA_ARGS__) VTABLE_NAME(exception_name) @= { \
+                &VTABLE_NAME(parent_name), sizeof(exception_name(__VA_ARGS__)), \
+                _GLUE2(exception_name,_copy), ^?{}, \
+                _CLOSE

libcfa/src/executor.cfa

-              r07d867b
+              r22f94a4
 // buffer.
-#include <bits/containers.hfa>
 #include <thread.hfa>
 #include <stdio.h>
+#include <containers/list.hfa>
+forall( dtype T )
+monitor Buffer {                                        // unbounded buffer
+    __queue_t( T ) queue;                               // unbounded list of work requests
+    condition delay;
+}; // Buffer
+forall( dtype T | is_node(T) ) {
+    void insert( Buffer( T ) & mutex buf, T * elem ) with(buf) {
+        append( queue, elem );                          // insert element into buffer
+        signal( delay );                                // restart
+    } // insert
+forall( dtype T | $dlistable(T, T) ) {
+        monitor Buffer {                                                                        // unbounded buffer
+                dlist( T, T ) queue;                                                    // unbounded list of work requests
+                condition delay;
+        }; // Buffer
     T * remove( Buffer( T ) & mutex buf ) with(buf) {
         if ( queue.head != 0 ) wait( delay );                   // no request to process ? => wait
+//      return pop_head( queue );
+    } // remove
+} // distribution
+        void insert( Buffer(T) & mutex buf, T * elem ) with(buf) {
+                dlist( T, T ) * qptr = &queue;                                  // workaround https://cforall.uwaterloo.ca/trac/ticket/166
+                insert_last( *qptr, *elem );                                    // insert element into buffer
+                signal( delay );                                                                // restart
+        } // insert
+struct WRequest {                                       // client request, no return
+    void (* action)( void );
+    WRequest * next;                                    // intrusive queue field
+        T * remove( Buffer(T) & mutex buf ) with(buf) {
+                dlist( T, T ) * qptr = &queue;                                  // workaround https://cforall.uwaterloo.ca/trac/ticket/166
+                // if ( (*qptr)`is_empty ) wait( delay );                       // no request to process ? => wait
+          if ( (*qptr)`is_empty ) return 0p;                            // no request to process ? => wait
+                return &pop_first( *qptr );
+        } // remove
+} // forall
+struct WRequest {                                                                               // client request, no return
+        void (* action)( void );
+        DLISTED_MGD_IMPL_IN(WRequest)
 }; // WRequest
+DLISTED_MGD_IMPL_OUT(WRequest)
+WRequest *& get_next( WRequest & this ) { return this.next; }
+void ?{}( WRequest & req ) with(req) { action = 0; next = 0; }
+void ?{}( WRequest & req, void (* action)( void ) ) with(req) { req.action = action; next = 0; }
+void ?{}( WRequest & req ) with(req) { action = 0; }
+void ?{}( WRequest & req, void (* action)( void ) ) with(req) { req.action = action; }
 bool stop( WRequest & req ) { return req.action == 0; }
 void doit( WRequest & req ) { req.action(); }
 // Each worker has its own work buffer to reduce contention between client and server. Hence, work requests arrive and
 // are distributed into buffers in a roughly round-robin order.
+// Each worker has its own set (when requests buffers > workers) of work buffers to reduce contention between client
+// and server, where work requests arrive and are distributed into buffers in a roughly round-robin order.
 thread Worker {
+    Buffer( WRequest ) * requests;
+    unsigned int start, range;
+        Buffer(WRequest) * requests;
+        WRequest * request;
+        unsigned int start, range;
 }; // Worker
 void main( Worker & w ) with(w) {
     for ( int i = 0;; i = (i + 1) % range ) {
         WRequest * request = remove( requests[i + start] );
       if ( ! request ) { yield(); continue; }
       if ( stop( *request ) ) break;
         doit( *request );
         delete( request );
     } // for
+        for ( int i = 0;; i = (i + 1) % range ) {
+                request = remove( requests[i + start] );
+          if ( ! request ) { yield(); continue; }
+          if ( stop( *request ) ) break;
+                doit( *request );
+                delete( request );
+        } // for
 } // Worker::main
 void ?{}( Worker & worker, cluster * wc, Buffer( WRequest ) * requests, unsigned int start, unsigned int range ) {
+    (*get_thread(worker)){ *wc };                       // create on given cluster
     worker.[requests, start, range] = [requests, start, range];
+void ?{}( Worker & worker, cluster * wc, Buffer(WRequest) * requests, unsigned int start, unsigned int range ) {
+        ((thread &)worker){ *wc };
+        worker.[requests, request, start, range] = [requests, 0p, start, range];
 } // ?{}
+WRequest * current_request( Worker & worker ) { return worker.request; }
 struct Executor {
+    cluster * cluster;                                  // if workers execute on separate cluster
+    processor ** processors;                            // array of virtual processors adding parallelism for workers
+    Buffer( WRequest ) * requests;                      // list of work requests
+    Worker ** workers;                                  // array of workers executing work requests
+    unsigned int nprocessors, nworkers, nmailboxes;     // number of mailboxes/workers/processor tasks
+    bool sepClus;                                       // use same or separate cluster for executor
+        cluster * cluster;                                                                      // if workers execute on separate cluster
+        processor ** processors;                                                        // array of virtual processors adding parallelism for workers
+        Buffer(WRequest) * requests;                                            // list of work requests
+        Worker ** workers;                                                                      // array of workers executing work requests
+        unsigned int nprocessors, nworkers, nrqueues;           // number of processors/threads/request queues
+        bool sepClus;                                                                           // use same or separate cluster for executor
+        unsigned int next;                                                                      // demultiplexed across worker buffers
 }; // Executor
-static thread_local unsigned int next;                  // demultiplexed across worker buffers
 unsigned int tickets( Executor & ex ) with(ex) {
     //return uFetchAdd( next, 1 ) % nmailboxes;
     return next++ % nmailboxes;                         // no locking, interference randomizes
+        //return uFetchAdd( next, 1 ) % nrqueues;
+        return next++ % nrqueues;                                                       // no locking, interference randomizes
 } // tickets
 void ?{}( Executor & ex, unsigned int np, unsigned int nw, unsigned int nm, bool sc = false ) with(ex) {
     [nprocessors, nworkers, nmailboxes, sepClus] = [np, nw, nm, sc];
     assert( nmailboxes >= nworkers );
     cluster = sepClus ? new( "Executor" ) : active_cluster();
     processors = (processor **)anew( nprocessors );
     requests = anew( nmailboxes );
     workers = (Worker **)anew( nworkers );
+void ?{}( Executor & ex, unsigned int np, unsigned int nw, unsigned int nr, bool sc = false ) with(ex) {
+        [nprocessors, nworkers, nrqueues, sepClus] = [np, nw, nr, sc];
+        assert( nrqueues >= nworkers );
+        cluster = sepClus ? new( "Executor" ) : active_cluster();
+        processors = aalloc( nprocessors );
+        requests = anew( nrqueues );
+        workers = aalloc( nworkers );
     for ( i; nprocessors ) {
         processors[ i ] = new( *cluster );
     } // for
+        for ( i; nprocessors ) {
+                processors[i] = new( *cluster );
+        } // for
+    unsigned int reqPerWorker = nmailboxes / nworkers, extras = nmailboxes % nworkers;
+    for ( unsigned int i = 0, step = 0; i < nworkers; i += 1, step += reqPerWorker + ( i < extras ? 1 : 0 ) ) {
+        workers[ i ] = new( cluster, requests, step, reqPerWorker + ( i < extras ? 1 : 0 ) );
+    } // for
+        unsigned int reqPerWorker = nrqueues / nworkers, extras = nrqueues % nworkers;
+//      for ( unsigned int i = 0, start = 0, range; i < nworkers; i += 1, start += range ) {
+    for ( i; nworkers : start; 0u ~ @ ~ range : range; ) {
+            range = reqPerWorker + ( i < extras ? 1 : 0 );
+                workers[i] = new( cluster, requests, start, range );
+        } // for
 } // ?{}
 void ?{}( Executor & ex, unsigned int nprocessors, unsigned int nworkers, bool sepClus = false ) {
     ex{ nprocessors, nworkers, nworkers, sepClus };
+        ex{ nprocessors, nworkers, nworkers, sepClus };
+}
 void ?{}( Executor & ex, unsigned int nprocessors, bool sepClus = false ) {
     ex{ nprocessors, nprocessors, nprocessors, sepClus };
+        ex{ nprocessors, nprocessors, nprocessors, sepClus };
+}
 void ?{}( Executor & ex ) {                             // special for current cluster
     ex{ 0, active_cluster()->nprocessors, false };
+void ?{}( Executor & ex ) {                                                             // special for current cluster, no processors added
+        ex{ 0, active_cluster()->nprocessors, false };
+}
 void ^?{}( Executor & ex ) with(ex) {
     // Add one sentinel per worker to stop them. Since in destructor, no new work should be queued.  Cannot combine next
+    // two loops and only have a single sentinel because workers arrive in arbitrary order, so worker1 may take the
     // single sentinel while waiting for worker 0 to end.
+        // Add one sentinel per worker to stop them. Since in destructor, no new external work should be queued.  Cannot
+        // combine next two loops and only have a single sentinel because workers arrive in arbitrary order, so worker1 may
+        // take the single sentinel while waiting for worker 0 to end.
     WRequest sentinel[nworkers];
     unsigned int reqPerWorker = nmailboxes / nworkers;
     for ( unsigned int i = 0, step = 0; i < nworkers; i += 1, step += reqPerWorker ) {
         insert( requests[step], &sentinel[i] );         // force eventually termination
     } // for
     for ( i; nworkers ) {
         delete( workers[ i ] );
     } // for
     for ( i; nprocessors ) {
         delete( processors[ i ] );
     } // for
+        WRequest sentinel[nworkers];
+        unsigned int reqPerWorker = nrqueues / nworkers;
+        for ( unsigned int i = 0, step = 0; i < nworkers; i += 1, step += reqPerWorker ) {
+                insert( requests[step], &sentinel[i] );                 // force eventually termination
+        } // for
+        for ( i; nworkers ) {
+                delete( workers[i] );
+        } // for
+        for ( i; nprocessors ) {
+                delete( processors[i] );
+        } // for
+    delete( workers );
+    delete( requests );
+    delete( processors );
+    if ( sepClus ) { delete( cluster ); }
+        free( workers );
+//      adelete( nrqueues, requests );
+        for ( i; nrqueues ) ^?{}( requests[i] );                        // FIX ME: problem with resolver
+        free( requests );
+        free( processors );
+        if ( sepClus ) { delete( cluster ); }
 } // ^?{}
 void send( Executor & ex, void (* action)( void ) ) {   // asynchronous call, no return value
     WRequest * node = new( action );
     insert( ex.requests[tickets( ex )], node );
+        WRequest * node = new( action );
+        insert( ex.requests[tickets( ex )], node );
 } // send
 int counter = 0;
 void workie( void ) {
     __atomic_add_fetch( &counter, 1, __ATOMIC_SEQ_CST );
 //    fprintf( stderr, "workie\n" );
+void work( void ) {
+        __atomic_add_fetch( &counter, 1, __ATOMIC_SEQ_CST );
+        // fprintf( stderr, "workie\n" );
+}
+int main() {
+    {
+        Executor exector;
+        for ( i; 3000 ) {
+            send( exector, workie );
+            if ( i % 100 ) yield();
+        } // for
+    }
+    printf( "%d\n", counter );
+int main( int argc, char * argv[] ) {
+        int times = 1_000_000;
+        if ( argc == 2 ) times = atoi( argv[1] );
+        processor p[7];
+        {
+                Executor exector;
+                for ( i; times ) {
+                        send( exector, work );
+                        if ( i % 100 == 0 ) yield();
+                } // for
+        }
+        printf( "%d\n", counter );
+}
 // Local Variables: //
+// tab-width: 4" //
 // compile-command: "cfa executor.cfa" //
 // End: //

libcfa/src/fstream.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 19:01:01 2020
 // Update Count     : 363
+// Last Modified On : Fri Jun 19 16:24:54 2020
+// Update Count     : 384
 //
 …
 //*********************************** ofstream ***********************************
+// *********************************** ofstream ***********************************
 …
         #ifdef __CFA_DEBUG__
         if ( file == 0p ) {
+                abort | IO_MSG "open output file \"" | name | "\"" | nl | strerror( errno );
+                throw (Open_Failure){ os };
+                // abort | IO_MSG "open output file \"" | name | "\"" | nl | strerror( errno );
         } // if
         #endif // __CFA_DEBUG__
 …
 void close( ofstream & os ) {
+        if ( (FILE *)(os.$file) == stdout || (FILE *)(os.$file) == stderr ) return;
+  if ( (FILE *)(os.$file) == 0p ) return;
+  if ( (FILE *)(os.$file) == (FILE *)stdout || (FILE *)(os.$file) == (FILE *)stderr ) return;
         if ( fclose( (FILE *)(os.$file) ) == EOF ) {
                 abort | IO_MSG "close output" | nl | strerror( errno );
         } // if
+        os.$file = 0p;
 } // close
 …
 //*********************************** ifstream ***********************************
+// *********************************** ifstream ***********************************
 …
         #ifdef __CFA_DEBUG__
         if ( file == 0p ) {
+                abort | IO_MSG "open input file \"" | name | "\"" | nl | strerror( errno );
+                throw (Open_Failure){ is };
+                // abort | IO_MSG "open input file \"" | name | "\"" | nl | strerror( errno );
         } // if
         #endif // __CFA_DEBUG__
 …
 void close( ifstream & is ) {
+        if ( (FILE *)(is.$file) == stdin ) return;
+  if ( (FILE *)(is.$file) == 0p ) return;
+  if ( (FILE *)(is.$file) == (FILE *)stdin ) return;
         if ( fclose( (FILE *)(is.$file) ) == EOF ) {
                 abort | IO_MSG "close input" | nl | strerror( errno );
         } // if
+        is.$file = 0p;
 } // close
 …
 ifstream & sin = sinFile, & stdin = sinFile;
+// *********************************** exceptions ***********************************
+void ?{}( Open_Failure & this, ofstream & ostream ) {
+        VTABLE_INIT(this, Open_Failure);
+        this.ostream = &ostream;
+        this.tag = 1;
+}
+void ?{}( Open_Failure & this, ifstream & istream ) {
+        VTABLE_INIT(this, Open_Failure);
+        this.istream = &istream;
+        this.tag = 0;
+}
+const char * Open_Failure_msg(Open_Failure * this) {
+        return "Open_Failure";
+}
+VTABLE_INSTANCE(Open_Failure)(Open_Failure_msg);
+void throwOpen_Failure( ofstream & ostream ) {
+        Open_Failure exc = { ostream };
+}
+void throwOpen_Failure( ifstream & istream ) {
+        Open_Failure exc = { istream };
+}
 // Local Variables: //
 // tab-width: 4 //

libcfa/src/fstream.hfa

-              r07d867b
+              r22f94a4
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Feb 17 08:29:23 2020
 // Update Count     : 175
+// Last Modified On : Fri Jun 19 16:29:17 2020
+// Update Count     : 189
 //
 …
 #include "iostream.hfa"
+#include <exception.hfa>
 //*********************************** ofstream ***********************************
+// *********************************** ofstream ***********************************
 …
 //*********************************** ifstream ***********************************
+// *********************************** ifstream ***********************************
 …
 extern ifstream & sin, & stdin;                                                 // aliases
+// *********************************** exceptions ***********************************
+DATA_EXCEPTION(Open_Failure)(
+        union {
+                ofstream * ostream;
+                ifstream * istream;
+        };
+        // TEMPORARY: need polymorphic exceptions
+        int tag;                                                                                        // 1 => ostream; 0 => istream
+);
+void ?{}( Open_Failure & this, ofstream & ostream );
+void ?{}( Open_Failure & this, ifstream & istream );
 // Local Variables: //
 // mode: c //

libcfa/src/heap.cfa

-              r07d867b
+              r22f94a4
 // file "LICENCE" distributed with Cforall.
 //
 // heap.c --
+// heap.cfa --
 //
 // Author           : Peter A. Buhr
 // Created On       : Tue Dec 19 21:58:35 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed May  6 17:29:26 2020
 // Update Count     : 727
+// Last Modified On : Sun Aug  9 12:23:20 2020
+// Update Count     : 894
 //
 …
 #include <string.h>                                                                             // memset, memcpy
 #include <limits.h>                                                                             // ULONG_MAX
+extern "C" {
+#include <malloc.h>                                                                             // memalign, malloc_usable_size
 #include <sys/mman.h>                                                                   // mmap, munmap
+} // extern "C"
+#include "bits/align.hfa"                                                               // libPow2
+#include "bits/align.hfa"                                                               // libAlign
 #include "bits/defs.hfa"                                                                // likely, unlikely
 #include "bits/locks.hfa"                                                               // __spinlock_t
 #include "startup.hfa"                                                                  // STARTUP_PRIORITY_MEMORY
 //#include "stdlib.hfa"                                                                 // bsearchl
-#include "malloc.h"
 #include "bitmanip.hfa"                                                                 // ceiling
 …
 };
+size_t default_mmap_start() __attribute__(( weak )) {
+        return __CFA_DEFAULT_MMAP_START__;
+} // default_mmap_start
 size_t default_heap_expansion() __attribute__(( weak )) {
         return __CFA_DEFAULT_HEAP_EXPANSION__;
 } // default_heap_expansion
-size_t default_mmap_start() __attribute__(( weak )) {
-        return __CFA_DEFAULT_MMAP_START__;
-} // default_mmap_start
 #ifdef __CFA_DEBUG__
 static unsigned int allocFree;                                                  // running total of allocations minus frees
+static size_t allocUnfreed;                                                             // running total of allocations minus frees
 static void prtUnfreed() {
         if ( allocFree != 0 ) {
+        if ( allocUnfreed != 0 ) {
                 // DO NOT USE STREAMS AS THEY MAY BE UNAVAILABLE AT THIS POINT.
                 char helpText[512];
                 int len = snprintf( helpText, sizeof(helpText), "CFA warning (UNIX pid:%ld) : program terminating with %u(0x%x) bytes of storage allocated but not freed.\n"
+                int len = snprintf( helpText, sizeof(helpText), "CFA warning (UNIX pid:%ld) : program terminating with %zu(0x%zx) bytes of storage allocated but not freed.\n"
                                                         "Possible cause is unfreed storage allocated by the program or system/library routines called from the program.\n",
                                                         (long int)getpid(), allocFree, allocFree ); // always print the UNIX pid
+                                                        (long int)getpid(), allocUnfreed, allocUnfreed ); // always print the UNIX pid
                 __cfaabi_bits_write( STDERR_FILENO, helpText, len ); // print debug/nodebug
         } // if
 …
 extern "C" {
         void heapAppStart() {                                                           // called by __cfaabi_appready_startup
                 allocFree = 0;
+                allocUnfreed = 0;
         } // heapAppStart
 …
 #define LOCKFREE 1
 #define BUCKETLOCK SPINLOCK
+#if BUCKETLOCK == LOCKFREE
+#include <uStackLF.h>
+#if BUCKETLOCK == SPINLOCK
+#elif BUCKETLOCK == LOCKFREE
+#include <stackLockFree.hfa>
+#else
+        #error undefined lock type for bucket lock
 #endif // LOCKFREE
 …
 struct HeapManager {
-//      struct FreeHeader;                                                                      // forward declaration
         struct Storage {
                 struct Header {                                                                 // header
 …
                                                 struct {                                                // 4-byte word => 8-byte header, 8-byte word => 16-byte header
                                                         #if __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ && __SIZEOF_POINTER__ == 4
                                                         uint32_t padding;                       // unused, force home/blocksize to overlay alignment in fake header
+                                                        uint64_t padding;                       // unused, force home/blocksize to overlay alignment in fake header
                                                         #endif // __BYTE_ORDER__ == __ORDER_BIG_ENDIAN__ && __SIZEOF_POINTER__ == 4
                                                         union {
 //                                                              FreeHeader * home;              // allocated block points back to home locations (must overlay alignment)
+                                                                // FreeHeader * home;           // allocated block points back to home locations (must overlay alignment)
                                                                 // 2nd low-order bit => zero filled
                                                                 void * home;                    // allocated block points back to home locations (must overlay alignment)
                                                                 size_t blockSize;               // size for munmap (must overlay alignment)
                                                                 #if BUCKLOCK == SPINLOCK
+                                                                #if BUCKETLOCK == SPINLOCK
                                                                 Storage * next;                 // freed block points next freed block of same size
                                                                 #endif // SPINLOCK
                                                         };
+                                                        size_t size;                            // allocation size in bytes
                                                         #if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ && __SIZEOF_POINTER__ == 4
                                                         uint32_t padding;                       // unused, force home/blocksize to overlay alignment in fake header
+                                                        uint64_t padding;                       // unused, force home/blocksize to overlay alignment in fake header
                                                         #endif // __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__ && __SIZEOF_POINTER__ == 4
                                                 };
+                                                // future code
+                                                #if BUCKLOCK == LOCKFREE
+                                                Stack<Storage>::Link next;              // freed block points next freed block of same size (double-wide)
+                                                #if BUCKETLOCK == LOCKFREE
+                                                Link(Storage) next;                             // freed block points next freed block of same size (double-wide)
                                                 #endif // LOCKFREE
                                         };
                                 } real; // RealHeader
                                 struct FakeHeader {
                                         #if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
+                                        // 1st low-order bit => fake header & alignment
+                                        uint32_t alignment;
+                                        uint32_t alignment;                                     // 1st low-order bit => fake header & alignment
                                         #endif // __ORDER_LITTLE_ENDIAN__
 …
                                 } fake; // FakeHeader
                         } kind; // Kind
-                        size_t size;                                                            // allocation size in bytes
                 } header; // Header
                 char pad[libAlign() - sizeof( Header )];
 …
         struct FreeHeader {
                 #if BUCKLOCK == SPINLOCK
+                #if BUCKETLOCK == SPINLOCK
                 __spinlock_t lock;                                                              // must be first field for alignment
                 Storage * freeList;
-                #elif BUCKLOCK == LOCKFREE
-                // future code
-                StackLF<Storage> freeList;
                 #else
                         #error undefined lock type for bucket lock
                 #endif // SPINLOCK
+                StackLF(Storage) freeList;
+                #endif // BUCKETLOCK
                 size_t blockSize;                                                               // size of allocations on this list
         }; // FreeHeader
 …
 }; // HeapManager
+#if BUCKETLOCK == LOCKFREE
+static inline {
+        Link(HeapManager.Storage) * ?`next( HeapManager.Storage * this ) { return &this->header.kind.real.next; }
+        void ?{}( HeapManager.FreeHeader & ) {}
+        void ^?{}( HeapManager.FreeHeader & ) {}
+} // distribution
+#endif // LOCKFREE
 static inline size_t getKey( const HeapManager.FreeHeader & freeheader ) { return freeheader.blockSize; }
 …
 #define __STATISTICS__
+// Bucket size must be multiple of 16.
+// Powers of 2 are common allocation sizes, so make powers of 2 generate the minimum required size.
+// Size of array must harmonize with NoBucketSizes and individual bucket sizes must be multiple of 16.
+// Smaller multiples of 16 and powers of 2 are common allocation sizes, so make them generate the minimum required bucket size.
+// malloc(0) returns 0p, so no bucket is necessary for 0 bytes returning an address that can be freed.
 static const unsigned int bucketSizes[] @= {                    // different bucket sizes
 , 32, 48, 64 + sizeof(HeapManager.Storage), // 4
 , 112, 128 + sizeof(HeapManager.Storage), // 3
+ + sizeof(HeapManager.Storage), 32 + sizeof(HeapManager.Storage), 48 + sizeof(HeapManager.Storage), 64 + sizeof(HeapManager.Storage), // 4
+ + sizeof(HeapManager.Storage), 112 + sizeof(HeapManager.Storage), 128 + sizeof(HeapManager.Storage), // 3
 , 192, 224, 256 + sizeof(HeapManager.Storage), // 4
 , 384, 448, 512 + sizeof(HeapManager.Storage), // 4
 …
 };
 static_assert( NoBucketSizes == sizeof(bucketSizes) / sizeof(bucketSizes[0]), "size of bucket array wrong" );
+static_assert( NoBucketSizes == sizeof(bucketSizes) / sizeof(bucketSizes[0] ), "size of bucket array wrong" );
 #ifdef FASTLOOKUP
 …
 static bool heapBoot = 0;                                                               // detect recursion during boot
 #endif // __CFA_DEBUG__
+// The constructor for heapManager is called explicitly in memory_startup.
 static HeapManager heapManager __attribute__(( aligned (128) )) @= {}; // size of cache line to prevent false sharing
 …
 #ifdef __STATISTICS__
 // Heap statistics counters.
+static unsigned int malloc_calls;
+static unsigned long long int malloc_storage;
+static unsigned int aalloc_calls;
+static unsigned long long int aalloc_storage;
+static unsigned int calloc_calls;
+static unsigned long long int calloc_storage;
+static unsigned int memalign_calls;
+static unsigned long long int memalign_storage;
+static unsigned int amemalign_calls;
+static unsigned long long int amemalign_storage;
+static unsigned int cmemalign_calls;
+static unsigned long long int cmemalign_storage;
+static unsigned int resize_calls;
+static unsigned long long int resize_storage;
+static unsigned int realloc_calls;
+static unsigned long long int realloc_storage;
+static unsigned int free_calls;
+static unsigned long long int free_storage;
+static unsigned int mmap_calls;
 static unsigned long long int mmap_storage;
 static unsigned int mmap_calls;
+static unsigned int munmap_calls;
 static unsigned long long int munmap_storage;
 static unsigned int munmap_calls;
+static unsigned int sbrk_calls;
 static unsigned long long int sbrk_storage;
-static unsigned int sbrk_calls;
-static unsigned long long int malloc_storage;
-static unsigned int malloc_calls;
-static unsigned long long int free_storage;
-static unsigned int free_calls;
-static unsigned long long int aalloc_storage;
-static unsigned int aalloc_calls;
-static unsigned long long int calloc_storage;
-static unsigned int calloc_calls;
-static unsigned long long int memalign_storage;
-static unsigned int memalign_calls;
-static unsigned long long int amemalign_storage;
-static unsigned int amemalign_calls;
-static unsigned long long int cmemalign_storage;
-static unsigned int cmemalign_calls;
-static unsigned long long int resize_storage;
-static unsigned int resize_calls;
-static unsigned long long int realloc_storage;
-static unsigned int realloc_calls;
 // Statistics file descriptor (changed by malloc_stats_fd).
 static int statfd = STDERR_FILENO;                                              // default stderr
+static int stat_fd = STDERR_FILENO;                                             // default stderr
 // Use "write" because streams may be shutdown when calls are made.
 …
                                                                         "  sbrk: calls %u / storage %llu\n",
                                                                         malloc_calls, malloc_storage,
                                                                         aalloc_calls, calloc_storage,
+                                                                        aalloc_calls, aalloc_storage,
                                                                         calloc_calls, calloc_storage,
                                                                         memalign_calls, memalign_storage,
 …
-// static inline void noMemory() {
-//      abort( "Heap memory exhausted at %zu bytes.\n"
-//                 "Possible cause is very large memory allocation and/or large amount of unfreed storage allocated by the program or system/library routines.",
-//                 ((char *)(sbrk( 0 )) - (char *)(heapManager.heapBegin)) );
-// } // noMemory
 // thunk problem
 size_t Bsearchl( unsigned int key, const unsigned int * vals, size_t dim ) {
 …
 static inline void checkAlign( size_t alignment ) {
         if ( alignment < libAlign() || ! libPow2( alignment ) ) {
+        if ( alignment < libAlign() || ! is_pow2( alignment ) ) {
                 abort( "Alignment %zu for memory allocation is less than %d and/or not a power of 2.", alignment, libAlign() );
         } // if
 …
                 #endif // __CFA_DEBUG__
                 header = realHeader( header );                                  // backup from fake to real header
+        } else {
+                alignment = libAlign();                                                 // => no fake header
         } // if
 } // fakeHeader
+static inline bool headers( const char name[] __attribute__(( unused )), void * addr, HeapManager.Storage.Header *& header, HeapManager.FreeHeader *& freeElem, size_t & size, size_t & alignment ) with ( heapManager ) {
+static inline bool headers( const char name[] __attribute__(( unused )), void * addr, HeapManager.Storage.Header *& header, HeapManager.FreeHeader *& freeElem,
+                                                        size_t & size, size_t & alignment ) with( heapManager ) {
         header = headerAddr( addr );
         if ( unlikely( heapEnd < addr ) ) {                                     // mmapped ?
+  if ( unlikely( heapEnd < addr ) ) {                                   // mmapped ?
                 fakeHeader( header, alignment );
                 size = header->kind.real.blockSize & -3;                // mmap size
 …
 } // headers
+static inline void * extend( size_t size ) with ( heapManager ) {
+#define NO_MEMORY_MSG "insufficient heap memory available for allocating %zd new bytes."
+static inline void * extend( size_t size ) with( heapManager ) {
         lock( extlock __cfaabi_dbg_ctx2 );
         ptrdiff_t rem = heapRemaining - size;
 …
                 // If the size requested is bigger than the current remaining storage, increase the size of the heap.
                 size_t increase = libCeiling( size > heapExpand ? size : heapExpand, libAlign() );
                 if ( sbrk( increase ) == (void *)-1 ) {
+                size_t increase = ceiling2( size > heapExpand ? size : heapExpand, libAlign() );
+                if ( sbrk( increase ) == (void *)-1 ) {                 // failed, no memory ?
                         unlock( extlock );
+                        errno = ENOMEM;
+                        return 0p;
+                        abort( NO_MEMORY_MSG, size );                           // give up
                 } // if
                 #ifdef __STATISTICS__
 …
 static inline void * doMalloc( size_t size ) with ( heapManager ) {
+static inline void * doMalloc( size_t size ) with( heapManager ) {
         HeapManager.Storage * block;                                            // pointer to new block of storage
 …
                         posn = Bsearchl( (unsigned int)tsize, bucketSizes, (size_t)maxBucketsUsed );
                 HeapManager.FreeHeader * freeElem = &freeLists[posn];
+                // #ifdef FASTLOOKUP
+                // if ( tsize < LookupSizes )
+                //      freeElem = &freeLists[lookup[tsize]];
+                // else
+                // #endif // FASTLOOKUP
+                //      freeElem = bsearchl( tsize, freeLists, (size_t)maxBucketsUsed ); // binary search
+                // HeapManager.FreeHeader * freeElem =
+                //      #ifdef FASTLOOKUP
+                //      tsize < LookupSizes ? &freeLists[lookup[tsize]] :
+                //      #endif // FASTLOOKUP
+                //      bsearchl( tsize, freeLists, (size_t)maxBucketsUsed ); // binary search
+                assert( freeElem <= &freeLists[maxBucketsUsed] ); // subscripting error ?
+                assert( tsize <= freeElem->blockSize );                 // search failure ?
+                verify( freeElem <= &freeLists[maxBucketsUsed] ); // subscripting error ?
+                verify( tsize <= freeElem->blockSize );                 // search failure ?
                 tsize = freeElem->blockSize;                                    // total space needed for request
                 // Spin until the lock is acquired for this particular size of block.
                 #if defined( SPINLOCK )
+                #if BUCKETLOCK == SPINLOCK
                 lock( freeElem->lock __cfaabi_dbg_ctx2 );
                 block = freeElem->freeList;                                             // remove node from stack
                 #else
                 block = freeElem->freeList.pop();
                 #endif // SPINLOCK
+                block = pop( freeElem->freeList );
+                #endif // BUCKETLOCK
                 if ( unlikely( block == 0p ) ) {                                // no free block ?
                         #if defined( SPINLOCK )
+                        #if BUCKETLOCK == SPINLOCK
                         unlock( freeElem->lock );
                         #endif // SPINLOCK
+                        #endif // BUCKETLOCK
                         // Freelist for that size was empty, so carve it out of the heap if there's enough left, or get some more
 …
                         block = (HeapManager.Storage *)extend( tsize ); // mutual exclusion on call
+  if ( unlikely( block == 0p ) ) return 0p;
+                #if defined( SPINLOCK )
+                #if BUCKETLOCK == SPINLOCK
                 } else {
                         freeElem->freeList = block->header.kind.real.next;
                         unlock( freeElem->lock );
                 #endif // SPINLOCK
+                #endif // BUCKETLOCK
                 } // if
 …
         } else {                                                                                        // large size => mmap
   if ( unlikely( size > ULONG_MAX - pageSize ) ) return 0p;
                 tsize = libCeiling( tsize, pageSize );                  // must be multiple of page size
+                tsize = ceiling2( tsize, pageSize );                    // must be multiple of page size
                 #ifdef __STATISTICS__
                 __atomic_add_fetch( &mmap_calls, 1, __ATOMIC_SEQ_CST );
                 __atomic_add_fetch( &mmap_storage, tsize, __ATOMIC_SEQ_CST );
                 #endif // __STATISTICS__
                 block = (HeapManager.Storage *)mmap( 0, tsize, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS, mmapFd, 0 );
+                if ( block == (HeapManager.Storage *)MAP_FAILED ) {
+                if ( block == (HeapManager.Storage *)MAP_FAILED ) { // failed ?
+                        if ( errno == ENOMEM ) abort( NO_MEMORY_MSG, tsize ); // no memory
                         // Do not call strerror( errno ) as it may call malloc.
                         abort( "(HeapManager &)0x%p.doMalloc() : internal error, mmap failure, size:%zu error:%d.", &heapManager, tsize, errno );
                 } // if
+                } //if
                 #ifdef __CFA_DEBUG__
                 // Set new memory to garbage so subsequent uninitialized usages might fail.
 …
         } // if
         block->header.size = size;                                                      // store allocation size
+        block->header.kind.real.size = size;                            // store allocation size
         void * addr = &(block->data);                                           // adjust off header to user bytes
+        verify( ((uintptr_t)addr & (libAlign() - 1)) == 0 ); // minimum alignment ?
         #ifdef __CFA_DEBUG__
+        assert( ((uintptr_t)addr & (libAlign() - 1)) == 0 ); // minimum alignment ?
+        __atomic_add_fetch( &allocFree, tsize, __ATOMIC_SEQ_CST );
+        __atomic_add_fetch( &allocUnfreed, tsize, __ATOMIC_SEQ_CST );
         if ( traceHeap() ) {
                 enum { BufferSize = 64 };
                 char helpText[BufferSize];
                 int len = snprintf( helpText, BufferSize, "%p = Malloc( %zu ) (allocated %zu)\n", addr, size, tsize );
-                // int len = snprintf( helpText, BufferSize, "Malloc %p %zu\n", addr, size );
                 __cfaabi_bits_write( STDERR_FILENO, helpText, len ); // print debug/nodebug
         } // if
 …
 static inline void doFree( void * addr ) with ( heapManager ) {
+static inline void doFree( void * addr ) with( heapManager ) {
         #ifdef __CFA_DEBUG__
         if ( unlikely( heapManager.heapBegin == 0p ) ) {
 …
                 free_storage += size;
                 #endif // __STATISTICS__
                 #if defined( SPINLOCK )
+                #if BUCKETLOCK == SPINLOCK
                 lock( freeElem->lock __cfaabi_dbg_ctx2 );               // acquire spin lock
                 header->kind.real.next = freeElem->freeList;    // push on stack
 …
                 unlock( freeElem->lock );                                               // release spin lock
                 #else
                 freeElem->freeList.push( *(HeapManager.Storage *)header );
                 #endif // SPINLOCK
+                push( freeElem->freeList, *(HeapManager.Storage *)header );
+                #endif // BUCKETLOCK
         } // if
         #ifdef __CFA_DEBUG__
         __atomic_add_fetch( &allocFree, -size, __ATOMIC_SEQ_CST );
+        __atomic_add_fetch( &allocUnfreed, -size, __ATOMIC_SEQ_CST );
         if ( traceHeap() ) {
+                enum { BufferSize = 64 };
+                char helpText[BufferSize];
+                char helpText[64];
                 int len = snprintf( helpText, sizeof(helpText), "Free( %p ) size:%zu\n", addr, size );
                 __cfaabi_bits_write( STDERR_FILENO, helpText, len ); // print debug/nodebug
 …
 size_t prtFree( HeapManager & manager ) with ( manager ) {
+size_t prtFree( HeapManager & manager ) with( manager ) {
         size_t total = 0;
         #ifdef __STATISTICS__
 …
                 #endif // __STATISTICS__
                 #if defined( SPINLOCK )
+                #if BUCKETLOCK == SPINLOCK
                 for ( HeapManager.Storage * p = freeLists[i].freeList; p != 0p; p = p->header.kind.real.next ) {
                 #else
                 for ( HeapManager.Storage * p = freeLists[i].freeList.top(); p != 0p; p = p->header.kind.real.next.top ) {
                 #endif // SPINLOCK
+                for ( HeapManager.Storage * p = top( freeLists[i].freeList ); p != 0p; p = (p)`next->top ) {
+                #endif // BUCKETLOCK
                         total += size;
                         #ifdef __STATISTICS__
 …
 static void ?{}( HeapManager & manager ) with ( manager ) {
+static void ?{}( HeapManager & manager ) with( manager ) {
         pageSize = sysconf( _SC_PAGESIZE );
 …
         char * end = (char *)sbrk( 0 );
         heapBegin = heapEnd = sbrk( (char *)libCeiling( (long unsigned int)end, libAlign() ) - end ); // move start of heap to multiple of alignment
+        heapBegin = heapEnd = sbrk( (char *)ceiling2( (long unsigned int)end, libAlign() ) - end ); // move start of heap to multiple of alignment
 } // HeapManager
 …
         if ( traceHeapTerm() ) {
                 printStats();
                 // if ( prtfree() ) prtFree( heapManager, true );
+                // prtUnfreed() called in heapAppStop()
         } // if
         #endif // __STATISTICS__
 …
 void memory_startup( void ) {
         #ifdef __CFA_DEBUG__
         if ( unlikely( heapBoot ) ) {                                           // check for recursion during system boot
+        if ( heapBoot ) {                                                                       // check for recursion during system boot
                 // DO NOT USE STREAMS AS THEY MAY BE UNAVAILABLE AT THIS POINT.
                 abort( "boot() : internal error, recursively invoked during system boot." );
 …
         #endif // __CFA_DEBUG__
         //assert( heapManager.heapBegin != 0 );
+        //verify( heapManager.heapBegin != 0 );
         //heapManager{};
         if ( heapManager.heapBegin == 0p ) heapManager{};       // sanity check
 …
 static inline void * mallocNoStats( size_t size ) {             // necessary for malloc statistics
+        //assert( heapManager.heapBegin != 0 );
+        if ( unlikely( heapManager.heapBegin == 0p ) ) heapManager{}; // called before memory_startup ?
+        verify( heapManager.heapBegin != 0p );                          // called before memory_startup ?
+  if ( unlikely( size ) == 0 ) return 0p;                               // 0 BYTE ALLOCATION RETURNS NULL POINTER
 #if __SIZEOF_POINTER__ == 8
         verify( size < ((typeof(size_t))1 << 48) );
 #endif // __SIZEOF_POINTER__ == 8
+        void * addr = doMalloc( size );
+        if ( unlikely( addr == 0p ) ) errno = ENOMEM;           // POSIX
+        return addr;
+        return doMalloc( size );
 } // mallocNoStats
 …
 static inline void * callocNoStats( size_t dim, size_t elemSize ) {
         size_t size = dim * elemSize;
+  if ( unlikely( size ) == 0 ) return 0p;                               // 0 BYTE ALLOCATION RETURNS NULL POINTER
         char * addr = (char *)mallocNoStats( size );
-  if ( unlikely( addr == 0p ) ) return 0p;
         HeapManager.Storage.Header * header;
         HeapManager.FreeHeader * freeElem;
         size_t bsize, alignment;
-        bool mapped __attribute__(( unused )) = headers( "calloc", addr, header, freeElem, bsize, alignment );
         #ifndef __CFA_DEBUG__
+        bool mapped =
+        #endif // __CFA_DEBUG__
+                headers( "calloc", addr, header, freeElem, bsize, alignment );
+        #ifndef __CFA_DEBUG__
         // Mapped storage is zero filled, but in debug mode mapped memory is scrubbed in doMalloc, so it has to be reset to zero.
         if ( ! mapped )
         #endif // __CFA_DEBUG__
+                // Zero entire data space even when > than size => realloc without a new allocation and zero fill works.
+                // <-------00000000000000000000000000000000000000000000000000000> bsize (bucket size)
+                // <-------0000000000000000000000000000UUUUUUUUUUUUUUUUUUUUUUUUU> bsize (bucket size) U => undefined
                 // `-header`-addr                      `-size
                 memset( addr, '\0', bsize - sizeof(HeapManager.Storage) ); // set to zeros
+                memset( addr, '\0', size );                                             // set to zeros
         header->kind.real.blockSize |= 2;                                       // mark as zero filled
 …
+static inline void * memalignNoStats( size_t alignment, size_t size ) { // necessary for malloc statistics
+static inline void * memalignNoStats( size_t alignment, size_t size ) {
+  if ( unlikely( size ) == 0 ) return 0p;                               // 0 BYTE ALLOCATION RETURNS NULL POINTER
         #ifdef __CFA_DEBUG__
         checkAlign( alignment );                                                        // check alignment
 …
         // add sizeof(Storage) for fake header
         char * addr = (char *)mallocNoStats( size + alignment - libAlign() + sizeof(HeapManager.Storage) );
-  if ( unlikely( addr == 0p ) ) return addr;
         // address in the block of the "next" alignment address
         char * user = (char *)libCeiling( (uintptr_t)(addr + sizeof(HeapManager.Storage)), alignment );
+        char * user = (char *)ceiling2( (uintptr_t)(addr + sizeof(HeapManager.Storage)), alignment );
         // address of header from malloc
         HeapManager.Storage.Header * realHeader = headerAddr( addr );
+        realHeader->kind.real.size = size;                                      // correct size to eliminate above alignment offset
         // address of fake header * before* the alignment location
         HeapManager.Storage.Header * fakeHeader = headerAddr( user );
 …
 static inline void * cmemalignNoStats( size_t alignment, size_t dim, size_t elemSize ) {
         size_t size = dim * elemSize;
+  if ( unlikely( size ) == 0 ) return 0p;                               // 0 BYTE ALLOCATION RETURNS NULL POINTER
         char * addr = (char *)memalignNoStats( alignment, size );
+  if ( unlikely( addr == 0p ) ) return 0p;
         HeapManager.Storage.Header * header;
         HeapManager.FreeHeader * freeElem;
         size_t bsize;
-        bool mapped __attribute__(( unused )) = headers( "cmemalign", addr, header, freeElem, bsize, alignment );
         #ifndef __CFA_DEBUG__
+        bool mapped =
+        #endif // __CFA_DEBUG__
+                headers( "cmemalign", addr, header, freeElem, bsize, alignment );
         // Mapped storage is zero filled, but in debug mode mapped memory is scrubbed in doMalloc, so it has to be reset to zero.
+        #ifndef __CFA_DEBUG__
         if ( ! mapped )
         #endif // __CFA_DEBUG__
+                memset( addr, '\0', dataStorage( bsize, addr, header ) ); // set to zeros
+                // <-------0000000000000000000000000000UUUUUUUUUUUUUUUUUUUUUUUUU> bsize (bucket size) U => undefined
+                // `-header`-addr                      `-size
+                memset( addr, '\0', size );                                             // set to zeros
         header->kind.real.blockSize |= 2;                                       // mark as zero filled
         return addr;
 } // cmemalignNoStats
-// supported mallopt options
-#ifndef M_MMAP_THRESHOLD
-#define M_MMAP_THRESHOLD (-1)
-#endif // M_TOP_PAD
-#ifndef M_TOP_PAD
-#define M_TOP_PAD (-2)
-#endif // M_TOP_PAD
 …
         // Same as malloc() except size bytes is an array of dim elements each of elemSize bytes.
         void * aalloc( size_t dim, size_t elemSize ) {
+                size_t size = dim * elemSize;
                 #ifdef __STATISTICS__
                 __atomic_add_fetch( &aalloc_calls, 1, __ATOMIC_SEQ_CST );
                 __atomic_add_fetch( &aalloc_storage, dim * elemSize, __ATOMIC_SEQ_CST );
                 #endif // __STATISTICS__
                 return mallocNoStats( dim * elemSize );
+                __atomic_add_fetch( &aalloc_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                return mallocNoStats( size );
         } // aalloc
 …
                 return callocNoStats( dim, elemSize );
         } // calloc
         // Change the size of the memory block pointed to by oaddr to size bytes. The contents are undefined.  If oaddr is
 …
                 #ifdef __STATISTICS__
                 __atomic_add_fetch( &resize_calls, 1, __ATOMIC_SEQ_CST );
-                __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
                 #endif // __STATISTICS__
                 // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
+          if ( unlikely( size == 0 ) ) { free( oaddr ); return mallocNoStats( size ); } // special cases
+          if ( unlikely( oaddr == 0p ) ) return mallocNoStats( size );
+          if ( unlikely( size == 0 ) ) { free( oaddr ); return 0p; } // special cases
+          if ( unlikely( oaddr == 0p ) ) {
+                        #ifdef __STATISTICS__
+                        __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
+                        #endif // __STATISTICS__
+                        return mallocNoStats( size );
+                } // if
                 HeapManager.Storage.Header * header;
                 HeapManager.FreeHeader * freeElem;
                 size_t bsize, oalign = 0;
+                size_t bsize, oalign;
                 headers( "resize", oaddr, header, freeElem, bsize, oalign );
                 size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
                 // same size, DO NOT preserve STICKY PROPERTIES.
           if ( oalign == 0 && size <= odsize && odsize <= size * 2 ) { // allow 50% wasted storage for smaller size
+                if ( oalign <= libAlign() && size <= odsize && odsize <= size * 2 ) { // allow 50% wasted storage for smaller size
                         header->kind.real.blockSize &= -2;                      // no alignment and turn off 0 fill
+                        header->kind.real.size = size;                          // reset allocation size
                         return oaddr;
                 } // if
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
                 // change size, DO NOT preserve STICKY PROPERTIES.
                 free( oaddr );
+                void * naddr = mallocNoStats( size );                   // create new area
+                return naddr;
+                return mallocNoStats( size );                                   // create new area
         } // resize
 …
                 #ifdef __STATISTICS__
                 __atomic_add_fetch( &realloc_calls, 1, __ATOMIC_SEQ_CST );
-                __atomic_add_fetch( &realloc_storage, size, __ATOMIC_SEQ_CST );
                 #endif // __STATISTICS__
                 // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
+          if ( unlikely( size == 0 ) ) { free( oaddr ); return mallocNoStats( size ); } // special cases
+          if ( unlikely( oaddr == 0p ) ) return mallocNoStats( size );
+          if ( unlikely( size == 0 ) ) { free( oaddr ); return 0p; } // special cases
+          if ( unlikely( oaddr == 0p ) ) {
+                        #ifdef __STATISTICS__
+                        __atomic_add_fetch( &realloc_storage, size, __ATOMIC_SEQ_CST );
+                        #endif // __STATISTICS__
+                        return mallocNoStats( size );
+                } // if
                 HeapManager.Storage.Header * header;
                 HeapManager.FreeHeader * freeElem;
                 size_t bsize, oalign = 0;
+                size_t bsize, oalign;
                 headers( "realloc", oaddr, header, freeElem, bsize, oalign );
                 size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
+          if ( size <= odsize && odsize <= size * 2 ) { // allow up to 50% wasted storage in smaller size
+                        // Do not know size of original allocation => cannot do 0 fill for any additional space because do not know
+                        // where to start filling, i.e., do not overwrite existing values in space.
+                size_t osize = header->kind.real.size;                  // old allocation size
+                bool ozfill = (header->kind.real.blockSize & 2) != 0; // old allocation zero filled
+          if ( unlikely( size <= odsize ) && size > odsize / 2 ) { // allow up to 50% wasted storage
+                        header->kind.real.size = size;                          // reset allocation size
+                        if ( unlikely( ozfill ) && size > osize ) {     // previous request zero fill and larger ?
+                                memset( (char *)oaddr + osize, (int)'\0', size - osize ); // initialize added storage
+                        } // if
                         return oaddr;
                 } // if
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &realloc_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
                 // change size and copy old content to new storage
                 void * naddr;
+                if ( unlikely( oalign != 0 ) ) {                                // previous request memalign?
+                        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
+                                naddr = cmemalignNoStats( oalign, 1, size ); // create new aligned area
+                        } else {
+                                naddr = memalignNoStats( oalign, size ); // create new aligned area
+                if ( likely( oalign <= libAlign() ) ) {                 // previous request not aligned ?
+                        naddr = mallocNoStats( size );                          // create new area
+                } else {
+                        naddr = memalignNoStats( oalign, size );        // create new aligned area
+                } // if
+                headers( "realloc", naddr, header, freeElem, bsize, oalign );
+                memcpy( naddr, oaddr, MIN( osize, size ) );             // copy bytes
+                free( oaddr );
+                if ( unlikely( ozfill ) ) {                                             // previous request zero fill ?
+                        header->kind.real.blockSize |= 2;                       // mark new request as zero filled
+                        if ( size > osize ) {                                           // previous request larger ?
+                                memset( (char *)naddr + osize, (int)'\0', size - osize ); // initialize added storage
                         } // if
+                } else {
+                        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
+                                naddr = callocNoStats( 1, size );               // create new area
+                        } else {
+                                naddr = mallocNoStats( size );                  // create new area
+                        } // if
+                } // if
+          if ( unlikely( naddr == 0p ) ) return 0p;
+                headers( "realloc", naddr, header, freeElem, bsize, oalign );
+                size_t ndsize = dataStorage( bsize, naddr, header ); // data storage avilable in bucket
+                // To preserve prior fill, the entire bucket must be copied versus the size.
+                memcpy( naddr, oaddr, MIN( odsize, ndsize ) );  // copy bytes
+                free( oaddr );
+                } // if
                 return naddr;
         } // realloc
         // Same as malloc() except the memory address is a multiple of alignment, which must be a power of two. (obsolete)
         void * memalign( size_t alignment, size_t size ) {
 …
         // Same as aalloc() with memory alignment.
         void * amemalign( size_t alignment, size_t dim, size_t elemSize ) {
+                size_t size = dim * elemSize;
                 #ifdef __STATISTICS__
                 __atomic_add_fetch( &cmemalign_calls, 1, __ATOMIC_SEQ_CST );
                 __atomic_add_fetch( &cmemalign_storage, dim * elemSize, __ATOMIC_SEQ_CST );
                 #endif // __STATISTICS__
                 return memalignNoStats( alignment, dim * elemSize );
+                __atomic_add_fetch( &cmemalign_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                return memalignNoStats( alignment, size );
         } // amemalign
 …
         // free(3).
         int posix_memalign( void ** memptr, size_t alignment, size_t size ) {
           if ( alignment < sizeof(void *) || ! libPow2( alignment ) ) return EINVAL; // check alignment
+          if ( alignment < libAlign() || ! is_pow2( alignment ) ) return EINVAL; // check alignment
                 * memptr = memalign( alignment, size );
-          if ( unlikely( * memptr == 0p ) ) return ENOMEM;
                 return 0;
         } // posix_memalign
 …
         // Same as valloc but rounds size to multiple of page size.
         void * pvalloc( size_t size ) {
                 return memalign( pageSize, libCeiling( size, pageSize ) );
+                return memalign( pageSize, ceiling2( size, pageSize ) );
         } // pvalloc
 …
         } // malloc_alignment
         // Set the alignment for an the allocation and return previous alignment or 0 if no alignment.
         size_t $malloc_alignment_set( void * addr, size_t alignment ) {
 …
         // Returns original total allocation size (not bucket size) => array size is dimension * sizeif(T).
         size_t malloc_size( void * addr ) {
           if ( unlikely( addr == 0p ) ) return false;           // null allocation is not zero fill
+          if ( unlikely( addr == 0p ) ) return 0;                       // null allocation has zero size
                 HeapManager.Storage.Header * header = headerAddr( addr );
                 if ( (header->kind.fake.alignment & 1) == 1 ) { // fake header ?
                         header = realHeader( header );                          // backup from fake to real header
                 } // if
                 return header->size;
+                return header->kind.real.size;
         } // malloc_size
         // Set allocation size and return previous size.
         size_t $malloc_size_set( void * addr, size_t size ) {
           if ( unlikely( addr == 0p ) ) return false;           // null allocation is not zero fill
+          if ( unlikely( addr == 0p ) ) return 0;                       // null allocation has 0 size
                 HeapManager.Storage.Header * header = headerAddr( addr );
                 if ( (header->kind.fake.alignment & 1) == 1 ) { // fake header ?
                         header = realHeader( header );                          // backup from fake to real header
                 } // if
                 size_t ret = header->size;
                 header->size = size;
+                size_t ret = header->kind.real.size;
+                header->kind.real.size = size;
                 return ret;
         } // $malloc_size_set
 …
                 headers( "malloc_usable_size", addr, header, freeElem, bsize, alignment );
                 return dataStorage( bsize, addr, header );      // data storage in bucket
+                return dataStorage( bsize, addr, header );              // data storage in bucket
         } // malloc_usable_size
 …
         } // malloc_stats
         // Changes the file descripter where malloc_stats() writes statistics.
         int malloc_stats_fd( int fd __attribute__(( unused )) ) {
                 #ifdef __STATISTICS__
                 int temp = statfd;
                 statfd = fd;
+                int temp = stat_fd;
+                stat_fd = fd;
                 return temp;
                 #else
 …
         } // mallopt
         // Attempt to release free memory at the top of the heap (by calling sbrk with a suitable argument).
         int malloc_trim( size_t ) {
 …
         // malloc).
         int malloc_info( int options, FILE * stream ) {
+                if ( options != 0 ) { errno = EINVAL; return -1; }
+          if ( options != 0 ) { errno = EINVAL; return -1; }
+                #ifdef __STATISTICS__
                 return printStatsXML( stream );
+                #else
+                return 0;                                                                               // unsupported
+                #endif // __STATISTICS__
         } // malloc_info
 …
         // Restores the state of all malloc internal bookkeeping variables to the values recorded in the opaque data
         // structure pointed to by state.
         int malloc_set_state( void * ptr ) {
+        int malloc_set_state( void * ) {
                 return 0;                                                                               // unsupported
         } // malloc_set_state
 …
         #ifdef __STATISTICS__
         __atomic_add_fetch( &resize_calls, 1, __ATOMIC_SEQ_CST );
-        __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
         #endif // __STATISTICS__
         // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
+  if ( unlikely( size == 0 ) ) { free( oaddr ); return memalignNoStats( nalign, size ); } // special cases
+  if ( unlikely( oaddr == 0p ) ) return memalignNoStats( nalign, size );
+        if ( unlikely( nalign == 0 ) ) nalign = libAlign();     // reset alignment to minimum
+  if ( unlikely( size == 0 ) ) { free( oaddr ); return 0p; } // special cases
+  if ( unlikely( oaddr == 0p ) ) {
+                #ifdef __STATISTICS__
+                __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
+                #endif // __STATISTICS__
+                return memalignNoStats( nalign, size );
+        } // if
+        if ( unlikely( nalign < libAlign() ) ) nalign = libAlign(); // reset alignment to minimum
         #ifdef __CFA_DEBUG__
         else
 …
         HeapManager.Storage.Header * header;
         HeapManager.FreeHeader * freeElem;
         size_t bsize, oalign = 0;
+        size_t bsize, oalign;
         headers( "resize", oaddr, header, freeElem, bsize, oalign );
         size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
         if ( oalign <= nalign && (uintptr_t)oaddr % nalign == 0 ) { // <= alignment and new alignment happens to match
                 if ( oalign >= libAlign() ) {                                   // fake header ?
+                if ( oalign > libAlign() ) {                                    // fake header ?
                         headerAddr( oaddr )->kind.fake.alignment = nalign | 1; // update alignment (could be the same)
                 } // if
                 if ( size <= odsize && odsize <= size * 2 ) {   // allow 50% wasted storage for smaller size
                         header->kind.real.blockSize &= -2;                      // turn off 0 fill
+                        header->kind.real.size = size;                          // reset allocation size
                         return oaddr;
                 } // if
         } // if
+        // change size
+        void * naddr;
+        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
+                naddr = cmemalignNoStats( nalign, 1, size );    // create new aligned area
+        } else {
+                naddr = memalignNoStats( nalign, size );                // create new aligned area
+        } // if
+        #ifdef __STATISTICS__
+        __atomic_add_fetch( &resize_storage, size, __ATOMIC_SEQ_CST );
+        #endif // __STATISTICS__
+        // change size, DO NOT preserve STICKY PROPERTIES.
         free( oaddr );
         return naddr;
+        return memalignNoStats( nalign, size );                         // create new aligned area
 } // resize
 void * realloc( void * oaddr, size_t nalign, size_t size ) {
         if ( unlikely( nalign == 0 ) ) nalign = libAlign();     // reset alignment to minimum
+        if ( unlikely( nalign < libAlign() ) ) nalign = libAlign(); // reset alignment to minimum
         #ifdef __CFA_DEBUG__
         else
 …
         HeapManager.Storage.Header * header;
         HeapManager.FreeHeader * freeElem;
         size_t bsize, oalign = 0;
+        size_t bsize, oalign;
         headers( "realloc", oaddr, header, freeElem, bsize, oalign );
-        size_t odsize = dataStorage( bsize, oaddr, header ); // data storage available in bucket
         if ( oalign <= nalign && (uintptr_t)oaddr % nalign == 0 ) { // <= alignment and new alignment happens to match
                 if ( oalign >= libAlign() ) {                                   // fake header ?
+                if ( oalign > libAlign() ) {                                    // fake header ?
                         headerAddr( oaddr )->kind.fake.alignment = nalign | 1; // update alignment (could be the same)
                 } // if
 …
         // If size is equal to 0, either NULL or a pointer suitable to be passed to free() is returned.
   if ( unlikely( size == 0 ) ) { free( oaddr ); return memalignNoStats( nalign, size ); } // special cases
+  if ( unlikely( size == 0 ) ) { free( oaddr ); return 0p; } // special cases
   if ( unlikely( oaddr == 0p ) ) return memalignNoStats( nalign, size );
+        void * naddr;
+        if ( unlikely( header->kind.real.blockSize & 2 ) ) { // previous request zero fill
+                naddr = cmemalignNoStats( nalign, 1, size );    // create new aligned area
+        } else {
+                naddr = memalignNoStats( nalign, size );                // create new aligned area
+        } // if
+        size_t osize = header->kind.real.size;                          // old allocation size
+        bool ozfill = (header->kind.real.blockSize & 2) != 0; // old allocation zero filled
+        void * naddr = memalignNoStats( nalign, size );         // create new aligned area
         headers( "realloc", naddr, header, freeElem, bsize, oalign );
+        size_t ndsize = dataStorage( bsize, naddr, header ); // data storage available in bucket
+        // To preserve prior fill, the entire bucket must be copied versus the size.
+        memcpy( naddr, oaddr, MIN( odsize, ndsize ) );          // copy bytes
+        memcpy( naddr, oaddr, MIN( osize, size ) );                     // copy bytes
         free( oaddr );
+        if ( unlikely( ozfill ) ) {                                                     // previous request zero fill ?
+                header->kind.real.blockSize |= 2;                               // mark new request as zero filled
+                if ( size > osize ) {                                                   // previous request larger ?
+                        memset( (char *)naddr + osize, (int)'\0', size - osize ); // initialize added storage
+                } // if
+        } // if
         return naddr;
 } // realloc

libcfa/src/iostream.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat May  2 18:30:25 2020
 // Update Count     : 1017
+// Last Modified On : Mon Aug 10 09:32:14 2020
+// Update Count     : 1126
 //
 #include "iostream.hfa"
-extern "C" {
 #include <stdio.h>
 #include <stdbool.h>                                                                    // true/false
 #include <stdint.h>                                                                             // UINT64_MAX
+#include <float.h>                                                                              // DBL_DIG, LDBL_DIG
+#include <math.h>                                                                               // isfinite
+#include <complex.h>                                                                    // creal, cimag
 //#include <string.h>                                                                   // strlen, strcmp
+extern "C" {
 extern size_t strlen (const char *__s) __attribute__ ((__nothrow__ , __leaf__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));
 extern int strcmp (const char *__s1, const char *__s2) __attribute__ ((__nothrow__ , __leaf__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));
 extern char *strcpy (char *__restrict __dest, const char *__restrict __src) __attribute__ ((__nothrow__ , __leaf__)) __attribute__ ((__nonnull__ (1, 2)));
 extern void *memcpy (void *__restrict __dest, const void *__restrict __src, size_t __n) __attribute__ ((__nothrow__ , __leaf__)) __attribute__ ((__nonnull__ (1, 2)));
-#include <float.h>                                                                              // DBL_DIG, LDBL_DIG
-#include <math.h>                                                                               // isfinite
-#include <complex.h>                                                                    // creal, cimag
 } // extern "C"
 #include <bitmanip.hfa>                                                                 // fms
 //*********************************** ostream ***********************************
+#include <bitmanip.hfa>                                                                 // high1
+// *********************************** ostream ***********************************
 …
         #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 ) {
+        static inline void base10_128( ostype & os, unsigned int128 val ) {
+#if defined(__GNUC__) && __GNUC_PREREQ(7,0)                             // gcc version >= 7
+                if ( val > P10_UINT64 ) {
+#else
+                if ( (uint64_t)(val >> 64) != 0 || (uint64_t)val > P10_UINT64 ) { // patch gcc 5 & 6 -O3 bug
+#endif // __GNUC_PREREQ(7,0)
                         base10_128( os, val / P10_UINT64 );                     // recursive
                         fmt( os, "%.19lu", (uint64_t)(val % P10_UINT64) );
 …
         } // base10_128
         static void base10_128( ostype & os, int128 val ) {
+        static inline void base10_128( ostype & os, int128 val ) {
                 if ( val < 0 ) {
                         fmt( os, "-" );                                                         // leading negative sign
 …
 } // distribution
 //*********************************** manipulators ***********************************
 //*********************************** integral ***********************************
+// *********************************** manipulators ***********************************
+// *********************************** integral ***********************************
 static const char * shortbin[] = { "0", "1", "10", "11", "100", "101", "110", "111", "1000", "1001", "1010", "1011", "1100", "1101", "1110", "1111" };
 …
 // Default prefix for non-decimal prints is 0b, 0, 0x.
 #define IntegralFMTImpl( T, CODE, IFMTNP, IFMTP ) \
+#define IntegralFMTImpl( T, IFMTNP, IFMTP ) \
 forall( dtype ostype | ostream( ostype ) ) { \
         ostype & ?|?( ostype & os, _Ostream_Manip(T) f ) { \
 …
                         if ( f.flags.left && spaces > 0 ) fmt( os, "%*s", spaces, " " ); \
                         return os; \
                 } /* if  */ \
+                } /* if */ \
+\
                 char fmtstr[sizeof(IFMTP)];                                             /* sizeof includes '\0' */ \
 …
                 if ( ! f.flags.nobsdp ) { fmtstr[star] = '#'; star -= 1; } \
                 if ( f.flags.left ) { fmtstr[star] = '-'; star -= 1; } \
                 if ( f.flags.sign && f.base == CODE ) { fmtstr[star] = '+'; star -= 1; } \
+                if ( f.flags.sign ) { fmtstr[star] = '+'; star -= 1; } \
                 if ( f.flags.pad0 && ! f.flags.pc ) { fmtstr[star] = '0'; star -= 1; } \
                 fmtstr[star] = '%'; \
 …
                 if ( ! f.flags.pc ) {                                                   /* no precision */ \
                         fmtstr[sizeof(IFMTNP)-2] = f.base;                      /* sizeof includes '\0' */ \
                         /* printf( "%s %c %c\n", &fmtstr[star], f.base, CODE ); */ \
+                        /* printf( "%s %c\n", &fmtstr[star], f.base ); */ \
                         fmt( os, &fmtstr[star], f.wd, f.val ); \
                 } else {                                                                                /* precision */ \
                         fmtstr[sizeof(IFMTP)-2] = f.base;                       /* sizeof includes '\0' */ \
                         /* printf( "%s %c %c\n", &fmtstr[star], f.base, CODE ); */ \
+                        /* printf( "%s %c\n", &fmtstr[star], f.base ); */ \
                         fmt( os, &fmtstr[star], f.wd, f.pc, f.val ); \
                 } /* if */ \
 …
 } // distribution
 IntegralFMTImpl( signed char, 'd', "%    *hh ", "%    *.*hh " )
 IntegralFMTImpl( unsigned char, 'u', "%    *hh ", "%    *.*hh " )
 IntegralFMTImpl( signed short int, 'd', "%    *h ", "%    *.*h " )
 IntegralFMTImpl( unsigned short int, 'u', "%    *h ", "%    *.*h " )
 IntegralFMTImpl( signed int, 'd', "%    * ", "%    *.* " )
 IntegralFMTImpl( unsigned int, 'u', "%    * ", "%    *.* " )
 IntegralFMTImpl( signed long int, 'd', "%    *l ", "%    *.*l " )
 IntegralFMTImpl( unsigned long int, 'u', "%    *l ", "%    *.*l " )
 IntegralFMTImpl( signed long long int, 'd', "%    *ll ", "%    *.*ll " )
 IntegralFMTImpl( unsigned long long int, 'u', "%    *ll ", "%    *.*ll " )
+IntegralFMTImpl( signed char, "%    *hh ", "%    *.*hh " )
+IntegralFMTImpl( unsigned char, "%    *hh ", "%    *.*hh " )
+IntegralFMTImpl( signed short int, "%    *h ", "%    *.*h " )
+IntegralFMTImpl( unsigned short int, "%    *h ", "%    *.*h " )
+IntegralFMTImpl( signed int, "%    * ", "%    *.* " )
+IntegralFMTImpl( unsigned int, "%    * ", "%    *.* " )
+IntegralFMTImpl( signed long int, "%    *l ", "%    *.*l " )
+IntegralFMTImpl( unsigned long int, "%    *l ", "%    *.*l " )
+IntegralFMTImpl( signed long long int, "%    *ll ", "%    *.*ll " )
+IntegralFMTImpl( unsigned long long int, "%    *ll ", "%    *.*ll " )
+#if 0
 #if defined( __SIZEOF_INT128__ )
 // Default prefix for non-decimal prints is 0b, 0, 0x.
 …
                                                 sepOff( os ); \
                                                 fmt2.flags.left = true; \
                                                 int msigd = ceiling( high1( fmt.val ), 3 ); \
+                                                int msigd = ceiling_div( high1( fmt.val ), 3 ); \
                                                 fmt2.wd = f.wd - (fmt.pc > msigd ? fmt.pc : msigd); \
                                                 if ( ! fmt.flags.nobsdp ) fmt2.wd -= 1; /* compensate for 0 base specifier */ \
 …
 IntegralFMTImpl128( unsigned int128, unsigned, 'u', "%    *ll ", "%    *.*ll " )
 #endif // __SIZEOF_INT128__
+//*********************************** floating point ***********************************
+#endif // 0
+#if 1
+#if defined( __SIZEOF_INT128__ )
+// Default prefix for non-decimal prints is 0b, 0, 0x.
+forall( dtype ostype | ostream( ostype ) )
+static inline void base_128( ostype & os, unsigned int128 val, unsigned int128 power, _Ostream_Manip(uint64_t) & f, unsigned int maxdig, unsigned int bits, unsigned int cnt = 0 ) {
+        int wd = 1;                                                                                     // f.wd is never 0 because 0 implies left-pad
+        if ( val > power ) {                                                            // subdivide value into printable 64-bit values
+                base_128( os, val / power, power, f, maxdig, bits, cnt + 1 ); // recursive
+                f.val = val % power;
+                if ( cnt == 1 && f.flags.left ) { wd = f.wd; f.wd = maxdig; } // copy f.wd and reset for printing middle chunk
+                // printf( "R val:%#lx(%lu) wd:%u pc:%u base:%c neg:%d pc:%d left:%d nobsdp:%d sign:%d pad0:%d\n",
+                //              f.val, f.val, f.wd, f.pc, f.base, f.flags.neg, f.flags.pc, f.flags.left, f.flags.nobsdp, f.flags.sign, f.flags.pad0 );
+                (ostype &)(os | f);
+                if ( cnt == 1 ) {
+                        if ( f.flags.left ) { wd -= maxdig; f.wd = wd < 0 ? 1 : wd; } // update and restore f.wd for printing end chunk
+                        sepOff( os );                                                           // no seperator between chunks
+                } // if
+        } else {                                                                                        // print start chunk
+                f.val = val;
+                // f.pc is unsigned => use wd
+                if ( f.flags.pc && f.pc > maxdig * cnt ) { wd = f.pc - maxdig * cnt; f.pc = wd < 0 ? 0 : wd; }
+                else { f.flags.pc = false; f.pc = 0; }
+                if ( ! f.flags.left ) {                                                 // right justify
+                        wd = f.wd - maxdig * cnt;
+                        f.wd = wd < 0 ? 1 : wd;
+                        wd = maxdig;
+                } else {                                                                                // left justify
+                        if ( cnt != 0 ) {                                                       // value >= 2^64 ?
+                                unsigned int dig, bs = 0;
+                                // compute size of prefix digits and base
+                                if ( f.base == 'd' || f.base == 'u' ) { // no base prefix
+                                        dig = ceil( log10( f.val ) );           // use floating-point
+                                        if ( f.base == 'd' && (f.flags.neg || f.flags.sign) ) bs = 1; // sign ?
+                                } else {
+                                        dig = ceiling_div( high1( f.val ), bits );
+                                        if ( ! f.flags.nobsdp ) {                       // base prefix ?
+                                                if ( f.base == 'o' ) {
+                                                        // 0 prefix for octal is not added for precision with leading zero
+                                                        if ( f.pc <= dig ) bs = 1;      // 1 character prefix
+                                                } else bs = 2;                                  // 2 character prefix
+                                        } // if
+                                } // if
+                                wd = f.wd - (f.pc > dig ? f.pc : dig) - bs; // precision > leading digits ?
+                                if ( wd < 0 ) wd = 1;
+                                f.wd = 1;
+                        } // if
+                        // all manipulators handled implicitly for value < 2^64
+                } // if
+                // prior checks ensure wd not negative
+                if ( f.flags.neg ) f.val = -f.val;
+                // printf( "L val:%#lx(%lu) wd:%u pc:%u base:%c neg:%d pc:%d left:%d nobsdp:%d sign:%d pad0:%d\n",
+                //              f.val, f.val, f.wd, f.pc, f.base, f.flags.neg, f.flags.pc, f.flags.left, f.flags.nobsdp, f.flags.sign, f.flags.pad0 );
+                (ostype &)(os | f);
+                // remaining middle and end chunks are padded with 0s on the left
+                if ( ! f.flags.left ) { f.flags.pad0 = true; f.flags.pc = false; } // left pad with 0s
+                else { f.pc = maxdig; f.flags.pc = true; }              // left pad with precision
+                if ( cnt != 0 ) sepOff( os );                                   // no seperator between chunks
+                f.wd = wd;                                                                              // reset f.wd for next chunk
+                f.flags.sign = false;                                                   // no leading +/- sign
+                f.flags.nobsdp = true;                                                  // no leading base prefix
+        } // if
+} // base_128
+#define IntegralFMTImpl128( T ) \
+forall( dtype ostype | ostream( ostype ) ) { \
+        ostype & ?|?( ostype & os, _Ostream_Manip(T) f ) { \
+                _Ostream_Manip(uint64_t) fmt; \
+                fmt.[wd, pc, base, all] = f.[wd, pc, base, all]; \
+                if ( f.base == 'b' | f.base == 'B' ) { \
+                        base_128( os, f.val, (unsigned int128)1 << 64, fmt, 64, 1 ); \
+                } else if ( f.base == 'o' ) { \
+                        base_128( os, f.val, (unsigned int128)1 << 63, fmt, 21, 3 ); \
+                } else if ( f.base == 'd' || f.base == 'u' ) { \
+                        if ( f.base == 'd' && f.val < 0 ) { f.val = -f.val; fmt.flags.neg = true; } \
+                        base_128( os, f.val, (unsigned int128)10_000_000_000_000_000_000UL, fmt, 19, 0 ); \
+                } else { \
+                        base_128( os, f.val, (unsigned int128)1 << 64, fmt, 16, 4 ); \
+                } /* if */ \
+                return os; \
+        } /* ?|? */ \
+        void ?|?( ostype & os, _Ostream_Manip(T) f ) { (ostype &)(os | f); ends( os ); } \
+} // distribution
+IntegralFMTImpl128( int128 )
+IntegralFMTImpl128( unsigned int128 )
+#endif // __SIZEOF_INT128__
+#endif // 0
+// *********************************** floating point ***********************************
 #define PrintWithDP2( os, format, val, ... ) \
 …
 FloatingPointFMTImpl( long double, "%    *L ", "%    *.*L " )
 //*********************************** character ***********************************
+// *********************************** character ***********************************
 forall( dtype ostype | ostream( ostype ) ) {
 …
 } // distribution
 //*********************************** C string ***********************************
+// *********************************** C string ***********************************
 forall( dtype ostype | ostream( ostype ) ) {
 …
 //*********************************** istream ***********************************
+// *********************************** istream ***********************************
 …
         } // ?|?
+#if defined( __SIZEOF_INT128__ )
+        istype & ?|?( istype & is, int128 & i128 ) {
+                return (istype &)(is | (unsigned int128 &)i128);
+        } // ?|?
+        istype & ?|?( istype & is, unsigned int128 & ui128 ) {
+                char s[40];
+                bool sign = false;
+                if ( fmt( is, " %[-]", s ) == 1 ) sign = true;  // skip whitespace, negative sign ?
+                // If the input is too large, the value returned is undefined. If there is no input, no value is returned
+                if ( fmt( is, "%39[0-9]%*[0-9]", s ) == 1 ) {   // take first 39 characters, ignore remaining
+                        ui128 = 0;
+                        for ( unsigned int i = 0; s[i] != '\0'; i += 1 ) {
+                                ui128 = ui128 * 10 + s[i] - '0';
+                        } // for
+                        if ( sign ) ui128 = -ui128;
+                } else if ( sign ) ungetc( is, '-' );                   // return minus when no digits
+                return is;
+        } // ?|?
+#endif // __SIZEOF_INT128__
         istype & ?|?( istype & is, float & f ) {
 …
 } // distribution
 //*********************************** manipulators ***********************************
+// *********************************** manipulators ***********************************
 forall( dtype istype | istream( istype ) )

libcfa/src/iostream.hfa

-              r07d867b
+              r22f94a4
 // Created On       : Wed May 27 17:56:53 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Feb 20 15:30:56 2020
 // Update Count     : 337
+// Last Modified On : Thu Jul 16 07:43:32 2020
+// Update Count     : 348
 //
 …
 //*********************************** ostream ***********************************
+// *********************************** ostream ***********************************
 …
 } // distribution
 //*********************************** manipulators ***********************************
+// *********************************** manipulators ***********************************
 forall( otype T )
 …
                 unsigned char all;
                 struct {
+                        unsigned char neg:1;                                            // val is negative
                         unsigned char pc:1;                                                     // precision specified
                         unsigned char left:1;                                           // left justify
 …
 }; // _Ostream_Manip
 //*********************************** integral ***********************************
+// *********************************** integral ***********************************
 // See 6.7.9. 19) The initialization shall occur in initializer list order, each initializer provided for a particular
 …
 IntegralFMTDecl( int128, 'd' )
 IntegralFMTDecl( unsigned int128, 'u' )
 #endif
 //*********************************** floating point ***********************************
+#endif // __SIZEOF_INT128__
+// *********************************** floating point ***********************************
 // Default suffix for values with no fraction is "."
 …
 FloatingPointFMTDecl( long double )
 //*********************************** character ***********************************
+// *********************************** character ***********************************
 static inline {
 …
 } // ?|?
 //*********************************** C string ***********************************
+// *********************************** C string ***********************************
 static inline {
 …
 //*********************************** istream ***********************************
+// *********************************** istream ***********************************
 …
         istype & ?|?( istype &, unsigned int & );
         istype & ?|?( istype &, long int & );
+        istype & ?|?( istype &, unsigned long int & );
         istype & ?|?( istype &, long long int & );
-        istype & ?|?( istype &, unsigned long int & );
         istype & ?|?( istype &, unsigned long long int & );
+#if defined( __SIZEOF_INT128__ )
+        istype & ?|?( istype &, int128 & );
+        istype & ?|?( istype &, unsigned int128 & );
+#endif // __SIZEOF_INT128__
         istype & ?|?( istype &, float & );
 …
 } // distribution
 //*********************************** manipulators ***********************************
+// *********************************** manipulators ***********************************
 struct _Istream_Cstr {
 …
         _Istream_Cstr excl( const char scanset[], char * s ) { return (_Istream_Cstr){ s, scanset, -1, { .flags.inex : true } }; }
         _Istream_Cstr & excl( const char scanset[], _Istream_Cstr & fmt ) { fmt.scanset = scanset; fmt.flags.inex = true; return fmt; }
         _Istream_Cstr ignore( const char s[] ) { return (_Istream_Cstr)@{ s, 0p, -1, { .flags.ignore : true } }; }
+        _Istream_Cstr ignore( char s[] ) { return (_Istream_Cstr)@{ s, 0p, -1, { .flags.ignore : true } }; }
         _Istream_Cstr & ignore( _Istream_Cstr & fmt ) { fmt.flags.ignore = true; return fmt; }
         _Istream_Cstr wdi( unsigned int w, char s[] ) { return (_Istream_Cstr)@{ s, 0p, w, { .all : 0 } }; }
 …
 static inline {
         _Istream_Char ignore( const char c ) { return (_Istream_Char)@{ true }; }
+        _Istream_Char ignore( const char ) { return (_Istream_Char)@{ true }; }
         _Istream_Char & ignore( _Istream_Char & fmt ) { fmt.ignore = true; return fmt; }
 } // distribution
 forall( dtype istype | istream( istype ) ) istype & ?|?( istype & is, _Istream_Char f );
 forall( otype T )
+forall( dtype T | sized( T ) )
 struct _Istream_Manip {
         T & val;                                                                                        // polymorphic base-type
 …
 //*********************************** time ***********************************
+// *********************************** time ***********************************

libcfa/src/stdhdr/assert.h

-              r07d867b
+              r22f94a4
         #define verify(x) assert(x)
         #define verifyf(x, ...) assertf(x, __VA_ARGS__)
+        #define verifyfail(...)
         #define __CFA_WITH_VERIFY__
 #else
         #define verify(x)
         #define verifyf(x, ...)
+        #define verifyfail(...)
 #endif

libcfa/src/stdhdr/malloc.h

-              r07d867b
+              r22f94a4
 // Created On       : Thu Jul 20 15:58:16 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Apr 16 22:44:06 2020
 // Update Count     : 13
+// Last Modified On : Wed May 27 14:13:14 2020
+// Update Count     : 18
 //
-size_t default_mmap_start();                                                    // CFA extras
-size_t default_heap_expansion();
-bool traceHeap();
-bool traceHeapOn();
-bool traceHeapOff();
-bool traceHeapTerm();
-bool traceHeapTermOn();
-bool traceHeapTermOff();
-bool checkFree();
-bool checkFreeOn();
-bool checkFreeOff();
-extern "C" {
-void * aalloc( size_t noOfElems, size_t elemSize );
-void * amemalign( size_t alignment, size_t noOfElems, size_t elemSize );
-void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize );
-size_t malloc_alignment( void * );
-bool malloc_zero_fill( void * );
-size_t malloc_size( void * );
-int malloc_stats_fd( int fd );
-} // extern "C"
 extern "C" {
 #include_next <malloc.h>                                                                // has internal check for multiple expansion
 } // extern "C"
+#include <heap.hfa>
 // Local Variables: //

libcfa/src/stdlib.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Thu Jan 28 17:10:29 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Apr 16 22:43:33 2020
 // Update Count     : 498
+// Last Modified On : Sun Jul 19 15:05:28 2020
+// Update Count     : 501
 //
 …
 //---------------------------------------
-forall( dtype T | sized(T) ) {
-        T * alloc_set( T ptr[], size_t dim, char fill ) {       // realloc array with fill
-                size_t olen = malloc_usable_size( ptr );                // current allocation
-                void * nptr = (void *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
-                size_t nlen = malloc_usable_size( nptr );               // new allocation
-                if ( nlen > olen ) {                                                    // larger ?
-                        memset( (char *)nptr + olen, (int)fill, nlen - olen ); // initialize added storage
-                } // if
-                return (T *)nptr;
-        } // alloc_set
-        T * alloc_set( T ptr[], size_t dim, T fill ) {          // realloc array with fill
-                size_t olen = malloc_usable_size( ptr );                // current allocation
-                void * nptr = (void *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
-                size_t nlen = malloc_usable_size( nptr );               // new allocation
-                if ( nlen > olen ) {                                                    // larger ?
-                        for ( i; malloc_size( ptr ) / sizeof(T) ~ dim ) {
-                                memcpy( &ptr[i], &fill, sizeof(T) );    // initialize with fill value
-                        } // for
-                } // if
-                return (T *)nptr;
-        } // alloc_align_set
-        T * alloc_align_set( T ptr[], size_t align, char fill ) { // aligned realloc with fill
-                size_t olen = malloc_usable_size( ptr );                // current allocation
-                void * nptr = (void *)realloc( (void *)ptr, align, sizeof(T) ); // CFA realloc
-                // char * nptr = alloc_align( ptr, align );
-                size_t nlen = malloc_usable_size( nptr );               // new allocation
-                if ( nlen > olen ) {                                                    // larger ?
-                        memset( (char *)nptr + olen, (int)fill, nlen - olen ); // initialize added storage
-                } // if
-                return (T *)nptr;
-        } // alloc_align_set
-        T * alloc_align_set( T ptr[], size_t align, size_t dim, T fill ) { // aligned realloc with fill
-                size_t olen = malloc_usable_size( ptr );                // current allocation
-                void * nptr = (void *)realloc( (void *)ptr, align, sizeof(T) ); // CFA realloc
-                // char * nptr = alloc_align( ptr, align );
-                size_t nlen = malloc_usable_size( nptr );               // new allocation
-                if ( nlen > olen ) {                                                    // larger ?
-                        for ( i; dim ) { memcpy( &ptr[i], &fill, sizeof(T) ); } // initialize with fill value
-                } // if
-                return (T *)nptr;
-        } // alloc_align_set
-} // distribution
 // allocation/deallocation and constructor/destructor, non-array types
 forall( dtype T | sized(T), ttype Params | { void ?{}( T &, Params ); } )
 …
 } // new
 forall( dtype T | sized(T) | { void ^?{}( T & ); } )
+forall( dtype T | { void ^?{}( T & ); } )
 void delete( T * ptr ) {
         if ( ptr ) {                                                                            // ignore null
 …
 } // delete
 forall( dtype T, ttype Params | sized(T) | { void ^?{}( T & ); void delete( Params ); } )
+forall( dtype T, ttype Params | { void ^?{}( T & ); void delete( Params ); } )
 void delete( T * ptr, Params rest ) {
+        if ( ptr ) {                                                                            // ignore null
+                ^(*ptr){};                                                                              // run destructor
+                free( ptr );
+        } // if
+        delete( ptr );
         delete( rest );
 } // delete
 …
 extern "C" {                                                                                    // override C version
         void srandom( unsigned int seed ) { srand48( (long int)seed ); }
         long int random( void ) { return mrand48(); }
+        long int random( void ) { return mrand48(); }           // GENERATES POSITIVE AND NEGATIVE VALUES
 } // extern "C"

libcfa/src/stdlib.hfa

-              r07d867b
+              r22f94a4
 // Created On       : Thu Jan 28 17:12:35 2016
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Apr 16 22:44:05 2020
 // Update Count     : 432
+// Last Modified On : Thu Jul 30 16:14:58 2020
+// Update Count     : 490
 //
 …
 #include <stdlib.h>                                                                             // *alloc, strto*, ato*
+#include <heap.hfa>
 // Reduce includes by explicitly defining these routines.
 extern "C" {
+        void * memalign( size_t align, size_t size );           // malloc.h
+        size_t malloc_usable_size( void * ptr );                        // malloc.h
+        size_t malloc_size( void * addr );                                      // CFA heap
+        void * cmemalign( size_t alignment, size_t noOfElems, size_t elemSize ); // CFA heap
+        void * memalign( size_t alignment, size_t size );       // malloc.h
+        void * pvalloc( size_t size );                                          // malloc.h
         void * memset( void * dest, int fill, size_t size ); // string.h
         void * memcpy( void * dest, const void * src, size_t size ); // string.h
-        void * resize( void * oaddr, size_t size );                     // CFA heap
 } // extern "C"
-void * resize( void * oaddr, size_t nalign, size_t size ); // CFA heap
-void * realloc( void * oaddr, size_t nalign, size_t size ); // CFA heap
 //---------------------------------------
 …
 //---------------------------------------
+// Macro because of returns
+#define $VAR_ALLOC( allocation, alignment ) \
+        if ( _Alignof(T) <= libAlign() ) return (T *)(void *)allocation( (size_t)sizeof(T) ); /* C allocation */ \
+        else return (T *)alignment( _Alignof(T), sizeof(T) )
+#define $ARRAY_ALLOC( allocation, alignment, dim ) \
+        if ( _Alignof(T) <= libAlign() ) return (T *)(void *)allocation( dim, (size_t)sizeof(T) ); /* C allocation */ \
+        else return (T *)alignment( _Alignof(T), dim, sizeof(T) )
+#define $RE_SPECIALS( ptr, size, allocation, alignment ) \
+        if ( unlikely( size == 0 ) || unlikely( ptr == 0p ) ) { \
+                if ( unlikely( size == 0 ) ) free( ptr ); \
+                $VAR_ALLOC( malloc, memalign ); \
+        } /* if */
 static inline forall( dtype T | sized(T) ) {
         // Cforall safe equivalents, i.e., implicit size specification
         T * malloc( void ) {
+                if ( _Alignof(T) <= libAlign() ) return (T *)(void *)malloc( (size_t)sizeof(T) ); // C malloc
+                else return (T *)memalign( _Alignof(T), sizeof(T) );
+                $VAR_ALLOC( malloc, memalign );
         } // malloc
+        T * aalloc( size_t dim ) {
+                $ARRAY_ALLOC( aalloc, amemalign, dim );
+        } // aalloc
         T * calloc( size_t dim ) {
+                if ( _Alignof(T) <= libAlign() )return (T *)(void *)calloc( dim, sizeof(T) ); // C calloc
+                else return (T *)cmemalign( _Alignof(T), dim, sizeof(T) );
+                $ARRAY_ALLOC( calloc, cmemalign, dim );
         } // calloc
+        T * resize( T * ptr, size_t size ) {                            // CFA resize, eliminate return-type cast
+                $RE_SPECIALS( ptr, size, malloc, memalign );
+                if ( _Alignof(T) <= libAlign() ) return (T *)(void *)resize( (void *)ptr, size ); // CFA resize
+                else return (T *)(void *)resize( (void *)ptr, _Alignof(T), size ); // CFA resize
+        } // resize
         T * realloc( T * ptr, size_t size ) {                           // CFA realloc, eliminate return-type cast
+                return (T *)(void *)realloc( (void *)ptr, size ); // C realloc
+                $RE_SPECIALS( ptr, size, malloc, memalign );
+                if ( _Alignof(T) <= libAlign() ) return (T *)(void *)realloc( (void *)ptr, size ); // C realloc
+                else return (T *)(void *)realloc( (void *)ptr, _Alignof(T), size ); // CFA realloc
         } // realloc
 …
         } // memalign
+        T * amemalign( size_t align, size_t dim ) {
+                return (T *)amemalign( align, dim, sizeof(T) ); // CFA amemalign
+        } // amemalign
         T * cmemalign( size_t align, size_t dim  ) {
                 return (T *)cmemalign( align, dim, sizeof(T) ); // CFA cmemalign
 …
                 return posix_memalign( (void **)ptr, align, sizeof(T) ); // C posix_memalign
         } // posix_memalign
+        T * valloc( void ) {
+                return (T *)valloc( sizeof(T) );                                // C valloc
+        } // valloc
+        T * pvalloc( void ) {
+                return (T *)pvalloc( sizeof(T) );                               // C pvalloc
+        } // pvalloc
 } // distribution
 …
         T * alloc( size_t dim ) {
+                if ( _Alignof(T) <= libAlign() ) return (T *)(void *)malloc( dim * (size_t)sizeof(T) );
+                else return (T *)memalign( _Alignof(T), dim * sizeof(T) );
+                return aalloc( dim );
         } // alloc
         forall( dtype S | sized(S) )
         T * alloc( S ptr[], size_t dim = 1 ) {                          // singleton/array resize
+                size_t len = malloc_usable_size( ptr );                 // current bucket size
+                if ( sizeof(T) * dim > len ) {                                  // not enough space ?
+                        T * temp = alloc( dim );                                        // new storage
+                        free( ptr );                                                            // free old storage
+                        return temp;
+                return resize( (T *)ptr, dim * sizeof(T) );             // CFA resize
+        } // alloc
+        T * alloc( T ptr[], size_t dim = 1, bool copy = true ) {
+                if ( copy ) {
+                        return realloc( ptr, dim * sizeof(T) );         // CFA realloc
                 } else {
+                        return (T *)ptr;
+                } // if
+        } // alloc
+        T * alloc( T ptr[], size_t dim, bool copy = true ) {
+                if ( copy ) {                                                                   // realloc
+                        return (T *)(void *)realloc( (void *)ptr, dim * sizeof(T) ); // C realloc
+                } else {
+                        struct __Unknown {};
+                        return alloc( (__Unknown *)ptr, dim );          // reuse, cheat making T/S different types
+                        return resize( ptr, dim * sizeof(T) );          // CFA resize
                 } // if
         } // alloc
 …
                 return (T *)memcpy( (T *)alloc( dim ), fill, dim * sizeof(T) ); // initialize with fill value
         } // alloc
+} // distribution
+forall( dtype T | sized(T) ) {
+        T * alloc_set( T ptr[], size_t dim, char fill );        // realloc array with fill
+        T * alloc_set( T ptr[], size_t dim, T fill );           // realloc array with fill
+        T * alloc_set( T ptr[], size_t dim, char fill ) {       // realloc array with fill
+                size_t osize = malloc_size( ptr );                              // current allocation
+                size_t nsize = dim * sizeof(T);                                 // new allocation
+                T * nptr = realloc( ptr, nsize );                               // CFA realloc
+                if ( nsize > osize ) {                                                  // larger ?
+                        memset( (char *)nptr + osize, (int)fill, nsize - osize ); // initialize added storage
+                } // if
+                return nptr;
+        } // alloc_set
+        T * alloc_set( T ptr[], size_t dim, T & fill ) {        // realloc array with fill
+                size_t odim = malloc_size( ptr ) / sizeof(T);   // current dimension
+                size_t nsize = dim * sizeof(T);                                 // new allocation
+                size_t ndim = nsize / sizeof(T);                                // new dimension
+                T * nptr = realloc( ptr, nsize );                               // CFA realloc
+                if ( ndim > odim ) {                                                    // larger ?
+                        for ( i; odim ~ ndim ) {
+                                memcpy( &nptr[i], &fill, sizeof(T) );   // initialize with fill value
+                        } // for
+                } // if
+                return nptr;
+        } // alloc_align_set
 } // distribution
 …
         } // alloc_align
         T * alloc_align( T ptr[], size_t align ) {                      // aligned realloc array
                 return (T *)(void *)realloc( (void *)ptr, align, sizeof(T) ); // CFA realloc
+        T * alloc_align( T * ptr, size_t align ) {                      // aligned realloc array
+                return (T *)(void *)realloc( (void *)ptr, align, sizeof(T) ); // CFA C realloc
         } // alloc_align
 …
                 return (T *)memcpy( (T *)alloc_align( align, dim ), fill, dim * sizeof(T) );
         } // alloc_align
+} // distribution
+forall( dtype T | sized(T) ) {
+        T * alloc_align_set( T ptr[], size_t align, char fill ); // aligned realloc with fill
+        T * alloc_align_set( T ptr[], size_t align, T fill ); // aligned realloc with fill
+        T * alloc_align_set( T ptr[], size_t align, size_t dim, char fill ); // aligned realloc array with fill
+        T * alloc_align_set( T ptr[], size_t align, size_t dim, T fill ); // aligned realloc array with fill
+        T * alloc_align_set( T ptr[], size_t align, size_t dim, char fill ) {
+                size_t osize = malloc_size( ptr );                              // current allocation
+                size_t nsize = dim * sizeof(T);                                 // new allocation
+                T * nptr = alloc_align( ptr, align, nsize );
+                if ( nsize > osize ) {                                                  // larger ?
+                        memset( (char *)nptr + osize, (int)fill, nsize - osize ); // initialize added storage
+                } // if
+                return nptr;
+        } // alloc_align_set
+        T * alloc_align_set( T ptr[], size_t align, size_t dim, T & fill ) {
+                size_t odim = malloc_size( ptr ) / sizeof(T);   // current dimension
+                size_t nsize = dim * sizeof(T);                                 // new allocation
+                size_t ndim = nsize / sizeof(T);                                // new dimension
+                T * nptr = alloc_align( ptr, align, nsize );
+                if ( ndim > odim ) {                                                    // larger ?
+                        for ( i; odim ~ ndim ) {
+                                memcpy( &nptr[i], &fill, sizeof(T) );   // initialize with fill value
+                        } // for
+                } // if
+                return nptr;
+        } // alloc_align_set
 } // distribution
 …
 // Cforall allocation/deallocation and constructor/destructor, non-array types
 forall( dtype T | sized(T), ttype Params | { void ?{}( T &, Params ); } ) T * new( Params p );
 forall( dtype T | sized(T) | { void ^?{}( T & ); } ) void delete( T * ptr );
 forall( dtype T, ttype Params | sized(T) | { void ^?{}( T & ); void delete( Params ); } ) void delete( T * ptr, Params rest );
+forall( dtype T | { void ^?{}( T & ); } ) void delete( T * ptr );
+forall( dtype T, ttype Params | { void ^?{}( T & ); void delete( Params ); } ) void delete( T * ptr, Params rest );
 // Cforall allocation/deallocation and constructor/destructor, array types
 …
 extern "C" {                                                                                    // override C version
         void srandom( unsigned int seed );
+        long int random( void );
+        long int random( void );                                                        // GENERATES POSITIVE AND NEGATIVE VALUES
+        // For positive values, use unsigned int, e.g., unsigned int r = random() % 100U;
 } // extern "C"
 …
         long int random( long int u ) { if ( u < 0 ) return random( u, 0 ); else return random( 0, u ); } // [0,u)
         unsigned long int random( void ) { return lrand48(); }
+        unsigned long int random( unsigned long int u ) { return lrand48() % u; } // [0,u)
         unsigned long int random( unsigned long int l, unsigned long int u ) { if ( u < l ) [u, l] = [l, u]; return lrand48() % (u - l) + l; } // [l,u)
-        unsigned long int random( unsigned long int u ) { return lrand48() % u; } // [0,u)
         char random( void ) { return (unsigned long int)random(); }

libcfa/src/time.hfa

-              r07d867b
+              r22f94a4
 // Created On       : Wed Mar 14 23:18:57 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Feb  4 08:24:32 2020
 // Update Count     : 654
+// Last Modified On : Wed Jun 17 16:13:00 2020
+// Update Count     : 663
 //
 …
 #include <time.h>                                                                               // timespec
-extern "C" {
 #include <sys/time.h>                                                                   // timeval
+}
 #include <time_t.hfa>                                                                   // Duration/Time types
 …
         int64_t ?`w( Duration dur ) { return dur.tn / (7LL * 24LL * 60LL * 60LL * TIMEGRAN); }
+        double ?`dns( Duration dur ) { return dur.tn; }
+        double ?`dus( Duration dur ) { return dur.tn / ((double)TIMEGRAN / 1_000_000.); }
+        double ?`dms( Duration dur ) { return dur.tn / ((double)TIMEGRAN / 1_000.); }
+        double ?`ds( Duration dur ) { return dur.tn / (double)TIMEGRAN; }
+        double ?`dm( Duration dur ) { return dur.tn / (60. * TIMEGRAN); }
+        double ?`dh( Duration dur ) { return dur.tn / (60. * 60. * (double)TIMEGRAN); }
+        double ?`dd( Duration dur ) { return dur.tn / (24. * 60. * 60. * (double)TIMEGRAN); }
+        double ?`dw( Duration dur ) { return dur.tn / (7. * 24. * 60. * 60. * (double)TIMEGRAN); }
         Duration max( Duration lhs, Duration rhs ) { return  (lhs.tn < rhs.tn) ? rhs : lhs;}
         Duration min( Duration lhs, Duration rhs ) { return !(rhs.tn < lhs.tn) ? lhs : rhs;}

longrun_tests/Makefile.am

r07d867b	r22f94a4
18	18	ACLOCAL_AMFLAGS = -I automake
19	19
20		include $(top_srcdir)/~~src~~/cfa.make
	20	include $(top_srcdir)/tools/build/cfa.make
21	21
22	22	repeats=10

src/Common/PassVisitor.proto.h

r07d867b	r22f94a4
38	38	};
39	39
40		std::stack< cleanup_t, std::vector<~~cleanup_t~~> > cleanups;
	40	std::stack< cleanup_t, std::vector< cleanup_t > > cleanups;
41	41	};
42	42

src/Common/ScopedMap.h

r07d867b	r22f94a4
93	93
94	94	reference operator* () { return *it; }
95		pointer operator-> () { return it.operator->(); }
	95	pointer operator-> () const { return it.operator->(); }
96	96
97	97	iterator& operator++ () {

src/Concurrency/Keywords.cc

-              r07d867b
+              r22f94a4
                                                 new CastExpr(
                                                         new VariableExpr( func->get_functionType()->get_parameters().front() ),
+                                                        func->get_functionType()->get_parameters().front()->get_type()->stripReferences()->clone()
+                                                        func->get_functionType()->get_parameters().front()->get_type()->stripReferences()->clone(),
+                                                        false
+                                                )
+                                        )
 …
                         new SingleInit( new UntypedExpr(
                                 new NameExpr( "get_monitor" ),
                                 {  new CastExpr( new VariableExpr( args.front() ), arg_type ) }
+                                {  new CastExpr( new VariableExpr( args.front() ), arg_type, false ) }
                         ))
                 );
 …
+                                        {
                                                 new SingleInit( new AddressExpr( new VariableExpr( monitors ) ) ),
                                                 new SingleInit( new CastExpr( new VariableExpr( func ), generic_func->clone() ) )
+                                                new SingleInit( new CastExpr( new VariableExpr( func ), generic_func->clone(), false ) )
                                         },
                                         noDesignators,
 …
                                         return new SingleInit( new UntypedExpr(
                                                 new NameExpr( "get_monitor" ),
                                                 {  new CastExpr( new VariableExpr( var ), type ) }
+                                                {  new CastExpr( new VariableExpr( var ), type, false ) }
                                         ) );
                                 })
 …
                                                 new SingleInit( new VariableExpr( monitors ) ),
                                                 new SingleInit( new ConstantExpr( Constant::from_ulong( args.size() ) ) ),
                                                 new SingleInit( new CastExpr( new VariableExpr( func ), generic_func->clone() ) )
+                                                new SingleInit( new CastExpr( new VariableExpr( func ), generic_func->clone(), false ) )
                                         },
                                         noDesignators,

src/Concurrency/Waitfor.cc

-              r07d867b
+              r22f94a4
                                                                 decl_monitor
+                                                        )
+                                                )
+                                                ),
+                                                false
                                         );
 …
                         new CompoundStmt({
                                 makeAccStatement( acceptables, index, "is_dtor", detectIsDtor( clause.target.function )                                    , indexer ),
                                 makeAccStatement( acceptables, index, "func"   , new CastExpr( clause.target.function, fptr_t )                            , indexer ),
+                                makeAccStatement( acceptables, index, "func"   , new CastExpr( clause.target.function, fptr_t, false )                     , indexer ),
                                 makeAccStatement( acceptables, index, "data"   , new VariableExpr( monitors )                                              , indexer ),
                                 makeAccStatement( acceptables, index, "size"   , new ConstantExpr( Constant::from_ulong( clause.target.arguments.size() ) ), indexer ),
 …
                                                                 decl_mask
+                                                        )
+                                                )
+                                                ),
+                                                false
                                         ),
                                         timeout

src/ControlStruct/ExceptTranslate.cc

-              r07d867b
+              r22f94a4
 // Created On       : Wed Jun 14 16:49:00 2017
 // Last Modified By : Andrew Beach
 // Last Modified On : Fri Mar 27 11:58:00 2020
 // Update Count     : 13
+// Last Modified On : Wed Jun 24 11:18:00 2020
+// Update Count     : 17
 //
 …
+        }
+        class ExceptionMutatorCore : public WithGuards {
+                enum Context { NoHandler, TerHandler, ResHandler };
+                // Also need to handle goto, break & continue.
+                // They need to be cut off in a ResHandler, until we enter another
+                // loop, switch or the goto stays within the function.
+                Context cur_context;
+                // The current (innermost) termination handler exception declaration.
+                ObjectDecl * handler_except_decl;
+        class ThrowMutatorCore : public WithGuards {
+                ObjectDecl * terminate_handler_except;
+                enum Context { NoHandler, TerHandler, ResHandler } cur_context;
+                // The helper functions for code/syntree generation.
+                Statement * create_either_throw(
+                        ThrowStmt * throwStmt, const char * throwFunc );
+                Statement * create_terminate_rethrow( ThrowStmt * throwStmt );
+        public:
+                ThrowMutatorCore() :
+                        terminate_handler_except( nullptr ),
+                        cur_context( NoHandler )
+                {}
+                void premutate( CatchStmt *catchStmt );
+                Statement * postmutate( ThrowStmt *throwStmt );
+        };
+        // ThrowStmt Mutation Helpers
+        Statement * ThrowMutatorCore::create_either_throw(
+                        ThrowStmt * throwStmt, const char * throwFunc ) {
+                // `throwFunc`( `throwStmt->get_name()` );
+                UntypedExpr * call = new UntypedExpr( new NameExpr( throwFunc ) );
+                call->get_args().push_back( throwStmt->get_expr() );
+                throwStmt->set_expr( nullptr );
+                delete throwStmt;
+                return new ExprStmt( call );
+        }
+        Statement * ThrowMutatorCore::create_terminate_rethrow(
+                        ThrowStmt *throwStmt ) {
+                // { `terminate_handler_except` = 0p; __rethrow_terminate(); }
+                assert( nullptr == throwStmt->get_expr() );
+                assert( terminate_handler_except );
+                CompoundStmt * result = new CompoundStmt();
+                result->labels =  throwStmt->labels;
+                result->push_back( new ExprStmt( UntypedExpr::createAssign(
+                        nameOf( terminate_handler_except ),
+                        new ConstantExpr( Constant::null(
+                                terminate_handler_except->get_type()->clone()
+                                ) )
+                        ) ) );
+                result->push_back( new ExprStmt(
+                        new UntypedExpr( new NameExpr( "__cfaehm_rethrow_terminate" ) )
+                        ) );
+                delete throwStmt;
+                return result;
+        }
+        // Visiting/Mutating Functions
+        void ThrowMutatorCore::premutate( CatchStmt *catchStmt ) {
+                // Validate the statement's form.
+                ObjectDecl * decl = dynamic_cast<ObjectDecl *>( catchStmt->get_decl() );
+                // Also checking the type would be nice.
+                if ( !decl || !dynamic_cast<PointerType *>( decl->type ) ) {
+                        std::string kind = (CatchStmt::Terminate == catchStmt->kind) ? "catch" : "catchResume";
+                        SemanticError( catchStmt->location, kind + " must have pointer to an exception type" );
+                }
+                // Track the handler context.
+                GuardValue( cur_context );
+                if ( CatchStmt::Terminate == catchStmt->get_kind() ) {
+                        cur_context = TerHandler;
+                        GuardValue( terminate_handler_except );
+                        terminate_handler_except = decl;
+                } else {
+                        cur_context = ResHandler;
+                }
+        }
+        Statement * ThrowMutatorCore::postmutate( ThrowStmt *throwStmt ) {
+                // Ignoring throwStmt->get_target() for now.
+                if ( ThrowStmt::Terminate == throwStmt->get_kind() ) {
+                        if ( throwStmt->get_expr() ) {
+                                return create_either_throw( throwStmt, "$throw" );
+                        } else if ( TerHandler == cur_context ) {
+                                return create_terminate_rethrow( throwStmt );
+                        } else {
+                                abort("Invalid throw in %s at %i\n",
+                                        throwStmt->location.filename.c_str(),
+                                        throwStmt->location.first_line);
+                        }
+                } else {
+                        if ( throwStmt->get_expr() ) {
+                                return create_either_throw( throwStmt, "$throwResume" );
+                        } else if ( ResHandler == cur_context ) {
+                                // This has to be handled later.
+                                return throwStmt;
+                        } else {
+                                abort("Invalid throwResume in %s at %i\n",
+                                        throwStmt->location.filename.c_str(),
+                                        throwStmt->location.first_line);
+                        }
+                }
+        }
+        class TryMutatorCore {
                 // The built in types used in translation.
                 StructDecl * except_decl;
 …
                 // The many helper functions for code/syntree generation.
-                Statement * create_given_throw(
-                        const char * throwFunc, ThrowStmt * throwStmt );
-                Statement * create_terminate_throw( ThrowStmt * throwStmt );
-                Statement * create_terminate_rethrow( ThrowStmt * throwStmt );
-                Statement * create_resume_throw( ThrowStmt * throwStmt );
-                Statement * create_resume_rethrow( ThrowStmt * throwStmt );
                 CompoundStmt * take_try_block( TryStmt * tryStmt );
                 FunctionDecl * create_try_wrapper( CompoundStmt * body );
 …
                 FunctionDecl * create_finally_wrapper( TryStmt * tryStmt );
                 ObjectDecl * create_finally_hook( FunctionDecl * finally_wrapper );
+                Statement * create_resume_rethrow( ThrowStmt * throwStmt );
                 // Types used in translation, make sure to use clone.
 …
         public:
+                ExceptionMutatorCore() :
+                        cur_context( NoHandler ),
+                        handler_except_decl( nullptr ),
+                TryMutatorCore() :
                         except_decl( nullptr ), node_decl( nullptr ), hook_decl( nullptr ),
                         try_func_t( noQualifiers, false ),
 …
                 {}
-                void premutate( CatchStmt *catchStmt );
                 void premutate( StructDecl *structDecl );
+                Statement * postmutate( TryStmt *tryStmt );
                 Statement * postmutate( ThrowStmt *throwStmt );
-                Statement * postmutate( TryStmt *tryStmt );
         };
         void ExceptionMutatorCore::init_func_types() {
+        void TryMutatorCore::init_func_types() {
                 assert( except_decl );
 …
+        }
-        // ThrowStmt Mutation Helpers
-        Statement * ExceptionMutatorCore::create_given_throw(
-                        const char * throwFunc, ThrowStmt * throwStmt ) {
-                // `throwFunc`( `throwStmt->get_name` );
-                UntypedExpr * call = new UntypedExpr( new NameExpr( throwFunc ) );
-                call->get_args().push_back( throwStmt->get_expr() );
-                throwStmt->set_expr( nullptr );
-                delete throwStmt;
-                return new ExprStmt( call );
+        }
-        Statement * ExceptionMutatorCore::create_terminate_throw(
-                        ThrowStmt *throwStmt ) {
-                // __throw_terminate( `throwStmt->get_name()` ); }
-                return create_given_throw( "__cfaehm_throw_terminate", throwStmt );
+        }
-        Statement * ExceptionMutatorCore::create_terminate_rethrow(
-                        ThrowStmt *throwStmt ) {
-                // { `handler_except_decl` = NULL; __rethrow_terminate(); }
-                assert( nullptr == throwStmt->get_expr() );
-                assert( handler_except_decl );
-                CompoundStmt * result = new CompoundStmt();
-                result->labels =  throwStmt->labels;
-                result->push_back( new ExprStmt( UntypedExpr::createAssign(
-                        nameOf( handler_except_decl ),
-                        new ConstantExpr( Constant::null(
-                                new PointerType(
-                                        noQualifiers,
-                                        handler_except_decl->get_type()->clone()
+                                        )
-                                ) )
-                        ) ) );
-                result->push_back( new ExprStmt(
-                        new UntypedExpr( new NameExpr( "__cfaehm_rethrow_terminate" ) )
-                        ) );
-                delete throwStmt;
-                return result;
+        }
-        Statement * ExceptionMutatorCore::create_resume_throw(
-                        ThrowStmt *throwStmt ) {
-                // __throw_resume( `throwStmt->get_name` );
-                return create_given_throw( "__cfaehm_throw_resume", throwStmt );
+        }
-        Statement * ExceptionMutatorCore::create_resume_rethrow(
-                        ThrowStmt *throwStmt ) {
-                // return false;
-                Statement * result = new ReturnStmt(
-                        new ConstantExpr( Constant::from_bool( false ) )
-                        );
-                result->labels = throwStmt->labels;
-                delete throwStmt;
-                return result;
+        }
         // TryStmt Mutation Helpers
         CompoundStmt * ExceptionMutatorCore::take_try_block( TryStmt *tryStmt ) {
+        CompoundStmt * TryMutatorCore::take_try_block( TryStmt *tryStmt ) {
                 CompoundStmt * block = tryStmt->get_block();
                 tryStmt->set_block( nullptr );
 …
+        }
         FunctionDecl * ExceptionMutatorCore::create_try_wrapper(
+        FunctionDecl * TryMutatorCore::create_try_wrapper(
                         CompoundStmt *body ) {
 …
+        }
         FunctionDecl * ExceptionMutatorCore::create_terminate_catch(
+        FunctionDecl * TryMutatorCore::create_terminate_catch(
                         CatchList &handlers ) {
                 std::list<CaseStmt *> handler_wrappers;
 …
         // Create a single check from a moddified handler.
         // except_obj is referenced, modded_handler will be freed.
         CompoundStmt * ExceptionMutatorCore::create_single_matcher(
+        CompoundStmt * TryMutatorCore::create_single_matcher(
                         DeclarationWithType * except_obj, CatchStmt * modded_handler ) {
                 // {
 …
+        }
         FunctionDecl * ExceptionMutatorCore::create_terminate_match(
+        FunctionDecl * TryMutatorCore::create_terminate_match(
                         CatchList &handlers ) {
                 // int match(exception * except) {
 …
+        }
         CompoundStmt * ExceptionMutatorCore::create_terminate_caller(
+        CompoundStmt * TryMutatorCore::create_terminate_caller(
                         FunctionDecl * try_wrapper,
                         FunctionDecl * terminate_catch,
 …
+        }
         FunctionDecl * ExceptionMutatorCore::create_resume_handler(
+        FunctionDecl * TryMutatorCore::create_resume_handler(
                         CatchList &handlers ) {
                 // bool handle(exception * except) {
 …
+        }
         CompoundStmt * ExceptionMutatorCore::create_resume_wrapper(
+        CompoundStmt * TryMutatorCore::create_resume_wrapper(
                         Statement * wraps,
                         FunctionDecl * resume_handler ) {
 …
+        }
         FunctionDecl * ExceptionMutatorCore::create_finally_wrapper(
+        FunctionDecl * TryMutatorCore::create_finally_wrapper(
                         TryStmt * tryStmt ) {
                 // void finally() { <finally code> }
+                // void finally() { `finally->block` }
                 FinallyStmt * finally = tryStmt->get_finally();
                 CompoundStmt * body = finally->get_block();
 …
+        }
         ObjectDecl * ExceptionMutatorCore::create_finally_hook(
+        ObjectDecl * TryMutatorCore::create_finally_hook(
                         FunctionDecl * finally_wrapper ) {
                 // struct __cfaehm_cleanup_hook __finally_hook
                 //      __attribute__((cleanup( finally_wrapper )));
+                //      __attribute__((cleanup( `finally_wrapper` )));
                 // Make Cleanup Attribute.
 …
+        }
+        Statement * TryMutatorCore::create_resume_rethrow( ThrowStmt *throwStmt ) {
+                // return false;
+                Statement * result = new ReturnStmt(
+                        new ConstantExpr( Constant::from_bool( false ) )
+                        );
+                result->labels = throwStmt->labels;
+                delete throwStmt;
+                return result;
+        }
         // Visiting/Mutating Functions
+        void ExceptionMutatorCore::premutate( CatchStmt *catchStmt ) {
+                // Validate the Statement's form.
+                ObjectDecl * decl =
+                        dynamic_cast<ObjectDecl *>( catchStmt->get_decl() );
+                if ( decl && true /* check decl->get_type() */ ) {
+                        // Pass.
+                } else if ( CatchStmt::Terminate == catchStmt->get_kind() ) {
+                        SemanticError(catchStmt->location, "catch must have exception type");
+                } else {
+                        SemanticError(catchStmt->location, "catchResume must have exception type");
+                }
+                // Track the handler context.
+                GuardValue( cur_context );
+                if ( CatchStmt::Terminate == catchStmt->get_kind() ) {
+                        cur_context = TerHandler;
+                        GuardValue( handler_except_decl );
+                        handler_except_decl = decl;
+                } else {
+                        cur_context = ResHandler;
+                }
+        }
+        void ExceptionMutatorCore::premutate( StructDecl *structDecl ) {
+        void TryMutatorCore::premutate( StructDecl *structDecl ) {
                 if ( !structDecl->has_body() ) {
                         // Skip children?
 …
                         hook_decl = structDecl;
+                }
+                // Later we might get the exception type as well.
+        }
+        Statement * ExceptionMutatorCore::postmutate( ThrowStmt *throwStmt ) {
+                assert( except_decl );
+                // Ignoring throwStmt->get_target() for now.
+                if ( ThrowStmt::Terminate == throwStmt->get_kind() ) {
+                        if ( throwStmt->get_expr() ) {
+                                return create_terminate_throw( throwStmt );
+                        } else if ( TerHandler == cur_context ) {
+                                return create_terminate_rethrow( throwStmt );
+                        } else {
+                                abort("Invalid throw in %s at %i\n",
+                                        throwStmt->location.filename.c_str(),
+                                        throwStmt->location.first_line);
+                        }
+                } else {
+                        if ( throwStmt->get_expr() ) {
+                                return create_resume_throw( throwStmt );
+                        } else if ( ResHandler == cur_context ) {
+                                return create_resume_rethrow( throwStmt );
+                        } else {
+                                abort("Invalid throwResume in %s at %i\n",
+                                        throwStmt->location.filename.c_str(),
+                                        throwStmt->location.first_line);
+                        }
+                }
+        }
+        Statement * ExceptionMutatorCore::postmutate( TryStmt *tryStmt ) {
+        }
+        Statement * TryMutatorCore::postmutate( TryStmt *tryStmt ) {
                 assert( except_decl );
                 assert( node_decl );
 …
+        }
+        void translateEHM( std::list< Declaration *> & translationUnit ) {
+                PassVisitor<ExceptionMutatorCore> translator;
+        Statement * TryMutatorCore::postmutate( ThrowStmt *throwStmt ) {
+                // Only valid `throwResume;` statements should remain. (2/3 checks)
+                assert( ThrowStmt::Resume == throwStmt->kind && ! throwStmt->expr );
+                return create_resume_rethrow( throwStmt );
+        }
+        void translateThrows( std::list< Declaration *> & translationUnit ) {
+                PassVisitor<ThrowMutatorCore> translator;
                 mutateAll( translationUnit, translator );
+        }
+        void translateTries( std::list< Declaration *> & translationUnit ) {
+                PassVisitor<TryMutatorCore> translator;
+                mutateAll( translationUnit, translator );
+        }
+}

src/ControlStruct/ExceptTranslate.h

-              r07d867b
+              r22f94a4
 // Author           : Andrew Beach
 // Created On       : Tus Jun 06 10:13:00 2017
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Jul 22 09:19:23 2017
 // Update Count     : 4
+// Last Modified By : Andrew Beach
+// Last Modified On : Tus May 19 11:47:00 2020
+// Update Count     : 5
 //
 …
 namespace ControlStruct {
+        void translateEHM( std::list< Declaration *> & translationUnit );
+        // Converts exception handling structures into their underlying C code.  Translation does use the exception
+        // handling header, make sure it is visible wherever translation occurs.
+        void translateThrows( std::list< Declaration *> & translationUnit );
+        /* Replaces all throw & throwResume statements with function calls.
+         * These still need to be resolved, so call this before the reslover.
+         */
+        void translateTries( std::list< Declaration *> & translationUnit );
+        /* Replaces all try blocks (and their many clauses) with function definitions and calls.
+         * This uses the exception built-ins to produce typed output and should take place after
+         * the resolver. It also produces virtual casts and should happen before they are expanded.
+         */
+}

src/GenPoly/InstantiateGeneric.cc

-              r07d867b
+              r22f94a4
 // Author           : Aaron B. Moss
 // Created On       : Thu Aug 04 18:33:00 2016
 // Last Modified By : Aaron B. Moss
 // Last Modified On : Thu Aug 04 18:33:00 2016
 // Update Count     : 1
+// Last Modified By : Andrew Beach
+// Last Modified On : Wed Jul 16 10:17:00 2020
+// Update Count     : 2
 //
 #include "InstantiateGeneric.h"
 …
+        }
+        template< typename AggrInst >
+        static AggrInst * asForward( AggrInst * decl ) {
+                if ( !decl->body ) {
+                        return nullptr;
+                }
+                decl = decl->clone();
+                decl->body = false;
+                deleteAll( decl->members );
+                decl->members.clear();
+                return decl;
+        }
         void GenericInstantiator::stripDtypeParams( AggregateDecl *base, std::list< TypeDecl* >& baseParams, const std::list< TypeExpr* >& typeSubs ) {
                 substituteMembers( base->get_members(), baseParams, typeSubs );
 …
                                 concDecl->set_body( inst->get_baseStruct()->has_body() );
                                 substituteMembers( inst->get_baseStruct()->get_members(), *inst->get_baseParameters(), typeSubs, concDecl->get_members() );
+                                insert( inst, typeSubs, concDecl ); // must insert before recursion
+                                // Forward declare before recursion. (TODO: Only when needed, #199.)
+                                insert( inst, typeSubs, concDecl );
+                                if ( StructDecl *forwardDecl = asForward( concDecl ) ) {
+                                        declsToAddBefore.push_back( forwardDecl );
+                                }
                                 concDecl->acceptMutator( *visitor ); // recursively instantiate members
                                 declsToAddBefore.push_back( concDecl ); // must occur before declaration is added so that member instantiations appear first
 …
                                 concDecl->set_body( inst->get_baseUnion()->has_body() );
                                 substituteMembers( inst->get_baseUnion()->get_members(), *inst->get_baseParameters(), typeSubs, concDecl->get_members() );
+                                insert( inst, typeSubs, concDecl ); // must insert before recursion
+                                // Forward declare before recursion. (TODO: Only when needed, #199.)
+                                insert( inst, typeSubs, concDecl );
+                                if ( UnionDecl *forwardDecl = asForward( concDecl ) ) {
+                                        declsToAddBefore.push_back( forwardDecl );
+                                }
                                 concDecl->acceptMutator( *visitor ); // recursively instantiate members
                                 declsToAddBefore.push_back( concDecl ); // must occur before declaration is added so that member instantiations appear first

src/GenPoly/Specialize.cc

-              r07d867b
+              r22f94a4
 // Author           : Richard C. Bilson
 // Created On       : Mon May 18 07:44:20 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Dec 13 23:40:49 2019
 // Update Count     : 32
+// Last Modified By : Andrew Beach
+// Last Modified On : Thr Jul  2 17:42:00 2020
+// Update Count     : 33
 //
 …
 namespace GenPoly {
+        struct Specialize final : public WithConstTypeSubstitution, public WithStmtsToAdd, public WithVisitorRef<Specialize> {
+        struct Specialize final : public WithConstTypeSubstitution,
+                        public WithDeclsToAdd, public WithVisitorRef<Specialize> {
                 Expression * postmutate( ApplicationExpr *applicationExpr );
                 Expression * postmutate( CastExpr *castExpr );
 …
                 } // if
+                // handle any specializations that may still be present
+                std::string oldParamPrefix = paramPrefix;
+                paramPrefix += "p";
+                // save stmtsToAddBefore in oldStmts
+                std::list< Statement* > oldStmts;
+                oldStmts.splice( oldStmts.end(), stmtsToAddBefore );
+                appExpr->acceptMutator( *visitor );
+                paramPrefix = oldParamPrefix;
+                // write any statements added for recursive specializations into the thunk body
+                thunkFunc->statements->kids.splice( thunkFunc->statements->kids.end(), stmtsToAddBefore );
+                // restore oldStmts into stmtsToAddBefore
+                stmtsToAddBefore.splice( stmtsToAddBefore.end(), oldStmts );
+                // Handle any specializations that may still be present.
+                {
+                        std::string oldParamPrefix = paramPrefix;
+                        paramPrefix += "p";
+                        std::list< Declaration * > oldDecls;
+                        oldDecls.splice( oldDecls.end(), declsToAddBefore );
+                        appExpr->acceptMutator( *visitor );
+                        // Write recursive specializations into the thunk body.
+                        for ( Declaration * decl : declsToAddBefore ) {
+                                thunkFunc->statements->kids.push_back( new DeclStmt( decl ) );
+                        }
+                        declsToAddBefore = std::move( oldDecls );
+                        paramPrefix = oldParamPrefix;
+                }
                 // add return (or valueless expression) to the thunk
 …
                 thunkFunc->statements->kids.push_back( appStmt );
                 // add thunk definition to queue of statements to add
                 stmtsToAddBefore.push_back( new DeclStmt( thunkFunc ) );
+                // Add the thunk definition (converted to DeclStmt if appproprate).
+                declsToAddBefore.push_back( thunkFunc );
                 // return address of thunk function as replacement expression
                 return new AddressExpr( new VariableExpr( thunkFunc ) );

src/Parser/DeclarationNode.cc

-              r07d867b
+              r22f94a4
 // Created On       : Sat May 16 12:34:05 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Dec 16 15:32:22 2019
 // Update Count     : 1133
+// Last Modified On : Tue Jun  9 20:26:55 2020
+// Update Count     : 1134
 //
 …
         // SUE's cannot have function specifiers, either
         //
         //    inlne _Noreturn struct S { ... };         // disallowed
         //    inlne _Noreturn enum   E { ... };         // disallowed
+        //    inline _Noreturn struct S { ... };                // disallowed
+        //    inline _Noreturn enum   E { ... };                // disallowed
         if ( funcSpecs.any() ) {
                 SemanticError( this, "invalid function specifier for " );

src/Parser/ExpressionNode.cc

-              r07d867b
+              r22f94a4
 // Created On       : Sat May 16 13:17:07 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Dec 18 21:14:58 2019
 // Update Count     : 981
+// Last Modified On : Wed Jul 15 08:24:08 2020
+// Update Count     : 1046
 //
 …
         } // if
         // remove "lL" for these cases because it may not imply long
         str.erase( posn );                                                                      // remove length
+        str.erase( posn );                                                                      // remove length suffix and "uU"
 } // lnthSuffix
 …
 } // valueToType
+static void scanbin( string & str, unsigned long long int & v ) {
+        v = 0;
+        size_t last = str.length() - 1;                                         // last subscript of constant
+        for ( unsigned int i = 2;; ) {                                          // ignore prefix
+                if ( str[i] == '1' ) v |= 1;
+                i += 1;
+          if ( i == last - 1 || (str[i] != '0' && str[i] != '1') ) break;
+                v <<= 1;
+        } // for
+} // scanbin
 Expression * build_constantInteger( string & str ) {
         static const BasicType::Kind kind[2][6] = {
                 // short (h) must be before char (hh) because shorter type has the longer suffix
                 { BasicType::ShortSignedInt, BasicType::SignedChar, BasicType::SignedInt, BasicType::LongSignedInt, BasicType::LongLongSignedInt, BasicType::SignedInt128, },
                 { BasicType::ShortUnsignedInt, BasicType::UnsignedChar, BasicType::UnsignedInt, BasicType::LongUnsignedInt, BasicType::LongLongUnsignedInt, BasicType::UnsignedInt128, },
+                { BasicType::ShortSignedInt, BasicType::SignedChar, BasicType::SignedInt, BasicType::LongSignedInt, BasicType::LongLongSignedInt, /* BasicType::SignedInt128 */ BasicType::LongLongSignedInt, },
+                { BasicType::ShortUnsignedInt, BasicType::UnsignedChar, BasicType::UnsignedInt, BasicType::LongUnsignedInt, BasicType::LongLongUnsignedInt, /* BasicType::UnsignedInt128 */ BasicType::LongLongUnsignedInt, },
         };
 …
         }; // lnthsInt
+        unsigned long long int v;                                                       // converted integral value
+        size_t last = str.length() - 1;                                         // last subscript of constant
+        Expression * ret;
+        //string fred( str );
+        string str2( "0x0" );
+        unsigned long long int v, v2 = 0;                                       // converted integral value
+        Expression * ret, * ret2;
         int type = -1;                                                                          // 0 => short, 1 => char, 2 => int, 3 => long int, 4 => long long int, 5 => int128
 …
         } // if
+        string::size_type posn;
+        // 'u' can appear before or after length suffix
+        if ( str.find_last_of( "uU" ) != string::npos ) Unsigned = true;
+        if ( isdigit( str[str.length() - 1] ) ) {                       // no suffix ?
+                lnthSuffix( str, type, ltype );                                 // could have length suffix
+                if ( type == 5 && Unsigned ) str.erase( str.length() - 1 ); // L128 and terminating "uU" ?
+        } else {
+                // At least one digit in integer constant, so safe to backup while looking for suffix.
+                posn = str.find_last_of( "pP" );                                // pointer value
+                if ( posn != string::npos ) { ltype = 5; str.erase( posn, 1 ); goto FINI; }
+                posn = str.find_last_of( "zZ" );                                // size_t
+                if ( posn != string::npos ) { Unsigned = true; type = 2; ltype = 4; str.erase( posn, 1 ); goto FINI; }
+                posn = str.rfind( "hh" );                                               // char
+                if ( posn != string::npos ) { type = 1; str.erase( posn, 2 ); goto FINI; }
+                posn = str.rfind( "HH" );                                               // char
+                if ( posn != string::npos ) { type = 1; str.erase( posn, 2 ); goto FINI; }
+                posn = str.find_last_of( "hH" );                                // short
+                if ( posn != string::npos ) { type = 0; str.erase( posn, 1 ); goto FINI; }
+                posn = str.find_last_of( "nN" );                                // int (natural number)
+                if ( posn != string::npos ) { type = 2; str.erase( posn, 1 ); goto FINI; }
+                if ( str.rfind( "ll" ) != string::npos || str.rfind( "LL" ) != string::npos ) { type = 4; goto FINI; }
+                lnthSuffix( str, type, ltype );                                 // must be after check for "ll"
+          FINI: ;
+        } // if
         // Cannot be just "0"/"1"; sscanf stops at the suffix, if any; value goes over the wall => always generate
+#if ! defined(__SIZEOF_INT128__)
+        if ( type == 5 ) SemanticError( yylloc, "int128 constant is not supported on this target " + str );
+#endif // ! __SIZEOF_INT128__
         if ( str[0] == '0' ) {                                                          // radix character ?
                 dec = false;
                 if ( checkX( str[1] ) ) {                                               // hex constant ?
+                        sscanf( (char *)str.c_str(), "%llx", &v );
+                        if ( type < 5 ) {                                                       // not L128 ?
+                                sscanf( (char *)str.c_str(), "%llx", &v );
+                        } else {                                                                        // hex int128 constant
+                                unsigned int len = str.length();
+                                if ( len > (2 + 16 + 16) ) SemanticError( yylloc, "128-bit hexadecimal constant to large " + str );
+                          if ( len <= (2 + 16) ) goto FHEX1;            // hex digits < 2^64
+                                str2 = "0x" + str.substr( len - 16 );
+                                sscanf( (char *)str2.c_str(), "%llx", &v2 );
+                                str = str.substr( 0, len - 16 );
+                          FHEX1: ;
+                                sscanf( (char *)str.c_str(), "%llx", &v );
+                        } // if
                         //printf( "%llx %llu\n", v, v );
                 } else if ( checkB( str[1] ) ) {                                // binary constant ?
+                        v = 0;                                                                          // compute value
+                        for ( unsigned int i = 2;; ) {                          // ignore prefix
+                                if ( str[i] == '1' ) v |= 1;
+                                i += 1;
+                          if ( i == last - 1 || (str[i] != '0' && str[i] != '1') ) break;
+                                v <<= 1;
+                        } // for
+                        unsigned int len = str.length();
+                        if ( type == 5 && len > 2 + 64 ) {
+                                if ( len > 2 + 64 + 64 ) SemanticError( yylloc, "128-bit binary constant to large " + str );
+                                str2 = "0b" + str.substr( len - 64 );
+                                str = str.substr( 0, len - 64 );
+                                scanbin( str2, v2 );
+                        } // if
+                        scanbin( str, v );
                         //printf( "%#llx %llu\n", v, v );
                 } else {                                                                                // octal constant
+                        sscanf( (char *)str.c_str(), "%llo", &v );
+                        if ( type < 5 ) {                                                       // not L128 ?
+                                sscanf( (char *)str.c_str(), "%llo", &v );
+#if defined(__SIZEOF_INT128__)
+                        } else {                                                                        // octal int128 constant
+                                unsigned int len = str.length();
+                                if ( len > 1 + 43 || (len == 1 + 43 && str[0] > '3') ) SemanticError( yylloc, "128-bit octal constant to large " + str );
+                                if ( len <= 1 + 21 ) {                                  // value < 21 octal digitis
+                                        sscanf( (char *)str.c_str(), "%llo", &v ); // leave value in octal
+                                } else {
+                                        sscanf( &str[len - 21], "%llo", &v );
+                                        __int128 val = v;                                       // accumulate bits
+                                        str[len - 21] ='\0';                            // shorten string
+                                        sscanf( &str[len == 43 ? 1 : 0], "%llo", &v );
+                                        val |= (__int128)v << 63;                       // store bits
+                                        if ( len == 1 + 43 ) {                          // most significant 2 bits ?
+                                                str[2] = '\0';                                  // shorten string
+                                                sscanf( &str[1], "%llo", &v );  // process most significant 2 bits
+                                                val |= (__int128)v << 126;              // store bits
+                                        } // if
+                                        v = val >> 64; v2 = (uint64_t)val;      // replace octal constant with 2 hex constants
+                                        char buf[32];
+                                        sprintf( buf, "%#llx", v2 );
+                                        str2 = buf;
+                                        sprintf( buf, "%#llx", v );
+                                        str = buf;
+                                } // if
+#endif // __SIZEOF_INT128__
+                        } // if
                         //printf( "%#llo %llu\n", v, v );
                 } // if
         } else {                                                                                        // decimal constant ?
+                sscanf( (char *)str.c_str(), "%llu", &v );
+                if ( type < 5 ) {                                                               // not L128 ?
+                        sscanf( (char *)str.c_str(), "%llu", &v );
+#if defined(__SIZEOF_INT128__)
+                } else {                                                                                // decimal int128 constant
+                        #define P10_UINT64 10'000'000'000'000'000'000ULL // 19 zeroes
+                        unsigned int len = str.length();
+                        if ( str.length() == 39 && str > (Unsigned ? "340282366920938463463374607431768211455" : "170141183460469231731687303715884105727") )
+                                SemanticError( yylloc, "128-bit decimal constant to large " + str );
+                        if ( len <= 19 ) {                                                      // value < 19 decimal digitis
+                                sscanf( (char *)str.c_str(), "%llu", &v ); // leave value in decimal
+                        } else {
+                                sscanf( &str[len - 19], "%llu", &v );
+                                __int128 val = v;                                               // accumulate bits
+                                str[len - 19] ='\0';                                    // shorten string
+                                sscanf( &str[len == 39 ? 1 : 0], "%llu", &v );
+                                val += (__int128)v * (__int128)P10_UINT64; // store bits
+                                if ( len == 39 ) {                                              // most significant 2 bits ?
+                                        str[1] = '\0';                                          // shorten string
+                                        sscanf( &str[0], "%llu", &v );          // process most significant 2 bits
+                                        val += (__int128)v * (__int128)P10_UINT64 * (__int128)P10_UINT64; // store bits
+                                } // if
+                                v = val >> 64; v2 = (uint64_t)val;              // replace decimal constant with 2 hex constants
+                                char buf[32];
+                                sprintf( buf, "%#llx", v2 );
+                                str2 = buf;
+                                sprintf( buf, "%#llx", v );
+                                str = buf;
+                        } // if
+#endif // __SIZEOF_INT128__
+                } // if
                 //printf( "%llu\n", v );
         } // if
+        string::size_type posn;
+        if ( isdigit( str[last] ) ) {                                           // no suffix ?
+                lnthSuffix( str, type, ltype );                                 // could have length suffix
+                if ( type == -1 ) {                                                             // no suffix
+                        valueToType( v, dec, type, Unsigned );
+                } // if
+        } else {
+                // At least one digit in integer constant, so safe to backup while looking for suffix.
+                posn = str.find_last_of( "pP" );
+                if ( posn != string::npos ) { valueToType( v, dec, type, Unsigned ); ltype = 5; str.erase( posn, 1 ); goto FINI; }
+                posn = str.find_last_of( "zZ" );
+                if ( posn != string::npos ) { Unsigned = true; type = 2; ltype = 4; str.erase( posn, 1 ); goto FINI; }
+                // 'u' can appear before or after length suffix
+                if ( str.find_last_of( "uU" ) != string::npos ) Unsigned = true;
+                posn = str.rfind( "hh" );
+                if ( posn != string::npos ) { type = 1; str.erase( posn, 2 ); goto FINI; }
+                posn = str.rfind( "HH" );
+                if ( posn != string::npos ) { type = 1; str.erase( posn, 2 ); goto FINI; }
+                posn = str.find_last_of( "hH" );
+                if ( posn != string::npos ) { type = 0; str.erase( posn, 1 ); goto FINI; }
+                posn = str.find_last_of( "nN" );
+                if ( posn != string::npos ) { type = 2; str.erase( posn, 1 ); goto FINI; }
+                if ( str.rfind( "ll" ) != string::npos || str.rfind( "LL" ) != string::npos ) { type = 4; goto FINI; }
+                lnthSuffix( str, type, ltype );                                 // must be after check for "ll"
+                if ( type == -1 ) {                                                             // only 'u' suffix ?
+                        valueToType( v, dec, type, Unsigned );
+                } // if
+          FINI: ;
+        } // if
+        if ( type == -1 ) {                                                                     // no suffix => determine type from value size
+                valueToType( v, dec, type, Unsigned );
+        } // if
+        /* printf( "%s %llo %s %llo\n", str.c_str(), v, str2.c_str(), v2 ); */
         //if ( !( 0 <= type && type <= 6 ) ) { printf( "%s %lu %d %s\n", fred.c_str(), fred.length(), type, str.c_str() ); }
 …
         } else if ( ltype != -1 ) {                                                     // explicit length ?
                 if ( ltype == 6 ) {                                                             // int128, (int128)constant
+                        ret = new CastExpr( ret, new BasicType( Type::Qualifiers(), kind[Unsigned][type] ), false );
+//                      ret = new CastExpr( ret, new BasicType( Type::Qualifiers(), kind[Unsigned][type] ), false );
+                        ret2 = new ConstantExpr( Constant( new BasicType( noQualifiers, BasicType::LongLongSignedInt ), str2, v2 ) );
+                        ret = build_compoundLiteral( DeclarationNode::newBasicType( DeclarationNode::Int128 )->addType( DeclarationNode::newSignedNess( DeclarationNode::Unsigned ) ),
+                                                                                 new InitializerNode( (InitializerNode *)(new InitializerNode( new ExpressionNode( v2 == 0 ? ret2 : ret ) ))->set_last( new InitializerNode( new ExpressionNode( v2 == 0 ? ret : ret2 ) ) ), true ) );
                 } else {                                                                                // explicit length, (length_type)constant
                         ret = new CastExpr( ret, new TypeInstType( Type::Qualifiers(), lnthsInt[Unsigned][ltype], false ), false );
 …
                 if ( str[1] == '8' ) goto Default;                              // utf-8 characters => array of char
                 // lookup type of associated typedef
                 strtype = new TypeInstType( Type::Qualifiers( Type::Const ), "char16_t", false );
+                strtype = new TypeInstType( Type::Qualifiers( ), "char16_t", false );
                 break;
           case 'U':
                 strtype = new TypeInstType( Type::Qualifiers( Type::Const ), "char32_t", false );
+                strtype = new TypeInstType( Type::Qualifiers( ), "char32_t", false );
                 break;
           case 'L':
                 strtype = new TypeInstType( Type::Qualifiers( Type::Const ), "wchar_t", false );
+                strtype = new TypeInstType( Type::Qualifiers( ), "wchar_t", false );
                 break;
           Default:                                                                                      // char default string type
           default:
                 strtype = new BasicType( Type::Qualifiers( Type::Const ), BasicType::Char );
+                strtype = new BasicType( Type::Qualifiers( ), BasicType::Char );
         } // switch
         ArrayType * at = new ArrayType( noQualifiers, strtype,

src/Parser/ParseNode.h

-              r07d867b
+              r22f94a4
 // Created On       : Sat May 16 13:28:16 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Fri Feb  7 17:56:02 2020
 // Update Count     : 891
+// Last Modified On : Mon Jul  6 09:33:32 2020
+// Update Count     : 892
 //
 …
 class InitializerNode : public ParseNode {
   public:
         InitializerNode( ExpressionNode *, bool aggrp = false,  ExpressionNode * des = nullptr );
+        InitializerNode( ExpressionNode *, bool aggrp = false, ExpressionNode * des = nullptr );
         InitializerNode( InitializerNode *, bool aggrp = false, ExpressionNode * des = nullptr );
         InitializerNode( bool isDelete );

src/Parser/module.mk

r07d867b	r22f94a4
17	17	BUILT_SOURCES = Parser/parser.hh
18	18
19		AM_YFLAGS = -d -t -v
	19	AM_YFLAGS = -d -t -v -Wno-yacc
20	20
21	21	SRC += \

src/Parser/parser.yy

-              r07d867b
+              r22f94a4
 // Created On       : Sat Sep  1 20:22:55 2001
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Mon Apr 27 12:25:42 2020
 // Update Count     : 4483
+// Last Modified On : Thu May 28 12:11:45 2020
+// Update Count     : 4500
 //
 …
 %type<en> conditional_expression                constant_expression                     assignment_expression           assignment_expression_opt
 %type<en> comma_expression                              comma_expression_opt
 %type<en> argument_expression_list              argument_expression                     default_initialize_opt
+%type<en> argument_expression_list_opt          argument_expression                     default_initialize_opt
 %type<ifctl> if_control_expression
 %type<fctl> for_control_expression              for_control_expression_list
 …
                 // equivalent to the old x[i,j].
                 { $$ = new ExpressionNode( build_binary_val( OperKinds::Index, $1, $3 ) ); }
         | postfix_expression '{' argument_expression_list '}' // CFA, constructor call
+        | postfix_expression '{' argument_expression_list_opt '}' // CFA, constructor call
+                {
                         Token fn;
 …
                         $$ = new ExpressionNode( new ConstructorExpr( build_func( new ExpressionNode( build_varref( fn ) ), (ExpressionNode *)( $1 )->set_last( $3 ) ) ) );
+                }
         | postfix_expression '(' argument_expression_list ')'
+        | postfix_expression '(' argument_expression_list_opt ')'
                 { $$ = new ExpressionNode( build_func( $1, $3 ) ); }
         | postfix_expression '`' identifier                                     // CFA, postfix call
 …
         | '(' type_no_function ')' '@' '{' initializer_list_opt comma_opt '}' // CFA, explicit C compound-literal
                 { $$ = new ExpressionNode( build_compoundLiteral( $2, (new InitializerNode( $6, true ))->set_maybeConstructed( false ) ) ); }
         | '^' primary_expression '{' argument_expression_list '}' // CFA, destructor call
+        | '^' primary_expression '{' argument_expression_list_opt '}' // CFA, destructor call
+                {
                         Token fn;
 …
+        ;
 argument_expression_list:
+argument_expression_list_opt:
         // empty
                 { $$ = nullptr; }
         | argument_expression
         | argument_expression_list ',' argument_expression
+        | argument_expression_list_opt ',' argument_expression
                 { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+        ;
 …
                 { $$ = forCtrl( $1, new string( DeclarationNode::anonymous.newName() ), new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
                                                 OperKinds::LThan, $1->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
         | '=' comma_expression                                                                  // CFA
+        | '=' comma_expression                                                          // CFA
                 { $$ = forCtrl( $2, new string( DeclarationNode::anonymous.newName() ), new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
                                                 OperKinds::LEThan, $2->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
 …
         | comma_expression inclexcl comma_expression '~' comma_expression // CFA
                 { $$ = forCtrl( $1, new string( DeclarationNode::anonymous.newName() ), $1->clone(), $2, $3, $5 ); }
+        | comma_expression ';'                                                          // CFA
+                { $$ = forCtrl( new ExpressionNode( build_constantInteger( *new string( "0u" ) ) ), $1, nullptr, OperKinds::LThan, nullptr, nullptr ); }
         | comma_expression ';' comma_expression                         // CFA
                 { $$ = forCtrl( $3, $1, new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
                                                 OperKinds::LThan, $3->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
         | comma_expression ';' '=' comma_expression                             // CFA
+        | comma_expression ';' '=' comma_expression                     // CFA
                 { $$ = forCtrl( $4, $1, new ExpressionNode( build_constantInteger( *new string( "0" ) ) ),
                                                 OperKinds::LEThan, $4->clone(), new ExpressionNode( build_constantInteger( *new string( "1" ) ) ) ); }
 …
 // If MUTEX becomes a general qualifier, there are shift/reduce conflicts, so change syntax to "with mutex".
 mutex_statement:
         MUTEX '(' argument_expression_list ')' statement
+        MUTEX '(' argument_expression_list_opt ')' statement
                 { SemanticError( yylloc, "Mutex statement is currently unimplemented." ); $$ = nullptr; }
+        ;
 …
         WAITFOR '(' cast_expression ')'
                 { $$ = $3; }
 //      | WAITFOR '(' cast_expression ',' argument_expression_list ')'
+//      | WAITFOR '(' cast_expression ',' argument_expression_list_opt ')'
 //              { $$ = (ExpressionNode *)$3->set_last( $5 ); }
         | WAITFOR '(' cast_expression_list ':' argument_expression_list ')'
+        | WAITFOR '(' cast_expression_list ':' argument_expression_list_opt ')'
                 { $$ = (ExpressionNode *)($3->set_last( $5 )); }
+        ;
 …
         cast_expression
         | cast_expression_list ',' cast_expression
+                { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+                // { $$ = (ExpressionNode *)($1->set_last( $3 )); }
+                { SemanticError( yylloc, "List of mutex member is currently unimplemented." ); $$ = nullptr; }
+        ;
 …
 aggregate_control:                                                                              // CFA
+        GENERATOR
+        MONITOR
+                { yyy = true; $$ = AggregateDecl::Monitor; }
+        | MUTEX STRUCT
+                { yyy = true; $$ = AggregateDecl::Monitor; }
+        | GENERATOR
                 { yyy = true; $$ = AggregateDecl::Generator; }
         | MONITOR GENERATOR
+        | MUTEX GENERATOR
                 { SemanticError( yylloc, "monitor generator is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
         | COROUTINE
                 { yyy = true; $$ = AggregateDecl::Coroutine; }
+        | MONITOR
+                { yyy = true; $$ = AggregateDecl::Monitor; }
+        | MONITOR COROUTINE
+        | MUTEX COROUTINE
                 { SemanticError( yylloc, "monitor coroutine is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
         | THREAD
                 { yyy = true; $$ = AggregateDecl::Thread; }
         | MONITOR THREAD
+        | MUTEX THREAD
                 { SemanticError( yylloc, "monitor thread is currently unimplemented." ); $$ = AggregateDecl::NoAggregate; }
+        ;
 …
         | attr_name
                 { $$ = DeclarationNode::newAttribute( $1 ); }
         | attr_name '(' argument_expression_list ')'
+        | attr_name '(' argument_expression_list_opt ')'
                 { $$ = DeclarationNode::newAttribute( $1, $3 ); }
+        ;

src/ResolvExpr/AlternativeFinder.cc

-              r07d867b
+              r22f94a4
                         unify( castExpr->result, alt.expr->result, alt.env, needAssertions,
                                 haveAssertions, openVars, indexer );
+                        Cost thisCost = castCost( alt.expr->result, castExpr->result, alt.expr->get_lvalue(),
+                                indexer, alt.env );
+                        Cost thisCost =
+                                castExpr->isGenerated
+                                ? conversionCost( alt.expr->result, castExpr->result, alt.expr->get_lvalue(),   indexer, alt.env )
+                                : castCost( alt.expr->result, castExpr->result, alt.expr->get_lvalue(), indexer, alt.env );
                         PRINT(
                                 std::cerr << "working on cast with result: " << castExpr->result << std::endl;
 …
                                 // unification run for side-effects
+                                unify( toType, alt.expr->result, newEnv, need, have, openVars, indexer );
+                                bool canUnify = unify( toType, alt.expr->result, newEnv, need, have, openVars, indexer );
+                                (void) canUnify;
                                 // xxx - do some inspecting on this line... why isn't result bound to initAlt.type?
                                 Cost thisCost = castCost( alt.expr->result, toType, alt.expr->get_lvalue(),
+                                Cost thisCost = computeConversionCost( alt.expr->result, toType, alt.expr->get_lvalue(),
                                         indexer, newEnv );
+                                PRINT(
+                                        Cost legacyCost = castCost( alt.expr->result, toType, alt.expr->get_lvalue(),
+                                                indexer, newEnv );
+                                        std::cerr << "Considering initialization:";
+                                        std::cerr << std::endl << "  FROM: "; alt.expr->result->print(std::cerr);
+                                        std::cerr << std::endl << "  TO: ";   toType          ->print(std::cerr);
+                                        std::cerr << std::endl << "  Unification " << (canUnify ? "succeeded" : "failed");
+                                        std::cerr << std::endl << "  Legacy cost " << legacyCost;
+                                        std::cerr << std::endl << "  New cost " << thisCost;
+                                        std::cerr << std::endl;
+                                )
                                 if ( thisCost != Cost::infinity ) {
                                         // count one safe conversion for each value that is thrown away

src/ResolvExpr/ConversionCost.cc

-              r07d867b
+              r22f94a4
 // Created On       : Sun May 17 07:06:19 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Fri Oct  4 14:45:00 2019
+// Last Modified On : Wed Jul 29 16:11:00 2020
 // Update Count     : 28
 //
 …
         void ConversionCost::postvisit( const FunctionType * ) {}
-        void ConversionCost::postvisit( const StructInstType * inst ) {
-                /*
-                if ( const StructInstType * destAsInst = dynamic_cast< const StructInstType * >( dest ) ) {
-                        if ( inst->name == destAsInst->name ) {
-                                cost = Cost::zero;
-                        } // if
-                } // if
-                */
+        }
-        void ConversionCost::postvisit( const UnionInstType * inst ) {
-                /*
-                if ( const UnionInstType * destAsInst = dynamic_cast< const UnionInstType * >( dest ) ) {
-                        if ( inst->name == destAsInst->name ) {
-                                cost = Cost::zero;
-                        } // if
-                } // if
-                */
+        }
         void ConversionCost::postvisit( const EnumInstType * ) {
                 static Type::Qualifiers q;
 …
+}
-void ConversionCost_new::postvisit( const ast::StructInstType * structInstType ) {
-        /*
-        if ( const ast::StructInstType * dstAsInst =
-                        dynamic_cast< const ast::StructInstType * >( dst ) ) {
-                if ( structInstType->name == dstAsInst->name ) {
-                        cost = Cost::zero;
+                }
+        }
-        */
+}
-void ConversionCost_new::postvisit( const ast::UnionInstType * unionInstType ) {
-        /*
-        if ( const ast::UnionInstType * dstAsInst =
-                        dynamic_cast< const ast::UnionInstType * >( dst ) ) {
-                if ( unionInstType->name == dstAsInst->name ) {
-                        cost = Cost::zero;
+                }
+        }
-        */
+}
 void ConversionCost_new::postvisit( const ast::EnumInstType * enumInstType ) {
         (void)enumInstType;

src/ResolvExpr/ConversionCost.h

-              r07d867b
+              r22f94a4
 // Created On       : Sun May 17 09:37:28 2015
 // Last Modified By : Andrew Beach
 // Last Modified On : Tue Oct  4 14:59:00 2019
+// Last Modified On : Wed Jul 29 16:12:00 2020
 // Update Count     : 7
 //
 …
                 void postvisit( const ReferenceType * refType );
                 void postvisit( const FunctionType * functionType );
-                void postvisit( const StructInstType * aggregateUseType );
-                void postvisit( const UnionInstType * aggregateUseType );
                 void postvisit( const EnumInstType * aggregateUseType );
                 void postvisit( const TraitInstType * aggregateUseType );
 …
         void postvisit( const ast::ReferenceType * refType );
         void postvisit( const ast::FunctionType * functionType );
-        void postvisit( const ast::StructInstType * structInstType );
-        void postvisit( const ast::UnionInstType * unionInstType );
         void postvisit( const ast::EnumInstType * enumInstType );
         void postvisit( const ast::TraitInstType * traitInstType );

src/SynTree/Expression.h

-              r07d867b
+              r22f94a4
   public:
         Expression * arg;
+        bool isGenerated = true; // cast generated implicitly by code generation or explicit in program
+        // Inidicates cast is introduced by the CFA type system.
+        // true for casts that the resolver introduces to force a return type
+        // false for casts from user code
+        // false for casts from desugaring advanced CFA features into simpler CFA
+        // example
+        //   int * p;     // declaration
+        //   (float *) p; // use, with subject cast
+        // subject cast isGenerated means we are considering an interpretation with a type mismatch
+        // subject cast not isGenerated means someone in charge wants it that way
+        bool isGenerated = true;
         CastExpr( Expression * arg, bool isGenerated = true );

src/Virtual/ExpandCasts.cc

-              r07d867b
+              r22f94a4
 // Created On       : Mon Jul 24 13:59:00 2017
 // Last Modified By : Andrew Beach
 // Last Modified On : Tus Aug  2 14:59:00 2017
 // Update Count     : 1
+// Last Modified On : Fri Jul 31 10:29:00 2020
+// Update Count     : 4
 //
 …
 #include <cassert>                 // for assert, assertf
 #include <iterator>                // for back_inserter, inserter
-#include <map>                     // for map, _Rb_tree_iterator, map<>::ite...
 #include <string>                  // for string, allocator, operator==, ope...
-#include <utility>                 // for pair
 #include "Common/PassVisitor.h"    // for PassVisitor
+#include "Common/ScopedMap.h"      // for ScopedMap
 #include "Common/SemanticError.h"  // for SemanticError
+#include "SymTab/Mangler.h"        // for mangleType
 #include "SynTree/Declaration.h"   // for ObjectDecl, StructDecl, FunctionDecl
 #include "SynTree/Expression.h"    // for VirtualCastExpr, CastExpr, Address...
 …
 namespace Virtual {
+        // Indented until the new ast code gets added.
+        /// Maps virtual table types the instance for that type.
+        class VirtualTableMap final {
+                ScopedMap<std::string, ObjectDecl *> vtable_instances;
+        public:
+                void enterScope() {
+                        vtable_instances.beginScope();
+                }
+                void leaveScope() {
+                        vtable_instances.endScope();
+                }
+                ObjectDecl * insert( ObjectDecl * vtableDecl ) {
+                        std::string const & mangledName = SymTab::Mangler::mangleType( vtableDecl->type );
+                        ObjectDecl *& value = vtable_instances[ mangledName ];
+                        if ( value ) {
+                                if ( vtableDecl->storageClasses.is_extern ) {
+                                        return nullptr;
+                                } else if ( ! value->storageClasses.is_extern ) {
+                                        return value;
+                                }
+                        }
+                        value = vtableDecl;
+                        return nullptr;
+                }
+                ObjectDecl * lookup( const Type * vtableType ) {
+                        std::string const & mangledName = SymTab::Mangler::mangleType( vtableType );
+                        const auto it = vtable_instances.find( mangledName );
+                        return ( vtable_instances.end() == it ) ? nullptr : it->second;
+                }
+        };
         /* Currently virtual depends on the rather brittle name matching between
 …
          */
+        namespace {
         std::string get_vtable_name( std::string const & name ) {
                 return name + "_vtable";
 …
         std::string get_vtable_inst_name_root( std::string const & name ) {
                 return get_vtable_name_root( name.substr(1, name.size() - 10 ) );
+        }
-        bool is_vtable_name( std::string const & name ) {
-                return (name.substr( name.size() - 7 ) == "_vtable" );
+        }
 …
+        }
+        } // namespace
         class VirtualCastCore {
-        std::map<std::string, ObjectDecl *> vtable_instances;
-        FunctionDecl *vcast_decl;
-        StructDecl *pvt_decl;
                 Type * pointer_to_pvt(int level_of_indirection) {
                         Type * type = new StructInstType(
 …
         public:
                 VirtualCastCore() :
                         vtable_instances(), vcast_decl( nullptr ), pvt_decl( nullptr )
+                        indexer(), vcast_decl( nullptr ), pvt_decl( nullptr )
                 {}
 …
                 Expression * postmutate( VirtualCastExpr * castExpr );
+                VirtualTableMap indexer;
+        private:
+                FunctionDecl *vcast_decl;
+                StructDecl *pvt_decl;
         };
 …
         void VirtualCastCore::premutate( ObjectDecl * objectDecl ) {
                 if ( is_vtable_inst_name( objectDecl->get_name() ) ) {
+                        vtable_instances[objectDecl->get_name()] = objectDecl;
+                }
+        }
+                        if ( ObjectDecl * existing = indexer.insert( objectDecl ) ) {
+                                std::string msg = "Repeated instance of virtual table, original found at: ";
+                                msg += existing->location.filename;
+                                msg += ":" + toString( existing->location.first_line );
+                                SemanticError( objectDecl->location, msg );
+                        }
+                }
+        }
+        namespace {
+        /// Better error locations for generated casts.
+        CodeLocation castLocation( const VirtualCastExpr * castExpr ) {
+                if ( castExpr->location.isSet() ) {
+                        return castExpr->location;
+                } else if ( castExpr->arg->location.isSet() ) {
+                        return castExpr->arg->location;
+                } else if ( castExpr->result->location.isSet() ) {
+                        return castExpr->result->location;
+                } else {
+                        return CodeLocation();
+                }
+        }
+        [[noreturn]] void castError( const VirtualCastExpr * castExpr, std::string const & message ) {
+                SemanticError( castLocation( castExpr ), message );
+        }
+        /// Get the virtual table type used in a virtual cast.
+        Type * getVirtualTableType( const VirtualCastExpr * castExpr ) {
+                const Type * objectType;
+                if ( auto target = dynamic_cast<const PointerType *>( castExpr->result ) ) {
+                        objectType = target->base;
+                } else if ( auto target = dynamic_cast<const ReferenceType *>( castExpr->result ) ) {
+                        objectType = target->base;
+                } else {
+                        castError( castExpr, "Virtual cast type must be a pointer or reference type." );
+                }
+                assert( objectType );
+                const StructInstType * structType = dynamic_cast<const StructInstType *>( objectType );
+                if ( nullptr == structType ) {
+                        castError( castExpr, "Virtual cast type must refer to a structure type." );
+                }
+                const StructDecl * structDecl = structType->baseStruct;
+                assert( structDecl );
+                const ObjectDecl * fieldDecl = nullptr;
+                if ( 0 < structDecl->members.size() ) {
+                        const Declaration * memberDecl = structDecl->members.front();
+                        assert( memberDecl );
+                        fieldDecl = dynamic_cast<const ObjectDecl *>( memberDecl );
+                        if ( fieldDecl && fieldDecl->name != "virtual_table" ) {
+                                fieldDecl = nullptr;
+                        }
+                }
+                if ( nullptr == fieldDecl ) {
+                        castError( castExpr, "Virtual cast type must have a leading virtual_table field." );
+                }
+                const PointerType * fieldType = dynamic_cast<const PointerType *>( fieldDecl->type );
+                if ( nullptr == fieldType ) {
+                        castError( castExpr, "Virtual cast type virtual_table field is not a pointer." );
+                }
+                assert( fieldType->base );
+                auto virtualStructType = dynamic_cast<const StructInstType *>( fieldType->base );
+                assert( virtualStructType );
+                // Here is the type, but if it is polymorphic it will have lost information.
+                // (Always a clone so that it may always be deleted.)
+                StructInstType * virtualType = virtualStructType->clone();
+                if ( ! structType->parameters.empty() ) {
+                        deleteAll( virtualType->parameters );
+                        virtualType->parameters.clear();
+                        cloneAll( structType->parameters, virtualType->parameters );
+                }
+                return virtualType;
+        }
+        } // namespace
         Expression * VirtualCastCore::postmutate( VirtualCastExpr * castExpr ) {
                 assertf( castExpr->get_result(), "Virtual Cast target not found before expansion." );
+                assertf( castExpr->result, "Virtual Cast target not found before expansion." );
                 assert( vcast_decl );
                 assert( pvt_decl );
+                // May only cast to a pointer or reference type.
+                // A earlier validation should give a syntax error, this is
+                // just to make sure errors don't creep during translation.
+                // Move to helper with more detailed error messages.
+                PointerType * target_type =
+                        dynamic_cast<PointerType *>( castExpr->get_result() );
+                assert( target_type );
+                StructInstType * target_struct =
+                        dynamic_cast<StructInstType *>( target_type->get_base() );
+                assert( target_struct );
+                StructDecl * target_decl = target_struct->get_baseStruct();
+                std::map<std::string, ObjectDecl *>::iterator found =
+                        vtable_instances.find(
+                                get_vtable_inst_name( target_decl->get_name() ) );
+                if ( vtable_instances.end() == found ) {
+                        assertf( false, "virtual table instance not found." );
+                }
+                ObjectDecl * table = found->second;
+                const Type * vtable_type = getVirtualTableType( castExpr );
+                ObjectDecl * table = indexer.lookup( vtable_type );
+                if ( nullptr == table ) {
+                        SemanticError( castLocation( castExpr ),
+                                "Could not find virtual table instance." );
+                }
                 Expression * result = new CastExpr(
-                        //new ApplicationExpr(
-                                //new AddressExpr( new VariableExpr( vcast_decl ) ),
-                                //new CastExpr( new VariableExpr( vcast_decl ),
-                                //      new PointerType( noQualifiers,
-                                //              vcast_decl->get_type()->clone()
-                                //              )
-                                //      ),
                         new ApplicationExpr( VariableExpr::functionPointer( vcast_decl ), {
                                         new CastExpr(
 …
                 castExpr->set_result( nullptr );
                 delete castExpr;
+                delete vtable_type;
                 return result;
+        }

src/main.cc

-              r07d867b
+              r22f94a4
 // Author           : Peter Buhr and Rob Schluntz
 // Created On       : Fri May 15 23:12:02 2015
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sat Feb  8 08:33:50 2020
 // Update Count     : 633
+// Last Modified By : Andrew Beach
+// Last Modified On : Tue May 19 12:03:00 2020
+// Update Count     : 634
 //
 …
                 } // if
+                PASS( "Translate Throws", ControlStruct::translateThrows( translationUnit ) );
                 PASS( "Fix Labels", ControlStruct::fixLabels( translationUnit ) );
                 PASS( "Fix Names", CodeGen::fixNames( translationUnit ) );
 …
                 PASS( "Expand Unique Expr", Tuples::expandUniqueExpr( translationUnit ) ); // xxx - is this the right place for this? want to expand ASAP so tha, sequent passes don't need to worry about double-visiting a unique expr - needs to go after InitTweak::fix so that copy constructed return declarations are reused
                 PASS( "Translate EHM" , ControlStruct::translateEHM( translationUnit ) );
+                PASS( "Translate Tries" , ControlStruct::translateTries( translationUnit ) );
                 PASS( "Gen Waitfor" , Concurrency::generateWaitFor( translationUnit ) );

tests/.expect/alloc.txt

-              r07d867b
+              r22f94a4
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef
 CFA realloc array alloc, fill
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
+xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
 CFA realloc array alloc, fill
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef
 CFA realloc array alloc, fill
 xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0x1010101 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
+xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdeadbeef 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede 0xdededede
 C   memalign 42 42.5

tests/.expect/bitmanip3.x64.txt

-              r07d867b
+              r22f94a4
 signed char
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned char
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(254, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(253, 1) = 253
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(252, 2) = 252
 …
 short int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned short int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(65534, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(65533, 1) = 65533
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(65532, 2) = 65532
 …
 int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(4294967294, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(4294967293, 1) = 4294967293
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(4294967292, 2) = 4294967292
 …
 long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(18446744073709551614, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(18446744073709551613, 1) = 18446744073709551613
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(18446744073709551612, 2) = 18446744073709551612
 …
 long long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned long long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(18446744073709551614, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(18446744073709551613, 1) = 18446744073709551613
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(18446744073709551612, 2) = 18446744073709551612
 …
 signed char
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned char
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(254, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(253, 1) = 253
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(252, 2) = 252
 …
 short int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned short int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(65534, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(65533, 1) = 65533
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(65532, 2) = 65532
 …
 int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(4294967294, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(4294967293, 1) = 4294967293
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(4294967292, 2) = 4294967292
 …
 long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(18446744073709551614, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(18446744073709551613, 1) = 18446744073709551613
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(18446744073709551612, 2) = 18446744073709551612
 …
 long long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned long long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(18446744073709551614, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(18446744073709551613, 1) = 18446744073709551613
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(18446744073709551612, 2) = 18446744073709551612
 …
+ceiling_div
+signed char
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(-128, -128) = 2, ceiling_div(-126, -128) = 1, ceiling_div(126, -128) = 0
+unsigned char
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(253, 1) = 253
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(252, 2) = 126
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(250, 4) = 63
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(246, 8) = 31
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(238, 16) = 15
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(222, 32) = 7
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(190, 64) = 3
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(126, 128) = 1
+short int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(-32768, -32768) = 2, ceiling_div(-32766, -32768) = 1, ceiling_div(32766, -32768) = 0
+unsigned short int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(65533, 1) = 65533
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(65532, 2) = 32766
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(65530, 4) = 16383
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(65526, 8) = 8191
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(65518, 16) = 4095
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(65502, 32) = 2047
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(65470, 64) = 1023
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(65406, 128) = 511
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(65278, 256) = 255
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(65022, 512) = 127
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(64510, 1024) = 63
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(63486, 2048) = 31
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(61438, 4096) = 15
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(57342, 8192) = 7
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(49150, 16384) = 3
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(32766, 32768) = 1
+int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(-32770, 32768) = 0
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(-65538, 65536) = 0
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(-131074, 131072) = 0
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(-262146, 262144) = 0
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(-524290, 524288) = 0
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(-1048578, 1048576) = 0
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(-2097154, 2097152) = 0
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(-4194306, 4194304) = 0
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(-8388610, 8388608) = 0
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(-16777218, 16777216) = 0
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(-33554434, 33554432) = 0
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(-67108866, 67108864) = 0
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(-134217730, 134217728) = 0
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(-268435458, 268435456) = 0
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(-536870914, 536870912) = 0
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = -1, ceiling_div(-1073741826, 1073741824) = 0
+ceiling_div(-2147483648, -2147483648) = 0, ceiling_div(-2147483646, -2147483648) = 0, ceiling_div(2147483646, -2147483648) = 0
+unsigned int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(4294967293, 1) = 4294967293
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(4294967292, 2) = 2147483646
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(4294967290, 4) = 1073741823
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(4294967286, 8) = 536870911
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(4294967278, 16) = 268435455
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(4294967262, 32) = 134217727
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(4294967230, 64) = 67108863
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(4294967166, 128) = 33554431
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(4294967038, 256) = 16777215
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(4294966782, 512) = 8388607
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(4294966270, 1024) = 4194303
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(4294965246, 2048) = 2097151
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(4294963198, 4096) = 1048575
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(4294959102, 8192) = 524287
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(4294950910, 16384) = 262143
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(4294934526, 32768) = 131071
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(4294901758, 65536) = 65535
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(4294836222, 131072) = 32767
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(4294705150, 262144) = 16383
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(4294443006, 524288) = 8191
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(4293918718, 1048576) = 4095
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(4292870142, 2097152) = 2047
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(4290772990, 4194304) = 1023
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(4286578686, 8388608) = 511
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(4278190078, 16777216) = 255
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(4261412862, 33554432) = 127
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(4227858430, 67108864) = 63
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(4160749566, 134217728) = 31
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(4026531838, 268435456) = 15
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(3758096382, 536870912) = 7
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(3221225470, 1073741824) = 3
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 0, ceiling_div(2147483646, 2147483648) = 1
+long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(-32770, 32768) = 0
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(-65538, 65536) = 0
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(-131074, 131072) = 0
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(-262146, 262144) = 0
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(-524290, 524288) = 0
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(-1048578, 1048576) = 0
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(-2097154, 2097152) = 0
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(-4194306, 4194304) = 0
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(-8388610, 8388608) = 0
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(-16777218, 16777216) = 0
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(-33554434, 33554432) = 0
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(-67108866, 67108864) = 0
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(-134217730, 134217728) = 0
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(-268435458, 268435456) = 0
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(-536870914, 536870912) = 0
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(-1073741826, 1073741824) = 0
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 2, ceiling_div(-2147483650, 2147483648) = 0
+ceiling_div(4294967296, 4294967296) = 1, ceiling_div(4294967298, 4294967296) = 2, ceiling_div(-4294967298, 4294967296) = 0
+ceiling_div(8589934592, 8589934592) = 1, ceiling_div(8589934594, 8589934592) = 2, ceiling_div(-8589934594, 8589934592) = 0
+ceiling_div(17179869184, 17179869184) = 1, ceiling_div(17179869186, 17179869184) = 2, ceiling_div(-17179869186, 17179869184) = 0
+ceiling_div(34359738368, 34359738368) = 1, ceiling_div(34359738370, 34359738368) = 2, ceiling_div(-34359738370, 34359738368) = 0
+ceiling_div(68719476736, 68719476736) = 1, ceiling_div(68719476738, 68719476736) = 2, ceiling_div(-68719476738, 68719476736) = 0
+ceiling_div(137438953472, 137438953472) = 1, ceiling_div(137438953474, 137438953472) = 2, ceiling_div(-137438953474, 137438953472) = 0
+ceiling_div(274877906944, 274877906944) = 1, ceiling_div(274877906946, 274877906944) = 2, ceiling_div(-274877906946, 274877906944) = 0
+ceiling_div(549755813888, 549755813888) = 1, ceiling_div(549755813890, 549755813888) = 2, ceiling_div(-549755813890, 549755813888) = 0
+ceiling_div(1099511627776, 1099511627776) = 1, ceiling_div(1099511627778, 1099511627776) = 2, ceiling_div(-1099511627778, 1099511627776) = 0
+ceiling_div(2199023255552, 2199023255552) = 1, ceiling_div(2199023255554, 2199023255552) = 2, ceiling_div(-2199023255554, 2199023255552) = 0
+ceiling_div(4398046511104, 4398046511104) = 1, ceiling_div(4398046511106, 4398046511104) = 2, ceiling_div(-4398046511106, 4398046511104) = 0
+ceiling_div(8796093022208, 8796093022208) = 1, ceiling_div(8796093022210, 8796093022208) = 2, ceiling_div(-8796093022210, 8796093022208) = 0
+ceiling_div(17592186044416, 17592186044416) = 1, ceiling_div(17592186044418, 17592186044416) = 2, ceiling_div(-17592186044418, 17592186044416) = 0
+ceiling_div(35184372088832, 35184372088832) = 1, ceiling_div(35184372088834, 35184372088832) = 2, ceiling_div(-35184372088834, 35184372088832) = 0
+ceiling_div(70368744177664, 70368744177664) = 1, ceiling_div(70368744177666, 70368744177664) = 2, ceiling_div(-70368744177666, 70368744177664) = 0
+ceiling_div(140737488355328, 140737488355328) = 1, ceiling_div(140737488355330, 140737488355328) = 2, ceiling_div(-140737488355330, 140737488355328) = 0
+ceiling_div(281474976710656, 281474976710656) = 1, ceiling_div(281474976710658, 281474976710656) = 2, ceiling_div(-281474976710658, 281474976710656) = 0
+ceiling_div(562949953421312, 562949953421312) = 1, ceiling_div(562949953421314, 562949953421312) = 2, ceiling_div(-562949953421314, 562949953421312) = 0
+ceiling_div(1125899906842624, 1125899906842624) = 1, ceiling_div(1125899906842626, 1125899906842624) = 2, ceiling_div(-1125899906842626, 1125899906842624) = 0
+ceiling_div(2251799813685248, 2251799813685248) = 1, ceiling_div(2251799813685250, 2251799813685248) = 2, ceiling_div(-2251799813685250, 2251799813685248) = 0
+ceiling_div(4503599627370496, 4503599627370496) = 1, ceiling_div(4503599627370498, 4503599627370496) = 2, ceiling_div(-4503599627370498, 4503599627370496) = 0
+ceiling_div(9007199254740992, 9007199254740992) = 1, ceiling_div(9007199254740994, 9007199254740992) = 2, ceiling_div(-9007199254740994, 9007199254740992) = 0
+ceiling_div(18014398509481984, 18014398509481984) = 1, ceiling_div(18014398509481986, 18014398509481984) = 2, ceiling_div(-18014398509481986, 18014398509481984) = 0
+ceiling_div(36028797018963968, 36028797018963968) = 1, ceiling_div(36028797018963970, 36028797018963968) = 2, ceiling_div(-36028797018963970, 36028797018963968) = 0
+ceiling_div(72057594037927936, 72057594037927936) = 1, ceiling_div(72057594037927938, 72057594037927936) = 2, ceiling_div(-72057594037927938, 72057594037927936) = 0
+ceiling_div(144115188075855872, 144115188075855872) = 1, ceiling_div(144115188075855874, 144115188075855872) = 2, ceiling_div(-144115188075855874, 144115188075855872) = 0
+ceiling_div(288230376151711744, 288230376151711744) = 1, ceiling_div(288230376151711746, 288230376151711744) = 2, ceiling_div(-288230376151711746, 288230376151711744) = 0
+ceiling_div(576460752303423488, 576460752303423488) = 1, ceiling_div(576460752303423490, 576460752303423488) = 2, ceiling_div(-576460752303423490, 576460752303423488) = 0
+ceiling_div(1152921504606846976, 1152921504606846976) = 1, ceiling_div(1152921504606846978, 1152921504606846976) = 2, ceiling_div(-1152921504606846978, 1152921504606846976) = 0
+ceiling_div(2305843009213693952, 2305843009213693952) = 1, ceiling_div(2305843009213693954, 2305843009213693952) = 2, ceiling_div(-2305843009213693954, 2305843009213693952) = 0
+ceiling_div(4611686018427387904, 4611686018427387904) = 1, ceiling_div(4611686018427387906, 4611686018427387904) = -1, ceiling_div(-4611686018427387906, 4611686018427387904) = 0
+ceiling_div(-9223372036854775808, -9223372036854775808) = 0, ceiling_div(-9223372036854775806, -9223372036854775808) = 0, ceiling_div(9223372036854775806, -9223372036854775808) = 0
+unsigned long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(18446744073709551613, 1) = 18446744073709551613
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(18446744073709551612, 2) = 9223372036854775806
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(18446744073709551610, 4) = 4611686018427387903
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(18446744073709551606, 8) = 2305843009213693951
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(18446744073709551598, 16) = 1152921504606846975
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(18446744073709551582, 32) = 576460752303423487
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(18446744073709551550, 64) = 288230376151711743
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(18446744073709551486, 128) = 144115188075855871
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(18446744073709551358, 256) = 72057594037927935
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(18446744073709551102, 512) = 36028797018963967
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(18446744073709550590, 1024) = 18014398509481983
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(18446744073709549566, 2048) = 9007199254740991
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(18446744073709547518, 4096) = 4503599627370495
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(18446744073709543422, 8192) = 2251799813685247
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(18446744073709535230, 16384) = 1125899906842623
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(18446744073709518846, 32768) = 562949953421311
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(18446744073709486078, 65536) = 281474976710655
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(18446744073709420542, 131072) = 140737488355327
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(18446744073709289470, 262144) = 70368744177663
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(18446744073709027326, 524288) = 35184372088831
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(18446744073708503038, 1048576) = 17592186044415
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(18446744073707454462, 2097152) = 8796093022207
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(18446744073705357310, 4194304) = 4398046511103
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(18446744073701163006, 8388608) = 2199023255551
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(18446744073692774398, 16777216) = 1099511627775
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(18446744073675997182, 33554432) = 549755813887
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(18446744073642442750, 67108864) = 274877906943
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(18446744073575333886, 134217728) = 137438953471
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(18446744073441116158, 268435456) = 68719476735
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(18446744073172680702, 536870912) = 34359738367
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(18446744072635809790, 1073741824) = 17179869183
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 2, ceiling_div(18446744071562067966, 2147483648) = 8589934591
+ceiling_div(4294967296, 4294967296) = 1, ceiling_div(4294967298, 4294967296) = 2, ceiling_div(18446744069414584318, 4294967296) = 4294967295
+ceiling_div(8589934592, 8589934592) = 1, ceiling_div(8589934594, 8589934592) = 2, ceiling_div(18446744065119617022, 8589934592) = 2147483647
+ceiling_div(17179869184, 17179869184) = 1, ceiling_div(17179869186, 17179869184) = 2, ceiling_div(18446744056529682430, 17179869184) = 1073741823
+ceiling_div(34359738368, 34359738368) = 1, ceiling_div(34359738370, 34359738368) = 2, ceiling_div(18446744039349813246, 34359738368) = 536870911
+ceiling_div(68719476736, 68719476736) = 1, ceiling_div(68719476738, 68719476736) = 2, ceiling_div(18446744004990074878, 68719476736) = 268435455
+ceiling_div(137438953472, 137438953472) = 1, ceiling_div(137438953474, 137438953472) = 2, ceiling_div(18446743936270598142, 137438953472) = 134217727
+ceiling_div(274877906944, 274877906944) = 1, ceiling_div(274877906946, 274877906944) = 2, ceiling_div(18446743798831644670, 274877906944) = 67108863
+ceiling_div(549755813888, 549755813888) = 1, ceiling_div(549755813890, 549755813888) = 2, ceiling_div(18446743523953737726, 549755813888) = 33554431
+ceiling_div(1099511627776, 1099511627776) = 1, ceiling_div(1099511627778, 1099511627776) = 2, ceiling_div(18446742974197923838, 1099511627776) = 16777215
+ceiling_div(2199023255552, 2199023255552) = 1, ceiling_div(2199023255554, 2199023255552) = 2, ceiling_div(18446741874686296062, 2199023255552) = 8388607
+ceiling_div(4398046511104, 4398046511104) = 1, ceiling_div(4398046511106, 4398046511104) = 2, ceiling_div(18446739675663040510, 4398046511104) = 4194303
+ceiling_div(8796093022208, 8796093022208) = 1, ceiling_div(8796093022210, 8796093022208) = 2, ceiling_div(18446735277616529406, 8796093022208) = 2097151
+ceiling_div(17592186044416, 17592186044416) = 1, ceiling_div(17592186044418, 17592186044416) = 2, ceiling_div(18446726481523507198, 17592186044416) = 1048575
+ceiling_div(35184372088832, 35184372088832) = 1, ceiling_div(35184372088834, 35184372088832) = 2, ceiling_div(18446708889337462782, 35184372088832) = 524287
+ceiling_div(70368744177664, 70368744177664) = 1, ceiling_div(70368744177666, 70368744177664) = 2, ceiling_div(18446673704965373950, 70368744177664) = 262143
+ceiling_div(140737488355328, 140737488355328) = 1, ceiling_div(140737488355330, 140737488355328) = 2, ceiling_div(18446603336221196286, 140737488355328) = 131071
+ceiling_div(281474976710656, 281474976710656) = 1, ceiling_div(281474976710658, 281474976710656) = 2, ceiling_div(18446462598732840958, 281474976710656) = 65535
+ceiling_div(562949953421312, 562949953421312) = 1, ceiling_div(562949953421314, 562949953421312) = 2, ceiling_div(18446181123756130302, 562949953421312) = 32767
+ceiling_div(1125899906842624, 1125899906842624) = 1, ceiling_div(1125899906842626, 1125899906842624) = 2, ceiling_div(18445618173802708990, 1125899906842624) = 16383
+ceiling_div(2251799813685248, 2251799813685248) = 1, ceiling_div(2251799813685250, 2251799813685248) = 2, ceiling_div(18444492273895866366, 2251799813685248) = 8191
+ceiling_div(4503599627370496, 4503599627370496) = 1, ceiling_div(4503599627370498, 4503599627370496) = 2, ceiling_div(18442240474082181118, 4503599627370496) = 4095
+ceiling_div(9007199254740992, 9007199254740992) = 1, ceiling_div(9007199254740994, 9007199254740992) = 2, ceiling_div(18437736874454810622, 9007199254740992) = 2047
+ceiling_div(18014398509481984, 18014398509481984) = 1, ceiling_div(18014398509481986, 18014398509481984) = 2, ceiling_div(18428729675200069630, 18014398509481984) = 1023
+ceiling_div(36028797018963968, 36028797018963968) = 1, ceiling_div(36028797018963970, 36028797018963968) = 2, ceiling_div(18410715276690587646, 36028797018963968) = 511
+ceiling_div(72057594037927936, 72057594037927936) = 1, ceiling_div(72057594037927938, 72057594037927936) = 2, ceiling_div(18374686479671623678, 72057594037927936) = 255
+ceiling_div(144115188075855872, 144115188075855872) = 1, ceiling_div(144115188075855874, 144115188075855872) = 2, ceiling_div(18302628885633695742, 144115188075855872) = 127
+ceiling_div(288230376151711744, 288230376151711744) = 1, ceiling_div(288230376151711746, 288230376151711744) = 2, ceiling_div(18158513697557839870, 288230376151711744) = 63
+ceiling_div(576460752303423488, 576460752303423488) = 1, ceiling_div(576460752303423490, 576460752303423488) = 2, ceiling_div(17870283321406128126, 576460752303423488) = 31
+ceiling_div(1152921504606846976, 1152921504606846976) = 1, ceiling_div(1152921504606846978, 1152921504606846976) = 2, ceiling_div(17293822569102704638, 1152921504606846976) = 15
+ceiling_div(2305843009213693952, 2305843009213693952) = 1, ceiling_div(2305843009213693954, 2305843009213693952) = 2, ceiling_div(16140901064495857662, 2305843009213693952) = 7
+ceiling_div(4611686018427387904, 4611686018427387904) = 1, ceiling_div(4611686018427387906, 4611686018427387904) = 2, ceiling_div(13835058055282163710, 4611686018427387904) = 3
+ceiling_div(9223372036854775808, 9223372036854775808) = 1, ceiling_div(9223372036854775810, 9223372036854775808) = 0, ceiling_div(9223372036854775806, 9223372036854775808) = 1
+long long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(-32770, 32768) = 0
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(-65538, 65536) = 0
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(-131074, 131072) = 0
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(-262146, 262144) = 0
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(-524290, 524288) = 0
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(-1048578, 1048576) = 0
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(-2097154, 2097152) = 0
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(-4194306, 4194304) = 0
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(-8388610, 8388608) = 0
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(-16777218, 16777216) = 0
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(-33554434, 33554432) = 0
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(-67108866, 67108864) = 0
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(-134217730, 134217728) = 0
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(-268435458, 268435456) = 0
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(-536870914, 536870912) = 0
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(-1073741826, 1073741824) = 0
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 2, ceiling_div(-2147483650, 2147483648) = 0
+ceiling_div(4294967296, 4294967296) = 1, ceiling_div(4294967298, 4294967296) = 2, ceiling_div(-4294967298, 4294967296) = 0
+ceiling_div(8589934592, 8589934592) = 1, ceiling_div(8589934594, 8589934592) = 2, ceiling_div(-8589934594, 8589934592) = 0
+ceiling_div(17179869184, 17179869184) = 1, ceiling_div(17179869186, 17179869184) = 2, ceiling_div(-17179869186, 17179869184) = 0
+ceiling_div(34359738368, 34359738368) = 1, ceiling_div(34359738370, 34359738368) = 2, ceiling_div(-34359738370, 34359738368) = 0
+ceiling_div(68719476736, 68719476736) = 1, ceiling_div(68719476738, 68719476736) = 2, ceiling_div(-68719476738, 68719476736) = 0
+ceiling_div(137438953472, 137438953472) = 1, ceiling_div(137438953474, 137438953472) = 2, ceiling_div(-137438953474, 137438953472) = 0
+ceiling_div(274877906944, 274877906944) = 1, ceiling_div(274877906946, 274877906944) = 2, ceiling_div(-274877906946, 274877906944) = 0
+ceiling_div(549755813888, 549755813888) = 1, ceiling_div(549755813890, 549755813888) = 2, ceiling_div(-549755813890, 549755813888) = 0
+ceiling_div(1099511627776, 1099511627776) = 1, ceiling_div(1099511627778, 1099511627776) = 2, ceiling_div(-1099511627778, 1099511627776) = 0
+ceiling_div(2199023255552, 2199023255552) = 1, ceiling_div(2199023255554, 2199023255552) = 2, ceiling_div(-2199023255554, 2199023255552) = 0
+ceiling_div(4398046511104, 4398046511104) = 1, ceiling_div(4398046511106, 4398046511104) = 2, ceiling_div(-4398046511106, 4398046511104) = 0
+ceiling_div(8796093022208, 8796093022208) = 1, ceiling_div(8796093022210, 8796093022208) = 2, ceiling_div(-8796093022210, 8796093022208) = 0
+ceiling_div(17592186044416, 17592186044416) = 1, ceiling_div(17592186044418, 17592186044416) = 2, ceiling_div(-17592186044418, 17592186044416) = 0
+ceiling_div(35184372088832, 35184372088832) = 1, ceiling_div(35184372088834, 35184372088832) = 2, ceiling_div(-35184372088834, 35184372088832) = 0
+ceiling_div(70368744177664, 70368744177664) = 1, ceiling_div(70368744177666, 70368744177664) = 2, ceiling_div(-70368744177666, 70368744177664) = 0
+ceiling_div(140737488355328, 140737488355328) = 1, ceiling_div(140737488355330, 140737488355328) = 2, ceiling_div(-140737488355330, 140737488355328) = 0
+ceiling_div(281474976710656, 281474976710656) = 1, ceiling_div(281474976710658, 281474976710656) = 2, ceiling_div(-281474976710658, 281474976710656) = 0
+ceiling_div(562949953421312, 562949953421312) = 1, ceiling_div(562949953421314, 562949953421312) = 2, ceiling_div(-562949953421314, 562949953421312) = 0
+ceiling_div(1125899906842624, 1125899906842624) = 1, ceiling_div(1125899906842626, 1125899906842624) = 2, ceiling_div(-1125899906842626, 1125899906842624) = 0
+ceiling_div(2251799813685248, 2251799813685248) = 1, ceiling_div(2251799813685250, 2251799813685248) = 2, ceiling_div(-2251799813685250, 2251799813685248) = 0
+ceiling_div(4503599627370496, 4503599627370496) = 1, ceiling_div(4503599627370498, 4503599627370496) = 2, ceiling_div(-4503599627370498, 4503599627370496) = 0
+ceiling_div(9007199254740992, 9007199254740992) = 1, ceiling_div(9007199254740994, 9007199254740992) = 2, ceiling_div(-9007199254740994, 9007199254740992) = 0
+ceiling_div(18014398509481984, 18014398509481984) = 1, ceiling_div(18014398509481986, 18014398509481984) = 2, ceiling_div(-18014398509481986, 18014398509481984) = 0
+ceiling_div(36028797018963968, 36028797018963968) = 1, ceiling_div(36028797018963970, 36028797018963968) = 2, ceiling_div(-36028797018963970, 36028797018963968) = 0
+ceiling_div(72057594037927936, 72057594037927936) = 1, ceiling_div(72057594037927938, 72057594037927936) = 2, ceiling_div(-72057594037927938, 72057594037927936) = 0
+ceiling_div(144115188075855872, 144115188075855872) = 1, ceiling_div(144115188075855874, 144115188075855872) = 2, ceiling_div(-144115188075855874, 144115188075855872) = 0
+ceiling_div(288230376151711744, 288230376151711744) = 1, ceiling_div(288230376151711746, 288230376151711744) = 2, ceiling_div(-288230376151711746, 288230376151711744) = 0
+ceiling_div(576460752303423488, 576460752303423488) = 1, ceiling_div(576460752303423490, 576460752303423488) = 2, ceiling_div(-576460752303423490, 576460752303423488) = 0
+ceiling_div(1152921504606846976, 1152921504606846976) = 1, ceiling_div(1152921504606846978, 1152921504606846976) = 2, ceiling_div(-1152921504606846978, 1152921504606846976) = 0
+ceiling_div(2305843009213693952, 2305843009213693952) = 1, ceiling_div(2305843009213693954, 2305843009213693952) = 2, ceiling_div(-2305843009213693954, 2305843009213693952) = 0
+ceiling_div(4611686018427387904, 4611686018427387904) = 1, ceiling_div(4611686018427387906, 4611686018427387904) = -1, ceiling_div(-4611686018427387906, 4611686018427387904) = 0
+ceiling_div(-9223372036854775808, -9223372036854775808) = 0, ceiling_div(-9223372036854775806, -9223372036854775808) = 0, ceiling_div(9223372036854775806, -9223372036854775808) = 0
+unsigned long long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(18446744073709551613, 1) = 18446744073709551613
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(18446744073709551612, 2) = 9223372036854775806
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(18446744073709551610, 4) = 4611686018427387903
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(18446744073709551606, 8) = 2305843009213693951
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(18446744073709551598, 16) = 1152921504606846975
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(18446744073709551582, 32) = 576460752303423487
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(18446744073709551550, 64) = 288230376151711743
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(18446744073709551486, 128) = 144115188075855871
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(18446744073709551358, 256) = 72057594037927935
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(18446744073709551102, 512) = 36028797018963967
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(18446744073709550590, 1024) = 18014398509481983
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(18446744073709549566, 2048) = 9007199254740991
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(18446744073709547518, 4096) = 4503599627370495
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(18446744073709543422, 8192) = 2251799813685247
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(18446744073709535230, 16384) = 1125899906842623
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(18446744073709518846, 32768) = 562949953421311
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(18446744073709486078, 65536) = 281474976710655
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(18446744073709420542, 131072) = 140737488355327
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(18446744073709289470, 262144) = 70368744177663
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(18446744073709027326, 524288) = 35184372088831
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(18446744073708503038, 1048576) = 17592186044415
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(18446744073707454462, 2097152) = 8796093022207
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(18446744073705357310, 4194304) = 4398046511103
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(18446744073701163006, 8388608) = 2199023255551
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(18446744073692774398, 16777216) = 1099511627775
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(18446744073675997182, 33554432) = 549755813887
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(18446744073642442750, 67108864) = 274877906943
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(18446744073575333886, 134217728) = 137438953471
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(18446744073441116158, 268435456) = 68719476735
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(18446744073172680702, 536870912) = 34359738367
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(18446744072635809790, 1073741824) = 17179869183
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 2, ceiling_div(18446744071562067966, 2147483648) = 8589934591
+ceiling_div(4294967296, 4294967296) = 1, ceiling_div(4294967298, 4294967296) = 2, ceiling_div(18446744069414584318, 4294967296) = 4294967295
+ceiling_div(8589934592, 8589934592) = 1, ceiling_div(8589934594, 8589934592) = 2, ceiling_div(18446744065119617022, 8589934592) = 2147483647
+ceiling_div(17179869184, 17179869184) = 1, ceiling_div(17179869186, 17179869184) = 2, ceiling_div(18446744056529682430, 17179869184) = 1073741823
+ceiling_div(34359738368, 34359738368) = 1, ceiling_div(34359738370, 34359738368) = 2, ceiling_div(18446744039349813246, 34359738368) = 536870911
+ceiling_div(68719476736, 68719476736) = 1, ceiling_div(68719476738, 68719476736) = 2, ceiling_div(18446744004990074878, 68719476736) = 268435455
+ceiling_div(137438953472, 137438953472) = 1, ceiling_div(137438953474, 137438953472) = 2, ceiling_div(18446743936270598142, 137438953472) = 134217727
+ceiling_div(274877906944, 274877906944) = 1, ceiling_div(274877906946, 274877906944) = 2, ceiling_div(18446743798831644670, 274877906944) = 67108863
+ceiling_div(549755813888, 549755813888) = 1, ceiling_div(549755813890, 549755813888) = 2, ceiling_div(18446743523953737726, 549755813888) = 33554431
+ceiling_div(1099511627776, 1099511627776) = 1, ceiling_div(1099511627778, 1099511627776) = 2, ceiling_div(18446742974197923838, 1099511627776) = 16777215
+ceiling_div(2199023255552, 2199023255552) = 1, ceiling_div(2199023255554, 2199023255552) = 2, ceiling_div(18446741874686296062, 2199023255552) = 8388607
+ceiling_div(4398046511104, 4398046511104) = 1, ceiling_div(4398046511106, 4398046511104) = 2, ceiling_div(18446739675663040510, 4398046511104) = 4194303
+ceiling_div(8796093022208, 8796093022208) = 1, ceiling_div(8796093022210, 8796093022208) = 2, ceiling_div(18446735277616529406, 8796093022208) = 2097151
+ceiling_div(17592186044416, 17592186044416) = 1, ceiling_div(17592186044418, 17592186044416) = 2, ceiling_div(18446726481523507198, 17592186044416) = 1048575
+ceiling_div(35184372088832, 35184372088832) = 1, ceiling_div(35184372088834, 35184372088832) = 2, ceiling_div(18446708889337462782, 35184372088832) = 524287
+ceiling_div(70368744177664, 70368744177664) = 1, ceiling_div(70368744177666, 70368744177664) = 2, ceiling_div(18446673704965373950, 70368744177664) = 262143
+ceiling_div(140737488355328, 140737488355328) = 1, ceiling_div(140737488355330, 140737488355328) = 2, ceiling_div(18446603336221196286, 140737488355328) = 131071
+ceiling_div(281474976710656, 281474976710656) = 1, ceiling_div(281474976710658, 281474976710656) = 2, ceiling_div(18446462598732840958, 281474976710656) = 65535
+ceiling_div(562949953421312, 562949953421312) = 1, ceiling_div(562949953421314, 562949953421312) = 2, ceiling_div(18446181123756130302, 562949953421312) = 32767
+ceiling_div(1125899906842624, 1125899906842624) = 1, ceiling_div(1125899906842626, 1125899906842624) = 2, ceiling_div(18445618173802708990, 1125899906842624) = 16383
+ceiling_div(2251799813685248, 2251799813685248) = 1, ceiling_div(2251799813685250, 2251799813685248) = 2, ceiling_div(18444492273895866366, 2251799813685248) = 8191
+ceiling_div(4503599627370496, 4503599627370496) = 1, ceiling_div(4503599627370498, 4503599627370496) = 2, ceiling_div(18442240474082181118, 4503599627370496) = 4095
+ceiling_div(9007199254740992, 9007199254740992) = 1, ceiling_div(9007199254740994, 9007199254740992) = 2, ceiling_div(18437736874454810622, 9007199254740992) = 2047
+ceiling_div(18014398509481984, 18014398509481984) = 1, ceiling_div(18014398509481986, 18014398509481984) = 2, ceiling_div(18428729675200069630, 18014398509481984) = 1023
+ceiling_div(36028797018963968, 36028797018963968) = 1, ceiling_div(36028797018963970, 36028797018963968) = 2, ceiling_div(18410715276690587646, 36028797018963968) = 511
+ceiling_div(72057594037927936, 72057594037927936) = 1, ceiling_div(72057594037927938, 72057594037927936) = 2, ceiling_div(18374686479671623678, 72057594037927936) = 255
+ceiling_div(144115188075855872, 144115188075855872) = 1, ceiling_div(144115188075855874, 144115188075855872) = 2, ceiling_div(18302628885633695742, 144115188075855872) = 127
+ceiling_div(288230376151711744, 288230376151711744) = 1, ceiling_div(288230376151711746, 288230376151711744) = 2, ceiling_div(18158513697557839870, 288230376151711744) = 63
+ceiling_div(576460752303423488, 576460752303423488) = 1, ceiling_div(576460752303423490, 576460752303423488) = 2, ceiling_div(17870283321406128126, 576460752303423488) = 31
+ceiling_div(1152921504606846976, 1152921504606846976) = 1, ceiling_div(1152921504606846978, 1152921504606846976) = 2, ceiling_div(17293822569102704638, 1152921504606846976) = 15
+ceiling_div(2305843009213693952, 2305843009213693952) = 1, ceiling_div(2305843009213693954, 2305843009213693952) = 2, ceiling_div(16140901064495857662, 2305843009213693952) = 7
+ceiling_div(4611686018427387904, 4611686018427387904) = 1, ceiling_div(4611686018427387906, 4611686018427387904) = 2, ceiling_div(13835058055282163710, 4611686018427387904) = 3
+ceiling_div(9223372036854775808, 9223372036854775808) = 1, ceiling_div(9223372036854775810, 9223372036854775808) = 0, ceiling_div(9223372036854775806, 9223372036854775808) = 1
 ceiling
 signed char
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(-128, -128) = 2, ceiling(-126, -128) = 1, ceiling(126, -128) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = -128, ceiling(-66, 64) = 0
+ceiling(-128, -128) = -128, ceiling(-126, -128) = -128, ceiling(126, -128) = 0
 unsigned char
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(253, 1) = 253
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(252, 2) = 126
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(250, 4) = 63
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(246, 8) = 31
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(238, 16) = 15
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(222, 32) = 7
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(190, 64) = 3
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(126, 128) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(252, 2) = 252
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(250, 4) = 252
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(246, 8) = 248
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(238, 16) = 240
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(222, 32) = 224
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(190, 64) = 192
+ceiling(128, 128) = 128, ceiling(130, 128) = 0, ceiling(126, 128) = 128
 short int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(-32768, -32768) = 2, ceiling(-32766, -32768) = 1, ceiling(32766, -32768) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = -32768, ceiling(-16386, 16384) = 0
+ceiling(-32768, -32768) = -32768, ceiling(-32766, -32768) = -32768, ceiling(32766, -32768) = 0
 unsigned short int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(65533, 1) = 65533
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(65532, 2) = 32766
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(65530, 4) = 16383
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(65526, 8) = 8191
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(65518, 16) = 4095
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(65502, 32) = 2047
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(65470, 64) = 1023
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(65406, 128) = 511
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(65278, 256) = 255
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(65022, 512) = 127
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(64510, 1024) = 63
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(63486, 2048) = 31
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(61438, 4096) = 15
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(57342, 8192) = 7
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(49150, 16384) = 3
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(32766, 32768) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(65532, 2) = 65532
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(65530, 4) = 65532
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(65526, 8) = 65528
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(65518, 16) = 65520
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(65502, 32) = 65504
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(65470, 64) = 65472
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(65406, 128) = 65408
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(65278, 256) = 65280
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(65022, 512) = 65024
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(64510, 1024) = 64512
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(63486, 2048) = 63488
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(61438, 4096) = 61440
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(57342, 8192) = 57344
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(49150, 16384) = 49152
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 0, ceiling(32766, 32768) = 32768
 int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(-32770, 32768) = 0
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(-65538, 65536) = 0
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(-131074, 131072) = 0
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(-262146, 262144) = 0
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(-524290, 524288) = 0
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(-1048578, 1048576) = 0
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(-2097154, 2097152) = 0
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(-4194306, 4194304) = 0
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(-8388610, 8388608) = 0
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(-16777218, 16777216) = 0
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(-33554434, 33554432) = 0
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(-67108866, 67108864) = 0
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(-134217730, 134217728) = 0
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(-268435458, 268435456) = 0
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(-536870914, 536870912) = 0
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = -1, ceiling(-1073741826, 1073741824) = 0
 ceiling(-2147483648, -2147483648) = 0, ceiling(-2147483646, -2147483648) = 0, ceiling(2147483646, -2147483648) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(-16386, 16384) = 0
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(-32770, 32768) = 0
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(-65538, 65536) = 0
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(-131074, 131072) = 0
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(-262146, 262144) = 0
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(-524290, 524288) = 0
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(-1048578, 1048576) = 0
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(-2097154, 2097152) = 0
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(-4194306, 4194304) = 0
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(-8388610, 8388608) = 0
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(-16777218, 16777216) = 0
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(-33554434, 33554432) = 0
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(-67108866, 67108864) = 0
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(-134217730, 134217728) = 0
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(-268435458, 268435456) = 0
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(-536870914, 536870912) = 0
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = -1073741824, ceiling(-1073741826, 1073741824) = 0
+ceiling(-2147483648, -2147483648) = -2147483648, ceiling(-2147483646, -2147483648) = 0, ceiling(2147483646, -2147483648) = 0
 unsigned int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(4294967293, 1) = 4294967293
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(4294967292, 2) = 2147483646
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(4294967290, 4) = 1073741823
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(4294967286, 8) = 536870911
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(4294967278, 16) = 268435455
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(4294967262, 32) = 134217727
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(4294967230, 64) = 67108863
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(4294967166, 128) = 33554431
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(4294967038, 256) = 16777215
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(4294966782, 512) = 8388607
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(4294966270, 1024) = 4194303
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(4294965246, 2048) = 2097151
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(4294963198, 4096) = 1048575
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(4294959102, 8192) = 524287
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(4294950910, 16384) = 262143
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(4294934526, 32768) = 131071
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(4294901758, 65536) = 65535
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(4294836222, 131072) = 32767
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(4294705150, 262144) = 16383
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(4294443006, 524288) = 8191
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(4293918718, 1048576) = 4095
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(4292870142, 2097152) = 2047
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(4290772990, 4194304) = 1023
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(4286578686, 8388608) = 511
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(4278190078, 16777216) = 255
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(4261412862, 33554432) = 127
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(4227858430, 67108864) = 63
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(4160749566, 134217728) = 31
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(4026531838, 268435456) = 15
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(3758096382, 536870912) = 7
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(3221225470, 1073741824) = 3
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 0, ceiling(2147483646, 2147483648) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(4294967292, 2) = 4294967292
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(4294967290, 4) = 4294967292
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(4294967286, 8) = 4294967288
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(4294967278, 16) = 4294967280
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(4294967262, 32) = 4294967264
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(4294967230, 64) = 4294967232
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(4294967166, 128) = 4294967168
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(4294967038, 256) = 4294967040
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(4294966782, 512) = 4294966784
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(4294966270, 1024) = 4294966272
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(4294965246, 2048) = 4294965248
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(4294963198, 4096) = 4294963200
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(4294959102, 8192) = 4294959104
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(4294950910, 16384) = 4294950912
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(4294934526, 32768) = 4294934528
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(4294901758, 65536) = 4294901760
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(4294836222, 131072) = 4294836224
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(4294705150, 262144) = 4294705152
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(4294443006, 524288) = 4294443008
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(4293918718, 1048576) = 4293918720
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(4292870142, 2097152) = 4292870144
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(4290772990, 4194304) = 4290772992
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(4286578686, 8388608) = 4286578688
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(4278190078, 16777216) = 4278190080
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(4261412862, 33554432) = 4261412864
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(4227858430, 67108864) = 4227858432
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(4160749566, 134217728) = 4160749568
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(4026531838, 268435456) = 4026531840
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(3758096382, 536870912) = 3758096384
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(3221225470, 1073741824) = 3221225472
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 0, ceiling(2147483646, 2147483648) = 2147483648
 long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(-32770, 32768) = 0
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(-65538, 65536) = 0
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(-131074, 131072) = 0
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(-262146, 262144) = 0
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(-524290, 524288) = 0
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(-1048578, 1048576) = 0
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(-2097154, 2097152) = 0
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(-4194306, 4194304) = 0
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(-8388610, 8388608) = 0
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(-16777218, 16777216) = 0
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(-33554434, 33554432) = 0
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(-67108866, 67108864) = 0
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(-134217730, 134217728) = 0
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(-268435458, 268435456) = 0
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(-536870914, 536870912) = 0
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(-1073741826, 1073741824) = 0
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 2, ceiling(-2147483650, 2147483648) = 0
 ceiling(4294967296, 4294967296) = 1, ceiling(4294967298, 4294967296) = 2, ceiling(-4294967298, 4294967296) = 0
 ceiling(8589934592, 8589934592) = 1, ceiling(8589934594, 8589934592) = 2, ceiling(-8589934594, 8589934592) = 0
 ceiling(17179869184, 17179869184) = 1, ceiling(17179869186, 17179869184) = 2, ceiling(-17179869186, 17179869184) = 0
 ceiling(34359738368, 34359738368) = 1, ceiling(34359738370, 34359738368) = 2, ceiling(-34359738370, 34359738368) = 0
 ceiling(68719476736, 68719476736) = 1, ceiling(68719476738, 68719476736) = 2, ceiling(-68719476738, 68719476736) = 0
 ceiling(137438953472, 137438953472) = 1, ceiling(137438953474, 137438953472) = 2, ceiling(-137438953474, 137438953472) = 0
 ceiling(274877906944, 274877906944) = 1, ceiling(274877906946, 274877906944) = 2, ceiling(-274877906946, 274877906944) = 0
 ceiling(549755813888, 549755813888) = 1, ceiling(549755813890, 549755813888) = 2, ceiling(-549755813890, 549755813888) = 0
 ceiling(1099511627776, 1099511627776) = 1, ceiling(1099511627778, 1099511627776) = 2, ceiling(-1099511627778, 1099511627776) = 0
 ceiling(2199023255552, 2199023255552) = 1, ceiling(2199023255554, 2199023255552) = 2, ceiling(-2199023255554, 2199023255552) = 0
 ceiling(4398046511104, 4398046511104) = 1, ceiling(4398046511106, 4398046511104) = 2, ceiling(-4398046511106, 4398046511104) = 0
 ceiling(8796093022208, 8796093022208) = 1, ceiling(8796093022210, 8796093022208) = 2, ceiling(-8796093022210, 8796093022208) = 0
 ceiling(17592186044416, 17592186044416) = 1, ceiling(17592186044418, 17592186044416) = 2, ceiling(-17592186044418, 17592186044416) = 0
 ceiling(35184372088832, 35184372088832) = 1, ceiling(35184372088834, 35184372088832) = 2, ceiling(-35184372088834, 35184372088832) = 0
 ceiling(70368744177664, 70368744177664) = 1, ceiling(70368744177666, 70368744177664) = 2, ceiling(-70368744177666, 70368744177664) = 0
 ceiling(140737488355328, 140737488355328) = 1, ceiling(140737488355330, 140737488355328) = 2, ceiling(-140737488355330, 140737488355328) = 0
 ceiling(281474976710656, 281474976710656) = 1, ceiling(281474976710658, 281474976710656) = 2, ceiling(-281474976710658, 281474976710656) = 0
 ceiling(562949953421312, 562949953421312) = 1, ceiling(562949953421314, 562949953421312) = 2, ceiling(-562949953421314, 562949953421312) = 0
 ceiling(1125899906842624, 1125899906842624) = 1, ceiling(1125899906842626, 1125899906842624) = 2, ceiling(-1125899906842626, 1125899906842624) = 0
 ceiling(2251799813685248, 2251799813685248) = 1, ceiling(2251799813685250, 2251799813685248) = 2, ceiling(-2251799813685250, 2251799813685248) = 0
 ceiling(4503599627370496, 4503599627370496) = 1, ceiling(4503599627370498, 4503599627370496) = 2, ceiling(-4503599627370498, 4503599627370496) = 0
 ceiling(9007199254740992, 9007199254740992) = 1, ceiling(9007199254740994, 9007199254740992) = 2, ceiling(-9007199254740994, 9007199254740992) = 0
 ceiling(18014398509481984, 18014398509481984) = 1, ceiling(18014398509481986, 18014398509481984) = 2, ceiling(-18014398509481986, 18014398509481984) = 0
 ceiling(36028797018963968, 36028797018963968) = 1, ceiling(36028797018963970, 36028797018963968) = 2, ceiling(-36028797018963970, 36028797018963968) = 0
 ceiling(72057594037927936, 72057594037927936) = 1, ceiling(72057594037927938, 72057594037927936) = 2, ceiling(-72057594037927938, 72057594037927936) = 0
 ceiling(144115188075855872, 144115188075855872) = 1, ceiling(144115188075855874, 144115188075855872) = 2, ceiling(-144115188075855874, 144115188075855872) = 0
 ceiling(288230376151711744, 288230376151711744) = 1, ceiling(288230376151711746, 288230376151711744) = 2, ceiling(-288230376151711746, 288230376151711744) = 0
 ceiling(576460752303423488, 576460752303423488) = 1, ceiling(576460752303423490, 576460752303423488) = 2, ceiling(-576460752303423490, 576460752303423488) = 0
 ceiling(1152921504606846976, 1152921504606846976) = 1, ceiling(1152921504606846978, 1152921504606846976) = 2, ceiling(-1152921504606846978, 1152921504606846976) = 0
 ceiling(2305843009213693952, 2305843009213693952) = 1, ceiling(2305843009213693954, 2305843009213693952) = 2, ceiling(-2305843009213693954, 2305843009213693952) = 0
 ceiling(4611686018427387904, 4611686018427387904) = 1, ceiling(4611686018427387906, 4611686018427387904) = -1, ceiling(-4611686018427387906, 4611686018427387904) = 0
 ceiling(-9223372036854775808, -9223372036854775808) = 0, ceiling(-9223372036854775806, -9223372036854775808) = 0, ceiling(9223372036854775806, -9223372036854775808) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(-16386, 16384) = 0
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(-32770, 32768) = 0
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(-65538, 65536) = 0
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(-131074, 131072) = 0
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(-262146, 262144) = 0
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(-524290, 524288) = 0
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(-1048578, 1048576) = 0
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(-2097154, 2097152) = 0
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(-4194306, 4194304) = 0
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(-8388610, 8388608) = 0
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(-16777218, 16777216) = 0
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(-33554434, 33554432) = 0
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(-67108866, 67108864) = 0
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(-134217730, 134217728) = 0
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(-268435458, 268435456) = 0
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(-536870914, 536870912) = 0
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(-1073741826, 1073741824) = 0
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 4294967296, ceiling(-2147483650, 2147483648) = 0
+ceiling(4294967296, 4294967296) = 4294967296, ceiling(4294967298, 4294967296) = 8589934592, ceiling(-4294967298, 4294967296) = 0
+ceiling(8589934592, 8589934592) = 8589934592, ceiling(8589934594, 8589934592) = 17179869184, ceiling(-8589934594, 8589934592) = 0
+ceiling(17179869184, 17179869184) = 17179869184, ceiling(17179869186, 17179869184) = 34359738368, ceiling(-17179869186, 17179869184) = 0
+ceiling(34359738368, 34359738368) = 34359738368, ceiling(34359738370, 34359738368) = 68719476736, ceiling(-34359738370, 34359738368) = 0
+ceiling(68719476736, 68719476736) = 68719476736, ceiling(68719476738, 68719476736) = 137438953472, ceiling(-68719476738, 68719476736) = 0
+ceiling(137438953472, 137438953472) = 137438953472, ceiling(137438953474, 137438953472) = 274877906944, ceiling(-137438953474, 137438953472) = 0
+ceiling(274877906944, 274877906944) = 274877906944, ceiling(274877906946, 274877906944) = 549755813888, ceiling(-274877906946, 274877906944) = 0
+ceiling(549755813888, 549755813888) = 549755813888, ceiling(549755813890, 549755813888) = 1099511627776, ceiling(-549755813890, 549755813888) = 0
+ceiling(1099511627776, 1099511627776) = 1099511627776, ceiling(1099511627778, 1099511627776) = 2199023255552, ceiling(-1099511627778, 1099511627776) = 0
+ceiling(2199023255552, 2199023255552) = 2199023255552, ceiling(2199023255554, 2199023255552) = 4398046511104, ceiling(-2199023255554, 2199023255552) = 0
+ceiling(4398046511104, 4398046511104) = 4398046511104, ceiling(4398046511106, 4398046511104) = 8796093022208, ceiling(-4398046511106, 4398046511104) = 0
+ceiling(8796093022208, 8796093022208) = 8796093022208, ceiling(8796093022210, 8796093022208) = 17592186044416, ceiling(-8796093022210, 8796093022208) = 0
+ceiling(17592186044416, 17592186044416) = 17592186044416, ceiling(17592186044418, 17592186044416) = 35184372088832, ceiling(-17592186044418, 17592186044416) = 0
+ceiling(35184372088832, 35184372088832) = 35184372088832, ceiling(35184372088834, 35184372088832) = 70368744177664, ceiling(-35184372088834, 35184372088832) = 0
+ceiling(70368744177664, 70368744177664) = 70368744177664, ceiling(70368744177666, 70368744177664) = 140737488355328, ceiling(-70368744177666, 70368744177664) = 0
+ceiling(140737488355328, 140737488355328) = 140737488355328, ceiling(140737488355330, 140737488355328) = 281474976710656, ceiling(-140737488355330, 140737488355328) = 0
+ceiling(281474976710656, 281474976710656) = 281474976710656, ceiling(281474976710658, 281474976710656) = 562949953421312, ceiling(-281474976710658, 281474976710656) = 0
+ceiling(562949953421312, 562949953421312) = 562949953421312, ceiling(562949953421314, 562949953421312) = 1125899906842624, ceiling(-562949953421314, 562949953421312) = 0
+ceiling(1125899906842624, 1125899906842624) = 1125899906842624, ceiling(1125899906842626, 1125899906842624) = 2251799813685248, ceiling(-1125899906842626, 1125899906842624) = 0
+ceiling(2251799813685248, 2251799813685248) = 2251799813685248, ceiling(2251799813685250, 2251799813685248) = 4503599627370496, ceiling(-2251799813685250, 2251799813685248) = 0
+ceiling(4503599627370496, 4503599627370496) = 4503599627370496, ceiling(4503599627370498, 4503599627370496) = 9007199254740992, ceiling(-4503599627370498, 4503599627370496) = 0
+ceiling(9007199254740992, 9007199254740992) = 9007199254740992, ceiling(9007199254740994, 9007199254740992) = 18014398509481984, ceiling(-9007199254740994, 9007199254740992) = 0
+ceiling(18014398509481984, 18014398509481984) = 18014398509481984, ceiling(18014398509481986, 18014398509481984) = 36028797018963968, ceiling(-18014398509481986, 18014398509481984) = 0
+ceiling(36028797018963968, 36028797018963968) = 36028797018963968, ceiling(36028797018963970, 36028797018963968) = 72057594037927936, ceiling(-36028797018963970, 36028797018963968) = 0
+ceiling(72057594037927936, 72057594037927936) = 72057594037927936, ceiling(72057594037927938, 72057594037927936) = 144115188075855872, ceiling(-72057594037927938, 72057594037927936) = 0
+ceiling(144115188075855872, 144115188075855872) = 144115188075855872, ceiling(144115188075855874, 144115188075855872) = 288230376151711744, ceiling(-144115188075855874, 144115188075855872) = 0
+ceiling(288230376151711744, 288230376151711744) = 288230376151711744, ceiling(288230376151711746, 288230376151711744) = 576460752303423488, ceiling(-288230376151711746, 288230376151711744) = 0
+ceiling(576460752303423488, 576460752303423488) = 576460752303423488, ceiling(576460752303423490, 576460752303423488) = 1152921504606846976, ceiling(-576460752303423490, 576460752303423488) = 0
+ceiling(1152921504606846976, 1152921504606846976) = 1152921504606846976, ceiling(1152921504606846978, 1152921504606846976) = 2305843009213693952, ceiling(-1152921504606846978, 1152921504606846976) = 0
+ceiling(2305843009213693952, 2305843009213693952) = 2305843009213693952, ceiling(2305843009213693954, 2305843009213693952) = 4611686018427387904, ceiling(-2305843009213693954, 2305843009213693952) = 0
+ceiling(4611686018427387904, 4611686018427387904) = 4611686018427387904, ceiling(4611686018427387906, 4611686018427387904) = -4611686018427387904, ceiling(-4611686018427387906, 4611686018427387904) = 0
+ceiling(-9223372036854775808, -9223372036854775808) = -9223372036854775808, ceiling(-9223372036854775806, -9223372036854775808) = 0, ceiling(9223372036854775806, -9223372036854775808) = 0
 unsigned long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(18446744073709551613, 1) = 18446744073709551613
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(18446744073709551612, 2) = 9223372036854775806
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(18446744073709551610, 4) = 4611686018427387903
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(18446744073709551606, 8) = 2305843009213693951
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(18446744073709551598, 16) = 1152921504606846975
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(18446744073709551582, 32) = 576460752303423487
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(18446744073709551550, 64) = 288230376151711743
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(18446744073709551486, 128) = 144115188075855871
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(18446744073709551358, 256) = 72057594037927935
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(18446744073709551102, 512) = 36028797018963967
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(18446744073709550590, 1024) = 18014398509481983
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(18446744073709549566, 2048) = 9007199254740991
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(18446744073709547518, 4096) = 4503599627370495
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(18446744073709543422, 8192) = 2251799813685247
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(18446744073709535230, 16384) = 1125899906842623
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(18446744073709518846, 32768) = 562949953421311
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(18446744073709486078, 65536) = 281474976710655
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(18446744073709420542, 131072) = 140737488355327
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(18446744073709289470, 262144) = 70368744177663
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(18446744073709027326, 524288) = 35184372088831
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(18446744073708503038, 1048576) = 17592186044415
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(18446744073707454462, 2097152) = 8796093022207
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(18446744073705357310, 4194304) = 4398046511103
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(18446744073701163006, 8388608) = 2199023255551
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(18446744073692774398, 16777216) = 1099511627775
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(18446744073675997182, 33554432) = 549755813887
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(18446744073642442750, 67108864) = 274877906943
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(18446744073575333886, 134217728) = 137438953471
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(18446744073441116158, 268435456) = 68719476735
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(18446744073172680702, 536870912) = 34359738367
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(18446744072635809790, 1073741824) = 17179869183
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 2, ceiling(18446744071562067966, 2147483648) = 8589934591
 ceiling(4294967296, 4294967296) = 1, ceiling(4294967298, 4294967296) = 2, ceiling(18446744069414584318, 4294967296) = 4294967295
 ceiling(8589934592, 8589934592) = 1, ceiling(8589934594, 8589934592) = 2, ceiling(18446744065119617022, 8589934592) = 2147483647
 ceiling(17179869184, 17179869184) = 1, ceiling(17179869186, 17179869184) = 2, ceiling(18446744056529682430, 17179869184) = 1073741823
 ceiling(34359738368, 34359738368) = 1, ceiling(34359738370, 34359738368) = 2, ceiling(18446744039349813246, 34359738368) = 536870911
 ceiling(68719476736, 68719476736) = 1, ceiling(68719476738, 68719476736) = 2, ceiling(18446744004990074878, 68719476736) = 268435455
 ceiling(137438953472, 137438953472) = 1, ceiling(137438953474, 137438953472) = 2, ceiling(18446743936270598142, 137438953472) = 134217727
 ceiling(274877906944, 274877906944) = 1, ceiling(274877906946, 274877906944) = 2, ceiling(18446743798831644670, 274877906944) = 67108863
 ceiling(549755813888, 549755813888) = 1, ceiling(549755813890, 549755813888) = 2, ceiling(18446743523953737726, 549755813888) = 33554431
 ceiling(1099511627776, 1099511627776) = 1, ceiling(1099511627778, 1099511627776) = 2, ceiling(18446742974197923838, 1099511627776) = 16777215
 ceiling(2199023255552, 2199023255552) = 1, ceiling(2199023255554, 2199023255552) = 2, ceiling(18446741874686296062, 2199023255552) = 8388607
 ceiling(4398046511104, 4398046511104) = 1, ceiling(4398046511106, 4398046511104) = 2, ceiling(18446739675663040510, 4398046511104) = 4194303
 ceiling(8796093022208, 8796093022208) = 1, ceiling(8796093022210, 8796093022208) = 2, ceiling(18446735277616529406, 8796093022208) = 2097151
 ceiling(17592186044416, 17592186044416) = 1, ceiling(17592186044418, 17592186044416) = 2, ceiling(18446726481523507198, 17592186044416) = 1048575
 ceiling(35184372088832, 35184372088832) = 1, ceiling(35184372088834, 35184372088832) = 2, ceiling(18446708889337462782, 35184372088832) = 524287
 ceiling(70368744177664, 70368744177664) = 1, ceiling(70368744177666, 70368744177664) = 2, ceiling(18446673704965373950, 70368744177664) = 262143
 ceiling(140737488355328, 140737488355328) = 1, ceiling(140737488355330, 140737488355328) = 2, ceiling(18446603336221196286, 140737488355328) = 131071
 ceiling(281474976710656, 281474976710656) = 1, ceiling(281474976710658, 281474976710656) = 2, ceiling(18446462598732840958, 281474976710656) = 65535
 ceiling(562949953421312, 562949953421312) = 1, ceiling(562949953421314, 562949953421312) = 2, ceiling(18446181123756130302, 562949953421312) = 32767
 ceiling(1125899906842624, 1125899906842624) = 1, ceiling(1125899906842626, 1125899906842624) = 2, ceiling(18445618173802708990, 1125899906842624) = 16383
 ceiling(2251799813685248, 2251799813685248) = 1, ceiling(2251799813685250, 2251799813685248) = 2, ceiling(18444492273895866366, 2251799813685248) = 8191
 ceiling(4503599627370496, 4503599627370496) = 1, ceiling(4503599627370498, 4503599627370496) = 2, ceiling(18442240474082181118, 4503599627370496) = 4095
 ceiling(9007199254740992, 9007199254740992) = 1, ceiling(9007199254740994, 9007199254740992) = 2, ceiling(18437736874454810622, 9007199254740992) = 2047
 ceiling(18014398509481984, 18014398509481984) = 1, ceiling(18014398509481986, 18014398509481984) = 2, ceiling(18428729675200069630, 18014398509481984) = 1023
 ceiling(36028797018963968, 36028797018963968) = 1, ceiling(36028797018963970, 36028797018963968) = 2, ceiling(18410715276690587646, 36028797018963968) = 511
 ceiling(72057594037927936, 72057594037927936) = 1, ceiling(72057594037927938, 72057594037927936) = 2, ceiling(18374686479671623678, 72057594037927936) = 255
 ceiling(144115188075855872, 144115188075855872) = 1, ceiling(144115188075855874, 144115188075855872) = 2, ceiling(18302628885633695742, 144115188075855872) = 127
 ceiling(288230376151711744, 288230376151711744) = 1, ceiling(288230376151711746, 288230376151711744) = 2, ceiling(18158513697557839870, 288230376151711744) = 63
 ceiling(576460752303423488, 576460752303423488) = 1, ceiling(576460752303423490, 576460752303423488) = 2, ceiling(17870283321406128126, 576460752303423488) = 31
 ceiling(1152921504606846976, 1152921504606846976) = 1, ceiling(1152921504606846978, 1152921504606846976) = 2, ceiling(17293822569102704638, 1152921504606846976) = 15
 ceiling(2305843009213693952, 2305843009213693952) = 1, ceiling(2305843009213693954, 2305843009213693952) = 2, ceiling(16140901064495857662, 2305843009213693952) = 7
 ceiling(4611686018427387904, 4611686018427387904) = 1, ceiling(4611686018427387906, 4611686018427387904) = 2, ceiling(13835058055282163710, 4611686018427387904) = 3
 ceiling(9223372036854775808, 9223372036854775808) = 1, ceiling(9223372036854775810, 9223372036854775808) = 0, ceiling(9223372036854775806, 9223372036854775808) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(18446744073709551612, 2) = 18446744073709551612
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(18446744073709551610, 4) = 18446744073709551612
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(18446744073709551606, 8) = 18446744073709551608
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(18446744073709551598, 16) = 18446744073709551600
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(18446744073709551582, 32) = 18446744073709551584
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(18446744073709551550, 64) = 18446744073709551552
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(18446744073709551486, 128) = 18446744073709551488
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(18446744073709551358, 256) = 18446744073709551360
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(18446744073709551102, 512) = 18446744073709551104
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(18446744073709550590, 1024) = 18446744073709550592
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(18446744073709549566, 2048) = 18446744073709549568
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(18446744073709547518, 4096) = 18446744073709547520
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(18446744073709543422, 8192) = 18446744073709543424
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(18446744073709535230, 16384) = 18446744073709535232
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(18446744073709518846, 32768) = 18446744073709518848
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(18446744073709486078, 65536) = 18446744073709486080
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(18446744073709420542, 131072) = 18446744073709420544
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(18446744073709289470, 262144) = 18446744073709289472
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(18446744073709027326, 524288) = 18446744073709027328
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(18446744073708503038, 1048576) = 18446744073708503040
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(18446744073707454462, 2097152) = 18446744073707454464
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(18446744073705357310, 4194304) = 18446744073705357312
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(18446744073701163006, 8388608) = 18446744073701163008
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(18446744073692774398, 16777216) = 18446744073692774400
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(18446744073675997182, 33554432) = 18446744073675997184
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(18446744073642442750, 67108864) = 18446744073642442752
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(18446744073575333886, 134217728) = 18446744073575333888
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(18446744073441116158, 268435456) = 18446744073441116160
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(18446744073172680702, 536870912) = 18446744073172680704
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(18446744072635809790, 1073741824) = 18446744072635809792
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 4294967296, ceiling(18446744071562067966, 2147483648) = 18446744071562067968
+ceiling(4294967296, 4294967296) = 4294967296, ceiling(4294967298, 4294967296) = 8589934592, ceiling(18446744069414584318, 4294967296) = 18446744069414584320
+ceiling(8589934592, 8589934592) = 8589934592, ceiling(8589934594, 8589934592) = 17179869184, ceiling(18446744065119617022, 8589934592) = 18446744065119617024
+ceiling(17179869184, 17179869184) = 17179869184, ceiling(17179869186, 17179869184) = 34359738368, ceiling(18446744056529682430, 17179869184) = 18446744056529682432
+ceiling(34359738368, 34359738368) = 34359738368, ceiling(34359738370, 34359738368) = 68719476736, ceiling(18446744039349813246, 34359738368) = 18446744039349813248
+ceiling(68719476736, 68719476736) = 68719476736, ceiling(68719476738, 68719476736) = 137438953472, ceiling(18446744004990074878, 68719476736) = 18446744004990074880
+ceiling(137438953472, 137438953472) = 137438953472, ceiling(137438953474, 137438953472) = 274877906944, ceiling(18446743936270598142, 137438953472) = 18446743936270598144
+ceiling(274877906944, 274877906944) = 274877906944, ceiling(274877906946, 274877906944) = 549755813888, ceiling(18446743798831644670, 274877906944) = 18446743798831644672
+ceiling(549755813888, 549755813888) = 549755813888, ceiling(549755813890, 549755813888) = 1099511627776, ceiling(18446743523953737726, 549755813888) = 18446743523953737728
+ceiling(1099511627776, 1099511627776) = 1099511627776, ceiling(1099511627778, 1099511627776) = 2199023255552, ceiling(18446742974197923838, 1099511627776) = 18446742974197923840
+ceiling(2199023255552, 2199023255552) = 2199023255552, ceiling(2199023255554, 2199023255552) = 4398046511104, ceiling(18446741874686296062, 2199023255552) = 18446741874686296064
+ceiling(4398046511104, 4398046511104) = 4398046511104, ceiling(4398046511106, 4398046511104) = 8796093022208, ceiling(18446739675663040510, 4398046511104) = 18446739675663040512
+ceiling(8796093022208, 8796093022208) = 8796093022208, ceiling(8796093022210, 8796093022208) = 17592186044416, ceiling(18446735277616529406, 8796093022208) = 18446735277616529408
+ceiling(17592186044416, 17592186044416) = 17592186044416, ceiling(17592186044418, 17592186044416) = 35184372088832, ceiling(18446726481523507198, 17592186044416) = 18446726481523507200
+ceiling(35184372088832, 35184372088832) = 35184372088832, ceiling(35184372088834, 35184372088832) = 70368744177664, ceiling(18446708889337462782, 35184372088832) = 18446708889337462784
+ceiling(70368744177664, 70368744177664) = 70368744177664, ceiling(70368744177666, 70368744177664) = 140737488355328, ceiling(18446673704965373950, 70368744177664) = 18446673704965373952
+ceiling(140737488355328, 140737488355328) = 140737488355328, ceiling(140737488355330, 140737488355328) = 281474976710656, ceiling(18446603336221196286, 140737488355328) = 18446603336221196288
+ceiling(281474976710656, 281474976710656) = 281474976710656, ceiling(281474976710658, 281474976710656) = 562949953421312, ceiling(18446462598732840958, 281474976710656) = 18446462598732840960
+ceiling(562949953421312, 562949953421312) = 562949953421312, ceiling(562949953421314, 562949953421312) = 1125899906842624, ceiling(18446181123756130302, 562949953421312) = 18446181123756130304
+ceiling(1125899906842624, 1125899906842624) = 1125899906842624, ceiling(1125899906842626, 1125899906842624) = 2251799813685248, ceiling(18445618173802708990, 1125899906842624) = 18445618173802708992
+ceiling(2251799813685248, 2251799813685248) = 2251799813685248, ceiling(2251799813685250, 2251799813685248) = 4503599627370496, ceiling(18444492273895866366, 2251799813685248) = 18444492273895866368
+ceiling(4503599627370496, 4503599627370496) = 4503599627370496, ceiling(4503599627370498, 4503599627370496) = 9007199254740992, ceiling(18442240474082181118, 4503599627370496) = 18442240474082181120
+ceiling(9007199254740992, 9007199254740992) = 9007199254740992, ceiling(9007199254740994, 9007199254740992) = 18014398509481984, ceiling(18437736874454810622, 9007199254740992) = 18437736874454810624
+ceiling(18014398509481984, 18014398509481984) = 18014398509481984, ceiling(18014398509481986, 18014398509481984) = 36028797018963968, ceiling(18428729675200069630, 18014398509481984) = 18428729675200069632
+ceiling(36028797018963968, 36028797018963968) = 36028797018963968, ceiling(36028797018963970, 36028797018963968) = 72057594037927936, ceiling(18410715276690587646, 36028797018963968) = 18410715276690587648
+ceiling(72057594037927936, 72057594037927936) = 72057594037927936, ceiling(72057594037927938, 72057594037927936) = 144115188075855872, ceiling(18374686479671623678, 72057594037927936) = 18374686479671623680
+ceiling(144115188075855872, 144115188075855872) = 144115188075855872, ceiling(144115188075855874, 144115188075855872) = 288230376151711744, ceiling(18302628885633695742, 144115188075855872) = 18302628885633695744
+ceiling(288230376151711744, 288230376151711744) = 288230376151711744, ceiling(288230376151711746, 288230376151711744) = 576460752303423488, ceiling(18158513697557839870, 288230376151711744) = 18158513697557839872
+ceiling(576460752303423488, 576460752303423488) = 576460752303423488, ceiling(576460752303423490, 576460752303423488) = 1152921504606846976, ceiling(17870283321406128126, 576460752303423488) = 17870283321406128128
+ceiling(1152921504606846976, 1152921504606846976) = 1152921504606846976, ceiling(1152921504606846978, 1152921504606846976) = 2305843009213693952, ceiling(17293822569102704638, 1152921504606846976) = 17293822569102704640
+ceiling(2305843009213693952, 2305843009213693952) = 2305843009213693952, ceiling(2305843009213693954, 2305843009213693952) = 4611686018427387904, ceiling(16140901064495857662, 2305843009213693952) = 16140901064495857664
+ceiling(4611686018427387904, 4611686018427387904) = 4611686018427387904, ceiling(4611686018427387906, 4611686018427387904) = 9223372036854775808, ceiling(13835058055282163710, 4611686018427387904) = 13835058055282163712
+ceiling(9223372036854775808, 9223372036854775808) = 9223372036854775808, ceiling(9223372036854775810, 9223372036854775808) = 0, ceiling(9223372036854775806, 9223372036854775808) = 9223372036854775808
 long long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(-32770, 32768) = 0
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(-65538, 65536) = 0
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(-131074, 131072) = 0
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(-262146, 262144) = 0
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(-524290, 524288) = 0
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(-1048578, 1048576) = 0
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(-2097154, 2097152) = 0
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(-4194306, 4194304) = 0
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(-8388610, 8388608) = 0
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(-16777218, 16777216) = 0
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(-33554434, 33554432) = 0
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(-67108866, 67108864) = 0
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(-134217730, 134217728) = 0
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(-268435458, 268435456) = 0
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(-536870914, 536870912) = 0
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(-1073741826, 1073741824) = 0
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 2, ceiling(-2147483650, 2147483648) = 0
 ceiling(4294967296, 4294967296) = 1, ceiling(4294967298, 4294967296) = 2, ceiling(-4294967298, 4294967296) = 0
 ceiling(8589934592, 8589934592) = 1, ceiling(8589934594, 8589934592) = 2, ceiling(-8589934594, 8589934592) = 0
 ceiling(17179869184, 17179869184) = 1, ceiling(17179869186, 17179869184) = 2, ceiling(-17179869186, 17179869184) = 0
 ceiling(34359738368, 34359738368) = 1, ceiling(34359738370, 34359738368) = 2, ceiling(-34359738370, 34359738368) = 0
 ceiling(68719476736, 68719476736) = 1, ceiling(68719476738, 68719476736) = 2, ceiling(-68719476738, 68719476736) = 0
 ceiling(137438953472, 137438953472) = 1, ceiling(137438953474, 137438953472) = 2, ceiling(-137438953474, 137438953472) = 0
 ceiling(274877906944, 274877906944) = 1, ceiling(274877906946, 274877906944) = 2, ceiling(-274877906946, 274877906944) = 0
 ceiling(549755813888, 549755813888) = 1, ceiling(549755813890, 549755813888) = 2, ceiling(-549755813890, 549755813888) = 0
 ceiling(1099511627776, 1099511627776) = 1, ceiling(1099511627778, 1099511627776) = 2, ceiling(-1099511627778, 1099511627776) = 0
 ceiling(2199023255552, 2199023255552) = 1, ceiling(2199023255554, 2199023255552) = 2, ceiling(-2199023255554, 2199023255552) = 0
 ceiling(4398046511104, 4398046511104) = 1, ceiling(4398046511106, 4398046511104) = 2, ceiling(-4398046511106, 4398046511104) = 0
 ceiling(8796093022208, 8796093022208) = 1, ceiling(8796093022210, 8796093022208) = 2, ceiling(-8796093022210, 8796093022208) = 0
 ceiling(17592186044416, 17592186044416) = 1, ceiling(17592186044418, 17592186044416) = 2, ceiling(-17592186044418, 17592186044416) = 0
 ceiling(35184372088832, 35184372088832) = 1, ceiling(35184372088834, 35184372088832) = 2, ceiling(-35184372088834, 35184372088832) = 0
 ceiling(70368744177664, 70368744177664) = 1, ceiling(70368744177666, 70368744177664) = 2, ceiling(-70368744177666, 70368744177664) = 0
 ceiling(140737488355328, 140737488355328) = 1, ceiling(140737488355330, 140737488355328) = 2, ceiling(-140737488355330, 140737488355328) = 0
 ceiling(281474976710656, 281474976710656) = 1, ceiling(281474976710658, 281474976710656) = 2, ceiling(-281474976710658, 281474976710656) = 0
 ceiling(562949953421312, 562949953421312) = 1, ceiling(562949953421314, 562949953421312) = 2, ceiling(-562949953421314, 562949953421312) = 0
 ceiling(1125899906842624, 1125899906842624) = 1, ceiling(1125899906842626, 1125899906842624) = 2, ceiling(-1125899906842626, 1125899906842624) = 0
 ceiling(2251799813685248, 2251799813685248) = 1, ceiling(2251799813685250, 2251799813685248) = 2, ceiling(-2251799813685250, 2251799813685248) = 0
 ceiling(4503599627370496, 4503599627370496) = 1, ceiling(4503599627370498, 4503599627370496) = 2, ceiling(-4503599627370498, 4503599627370496) = 0
 ceiling(9007199254740992, 9007199254740992) = 1, ceiling(9007199254740994, 9007199254740992) = 2, ceiling(-9007199254740994, 9007199254740992) = 0
 ceiling(18014398509481984, 18014398509481984) = 1, ceiling(18014398509481986, 18014398509481984) = 2, ceiling(-18014398509481986, 18014398509481984) = 0
 ceiling(36028797018963968, 36028797018963968) = 1, ceiling(36028797018963970, 36028797018963968) = 2, ceiling(-36028797018963970, 36028797018963968) = 0
 ceiling(72057594037927936, 72057594037927936) = 1, ceiling(72057594037927938, 72057594037927936) = 2, ceiling(-72057594037927938, 72057594037927936) = 0
 ceiling(144115188075855872, 144115188075855872) = 1, ceiling(144115188075855874, 144115188075855872) = 2, ceiling(-144115188075855874, 144115188075855872) = 0
 ceiling(288230376151711744, 288230376151711744) = 1, ceiling(288230376151711746, 288230376151711744) = 2, ceiling(-288230376151711746, 288230376151711744) = 0
 ceiling(576460752303423488, 576460752303423488) = 1, ceiling(576460752303423490, 576460752303423488) = 2, ceiling(-576460752303423490, 576460752303423488) = 0
 ceiling(1152921504606846976, 1152921504606846976) = 1, ceiling(1152921504606846978, 1152921504606846976) = 2, ceiling(-1152921504606846978, 1152921504606846976) = 0
 ceiling(2305843009213693952, 2305843009213693952) = 1, ceiling(2305843009213693954, 2305843009213693952) = 2, ceiling(-2305843009213693954, 2305843009213693952) = 0
 ceiling(4611686018427387904, 4611686018427387904) = 1, ceiling(4611686018427387906, 4611686018427387904) = -1, ceiling(-4611686018427387906, 4611686018427387904) = 0
 ceiling(-9223372036854775808, -9223372036854775808) = 0, ceiling(-9223372036854775806, -9223372036854775808) = 0, ceiling(9223372036854775806, -9223372036854775808) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(-16386, 16384) = 0
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(-32770, 32768) = 0
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(-65538, 65536) = 0
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(-131074, 131072) = 0
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(-262146, 262144) = 0
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(-524290, 524288) = 0
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(-1048578, 1048576) = 0
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(-2097154, 2097152) = 0
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(-4194306, 4194304) = 0
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(-8388610, 8388608) = 0
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(-16777218, 16777216) = 0
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(-33554434, 33554432) = 0
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(-67108866, 67108864) = 0
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(-134217730, 134217728) = 0
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(-268435458, 268435456) = 0
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(-536870914, 536870912) = 0
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(-1073741826, 1073741824) = 0
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 4294967296, ceiling(-2147483650, 2147483648) = 0
+ceiling(4294967296, 4294967296) = 4294967296, ceiling(4294967298, 4294967296) = 8589934592, ceiling(-4294967298, 4294967296) = 0
+ceiling(8589934592, 8589934592) = 8589934592, ceiling(8589934594, 8589934592) = 17179869184, ceiling(-8589934594, 8589934592) = 0
+ceiling(17179869184, 17179869184) = 17179869184, ceiling(17179869186, 17179869184) = 34359738368, ceiling(-17179869186, 17179869184) = 0
+ceiling(34359738368, 34359738368) = 34359738368, ceiling(34359738370, 34359738368) = 68719476736, ceiling(-34359738370, 34359738368) = 0
+ceiling(68719476736, 68719476736) = 68719476736, ceiling(68719476738, 68719476736) = 137438953472, ceiling(-68719476738, 68719476736) = 0
+ceiling(137438953472, 137438953472) = 137438953472, ceiling(137438953474, 137438953472) = 274877906944, ceiling(-137438953474, 137438953472) = 0
+ceiling(274877906944, 274877906944) = 274877906944, ceiling(274877906946, 274877906944) = 549755813888, ceiling(-274877906946, 274877906944) = 0
+ceiling(549755813888, 549755813888) = 549755813888, ceiling(549755813890, 549755813888) = 1099511627776, ceiling(-549755813890, 549755813888) = 0
+ceiling(1099511627776, 1099511627776) = 1099511627776, ceiling(1099511627778, 1099511627776) = 2199023255552, ceiling(-1099511627778, 1099511627776) = 0
+ceiling(2199023255552, 2199023255552) = 2199023255552, ceiling(2199023255554, 2199023255552) = 4398046511104, ceiling(-2199023255554, 2199023255552) = 0
+ceiling(4398046511104, 4398046511104) = 4398046511104, ceiling(4398046511106, 4398046511104) = 8796093022208, ceiling(-4398046511106, 4398046511104) = 0
+ceiling(8796093022208, 8796093022208) = 8796093022208, ceiling(8796093022210, 8796093022208) = 17592186044416, ceiling(-8796093022210, 8796093022208) = 0
+ceiling(17592186044416, 17592186044416) = 17592186044416, ceiling(17592186044418, 17592186044416) = 35184372088832, ceiling(-17592186044418, 17592186044416) = 0
+ceiling(35184372088832, 35184372088832) = 35184372088832, ceiling(35184372088834, 35184372088832) = 70368744177664, ceiling(-35184372088834, 35184372088832) = 0
+ceiling(70368744177664, 70368744177664) = 70368744177664, ceiling(70368744177666, 70368744177664) = 140737488355328, ceiling(-70368744177666, 70368744177664) = 0
+ceiling(140737488355328, 140737488355328) = 140737488355328, ceiling(140737488355330, 140737488355328) = 281474976710656, ceiling(-140737488355330, 140737488355328) = 0
+ceiling(281474976710656, 281474976710656) = 281474976710656, ceiling(281474976710658, 281474976710656) = 562949953421312, ceiling(-281474976710658, 281474976710656) = 0
+ceiling(562949953421312, 562949953421312) = 562949953421312, ceiling(562949953421314, 562949953421312) = 1125899906842624, ceiling(-562949953421314, 562949953421312) = 0
+ceiling(1125899906842624, 1125899906842624) = 1125899906842624, ceiling(1125899906842626, 1125899906842624) = 2251799813685248, ceiling(-1125899906842626, 1125899906842624) = 0
+ceiling(2251799813685248, 2251799813685248) = 2251799813685248, ceiling(2251799813685250, 2251799813685248) = 4503599627370496, ceiling(-2251799813685250, 2251799813685248) = 0
+ceiling(4503599627370496, 4503599627370496) = 4503599627370496, ceiling(4503599627370498, 4503599627370496) = 9007199254740992, ceiling(-4503599627370498, 4503599627370496) = 0
+ceiling(9007199254740992, 9007199254740992) = 9007199254740992, ceiling(9007199254740994, 9007199254740992) = 18014398509481984, ceiling(-9007199254740994, 9007199254740992) = 0
+ceiling(18014398509481984, 18014398509481984) = 18014398509481984, ceiling(18014398509481986, 18014398509481984) = 36028797018963968, ceiling(-18014398509481986, 18014398509481984) = 0
+ceiling(36028797018963968, 36028797018963968) = 36028797018963968, ceiling(36028797018963970, 36028797018963968) = 72057594037927936, ceiling(-36028797018963970, 36028797018963968) = 0
+ceiling(72057594037927936, 72057594037927936) = 72057594037927936, ceiling(72057594037927938, 72057594037927936) = 144115188075855872, ceiling(-72057594037927938, 72057594037927936) = 0
+ceiling(144115188075855872, 144115188075855872) = 144115188075855872, ceiling(144115188075855874, 144115188075855872) = 288230376151711744, ceiling(-144115188075855874, 144115188075855872) = 0
+ceiling(288230376151711744, 288230376151711744) = 288230376151711744, ceiling(288230376151711746, 288230376151711744) = 576460752303423488, ceiling(-288230376151711746, 288230376151711744) = 0
+ceiling(576460752303423488, 576460752303423488) = 576460752303423488, ceiling(576460752303423490, 576460752303423488) = 1152921504606846976, ceiling(-576460752303423490, 576460752303423488) = 0
+ceiling(1152921504606846976, 1152921504606846976) = 1152921504606846976, ceiling(1152921504606846978, 1152921504606846976) = 2305843009213693952, ceiling(-1152921504606846978, 1152921504606846976) = 0
+ceiling(2305843009213693952, 2305843009213693952) = 2305843009213693952, ceiling(2305843009213693954, 2305843009213693952) = 4611686018427387904, ceiling(-2305843009213693954, 2305843009213693952) = 0
+ceiling(4611686018427387904, 4611686018427387904) = 4611686018427387904, ceiling(4611686018427387906, 4611686018427387904) = -4611686018427387904, ceiling(-4611686018427387906, 4611686018427387904) = 0
+ceiling(-9223372036854775808, -9223372036854775808) = -9223372036854775808, ceiling(-9223372036854775806, -9223372036854775808) = 0, ceiling(9223372036854775806, -9223372036854775808) = 0
 unsigned long long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(18446744073709551613, 1) = 18446744073709551613
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(18446744073709551612, 2) = 9223372036854775806
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(18446744073709551610, 4) = 4611686018427387903
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(18446744073709551606, 8) = 2305843009213693951
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(18446744073709551598, 16) = 1152921504606846975
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(18446744073709551582, 32) = 576460752303423487
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(18446744073709551550, 64) = 288230376151711743
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(18446744073709551486, 128) = 144115188075855871
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(18446744073709551358, 256) = 72057594037927935
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(18446744073709551102, 512) = 36028797018963967
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(18446744073709550590, 1024) = 18014398509481983
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(18446744073709549566, 2048) = 9007199254740991
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(18446744073709547518, 4096) = 4503599627370495
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(18446744073709543422, 8192) = 2251799813685247
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(18446744073709535230, 16384) = 1125899906842623
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(18446744073709518846, 32768) = 562949953421311
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(18446744073709486078, 65536) = 281474976710655
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(18446744073709420542, 131072) = 140737488355327
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(18446744073709289470, 262144) = 70368744177663
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(18446744073709027326, 524288) = 35184372088831
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(18446744073708503038, 1048576) = 17592186044415
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(18446744073707454462, 2097152) = 8796093022207
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(18446744073705357310, 4194304) = 4398046511103
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(18446744073701163006, 8388608) = 2199023255551
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(18446744073692774398, 16777216) = 1099511627775
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(18446744073675997182, 33554432) = 549755813887
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(18446744073642442750, 67108864) = 274877906943
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(18446744073575333886, 134217728) = 137438953471
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(18446744073441116158, 268435456) = 68719476735
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(18446744073172680702, 536870912) = 34359738367
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(18446744072635809790, 1073741824) = 17179869183
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 2, ceiling(18446744071562067966, 2147483648) = 8589934591
 ceiling(4294967296, 4294967296) = 1, ceiling(4294967298, 4294967296) = 2, ceiling(18446744069414584318, 4294967296) = 4294967295
 ceiling(8589934592, 8589934592) = 1, ceiling(8589934594, 8589934592) = 2, ceiling(18446744065119617022, 8589934592) = 2147483647
 ceiling(17179869184, 17179869184) = 1, ceiling(17179869186, 17179869184) = 2, ceiling(18446744056529682430, 17179869184) = 1073741823
 ceiling(34359738368, 34359738368) = 1, ceiling(34359738370, 34359738368) = 2, ceiling(18446744039349813246, 34359738368) = 536870911
 ceiling(68719476736, 68719476736) = 1, ceiling(68719476738, 68719476736) = 2, ceiling(18446744004990074878, 68719476736) = 268435455
 ceiling(137438953472, 137438953472) = 1, ceiling(137438953474, 137438953472) = 2, ceiling(18446743936270598142, 137438953472) = 134217727
 ceiling(274877906944, 274877906944) = 1, ceiling(274877906946, 274877906944) = 2, ceiling(18446743798831644670, 274877906944) = 67108863
 ceiling(549755813888, 549755813888) = 1, ceiling(549755813890, 549755813888) = 2, ceiling(18446743523953737726, 549755813888) = 33554431
 ceiling(1099511627776, 1099511627776) = 1, ceiling(1099511627778, 1099511627776) = 2, ceiling(18446742974197923838, 1099511627776) = 16777215
 ceiling(2199023255552, 2199023255552) = 1, ceiling(2199023255554, 2199023255552) = 2, ceiling(18446741874686296062, 2199023255552) = 8388607
 ceiling(4398046511104, 4398046511104) = 1, ceiling(4398046511106, 4398046511104) = 2, ceiling(18446739675663040510, 4398046511104) = 4194303
 ceiling(8796093022208, 8796093022208) = 1, ceiling(8796093022210, 8796093022208) = 2, ceiling(18446735277616529406, 8796093022208) = 2097151
 ceiling(17592186044416, 17592186044416) = 1, ceiling(17592186044418, 17592186044416) = 2, ceiling(18446726481523507198, 17592186044416) = 1048575
 ceiling(35184372088832, 35184372088832) = 1, ceiling(35184372088834, 35184372088832) = 2, ceiling(18446708889337462782, 35184372088832) = 524287
 ceiling(70368744177664, 70368744177664) = 1, ceiling(70368744177666, 70368744177664) = 2, ceiling(18446673704965373950, 70368744177664) = 262143
 ceiling(140737488355328, 140737488355328) = 1, ceiling(140737488355330, 140737488355328) = 2, ceiling(18446603336221196286, 140737488355328) = 131071
 ceiling(281474976710656, 281474976710656) = 1, ceiling(281474976710658, 281474976710656) = 2, ceiling(18446462598732840958, 281474976710656) = 65535
 ceiling(562949953421312, 562949953421312) = 1, ceiling(562949953421314, 562949953421312) = 2, ceiling(18446181123756130302, 562949953421312) = 32767
 ceiling(1125899906842624, 1125899906842624) = 1, ceiling(1125899906842626, 1125899906842624) = 2, ceiling(18445618173802708990, 1125899906842624) = 16383
 ceiling(2251799813685248, 2251799813685248) = 1, ceiling(2251799813685250, 2251799813685248) = 2, ceiling(18444492273895866366, 2251799813685248) = 8191
 ceiling(4503599627370496, 4503599627370496) = 1, ceiling(4503599627370498, 4503599627370496) = 2, ceiling(18442240474082181118, 4503599627370496) = 4095
 ceiling(9007199254740992, 9007199254740992) = 1, ceiling(9007199254740994, 9007199254740992) = 2, ceiling(18437736874454810622, 9007199254740992) = 2047
 ceiling(18014398509481984, 18014398509481984) = 1, ceiling(18014398509481986, 18014398509481984) = 2, ceiling(18428729675200069630, 18014398509481984) = 1023
 ceiling(36028797018963968, 36028797018963968) = 1, ceiling(36028797018963970, 36028797018963968) = 2, ceiling(18410715276690587646, 36028797018963968) = 511
 ceiling(72057594037927936, 72057594037927936) = 1, ceiling(72057594037927938, 72057594037927936) = 2, ceiling(18374686479671623678, 72057594037927936) = 255
 ceiling(144115188075855872, 144115188075855872) = 1, ceiling(144115188075855874, 144115188075855872) = 2, ceiling(18302628885633695742, 144115188075855872) = 127
 ceiling(288230376151711744, 288230376151711744) = 1, ceiling(288230376151711746, 288230376151711744) = 2, ceiling(18158513697557839870, 288230376151711744) = 63
 ceiling(576460752303423488, 576460752303423488) = 1, ceiling(576460752303423490, 576460752303423488) = 2, ceiling(17870283321406128126, 576460752303423488) = 31
 ceiling(1152921504606846976, 1152921504606846976) = 1, ceiling(1152921504606846978, 1152921504606846976) = 2, ceiling(17293822569102704638, 1152921504606846976) = 15
 ceiling(2305843009213693952, 2305843009213693952) = 1, ceiling(2305843009213693954, 2305843009213693952) = 2, ceiling(16140901064495857662, 2305843009213693952) = 7
 ceiling(4611686018427387904, 4611686018427387904) = 1, ceiling(4611686018427387906, 4611686018427387904) = 2, ceiling(13835058055282163710, 4611686018427387904) = 3
 ceiling(9223372036854775808, 9223372036854775808) = 1, ceiling(9223372036854775810, 9223372036854775808) = 0, ceiling(9223372036854775806, 9223372036854775808) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(18446744073709551612, 2) = 18446744073709551612
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(18446744073709551610, 4) = 18446744073709551612
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(18446744073709551606, 8) = 18446744073709551608
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(18446744073709551598, 16) = 18446744073709551600
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(18446744073709551582, 32) = 18446744073709551584
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(18446744073709551550, 64) = 18446744073709551552
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(18446744073709551486, 128) = 18446744073709551488
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(18446744073709551358, 256) = 18446744073709551360
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(18446744073709551102, 512) = 18446744073709551104
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(18446744073709550590, 1024) = 18446744073709550592
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(18446744073709549566, 2048) = 18446744073709549568
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(18446744073709547518, 4096) = 18446744073709547520
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(18446744073709543422, 8192) = 18446744073709543424
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(18446744073709535230, 16384) = 18446744073709535232
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(18446744073709518846, 32768) = 18446744073709518848
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(18446744073709486078, 65536) = 18446744073709486080
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(18446744073709420542, 131072) = 18446744073709420544
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(18446744073709289470, 262144) = 18446744073709289472
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(18446744073709027326, 524288) = 18446744073709027328
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(18446744073708503038, 1048576) = 18446744073708503040
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(18446744073707454462, 2097152) = 18446744073707454464
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(18446744073705357310, 4194304) = 18446744073705357312
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(18446744073701163006, 8388608) = 18446744073701163008
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(18446744073692774398, 16777216) = 18446744073692774400
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(18446744073675997182, 33554432) = 18446744073675997184
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(18446744073642442750, 67108864) = 18446744073642442752
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(18446744073575333886, 134217728) = 18446744073575333888
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(18446744073441116158, 268435456) = 18446744073441116160
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(18446744073172680702, 536870912) = 18446744073172680704
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(18446744072635809790, 1073741824) = 18446744072635809792
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 4294967296, ceiling(18446744071562067966, 2147483648) = 18446744071562067968
+ceiling(4294967296, 4294967296) = 4294967296, ceiling(4294967298, 4294967296) = 8589934592, ceiling(18446744069414584318, 4294967296) = 18446744069414584320
+ceiling(8589934592, 8589934592) = 8589934592, ceiling(8589934594, 8589934592) = 17179869184, ceiling(18446744065119617022, 8589934592) = 18446744065119617024
+ceiling(17179869184, 17179869184) = 17179869184, ceiling(17179869186, 17179869184) = 34359738368, ceiling(18446744056529682430, 17179869184) = 18446744056529682432
+ceiling(34359738368, 34359738368) = 34359738368, ceiling(34359738370, 34359738368) = 68719476736, ceiling(18446744039349813246, 34359738368) = 18446744039349813248
+ceiling(68719476736, 68719476736) = 68719476736, ceiling(68719476738, 68719476736) = 137438953472, ceiling(18446744004990074878, 68719476736) = 18446744004990074880
+ceiling(137438953472, 137438953472) = 137438953472, ceiling(137438953474, 137438953472) = 274877906944, ceiling(18446743936270598142, 137438953472) = 18446743936270598144
+ceiling(274877906944, 274877906944) = 274877906944, ceiling(274877906946, 274877906944) = 549755813888, ceiling(18446743798831644670, 274877906944) = 18446743798831644672
+ceiling(549755813888, 549755813888) = 549755813888, ceiling(549755813890, 549755813888) = 1099511627776, ceiling(18446743523953737726, 549755813888) = 18446743523953737728
+ceiling(1099511627776, 1099511627776) = 1099511627776, ceiling(1099511627778, 1099511627776) = 2199023255552, ceiling(18446742974197923838, 1099511627776) = 18446742974197923840
+ceiling(2199023255552, 2199023255552) = 2199023255552, ceiling(2199023255554, 2199023255552) = 4398046511104, ceiling(18446741874686296062, 2199023255552) = 18446741874686296064
+ceiling(4398046511104, 4398046511104) = 4398046511104, ceiling(4398046511106, 4398046511104) = 8796093022208, ceiling(18446739675663040510, 4398046511104) = 18446739675663040512
+ceiling(8796093022208, 8796093022208) = 8796093022208, ceiling(8796093022210, 8796093022208) = 17592186044416, ceiling(18446735277616529406, 8796093022208) = 18446735277616529408
+ceiling(17592186044416, 17592186044416) = 17592186044416, ceiling(17592186044418, 17592186044416) = 35184372088832, ceiling(18446726481523507198, 17592186044416) = 18446726481523507200
+ceiling(35184372088832, 35184372088832) = 35184372088832, ceiling(35184372088834, 35184372088832) = 70368744177664, ceiling(18446708889337462782, 35184372088832) = 18446708889337462784
+ceiling(70368744177664, 70368744177664) = 70368744177664, ceiling(70368744177666, 70368744177664) = 140737488355328, ceiling(18446673704965373950, 70368744177664) = 18446673704965373952
+ceiling(140737488355328, 140737488355328) = 140737488355328, ceiling(140737488355330, 140737488355328) = 281474976710656, ceiling(18446603336221196286, 140737488355328) = 18446603336221196288
+ceiling(281474976710656, 281474976710656) = 281474976710656, ceiling(281474976710658, 281474976710656) = 562949953421312, ceiling(18446462598732840958, 281474976710656) = 18446462598732840960
+ceiling(562949953421312, 562949953421312) = 562949953421312, ceiling(562949953421314, 562949953421312) = 1125899906842624, ceiling(18446181123756130302, 562949953421312) = 18446181123756130304
+ceiling(1125899906842624, 1125899906842624) = 1125899906842624, ceiling(1125899906842626, 1125899906842624) = 2251799813685248, ceiling(18445618173802708990, 1125899906842624) = 18445618173802708992
+ceiling(2251799813685248, 2251799813685248) = 2251799813685248, ceiling(2251799813685250, 2251799813685248) = 4503599627370496, ceiling(18444492273895866366, 2251799813685248) = 18444492273895866368
+ceiling(4503599627370496, 4503599627370496) = 4503599627370496, ceiling(4503599627370498, 4503599627370496) = 9007199254740992, ceiling(18442240474082181118, 4503599627370496) = 18442240474082181120
+ceiling(9007199254740992, 9007199254740992) = 9007199254740992, ceiling(9007199254740994, 9007199254740992) = 18014398509481984, ceiling(18437736874454810622, 9007199254740992) = 18437736874454810624
+ceiling(18014398509481984, 18014398509481984) = 18014398509481984, ceiling(18014398509481986, 18014398509481984) = 36028797018963968, ceiling(18428729675200069630, 18014398509481984) = 18428729675200069632
+ceiling(36028797018963968, 36028797018963968) = 36028797018963968, ceiling(36028797018963970, 36028797018963968) = 72057594037927936, ceiling(18410715276690587646, 36028797018963968) = 18410715276690587648
+ceiling(72057594037927936, 72057594037927936) = 72057594037927936, ceiling(72057594037927938, 72057594037927936) = 144115188075855872, ceiling(18374686479671623678, 72057594037927936) = 18374686479671623680
+ceiling(144115188075855872, 144115188075855872) = 144115188075855872, ceiling(144115188075855874, 144115188075855872) = 288230376151711744, ceiling(18302628885633695742, 144115188075855872) = 18302628885633695744
+ceiling(288230376151711744, 288230376151711744) = 288230376151711744, ceiling(288230376151711746, 288230376151711744) = 576460752303423488, ceiling(18158513697557839870, 288230376151711744) = 18158513697557839872
+ceiling(576460752303423488, 576460752303423488) = 576460752303423488, ceiling(576460752303423490, 576460752303423488) = 1152921504606846976, ceiling(17870283321406128126, 576460752303423488) = 17870283321406128128
+ceiling(1152921504606846976, 1152921504606846976) = 1152921504606846976, ceiling(1152921504606846978, 1152921504606846976) = 2305843009213693952, ceiling(17293822569102704638, 1152921504606846976) = 17293822569102704640
+ceiling(2305843009213693952, 2305843009213693952) = 2305843009213693952, ceiling(2305843009213693954, 2305843009213693952) = 4611686018427387904, ceiling(16140901064495857662, 2305843009213693952) = 16140901064495857664
+ceiling(4611686018427387904, 4611686018427387904) = 4611686018427387904, ceiling(4611686018427387906, 4611686018427387904) = 9223372036854775808, ceiling(13835058055282163710, 4611686018427387904) = 13835058055282163712
+ceiling(9223372036854775808, 9223372036854775808) = 9223372036854775808, ceiling(9223372036854775810, 9223372036854775808) = 0, ceiling(9223372036854775806, 9223372036854775808) = 9223372036854775808

tests/.expect/bitmanip3.x86.txt

-              r07d867b
+              r22f94a4
 signed char
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned char
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(254, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(253, 1) = 253
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(252, 2) = 252
 …
 short int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned short int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(65534, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(65533, 1) = 65533
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(65532, 2) = 65532
 …
 int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(4294967294, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(4294967293, 1) = 4294967293
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(4294967292, 2) = 4294967292
 …
 long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(4294967294, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(4294967293, 1) = 4294967293
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(4294967292, 2) = 4294967292
 …
 long long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(-2, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(-3, 1) = -3
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(-4, 2) = -4
 …
 unsigned long long int
-floor2(0, 0) = 0, floor2(2, 0) = 0, floor2(18446744073709551614, 0) = 0
 floor2(1, 1) = 1, floor2(3, 1) = 3, floor2(18446744073709551613, 1) = 18446744073709551613
 floor2(2, 2) = 2, floor2(4, 2) = 4, floor2(18446744073709551612, 2) = 18446744073709551612
 …
 signed char
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned char
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(254, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(253, 1) = 253
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(252, 2) = 252
 …
 short int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned short int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(65534, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(65533, 1) = 65533
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(65532, 2) = 65532
 …
 int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(4294967294, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(4294967293, 1) = 4294967293
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(4294967292, 2) = 4294967292
 …
 long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(4294967294, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(4294967293, 1) = 4294967293
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(4294967292, 2) = 4294967292
 …
 long long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(-2, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(-3, 1) = -3
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(-4, 2) = -4
 …
 unsigned long long int
-ceiling2(0, 0) = 0, ceiling2(2, 0) = 0, ceiling2(18446744073709551614, 0) = 0
 ceiling2(1, 1) = 1, ceiling2(3, 1) = 3, ceiling2(18446744073709551613, 1) = 18446744073709551613
 ceiling2(2, 2) = 2, ceiling2(4, 2) = 4, ceiling2(18446744073709551612, 2) = 18446744073709551612
 …
+ceiling_div
+signed char
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(-128, -128) = 2, ceiling_div(-126, -128) = 1, ceiling_div(126, -128) = 0
+unsigned char
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(253, 1) = 253
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(252, 2) = 126
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(250, 4) = 63
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(246, 8) = 31
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(238, 16) = 15
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(222, 32) = 7
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(190, 64) = 3
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(126, 128) = 1
+short int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(-32768, -32768) = 2, ceiling_div(-32766, -32768) = 1, ceiling_div(32766, -32768) = 0
+unsigned short int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(65533, 1) = 65533
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(65532, 2) = 32766
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(65530, 4) = 16383
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(65526, 8) = 8191
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(65518, 16) = 4095
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(65502, 32) = 2047
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(65470, 64) = 1023
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(65406, 128) = 511
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(65278, 256) = 255
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(65022, 512) = 127
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(64510, 1024) = 63
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(63486, 2048) = 31
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(61438, 4096) = 15
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(57342, 8192) = 7
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(49150, 16384) = 3
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(32766, 32768) = 1
+int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(-32770, 32768) = 0
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(-65538, 65536) = 0
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(-131074, 131072) = 0
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(-262146, 262144) = 0
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(-524290, 524288) = 0
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(-1048578, 1048576) = 0
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(-2097154, 2097152) = 0
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(-4194306, 4194304) = 0
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(-8388610, 8388608) = 0
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(-16777218, 16777216) = 0
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(-33554434, 33554432) = 0
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(-67108866, 67108864) = 0
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(-134217730, 134217728) = 0
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(-268435458, 268435456) = 0
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(-536870914, 536870912) = 0
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = -1, ceiling_div(-1073741826, 1073741824) = 0
+ceiling_div(-2147483648, -2147483648) = 0, ceiling_div(-2147483646, -2147483648) = 0, ceiling_div(2147483646, -2147483648) = 0
+unsigned int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(4294967293, 1) = 4294967293
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(4294967292, 2) = 2147483646
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(4294967290, 4) = 1073741823
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(4294967286, 8) = 536870911
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(4294967278, 16) = 268435455
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(4294967262, 32) = 134217727
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(4294967230, 64) = 67108863
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(4294967166, 128) = 33554431
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(4294967038, 256) = 16777215
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(4294966782, 512) = 8388607
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(4294966270, 1024) = 4194303
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(4294965246, 2048) = 2097151
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(4294963198, 4096) = 1048575
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(4294959102, 8192) = 524287
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(4294950910, 16384) = 262143
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(4294934526, 32768) = 131071
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(4294901758, 65536) = 65535
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(4294836222, 131072) = 32767
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(4294705150, 262144) = 16383
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(4294443006, 524288) = 8191
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(4293918718, 1048576) = 4095
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(4292870142, 2097152) = 2047
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(4290772990, 4194304) = 1023
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(4286578686, 8388608) = 511
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(4278190078, 16777216) = 255
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(4261412862, 33554432) = 127
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(4227858430, 67108864) = 63
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(4160749566, 134217728) = 31
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(4026531838, 268435456) = 15
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(3758096382, 536870912) = 7
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(3221225470, 1073741824) = 3
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 0, ceiling_div(2147483646, 2147483648) = 1
+long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(-32770, 32768) = 0
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(-65538, 65536) = 0
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(-131074, 131072) = 0
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(-262146, 262144) = 0
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(-524290, 524288) = 0
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(-1048578, 1048576) = 0
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(-2097154, 2097152) = 0
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(-4194306, 4194304) = 0
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(-8388610, 8388608) = 0
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(-16777218, 16777216) = 0
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(-33554434, 33554432) = 0
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(-67108866, 67108864) = 0
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(-134217730, 134217728) = 0
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(-268435458, 268435456) = 0
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(-536870914, 536870912) = 0
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = -1, ceiling_div(-1073741826, 1073741824) = 0
+ceiling_div(-2147483648, -2147483648) = 0, ceiling_div(-2147483646, -2147483648) = 0, ceiling_div(2147483646, -2147483648) = 0
+unsigned long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(4294967293, 1) = 4294967293
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(4294967292, 2) = 2147483646
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(4294967290, 4) = 1073741823
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(4294967286, 8) = 536870911
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(4294967278, 16) = 268435455
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(4294967262, 32) = 134217727
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(4294967230, 64) = 67108863
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(4294967166, 128) = 33554431
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(4294967038, 256) = 16777215
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(4294966782, 512) = 8388607
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(4294966270, 1024) = 4194303
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(4294965246, 2048) = 2097151
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(4294963198, 4096) = 1048575
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(4294959102, 8192) = 524287
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(4294950910, 16384) = 262143
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(4294934526, 32768) = 131071
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(4294901758, 65536) = 65535
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(4294836222, 131072) = 32767
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(4294705150, 262144) = 16383
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(4294443006, 524288) = 8191
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(4293918718, 1048576) = 4095
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(4292870142, 2097152) = 2047
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(4290772990, 4194304) = 1023
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(4286578686, 8388608) = 511
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(4278190078, 16777216) = 255
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(4261412862, 33554432) = 127
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(4227858430, 67108864) = 63
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(4160749566, 134217728) = 31
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(4026531838, 268435456) = 15
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(3758096382, 536870912) = 7
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(3221225470, 1073741824) = 3
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 0, ceiling_div(2147483646, 2147483648) = 1
+long long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(-3, 1) = -3
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(-4, 2) = -1
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(-6, 4) = 0
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(-10, 8) = 0
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(-18, 16) = 0
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(-34, 32) = 0
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(-66, 64) = 0
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(-130, 128) = 0
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(-258, 256) = 0
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(-514, 512) = 0
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(-1026, 1024) = 0
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(-2050, 2048) = 0
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(-4098, 4096) = 0
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(-8194, 8192) = 0
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(-16386, 16384) = 0
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(-32770, 32768) = 0
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(-65538, 65536) = 0
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(-131074, 131072) = 0
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(-262146, 262144) = 0
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(-524290, 524288) = 0
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(-1048578, 1048576) = 0
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(-2097154, 2097152) = 0
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(-4194306, 4194304) = 0
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(-8388610, 8388608) = 0
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(-16777218, 16777216) = 0
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(-33554434, 33554432) = 0
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(-67108866, 67108864) = 0
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(-134217730, 134217728) = 0
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(-268435458, 268435456) = 0
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(-536870914, 536870912) = 0
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(-1073741826, 1073741824) = 0
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 2, ceiling_div(-2147483650, 2147483648) = 0
+ceiling_div(4294967296, 4294967296) = 1, ceiling_div(4294967298, 4294967296) = 2, ceiling_div(-4294967298, 4294967296) = 0
+ceiling_div(8589934592, 8589934592) = 1, ceiling_div(8589934594, 8589934592) = 2, ceiling_div(-8589934594, 8589934592) = 0
+ceiling_div(17179869184, 17179869184) = 1, ceiling_div(17179869186, 17179869184) = 2, ceiling_div(-17179869186, 17179869184) = 0
+ceiling_div(34359738368, 34359738368) = 1, ceiling_div(34359738370, 34359738368) = 2, ceiling_div(-34359738370, 34359738368) = 0
+ceiling_div(68719476736, 68719476736) = 1, ceiling_div(68719476738, 68719476736) = 2, ceiling_div(-68719476738, 68719476736) = 0
+ceiling_div(137438953472, 137438953472) = 1, ceiling_div(137438953474, 137438953472) = 2, ceiling_div(-137438953474, 137438953472) = 0
+ceiling_div(274877906944, 274877906944) = 1, ceiling_div(274877906946, 274877906944) = 2, ceiling_div(-274877906946, 274877906944) = 0
+ceiling_div(549755813888, 549755813888) = 1, ceiling_div(549755813890, 549755813888) = 2, ceiling_div(-549755813890, 549755813888) = 0
+ceiling_div(1099511627776, 1099511627776) = 1, ceiling_div(1099511627778, 1099511627776) = 2, ceiling_div(-1099511627778, 1099511627776) = 0
+ceiling_div(2199023255552, 2199023255552) = 1, ceiling_div(2199023255554, 2199023255552) = 2, ceiling_div(-2199023255554, 2199023255552) = 0
+ceiling_div(4398046511104, 4398046511104) = 1, ceiling_div(4398046511106, 4398046511104) = 2, ceiling_div(-4398046511106, 4398046511104) = 0
+ceiling_div(8796093022208, 8796093022208) = 1, ceiling_div(8796093022210, 8796093022208) = 2, ceiling_div(-8796093022210, 8796093022208) = 0
+ceiling_div(17592186044416, 17592186044416) = 1, ceiling_div(17592186044418, 17592186044416) = 2, ceiling_div(-17592186044418, 17592186044416) = 0
+ceiling_div(35184372088832, 35184372088832) = 1, ceiling_div(35184372088834, 35184372088832) = 2, ceiling_div(-35184372088834, 35184372088832) = 0
+ceiling_div(70368744177664, 70368744177664) = 1, ceiling_div(70368744177666, 70368744177664) = 2, ceiling_div(-70368744177666, 70368744177664) = 0
+ceiling_div(140737488355328, 140737488355328) = 1, ceiling_div(140737488355330, 140737488355328) = 2, ceiling_div(-140737488355330, 140737488355328) = 0
+ceiling_div(281474976710656, 281474976710656) = 1, ceiling_div(281474976710658, 281474976710656) = 2, ceiling_div(-281474976710658, 281474976710656) = 0
+ceiling_div(562949953421312, 562949953421312) = 1, ceiling_div(562949953421314, 562949953421312) = 2, ceiling_div(-562949953421314, 562949953421312) = 0
+ceiling_div(1125899906842624, 1125899906842624) = 1, ceiling_div(1125899906842626, 1125899906842624) = 2, ceiling_div(-1125899906842626, 1125899906842624) = 0
+ceiling_div(2251799813685248, 2251799813685248) = 1, ceiling_div(2251799813685250, 2251799813685248) = 2, ceiling_div(-2251799813685250, 2251799813685248) = 0
+ceiling_div(4503599627370496, 4503599627370496) = 1, ceiling_div(4503599627370498, 4503599627370496) = 2, ceiling_div(-4503599627370498, 4503599627370496) = 0
+ceiling_div(9007199254740992, 9007199254740992) = 1, ceiling_div(9007199254740994, 9007199254740992) = 2, ceiling_div(-9007199254740994, 9007199254740992) = 0
+ceiling_div(18014398509481984, 18014398509481984) = 1, ceiling_div(18014398509481986, 18014398509481984) = 2, ceiling_div(-18014398509481986, 18014398509481984) = 0
+ceiling_div(36028797018963968, 36028797018963968) = 1, ceiling_div(36028797018963970, 36028797018963968) = 2, ceiling_div(-36028797018963970, 36028797018963968) = 0
+ceiling_div(72057594037927936, 72057594037927936) = 1, ceiling_div(72057594037927938, 72057594037927936) = 2, ceiling_div(-72057594037927938, 72057594037927936) = 0
+ceiling_div(144115188075855872, 144115188075855872) = 1, ceiling_div(144115188075855874, 144115188075855872) = 2, ceiling_div(-144115188075855874, 144115188075855872) = 0
+ceiling_div(288230376151711744, 288230376151711744) = 1, ceiling_div(288230376151711746, 288230376151711744) = 2, ceiling_div(-288230376151711746, 288230376151711744) = 0
+ceiling_div(576460752303423488, 576460752303423488) = 1, ceiling_div(576460752303423490, 576460752303423488) = 2, ceiling_div(-576460752303423490, 576460752303423488) = 0
+ceiling_div(1152921504606846976, 1152921504606846976) = 1, ceiling_div(1152921504606846978, 1152921504606846976) = 2, ceiling_div(-1152921504606846978, 1152921504606846976) = 0
+ceiling_div(2305843009213693952, 2305843009213693952) = 1, ceiling_div(2305843009213693954, 2305843009213693952) = 2, ceiling_div(-2305843009213693954, 2305843009213693952) = 0
+ceiling_div(4611686018427387904, 4611686018427387904) = 1, ceiling_div(4611686018427387906, 4611686018427387904) = -1, ceiling_div(-4611686018427387906, 4611686018427387904) = 0
+ceiling_div(-9223372036854775808, -9223372036854775808) = 0, ceiling_div(-9223372036854775806, -9223372036854775808) = 0, ceiling_div(9223372036854775806, -9223372036854775808) = 0
+unsigned long long int
+ceiling_div(1, 1) = 1, ceiling_div(3, 1) = 3, ceiling_div(18446744073709551613, 1) = 18446744073709551613
+ceiling_div(2, 2) = 1, ceiling_div(4, 2) = 2, ceiling_div(18446744073709551612, 2) = 9223372036854775806
+ceiling_div(4, 4) = 1, ceiling_div(6, 4) = 2, ceiling_div(18446744073709551610, 4) = 4611686018427387903
+ceiling_div(8, 8) = 1, ceiling_div(10, 8) = 2, ceiling_div(18446744073709551606, 8) = 2305843009213693951
+ceiling_div(16, 16) = 1, ceiling_div(18, 16) = 2, ceiling_div(18446744073709551598, 16) = 1152921504606846975
+ceiling_div(32, 32) = 1, ceiling_div(34, 32) = 2, ceiling_div(18446744073709551582, 32) = 576460752303423487
+ceiling_div(64, 64) = 1, ceiling_div(66, 64) = 2, ceiling_div(18446744073709551550, 64) = 288230376151711743
+ceiling_div(128, 128) = 1, ceiling_div(130, 128) = 2, ceiling_div(18446744073709551486, 128) = 144115188075855871
+ceiling_div(256, 256) = 1, ceiling_div(258, 256) = 2, ceiling_div(18446744073709551358, 256) = 72057594037927935
+ceiling_div(512, 512) = 1, ceiling_div(514, 512) = 2, ceiling_div(18446744073709551102, 512) = 36028797018963967
+ceiling_div(1024, 1024) = 1, ceiling_div(1026, 1024) = 2, ceiling_div(18446744073709550590, 1024) = 18014398509481983
+ceiling_div(2048, 2048) = 1, ceiling_div(2050, 2048) = 2, ceiling_div(18446744073709549566, 2048) = 9007199254740991
+ceiling_div(4096, 4096) = 1, ceiling_div(4098, 4096) = 2, ceiling_div(18446744073709547518, 4096) = 4503599627370495
+ceiling_div(8192, 8192) = 1, ceiling_div(8194, 8192) = 2, ceiling_div(18446744073709543422, 8192) = 2251799813685247
+ceiling_div(16384, 16384) = 1, ceiling_div(16386, 16384) = 2, ceiling_div(18446744073709535230, 16384) = 1125899906842623
+ceiling_div(32768, 32768) = 1, ceiling_div(32770, 32768) = 2, ceiling_div(18446744073709518846, 32768) = 562949953421311
+ceiling_div(65536, 65536) = 1, ceiling_div(65538, 65536) = 2, ceiling_div(18446744073709486078, 65536) = 281474976710655
+ceiling_div(131072, 131072) = 1, ceiling_div(131074, 131072) = 2, ceiling_div(18446744073709420542, 131072) = 140737488355327
+ceiling_div(262144, 262144) = 1, ceiling_div(262146, 262144) = 2, ceiling_div(18446744073709289470, 262144) = 70368744177663
+ceiling_div(524288, 524288) = 1, ceiling_div(524290, 524288) = 2, ceiling_div(18446744073709027326, 524288) = 35184372088831
+ceiling_div(1048576, 1048576) = 1, ceiling_div(1048578, 1048576) = 2, ceiling_div(18446744073708503038, 1048576) = 17592186044415
+ceiling_div(2097152, 2097152) = 1, ceiling_div(2097154, 2097152) = 2, ceiling_div(18446744073707454462, 2097152) = 8796093022207
+ceiling_div(4194304, 4194304) = 1, ceiling_div(4194306, 4194304) = 2, ceiling_div(18446744073705357310, 4194304) = 4398046511103
+ceiling_div(8388608, 8388608) = 1, ceiling_div(8388610, 8388608) = 2, ceiling_div(18446744073701163006, 8388608) = 2199023255551
+ceiling_div(16777216, 16777216) = 1, ceiling_div(16777218, 16777216) = 2, ceiling_div(18446744073692774398, 16777216) = 1099511627775
+ceiling_div(33554432, 33554432) = 1, ceiling_div(33554434, 33554432) = 2, ceiling_div(18446744073675997182, 33554432) = 549755813887
+ceiling_div(67108864, 67108864) = 1, ceiling_div(67108866, 67108864) = 2, ceiling_div(18446744073642442750, 67108864) = 274877906943
+ceiling_div(134217728, 134217728) = 1, ceiling_div(134217730, 134217728) = 2, ceiling_div(18446744073575333886, 134217728) = 137438953471
+ceiling_div(268435456, 268435456) = 1, ceiling_div(268435458, 268435456) = 2, ceiling_div(18446744073441116158, 268435456) = 68719476735
+ceiling_div(536870912, 536870912) = 1, ceiling_div(536870914, 536870912) = 2, ceiling_div(18446744073172680702, 536870912) = 34359738367
+ceiling_div(1073741824, 1073741824) = 1, ceiling_div(1073741826, 1073741824) = 2, ceiling_div(18446744072635809790, 1073741824) = 17179869183
+ceiling_div(2147483648, 2147483648) = 1, ceiling_div(2147483650, 2147483648) = 2, ceiling_div(18446744071562067966, 2147483648) = 8589934591
+ceiling_div(4294967296, 4294967296) = 1, ceiling_div(4294967298, 4294967296) = 2, ceiling_div(18446744069414584318, 4294967296) = 4294967295
+ceiling_div(8589934592, 8589934592) = 1, ceiling_div(8589934594, 8589934592) = 2, ceiling_div(18446744065119617022, 8589934592) = 2147483647
+ceiling_div(17179869184, 17179869184) = 1, ceiling_div(17179869186, 17179869184) = 2, ceiling_div(18446744056529682430, 17179869184) = 1073741823
+ceiling_div(34359738368, 34359738368) = 1, ceiling_div(34359738370, 34359738368) = 2, ceiling_div(18446744039349813246, 34359738368) = 536870911
+ceiling_div(68719476736, 68719476736) = 1, ceiling_div(68719476738, 68719476736) = 2, ceiling_div(18446744004990074878, 68719476736) = 268435455
+ceiling_div(137438953472, 137438953472) = 1, ceiling_div(137438953474, 137438953472) = 2, ceiling_div(18446743936270598142, 137438953472) = 134217727
+ceiling_div(274877906944, 274877906944) = 1, ceiling_div(274877906946, 274877906944) = 2, ceiling_div(18446743798831644670, 274877906944) = 67108863
+ceiling_div(549755813888, 549755813888) = 1, ceiling_div(549755813890, 549755813888) = 2, ceiling_div(18446743523953737726, 549755813888) = 33554431
+ceiling_div(1099511627776, 1099511627776) = 1, ceiling_div(1099511627778, 1099511627776) = 2, ceiling_div(18446742974197923838, 1099511627776) = 16777215
+ceiling_div(2199023255552, 2199023255552) = 1, ceiling_div(2199023255554, 2199023255552) = 2, ceiling_div(18446741874686296062, 2199023255552) = 8388607
+ceiling_div(4398046511104, 4398046511104) = 1, ceiling_div(4398046511106, 4398046511104) = 2, ceiling_div(18446739675663040510, 4398046511104) = 4194303
+ceiling_div(8796093022208, 8796093022208) = 1, ceiling_div(8796093022210, 8796093022208) = 2, ceiling_div(18446735277616529406, 8796093022208) = 2097151
+ceiling_div(17592186044416, 17592186044416) = 1, ceiling_div(17592186044418, 17592186044416) = 2, ceiling_div(18446726481523507198, 17592186044416) = 1048575
+ceiling_div(35184372088832, 35184372088832) = 1, ceiling_div(35184372088834, 35184372088832) = 2, ceiling_div(18446708889337462782, 35184372088832) = 524287
+ceiling_div(70368744177664, 70368744177664) = 1, ceiling_div(70368744177666, 70368744177664) = 2, ceiling_div(18446673704965373950, 70368744177664) = 262143
+ceiling_div(140737488355328, 140737488355328) = 1, ceiling_div(140737488355330, 140737488355328) = 2, ceiling_div(18446603336221196286, 140737488355328) = 131071
+ceiling_div(281474976710656, 281474976710656) = 1, ceiling_div(281474976710658, 281474976710656) = 2, ceiling_div(18446462598732840958, 281474976710656) = 65535
+ceiling_div(562949953421312, 562949953421312) = 1, ceiling_div(562949953421314, 562949953421312) = 2, ceiling_div(18446181123756130302, 562949953421312) = 32767
+ceiling_div(1125899906842624, 1125899906842624) = 1, ceiling_div(1125899906842626, 1125899906842624) = 2, ceiling_div(18445618173802708990, 1125899906842624) = 16383
+ceiling_div(2251799813685248, 2251799813685248) = 1, ceiling_div(2251799813685250, 2251799813685248) = 2, ceiling_div(18444492273895866366, 2251799813685248) = 8191
+ceiling_div(4503599627370496, 4503599627370496) = 1, ceiling_div(4503599627370498, 4503599627370496) = 2, ceiling_div(18442240474082181118, 4503599627370496) = 4095
+ceiling_div(9007199254740992, 9007199254740992) = 1, ceiling_div(9007199254740994, 9007199254740992) = 2, ceiling_div(18437736874454810622, 9007199254740992) = 2047
+ceiling_div(18014398509481984, 18014398509481984) = 1, ceiling_div(18014398509481986, 18014398509481984) = 2, ceiling_div(18428729675200069630, 18014398509481984) = 1023
+ceiling_div(36028797018963968, 36028797018963968) = 1, ceiling_div(36028797018963970, 36028797018963968) = 2, ceiling_div(18410715276690587646, 36028797018963968) = 511
+ceiling_div(72057594037927936, 72057594037927936) = 1, ceiling_div(72057594037927938, 72057594037927936) = 2, ceiling_div(18374686479671623678, 72057594037927936) = 255
+ceiling_div(144115188075855872, 144115188075855872) = 1, ceiling_div(144115188075855874, 144115188075855872) = 2, ceiling_div(18302628885633695742, 144115188075855872) = 127
+ceiling_div(288230376151711744, 288230376151711744) = 1, ceiling_div(288230376151711746, 288230376151711744) = 2, ceiling_div(18158513697557839870, 288230376151711744) = 63
+ceiling_div(576460752303423488, 576460752303423488) = 1, ceiling_div(576460752303423490, 576460752303423488) = 2, ceiling_div(17870283321406128126, 576460752303423488) = 31
+ceiling_div(1152921504606846976, 1152921504606846976) = 1, ceiling_div(1152921504606846978, 1152921504606846976) = 2, ceiling_div(17293822569102704638, 1152921504606846976) = 15
+ceiling_div(2305843009213693952, 2305843009213693952) = 1, ceiling_div(2305843009213693954, 2305843009213693952) = 2, ceiling_div(16140901064495857662, 2305843009213693952) = 7
+ceiling_div(4611686018427387904, 4611686018427387904) = 1, ceiling_div(4611686018427387906, 4611686018427387904) = 2, ceiling_div(13835058055282163710, 4611686018427387904) = 3
+ceiling_div(9223372036854775808, 9223372036854775808) = 1, ceiling_div(9223372036854775810, 9223372036854775808) = 0, ceiling_div(9223372036854775806, 9223372036854775808) = 1
 ceiling
 signed char
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(-128, -128) = 2, ceiling(-126, -128) = 1, ceiling(126, -128) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = -128, ceiling(-66, 64) = 0
+ceiling(-128, -128) = -128, ceiling(-126, -128) = -128, ceiling(126, -128) = 0
 unsigned char
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(253, 1) = 253
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(252, 2) = 126
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(250, 4) = 63
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(246, 8) = 31
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(238, 16) = 15
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(222, 32) = 7
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(190, 64) = 3
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(126, 128) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(252, 2) = 252
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(250, 4) = 252
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(246, 8) = 248
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(238, 16) = 240
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(222, 32) = 224
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(190, 64) = 192
+ceiling(128, 128) = 128, ceiling(130, 128) = 0, ceiling(126, 128) = 128
 short int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(-32768, -32768) = 2, ceiling(-32766, -32768) = 1, ceiling(32766, -32768) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = -32768, ceiling(-16386, 16384) = 0
+ceiling(-32768, -32768) = -32768, ceiling(-32766, -32768) = -32768, ceiling(32766, -32768) = 0
 unsigned short int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(65533, 1) = 65533
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(65532, 2) = 32766
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(65530, 4) = 16383
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(65526, 8) = 8191
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(65518, 16) = 4095
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(65502, 32) = 2047
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(65470, 64) = 1023
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(65406, 128) = 511
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(65278, 256) = 255
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(65022, 512) = 127
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(64510, 1024) = 63
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(63486, 2048) = 31
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(61438, 4096) = 15
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(57342, 8192) = 7
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(49150, 16384) = 3
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(32766, 32768) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(65532, 2) = 65532
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(65530, 4) = 65532
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(65526, 8) = 65528
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(65518, 16) = 65520
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(65502, 32) = 65504
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(65470, 64) = 65472
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(65406, 128) = 65408
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(65278, 256) = 65280
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(65022, 512) = 65024
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(64510, 1024) = 64512
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(63486, 2048) = 63488
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(61438, 4096) = 61440
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(57342, 8192) = 57344
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(49150, 16384) = 49152
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 0, ceiling(32766, 32768) = 32768
 int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(-32770, 32768) = 0
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(-65538, 65536) = 0
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(-131074, 131072) = 0
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(-262146, 262144) = 0
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(-524290, 524288) = 0
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(-1048578, 1048576) = 0
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(-2097154, 2097152) = 0
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(-4194306, 4194304) = 0
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(-8388610, 8388608) = 0
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(-16777218, 16777216) = 0
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(-33554434, 33554432) = 0
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(-67108866, 67108864) = 0
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(-134217730, 134217728) = 0
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(-268435458, 268435456) = 0
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(-536870914, 536870912) = 0
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = -1, ceiling(-1073741826, 1073741824) = 0
 ceiling(-2147483648, -2147483648) = 0, ceiling(-2147483646, -2147483648) = 0, ceiling(2147483646, -2147483648) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(-16386, 16384) = 0
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(-32770, 32768) = 0
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(-65538, 65536) = 0
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(-131074, 131072) = 0
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(-262146, 262144) = 0
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(-524290, 524288) = 0
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(-1048578, 1048576) = 0
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(-2097154, 2097152) = 0
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(-4194306, 4194304) = 0
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(-8388610, 8388608) = 0
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(-16777218, 16777216) = 0
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(-33554434, 33554432) = 0
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(-67108866, 67108864) = 0
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(-134217730, 134217728) = 0
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(-268435458, 268435456) = 0
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(-536870914, 536870912) = 0
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = -1073741824, ceiling(-1073741826, 1073741824) = 0
+ceiling(-2147483648, -2147483648) = -2147483648, ceiling(-2147483646, -2147483648) = 0, ceiling(2147483646, -2147483648) = 0
 unsigned int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(4294967293, 1) = 4294967293
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(4294967292, 2) = 2147483646
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(4294967290, 4) = 1073741823
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(4294967286, 8) = 536870911
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(4294967278, 16) = 268435455
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(4294967262, 32) = 134217727
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(4294967230, 64) = 67108863
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(4294967166, 128) = 33554431
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(4294967038, 256) = 16777215
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(4294966782, 512) = 8388607
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(4294966270, 1024) = 4194303
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(4294965246, 2048) = 2097151
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(4294963198, 4096) = 1048575
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(4294959102, 8192) = 524287
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(4294950910, 16384) = 262143
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(4294934526, 32768) = 131071
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(4294901758, 65536) = 65535
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(4294836222, 131072) = 32767
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(4294705150, 262144) = 16383
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(4294443006, 524288) = 8191
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(4293918718, 1048576) = 4095
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(4292870142, 2097152) = 2047
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(4290772990, 4194304) = 1023
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(4286578686, 8388608) = 511
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(4278190078, 16777216) = 255
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(4261412862, 33554432) = 127
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(4227858430, 67108864) = 63
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(4160749566, 134217728) = 31
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(4026531838, 268435456) = 15
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(3758096382, 536870912) = 7
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(3221225470, 1073741824) = 3
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 0, ceiling(2147483646, 2147483648) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(4294967292, 2) = 4294967292
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(4294967290, 4) = 4294967292
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(4294967286, 8) = 4294967288
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(4294967278, 16) = 4294967280
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(4294967262, 32) = 4294967264
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(4294967230, 64) = 4294967232
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(4294967166, 128) = 4294967168
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(4294967038, 256) = 4294967040
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(4294966782, 512) = 4294966784
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(4294966270, 1024) = 4294966272
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(4294965246, 2048) = 4294965248
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(4294963198, 4096) = 4294963200
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(4294959102, 8192) = 4294959104
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(4294950910, 16384) = 4294950912
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(4294934526, 32768) = 4294934528
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(4294901758, 65536) = 4294901760
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(4294836222, 131072) = 4294836224
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(4294705150, 262144) = 4294705152
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(4294443006, 524288) = 4294443008
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(4293918718, 1048576) = 4293918720
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(4292870142, 2097152) = 4292870144
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(4290772990, 4194304) = 4290772992
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(4286578686, 8388608) = 4286578688
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(4278190078, 16777216) = 4278190080
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(4261412862, 33554432) = 4261412864
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(4227858430, 67108864) = 4227858432
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(4160749566, 134217728) = 4160749568
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(4026531838, 268435456) = 4026531840
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(3758096382, 536870912) = 3758096384
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(3221225470, 1073741824) = 3221225472
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 0, ceiling(2147483646, 2147483648) = 2147483648
 long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(-32770, 32768) = 0
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(-65538, 65536) = 0
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(-131074, 131072) = 0
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(-262146, 262144) = 0
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(-524290, 524288) = 0
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(-1048578, 1048576) = 0
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(-2097154, 2097152) = 0
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(-4194306, 4194304) = 0
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(-8388610, 8388608) = 0
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(-16777218, 16777216) = 0
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(-33554434, 33554432) = 0
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(-67108866, 67108864) = 0
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(-134217730, 134217728) = 0
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(-268435458, 268435456) = 0
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(-536870914, 536870912) = 0
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = -1, ceiling(-1073741826, 1073741824) = 0
 ceiling(-2147483648, -2147483648) = 0, ceiling(-2147483646, -2147483648) = 0, ceiling(2147483646, -2147483648) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(-16386, 16384) = 0
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(-32770, 32768) = 0
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(-65538, 65536) = 0
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(-131074, 131072) = 0
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(-262146, 262144) = 0
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(-524290, 524288) = 0
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(-1048578, 1048576) = 0
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(-2097154, 2097152) = 0
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(-4194306, 4194304) = 0
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(-8388610, 8388608) = 0
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(-16777218, 16777216) = 0
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(-33554434, 33554432) = 0
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(-67108866, 67108864) = 0
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(-134217730, 134217728) = 0
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(-268435458, 268435456) = 0
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(-536870914, 536870912) = 0
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = -1073741824, ceiling(-1073741826, 1073741824) = 0
+ceiling(-2147483648, -2147483648) = -2147483648, ceiling(-2147483646, -2147483648) = 0, ceiling(2147483646, -2147483648) = 0
 unsigned long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(4294967293, 1) = 4294967293
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(4294967292, 2) = 2147483646
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(4294967290, 4) = 1073741823
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(4294967286, 8) = 536870911
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(4294967278, 16) = 268435455
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(4294967262, 32) = 134217727
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(4294967230, 64) = 67108863
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(4294967166, 128) = 33554431
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(4294967038, 256) = 16777215
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(4294966782, 512) = 8388607
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(4294966270, 1024) = 4194303
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(4294965246, 2048) = 2097151
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(4294963198, 4096) = 1048575
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(4294959102, 8192) = 524287
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(4294950910, 16384) = 262143
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(4294934526, 32768) = 131071
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(4294901758, 65536) = 65535
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(4294836222, 131072) = 32767
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(4294705150, 262144) = 16383
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(4294443006, 524288) = 8191
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(4293918718, 1048576) = 4095
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(4292870142, 2097152) = 2047
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(4290772990, 4194304) = 1023
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(4286578686, 8388608) = 511
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(4278190078, 16777216) = 255
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(4261412862, 33554432) = 127
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(4227858430, 67108864) = 63
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(4160749566, 134217728) = 31
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(4026531838, 268435456) = 15
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(3758096382, 536870912) = 7
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(3221225470, 1073741824) = 3
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 0, ceiling(2147483646, 2147483648) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(4294967292, 2) = 4294967292
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(4294967290, 4) = 4294967292
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(4294967286, 8) = 4294967288
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(4294967278, 16) = 4294967280
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(4294967262, 32) = 4294967264
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(4294967230, 64) = 4294967232
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(4294967166, 128) = 4294967168
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(4294967038, 256) = 4294967040
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(4294966782, 512) = 4294966784
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(4294966270, 1024) = 4294966272
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(4294965246, 2048) = 4294965248
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(4294963198, 4096) = 4294963200
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(4294959102, 8192) = 4294959104
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(4294950910, 16384) = 4294950912
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(4294934526, 32768) = 4294934528
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(4294901758, 65536) = 4294901760
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(4294836222, 131072) = 4294836224
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(4294705150, 262144) = 4294705152
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(4294443006, 524288) = 4294443008
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(4293918718, 1048576) = 4293918720
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(4292870142, 2097152) = 4292870144
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(4290772990, 4194304) = 4290772992
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(4286578686, 8388608) = 4286578688
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(4278190078, 16777216) = 4278190080
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(4261412862, 33554432) = 4261412864
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(4227858430, 67108864) = 4227858432
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(4160749566, 134217728) = 4160749568
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(4026531838, 268435456) = 4026531840
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(3758096382, 536870912) = 3758096384
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(3221225470, 1073741824) = 3221225472
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 0, ceiling(2147483646, 2147483648) = 2147483648
 long long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(-3, 1) = -3
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(-4, 2) = -1
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(-6, 4) = 0
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(-10, 8) = 0
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(-18, 16) = 0
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(-34, 32) = 0
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(-66, 64) = 0
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(-130, 128) = 0
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(-258, 256) = 0
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(-514, 512) = 0
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(-1026, 1024) = 0
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(-2050, 2048) = 0
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(-4098, 4096) = 0
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(-8194, 8192) = 0
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(-16386, 16384) = 0
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(-32770, 32768) = 0
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(-65538, 65536) = 0
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(-131074, 131072) = 0
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(-262146, 262144) = 0
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(-524290, 524288) = 0
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(-1048578, 1048576) = 0
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(-2097154, 2097152) = 0
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(-4194306, 4194304) = 0
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(-8388610, 8388608) = 0
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(-16777218, 16777216) = 0
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(-33554434, 33554432) = 0
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(-67108866, 67108864) = 0
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(-134217730, 134217728) = 0
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(-268435458, 268435456) = 0
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(-536870914, 536870912) = 0
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(-1073741826, 1073741824) = 0
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 2, ceiling(-2147483650, 2147483648) = 0
 ceiling(4294967296, 4294967296) = 1, ceiling(4294967298, 4294967296) = 2, ceiling(-4294967298, 4294967296) = 0
 ceiling(8589934592, 8589934592) = 1, ceiling(8589934594, 8589934592) = 2, ceiling(-8589934594, 8589934592) = 0
 ceiling(17179869184, 17179869184) = 1, ceiling(17179869186, 17179869184) = 2, ceiling(-17179869186, 17179869184) = 0
 ceiling(34359738368, 34359738368) = 1, ceiling(34359738370, 34359738368) = 2, ceiling(-34359738370, 34359738368) = 0
 ceiling(68719476736, 68719476736) = 1, ceiling(68719476738, 68719476736) = 2, ceiling(-68719476738, 68719476736) = 0
 ceiling(137438953472, 137438953472) = 1, ceiling(137438953474, 137438953472) = 2, ceiling(-137438953474, 137438953472) = 0
 ceiling(274877906944, 274877906944) = 1, ceiling(274877906946, 274877906944) = 2, ceiling(-274877906946, 274877906944) = 0
 ceiling(549755813888, 549755813888) = 1, ceiling(549755813890, 549755813888) = 2, ceiling(-549755813890, 549755813888) = 0
 ceiling(1099511627776, 1099511627776) = 1, ceiling(1099511627778, 1099511627776) = 2, ceiling(-1099511627778, 1099511627776) = 0
 ceiling(2199023255552, 2199023255552) = 1, ceiling(2199023255554, 2199023255552) = 2, ceiling(-2199023255554, 2199023255552) = 0
 ceiling(4398046511104, 4398046511104) = 1, ceiling(4398046511106, 4398046511104) = 2, ceiling(-4398046511106, 4398046511104) = 0
 ceiling(8796093022208, 8796093022208) = 1, ceiling(8796093022210, 8796093022208) = 2, ceiling(-8796093022210, 8796093022208) = 0
 ceiling(17592186044416, 17592186044416) = 1, ceiling(17592186044418, 17592186044416) = 2, ceiling(-17592186044418, 17592186044416) = 0
 ceiling(35184372088832, 35184372088832) = 1, ceiling(35184372088834, 35184372088832) = 2, ceiling(-35184372088834, 35184372088832) = 0
 ceiling(70368744177664, 70368744177664) = 1, ceiling(70368744177666, 70368744177664) = 2, ceiling(-70368744177666, 70368744177664) = 0
 ceiling(140737488355328, 140737488355328) = 1, ceiling(140737488355330, 140737488355328) = 2, ceiling(-140737488355330, 140737488355328) = 0
 ceiling(281474976710656, 281474976710656) = 1, ceiling(281474976710658, 281474976710656) = 2, ceiling(-281474976710658, 281474976710656) = 0
 ceiling(562949953421312, 562949953421312) = 1, ceiling(562949953421314, 562949953421312) = 2, ceiling(-562949953421314, 562949953421312) = 0
 ceiling(1125899906842624, 1125899906842624) = 1, ceiling(1125899906842626, 1125899906842624) = 2, ceiling(-1125899906842626, 1125899906842624) = 0
 ceiling(2251799813685248, 2251799813685248) = 1, ceiling(2251799813685250, 2251799813685248) = 2, ceiling(-2251799813685250, 2251799813685248) = 0
 ceiling(4503599627370496, 4503599627370496) = 1, ceiling(4503599627370498, 4503599627370496) = 2, ceiling(-4503599627370498, 4503599627370496) = 0
 ceiling(9007199254740992, 9007199254740992) = 1, ceiling(9007199254740994, 9007199254740992) = 2, ceiling(-9007199254740994, 9007199254740992) = 0
 ceiling(18014398509481984, 18014398509481984) = 1, ceiling(18014398509481986, 18014398509481984) = 2, ceiling(-18014398509481986, 18014398509481984) = 0
 ceiling(36028797018963968, 36028797018963968) = 1, ceiling(36028797018963970, 36028797018963968) = 2, ceiling(-36028797018963970, 36028797018963968) = 0
 ceiling(72057594037927936, 72057594037927936) = 1, ceiling(72057594037927938, 72057594037927936) = 2, ceiling(-72057594037927938, 72057594037927936) = 0
 ceiling(144115188075855872, 144115188075855872) = 1, ceiling(144115188075855874, 144115188075855872) = 2, ceiling(-144115188075855874, 144115188075855872) = 0
 ceiling(288230376151711744, 288230376151711744) = 1, ceiling(288230376151711746, 288230376151711744) = 2, ceiling(-288230376151711746, 288230376151711744) = 0
 ceiling(576460752303423488, 576460752303423488) = 1, ceiling(576460752303423490, 576460752303423488) = 2, ceiling(-576460752303423490, 576460752303423488) = 0
 ceiling(1152921504606846976, 1152921504606846976) = 1, ceiling(1152921504606846978, 1152921504606846976) = 2, ceiling(-1152921504606846978, 1152921504606846976) = 0
 ceiling(2305843009213693952, 2305843009213693952) = 1, ceiling(2305843009213693954, 2305843009213693952) = 2, ceiling(-2305843009213693954, 2305843009213693952) = 0
 ceiling(4611686018427387904, 4611686018427387904) = 1, ceiling(4611686018427387906, 4611686018427387904) = -1, ceiling(-4611686018427387906, 4611686018427387904) = 0
 ceiling(-9223372036854775808, -9223372036854775808) = 0, ceiling(-9223372036854775806, -9223372036854775808) = 0, ceiling(9223372036854775806, -9223372036854775808) = 0
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(-4, 2) = -4
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(-6, 4) = 0
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(-10, 8) = 0
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(-18, 16) = 0
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(-34, 32) = 0
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(-66, 64) = 0
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(-130, 128) = 0
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(-258, 256) = 0
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(-514, 512) = 0
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(-1026, 1024) = 0
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(-2050, 2048) = 0
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(-4098, 4096) = 0
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(-8194, 8192) = 0
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(-16386, 16384) = 0
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(-32770, 32768) = 0
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(-65538, 65536) = 0
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(-131074, 131072) = 0
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(-262146, 262144) = 0
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(-524290, 524288) = 0
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(-1048578, 1048576) = 0
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(-2097154, 2097152) = 0
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(-4194306, 4194304) = 0
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(-8388610, 8388608) = 0
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(-16777218, 16777216) = 0
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(-33554434, 33554432) = 0
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(-67108866, 67108864) = 0
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(-134217730, 134217728) = 0
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(-268435458, 268435456) = 0
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(-536870914, 536870912) = 0
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(-1073741826, 1073741824) = 0
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 4294967296, ceiling(-2147483650, 2147483648) = 0
+ceiling(4294967296, 4294967296) = 4294967296, ceiling(4294967298, 4294967296) = 8589934592, ceiling(-4294967298, 4294967296) = 0
+ceiling(8589934592, 8589934592) = 8589934592, ceiling(8589934594, 8589934592) = 17179869184, ceiling(-8589934594, 8589934592) = 0
+ceiling(17179869184, 17179869184) = 17179869184, ceiling(17179869186, 17179869184) = 34359738368, ceiling(-17179869186, 17179869184) = 0
+ceiling(34359738368, 34359738368) = 34359738368, ceiling(34359738370, 34359738368) = 68719476736, ceiling(-34359738370, 34359738368) = 0
+ceiling(68719476736, 68719476736) = 68719476736, ceiling(68719476738, 68719476736) = 137438953472, ceiling(-68719476738, 68719476736) = 0
+ceiling(137438953472, 137438953472) = 137438953472, ceiling(137438953474, 137438953472) = 274877906944, ceiling(-137438953474, 137438953472) = 0
+ceiling(274877906944, 274877906944) = 274877906944, ceiling(274877906946, 274877906944) = 549755813888, ceiling(-274877906946, 274877906944) = 0
+ceiling(549755813888, 549755813888) = 549755813888, ceiling(549755813890, 549755813888) = 1099511627776, ceiling(-549755813890, 549755813888) = 0
+ceiling(1099511627776, 1099511627776) = 1099511627776, ceiling(1099511627778, 1099511627776) = 2199023255552, ceiling(-1099511627778, 1099511627776) = 0
+ceiling(2199023255552, 2199023255552) = 2199023255552, ceiling(2199023255554, 2199023255552) = 4398046511104, ceiling(-2199023255554, 2199023255552) = 0
+ceiling(4398046511104, 4398046511104) = 4398046511104, ceiling(4398046511106, 4398046511104) = 8796093022208, ceiling(-4398046511106, 4398046511104) = 0
+ceiling(8796093022208, 8796093022208) = 8796093022208, ceiling(8796093022210, 8796093022208) = 17592186044416, ceiling(-8796093022210, 8796093022208) = 0
+ceiling(17592186044416, 17592186044416) = 17592186044416, ceiling(17592186044418, 17592186044416) = 35184372088832, ceiling(-17592186044418, 17592186044416) = 0
+ceiling(35184372088832, 35184372088832) = 35184372088832, ceiling(35184372088834, 35184372088832) = 70368744177664, ceiling(-35184372088834, 35184372088832) = 0
+ceiling(70368744177664, 70368744177664) = 70368744177664, ceiling(70368744177666, 70368744177664) = 140737488355328, ceiling(-70368744177666, 70368744177664) = 0
+ceiling(140737488355328, 140737488355328) = 140737488355328, ceiling(140737488355330, 140737488355328) = 281474976710656, ceiling(-140737488355330, 140737488355328) = 0
+ceiling(281474976710656, 281474976710656) = 281474976710656, ceiling(281474976710658, 281474976710656) = 562949953421312, ceiling(-281474976710658, 281474976710656) = 0
+ceiling(562949953421312, 562949953421312) = 562949953421312, ceiling(562949953421314, 562949953421312) = 1125899906842624, ceiling(-562949953421314, 562949953421312) = 0
+ceiling(1125899906842624, 1125899906842624) = 1125899906842624, ceiling(1125899906842626, 1125899906842624) = 2251799813685248, ceiling(-1125899906842626, 1125899906842624) = 0
+ceiling(2251799813685248, 2251799813685248) = 2251799813685248, ceiling(2251799813685250, 2251799813685248) = 4503599627370496, ceiling(-2251799813685250, 2251799813685248) = 0
+ceiling(4503599627370496, 4503599627370496) = 4503599627370496, ceiling(4503599627370498, 4503599627370496) = 9007199254740992, ceiling(-4503599627370498, 4503599627370496) = 0
+ceiling(9007199254740992, 9007199254740992) = 9007199254740992, ceiling(9007199254740994, 9007199254740992) = 18014398509481984, ceiling(-9007199254740994, 9007199254740992) = 0
+ceiling(18014398509481984, 18014398509481984) = 18014398509481984, ceiling(18014398509481986, 18014398509481984) = 36028797018963968, ceiling(-18014398509481986, 18014398509481984) = 0
+ceiling(36028797018963968, 36028797018963968) = 36028797018963968, ceiling(36028797018963970, 36028797018963968) = 72057594037927936, ceiling(-36028797018963970, 36028797018963968) = 0
+ceiling(72057594037927936, 72057594037927936) = 72057594037927936, ceiling(72057594037927938, 72057594037927936) = 144115188075855872, ceiling(-72057594037927938, 72057594037927936) = 0
+ceiling(144115188075855872, 144115188075855872) = 144115188075855872, ceiling(144115188075855874, 144115188075855872) = 288230376151711744, ceiling(-144115188075855874, 144115188075855872) = 0
+ceiling(288230376151711744, 288230376151711744) = 288230376151711744, ceiling(288230376151711746, 288230376151711744) = 576460752303423488, ceiling(-288230376151711746, 288230376151711744) = 0
+ceiling(576460752303423488, 576460752303423488) = 576460752303423488, ceiling(576460752303423490, 576460752303423488) = 1152921504606846976, ceiling(-576460752303423490, 576460752303423488) = 0
+ceiling(1152921504606846976, 1152921504606846976) = 1152921504606846976, ceiling(1152921504606846978, 1152921504606846976) = 2305843009213693952, ceiling(-1152921504606846978, 1152921504606846976) = 0
+ceiling(2305843009213693952, 2305843009213693952) = 2305843009213693952, ceiling(2305843009213693954, 2305843009213693952) = 4611686018427387904, ceiling(-2305843009213693954, 2305843009213693952) = 0
+ceiling(4611686018427387904, 4611686018427387904) = 4611686018427387904, ceiling(4611686018427387906, 4611686018427387904) = -4611686018427387904, ceiling(-4611686018427387906, 4611686018427387904) = 0
+ceiling(-9223372036854775808, -9223372036854775808) = -9223372036854775808, ceiling(-9223372036854775806, -9223372036854775808) = 0, ceiling(9223372036854775806, -9223372036854775808) = 0
 unsigned long long int
 ceiling(1, 1) = 1, ceiling(3, 1) = 3, ceiling(18446744073709551613, 1) = 18446744073709551613
 ceiling(2, 2) = 1, ceiling(4, 2) = 2, ceiling(18446744073709551612, 2) = 9223372036854775806
 ceiling(4, 4) = 1, ceiling(6, 4) = 2, ceiling(18446744073709551610, 4) = 4611686018427387903
 ceiling(8, 8) = 1, ceiling(10, 8) = 2, ceiling(18446744073709551606, 8) = 2305843009213693951
 ceiling(16, 16) = 1, ceiling(18, 16) = 2, ceiling(18446744073709551598, 16) = 1152921504606846975
 ceiling(32, 32) = 1, ceiling(34, 32) = 2, ceiling(18446744073709551582, 32) = 576460752303423487
 ceiling(64, 64) = 1, ceiling(66, 64) = 2, ceiling(18446744073709551550, 64) = 288230376151711743
 ceiling(128, 128) = 1, ceiling(130, 128) = 2, ceiling(18446744073709551486, 128) = 144115188075855871
 ceiling(256, 256) = 1, ceiling(258, 256) = 2, ceiling(18446744073709551358, 256) = 72057594037927935
 ceiling(512, 512) = 1, ceiling(514, 512) = 2, ceiling(18446744073709551102, 512) = 36028797018963967
 ceiling(1024, 1024) = 1, ceiling(1026, 1024) = 2, ceiling(18446744073709550590, 1024) = 18014398509481983
 ceiling(2048, 2048) = 1, ceiling(2050, 2048) = 2, ceiling(18446744073709549566, 2048) = 9007199254740991
 ceiling(4096, 4096) = 1, ceiling(4098, 4096) = 2, ceiling(18446744073709547518, 4096) = 4503599627370495
 ceiling(8192, 8192) = 1, ceiling(8194, 8192) = 2, ceiling(18446744073709543422, 8192) = 2251799813685247
 ceiling(16384, 16384) = 1, ceiling(16386, 16384) = 2, ceiling(18446744073709535230, 16384) = 1125899906842623
 ceiling(32768, 32768) = 1, ceiling(32770, 32768) = 2, ceiling(18446744073709518846, 32768) = 562949953421311
 ceiling(65536, 65536) = 1, ceiling(65538, 65536) = 2, ceiling(18446744073709486078, 65536) = 281474976710655
 ceiling(131072, 131072) = 1, ceiling(131074, 131072) = 2, ceiling(18446744073709420542, 131072) = 140737488355327
 ceiling(262144, 262144) = 1, ceiling(262146, 262144) = 2, ceiling(18446744073709289470, 262144) = 70368744177663
 ceiling(524288, 524288) = 1, ceiling(524290, 524288) = 2, ceiling(18446744073709027326, 524288) = 35184372088831
 ceiling(1048576, 1048576) = 1, ceiling(1048578, 1048576) = 2, ceiling(18446744073708503038, 1048576) = 17592186044415
 ceiling(2097152, 2097152) = 1, ceiling(2097154, 2097152) = 2, ceiling(18446744073707454462, 2097152) = 8796093022207
 ceiling(4194304, 4194304) = 1, ceiling(4194306, 4194304) = 2, ceiling(18446744073705357310, 4194304) = 4398046511103
 ceiling(8388608, 8388608) = 1, ceiling(8388610, 8388608) = 2, ceiling(18446744073701163006, 8388608) = 2199023255551
 ceiling(16777216, 16777216) = 1, ceiling(16777218, 16777216) = 2, ceiling(18446744073692774398, 16777216) = 1099511627775
 ceiling(33554432, 33554432) = 1, ceiling(33554434, 33554432) = 2, ceiling(18446744073675997182, 33554432) = 549755813887
 ceiling(67108864, 67108864) = 1, ceiling(67108866, 67108864) = 2, ceiling(18446744073642442750, 67108864) = 274877906943
 ceiling(134217728, 134217728) = 1, ceiling(134217730, 134217728) = 2, ceiling(18446744073575333886, 134217728) = 137438953471
 ceiling(268435456, 268435456) = 1, ceiling(268435458, 268435456) = 2, ceiling(18446744073441116158, 268435456) = 68719476735
 ceiling(536870912, 536870912) = 1, ceiling(536870914, 536870912) = 2, ceiling(18446744073172680702, 536870912) = 34359738367
 ceiling(1073741824, 1073741824) = 1, ceiling(1073741826, 1073741824) = 2, ceiling(18446744072635809790, 1073741824) = 17179869183
 ceiling(2147483648, 2147483648) = 1, ceiling(2147483650, 2147483648) = 2, ceiling(18446744071562067966, 2147483648) = 8589934591
 ceiling(4294967296, 4294967296) = 1, ceiling(4294967298, 4294967296) = 2, ceiling(18446744069414584318, 4294967296) = 4294967295
 ceiling(8589934592, 8589934592) = 1, ceiling(8589934594, 8589934592) = 2, ceiling(18446744065119617022, 8589934592) = 2147483647
 ceiling(17179869184, 17179869184) = 1, ceiling(17179869186, 17179869184) = 2, ceiling(18446744056529682430, 17179869184) = 1073741823
 ceiling(34359738368, 34359738368) = 1, ceiling(34359738370, 34359738368) = 2, ceiling(18446744039349813246, 34359738368) = 536870911
 ceiling(68719476736, 68719476736) = 1, ceiling(68719476738, 68719476736) = 2, ceiling(18446744004990074878, 68719476736) = 268435455
 ceiling(137438953472, 137438953472) = 1, ceiling(137438953474, 137438953472) = 2, ceiling(18446743936270598142, 137438953472) = 134217727
 ceiling(274877906944, 274877906944) = 1, ceiling(274877906946, 274877906944) = 2, ceiling(18446743798831644670, 274877906944) = 67108863
 ceiling(549755813888, 549755813888) = 1, ceiling(549755813890, 549755813888) = 2, ceiling(18446743523953737726, 549755813888) = 33554431
 ceiling(1099511627776, 1099511627776) = 1, ceiling(1099511627778, 1099511627776) = 2, ceiling(18446742974197923838, 1099511627776) = 16777215
 ceiling(2199023255552, 2199023255552) = 1, ceiling(2199023255554, 2199023255552) = 2, ceiling(18446741874686296062, 2199023255552) = 8388607
 ceiling(4398046511104, 4398046511104) = 1, ceiling(4398046511106, 4398046511104) = 2, ceiling(18446739675663040510, 4398046511104) = 4194303
 ceiling(8796093022208, 8796093022208) = 1, ceiling(8796093022210, 8796093022208) = 2, ceiling(18446735277616529406, 8796093022208) = 2097151
 ceiling(17592186044416, 17592186044416) = 1, ceiling(17592186044418, 17592186044416) = 2, ceiling(18446726481523507198, 17592186044416) = 1048575
 ceiling(35184372088832, 35184372088832) = 1, ceiling(35184372088834, 35184372088832) = 2, ceiling(18446708889337462782, 35184372088832) = 524287
 ceiling(70368744177664, 70368744177664) = 1, ceiling(70368744177666, 70368744177664) = 2, ceiling(18446673704965373950, 70368744177664) = 262143
 ceiling(140737488355328, 140737488355328) = 1, ceiling(140737488355330, 140737488355328) = 2, ceiling(18446603336221196286, 140737488355328) = 131071
 ceiling(281474976710656, 281474976710656) = 1, ceiling(281474976710658, 281474976710656) = 2, ceiling(18446462598732840958, 281474976710656) = 65535
 ceiling(562949953421312, 562949953421312) = 1, ceiling(562949953421314, 562949953421312) = 2, ceiling(18446181123756130302, 562949953421312) = 32767
 ceiling(1125899906842624, 1125899906842624) = 1, ceiling(1125899906842626, 1125899906842624) = 2, ceiling(18445618173802708990, 1125899906842624) = 16383
 ceiling(2251799813685248, 2251799813685248) = 1, ceiling(2251799813685250, 2251799813685248) = 2, ceiling(18444492273895866366, 2251799813685248) = 8191
 ceiling(4503599627370496, 4503599627370496) = 1, ceiling(4503599627370498, 4503599627370496) = 2, ceiling(18442240474082181118, 4503599627370496) = 4095
 ceiling(9007199254740992, 9007199254740992) = 1, ceiling(9007199254740994, 9007199254740992) = 2, ceiling(18437736874454810622, 9007199254740992) = 2047
 ceiling(18014398509481984, 18014398509481984) = 1, ceiling(18014398509481986, 18014398509481984) = 2, ceiling(18428729675200069630, 18014398509481984) = 1023
 ceiling(36028797018963968, 36028797018963968) = 1, ceiling(36028797018963970, 36028797018963968) = 2, ceiling(18410715276690587646, 36028797018963968) = 511
 ceiling(72057594037927936, 72057594037927936) = 1, ceiling(72057594037927938, 72057594037927936) = 2, ceiling(18374686479671623678, 72057594037927936) = 255
 ceiling(144115188075855872, 144115188075855872) = 1, ceiling(144115188075855874, 144115188075855872) = 2, ceiling(18302628885633695742, 144115188075855872) = 127
 ceiling(288230376151711744, 288230376151711744) = 1, ceiling(288230376151711746, 288230376151711744) = 2, ceiling(18158513697557839870, 288230376151711744) = 63
 ceiling(576460752303423488, 576460752303423488) = 1, ceiling(576460752303423490, 576460752303423488) = 2, ceiling(17870283321406128126, 576460752303423488) = 31
 ceiling(1152921504606846976, 1152921504606846976) = 1, ceiling(1152921504606846978, 1152921504606846976) = 2, ceiling(17293822569102704638, 1152921504606846976) = 15
 ceiling(2305843009213693952, 2305843009213693952) = 1, ceiling(2305843009213693954, 2305843009213693952) = 2, ceiling(16140901064495857662, 2305843009213693952) = 7
 ceiling(4611686018427387904, 4611686018427387904) = 1, ceiling(4611686018427387906, 4611686018427387904) = 2, ceiling(13835058055282163710, 4611686018427387904) = 3
 ceiling(9223372036854775808, 9223372036854775808) = 1, ceiling(9223372036854775810, 9223372036854775808) = 0, ceiling(9223372036854775806, 9223372036854775808) = 1
+ceiling(2, 2) = 2, ceiling(4, 2) = 4, ceiling(18446744073709551612, 2) = 18446744073709551612
+ceiling(4, 4) = 4, ceiling(6, 4) = 8, ceiling(18446744073709551610, 4) = 18446744073709551612
+ceiling(8, 8) = 8, ceiling(10, 8) = 16, ceiling(18446744073709551606, 8) = 18446744073709551608
+ceiling(16, 16) = 16, ceiling(18, 16) = 32, ceiling(18446744073709551598, 16) = 18446744073709551600
+ceiling(32, 32) = 32, ceiling(34, 32) = 64, ceiling(18446744073709551582, 32) = 18446744073709551584
+ceiling(64, 64) = 64, ceiling(66, 64) = 128, ceiling(18446744073709551550, 64) = 18446744073709551552
+ceiling(128, 128) = 128, ceiling(130, 128) = 256, ceiling(18446744073709551486, 128) = 18446744073709551488
+ceiling(256, 256) = 256, ceiling(258, 256) = 512, ceiling(18446744073709551358, 256) = 18446744073709551360
+ceiling(512, 512) = 512, ceiling(514, 512) = 1024, ceiling(18446744073709551102, 512) = 18446744073709551104
+ceiling(1024, 1024) = 1024, ceiling(1026, 1024) = 2048, ceiling(18446744073709550590, 1024) = 18446744073709550592
+ceiling(2048, 2048) = 2048, ceiling(2050, 2048) = 4096, ceiling(18446744073709549566, 2048) = 18446744073709549568
+ceiling(4096, 4096) = 4096, ceiling(4098, 4096) = 8192, ceiling(18446744073709547518, 4096) = 18446744073709547520
+ceiling(8192, 8192) = 8192, ceiling(8194, 8192) = 16384, ceiling(18446744073709543422, 8192) = 18446744073709543424
+ceiling(16384, 16384) = 16384, ceiling(16386, 16384) = 32768, ceiling(18446744073709535230, 16384) = 18446744073709535232
+ceiling(32768, 32768) = 32768, ceiling(32770, 32768) = 65536, ceiling(18446744073709518846, 32768) = 18446744073709518848
+ceiling(65536, 65536) = 65536, ceiling(65538, 65536) = 131072, ceiling(18446744073709486078, 65536) = 18446744073709486080
+ceiling(131072, 131072) = 131072, ceiling(131074, 131072) = 262144, ceiling(18446744073709420542, 131072) = 18446744073709420544
+ceiling(262144, 262144) = 262144, ceiling(262146, 262144) = 524288, ceiling(18446744073709289470, 262144) = 18446744073709289472
+ceiling(524288, 524288) = 524288, ceiling(524290, 524288) = 1048576, ceiling(18446744073709027326, 524288) = 18446744073709027328
+ceiling(1048576, 1048576) = 1048576, ceiling(1048578, 1048576) = 2097152, ceiling(18446744073708503038, 1048576) = 18446744073708503040
+ceiling(2097152, 2097152) = 2097152, ceiling(2097154, 2097152) = 4194304, ceiling(18446744073707454462, 2097152) = 18446744073707454464
+ceiling(4194304, 4194304) = 4194304, ceiling(4194306, 4194304) = 8388608, ceiling(18446744073705357310, 4194304) = 18446744073705357312
+ceiling(8388608, 8388608) = 8388608, ceiling(8388610, 8388608) = 16777216, ceiling(18446744073701163006, 8388608) = 18446744073701163008
+ceiling(16777216, 16777216) = 16777216, ceiling(16777218, 16777216) = 33554432, ceiling(18446744073692774398, 16777216) = 18446744073692774400
+ceiling(33554432, 33554432) = 33554432, ceiling(33554434, 33554432) = 67108864, ceiling(18446744073675997182, 33554432) = 18446744073675997184
+ceiling(67108864, 67108864) = 67108864, ceiling(67108866, 67108864) = 134217728, ceiling(18446744073642442750, 67108864) = 18446744073642442752
+ceiling(134217728, 134217728) = 134217728, ceiling(134217730, 134217728) = 268435456, ceiling(18446744073575333886, 134217728) = 18446744073575333888
+ceiling(268435456, 268435456) = 268435456, ceiling(268435458, 268435456) = 536870912, ceiling(18446744073441116158, 268435456) = 18446744073441116160
+ceiling(536870912, 536870912) = 536870912, ceiling(536870914, 536870912) = 1073741824, ceiling(18446744073172680702, 536870912) = 18446744073172680704
+ceiling(1073741824, 1073741824) = 1073741824, ceiling(1073741826, 1073741824) = 2147483648, ceiling(18446744072635809790, 1073741824) = 18446744072635809792
+ceiling(2147483648, 2147483648) = 2147483648, ceiling(2147483650, 2147483648) = 4294967296, ceiling(18446744071562067966, 2147483648) = 18446744071562067968
+ceiling(4294967296, 4294967296) = 4294967296, ceiling(4294967298, 4294967296) = 8589934592, ceiling(18446744069414584318, 4294967296) = 18446744069414584320
+ceiling(8589934592, 8589934592) = 8589934592, ceiling(8589934594, 8589934592) = 17179869184, ceiling(18446744065119617022, 8589934592) = 18446744065119617024
+ceiling(17179869184, 17179869184) = 17179869184, ceiling(17179869186, 17179869184) = 34359738368, ceiling(18446744056529682430, 17179869184) = 18446744056529682432
+ceiling(34359738368, 34359738368) = 34359738368, ceiling(34359738370, 34359738368) = 68719476736, ceiling(18446744039349813246, 34359738368) = 18446744039349813248
+ceiling(68719476736, 68719476736) = 68719476736, ceiling(68719476738, 68719476736) = 137438953472, ceiling(18446744004990074878, 68719476736) = 18446744004990074880
+ceiling(137438953472, 137438953472) = 137438953472, ceiling(137438953474, 137438953472) = 274877906944, ceiling(18446743936270598142, 137438953472) = 18446743936270598144
+ceiling(274877906944, 274877906944) = 274877906944, ceiling(274877906946, 274877906944) = 549755813888, ceiling(18446743798831644670, 274877906944) = 18446743798831644672
+ceiling(549755813888, 549755813888) = 549755813888, ceiling(549755813890, 549755813888) = 1099511627776, ceiling(18446743523953737726, 549755813888) = 18446743523953737728
+ceiling(1099511627776, 1099511627776) = 1099511627776, ceiling(1099511627778, 1099511627776) = 2199023255552, ceiling(18446742974197923838, 1099511627776) = 18446742974197923840
+ceiling(2199023255552, 2199023255552) = 2199023255552, ceiling(2199023255554, 2199023255552) = 4398046511104, ceiling(18446741874686296062, 2199023255552) = 18446741874686296064
+ceiling(4398046511104, 4398046511104) = 4398046511104, ceiling(4398046511106, 4398046511104) = 8796093022208, ceiling(18446739675663040510, 4398046511104) = 18446739675663040512
+ceiling(8796093022208, 8796093022208) = 8796093022208, ceiling(8796093022210, 8796093022208) = 17592186044416, ceiling(18446735277616529406, 8796093022208) = 18446735277616529408
+ceiling(17592186044416, 17592186044416) = 17592186044416, ceiling(17592186044418, 17592186044416) = 35184372088832, ceiling(18446726481523507198, 17592186044416) = 18446726481523507200
+ceiling(35184372088832, 35184372088832) = 35184372088832, ceiling(35184372088834, 35184372088832) = 70368744177664, ceiling(18446708889337462782, 35184372088832) = 18446708889337462784
+ceiling(70368744177664, 70368744177664) = 70368744177664, ceiling(70368744177666, 70368744177664) = 140737488355328, ceiling(18446673704965373950, 70368744177664) = 18446673704965373952
+ceiling(140737488355328, 140737488355328) = 140737488355328, ceiling(140737488355330, 140737488355328) = 281474976710656, ceiling(18446603336221196286, 140737488355328) = 18446603336221196288
+ceiling(281474976710656, 281474976710656) = 281474976710656, ceiling(281474976710658, 281474976710656) = 562949953421312, ceiling(18446462598732840958, 281474976710656) = 18446462598732840960
+ceiling(562949953421312, 562949953421312) = 562949953421312, ceiling(562949953421314, 562949953421312) = 1125899906842624, ceiling(18446181123756130302, 562949953421312) = 18446181123756130304
+ceiling(1125899906842624, 1125899906842624) = 1125899906842624, ceiling(1125899906842626, 1125899906842624) = 2251799813685248, ceiling(18445618173802708990, 1125899906842624) = 18445618173802708992
+ceiling(2251799813685248, 2251799813685248) = 2251799813685248, ceiling(2251799813685250, 2251799813685248) = 4503599627370496, ceiling(18444492273895866366, 2251799813685248) = 18444492273895866368
+ceiling(4503599627370496, 4503599627370496) = 4503599627370496, ceiling(4503599627370498, 4503599627370496) = 9007199254740992, ceiling(18442240474082181118, 4503599627370496) = 18442240474082181120
+ceiling(9007199254740992, 9007199254740992) = 9007199254740992, ceiling(9007199254740994, 9007199254740992) = 18014398509481984, ceiling(18437736874454810622, 9007199254740992) = 18437736874454810624
+ceiling(18014398509481984, 18014398509481984) = 18014398509481984, ceiling(18014398509481986, 18014398509481984) = 36028797018963968, ceiling(18428729675200069630, 18014398509481984) = 18428729675200069632
+ceiling(36028797018963968, 36028797018963968) = 36028797018963968, ceiling(36028797018963970, 36028797018963968) = 72057594037927936, ceiling(18410715276690587646, 36028797018963968) = 18410715276690587648
+ceiling(72057594037927936, 72057594037927936) = 72057594037927936, ceiling(72057594037927938, 72057594037927936) = 144115188075855872, ceiling(18374686479671623678, 72057594037927936) = 18374686479671623680
+ceiling(144115188075855872, 144115188075855872) = 144115188075855872, ceiling(144115188075855874, 144115188075855872) = 288230376151711744, ceiling(18302628885633695742, 144115188075855872) = 18302628885633695744
+ceiling(288230376151711744, 288230376151711744) = 288230376151711744, ceiling(288230376151711746, 288230376151711744) = 576460752303423488, ceiling(18158513697557839870, 288230376151711744) = 18158513697557839872
+ceiling(576460752303423488, 576460752303423488) = 576460752303423488, ceiling(576460752303423490, 576460752303423488) = 1152921504606846976, ceiling(17870283321406128126, 576460752303423488) = 17870283321406128128
+ceiling(1152921504606846976, 1152921504606846976) = 1152921504606846976, ceiling(1152921504606846978, 1152921504606846976) = 2305843009213693952, ceiling(17293822569102704638, 1152921504606846976) = 17293822569102704640
+ceiling(2305843009213693952, 2305843009213693952) = 2305843009213693952, ceiling(2305843009213693954, 2305843009213693952) = 4611686018427387904, ceiling(16140901064495857662, 2305843009213693952) = 16140901064495857664
+ceiling(4611686018427387904, 4611686018427387904) = 4611686018427387904, ceiling(4611686018427387906, 4611686018427387904) = 9223372036854775808, ceiling(13835058055282163710, 4611686018427387904) = 13835058055282163712
+ceiling(9223372036854775808, 9223372036854775808) = 9223372036854775808, ceiling(9223372036854775810, 9223372036854775808) = 0, ceiling(9223372036854775806, 9223372036854775808) = 9223372036854775808

tests/.expect/castError.txt

-              r07d867b
+              r22f94a4
 castError.cfa:21:1 error: Cannot choose between 3 alternatives for expression
+castError.cfa:23:1 error: Cannot choose between 3 alternatives for expression
 Explicit Cast of:
   Name: f
 …
 castError.cfa:26:1 error: Cannot choose between 2 alternatives for expression
+castError.cfa:28:1 error: Cannot choose between 2 alternatives for expression
 Generated Cast of:
   Comma Expression:
 …
+castError.cfa:30:1 error: No reasonable alternatives for expression Explicit Cast of:
+  Name: sint
+... to:
+  instance of struct S with body 1
+  ... with parameters
+    char

tests/.expect/copyfile.txt

-              r07d867b
+              r22f94a4
 //
 // Author           : Peter A. Buhr
 // Created On       : Tue Jul 16 16:47:22 2019
+// Created On       : Fri Jun 19 13:44:05 2020
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Jul 17 18:04:44 2019
 // Update Count     : 26
+// Last Modified On : Fri Jun 19 17:58:03 2020
+// Update Count     : 4
 //
 #include <fstream.hfa>
 #include <stdlib.hfa>                                                                   // new/delete
+#include <exception.hfa>
 int main( int argc, char * argv[] ) {
+        ifstream * in  = &stdin;                                                        // default files
+        ofstream * out = &stdout;
+        ifstream in  = stdin;                                                           // copy default files
+        ofstream out = stdout;
         try {
                 choose ( argc ) {
                   case 2, 3:
                           in = new( (const char *)argv[1] );            // open input file first as output creates file
                           if ( argc == 3 ) out = new( (const char *)argv[2] ); // only open output if input opens as output created if nonexistent
                   case 1: ;                                     // use default files
+                        open( in, argv[1] );                                            // open input file first as output creates file
+                        if ( argc == 3 ) open( out, argv[2] );          // do not create output unless input opens
+                  case 1: ;                                                                             // use default files
                   default:
                           exit | "Usage [ input-file (default stdin) [ output-file (default stdout) ] ]";
+                        exit | "Usage" | argv[0] | "[ input-file (default stdin) [ output-file (default stdout) ] ]";
                 } // choose
+        } catch( Open_Failure * ex ; ex->istream == &in ) {
+                exit | "Unable to open input file" | argv[1];
+        } catch( Open_Failure * ex ; ex->ostream == &out ) {
+                close( in );                                                                    // optional
+                exit | "Unable to open output file" | argv[2];
+        } // try
+                char ch;
+                *out | nlOff;                                                                   // turn off auto newline
+                *in  | nlOn;                                                                    // turn on reading newline
+        out | nlOff;                                                                            // turn off auto newline
+        in  | nlOn;                                                                                     // turn on reading newline
+                for () {                                                                                // read all characters
+                        *in | ch;
+                  if ( eof( *in ) ) break;                                              // eof ?
+                        *out | ch;
+                } // for
+        } finally {
+                if ( in  != &stdin  ) delete( in );                             // close file, do not delete stdin!
+                if ( out != &stdout ) delete( out );                    // close file, do not delete stdout!
+        } // try
+        char ch;
+        for () {                                                                                        // read all characters
+                in | ch;
+          if ( eof( in ) ) break;                                                       // eof ?
+                out | ch;
+        } //for
 } // main

tests/.expect/functions.x64.txt

-              r07d867b
+              r22f94a4
+}
+struct _conc__tuple2_0;
 struct _conc__tuple2_0 {
     signed int field_0;
 …
+}
+struct _conc__tuple3_1;
 struct _conc__tuple3_1 {
     signed int field_0;
 …
     __attribute__ ((unused)) struct _conc__tuple3_1 _X9_retval_fT3iii_1 = {  };
+}
+struct _conc__tuple3_2;
 struct _conc__tuple3_2 {
     signed int field_0;
 …
     __attribute__ ((unused)) signed int *const _X10_retval_f3KPi_1;
+}
+struct _conc__tuple2_3;
 struct _conc__tuple2_3 {
     signed int *field_0;

tests/.expect/functions.x86.txt

-              r07d867b
+              r22f94a4
+}
+struct _conc__tuple2_0;
 struct _conc__tuple2_0 {
     signed int field_0;
 …
+}
+struct _conc__tuple3_1;
 struct _conc__tuple3_1 {
     signed int field_0;
 …
     __attribute__ ((unused)) struct _conc__tuple3_1 _X9_retval_fT3iii_1 = {  };
+}
+struct _conc__tuple3_2;
 struct _conc__tuple3_2 {
     signed int field_0;
 …
     __attribute__ ((unused)) signed int *const _X10_retval_f3KPi_1;
+}
+struct _conc__tuple2_3;
 struct _conc__tuple2_3 {
     signed int *field_0;

tests/.expect/time.txt

r07d867b	r22f94a4
1	1	10800 2 3.375 12 1.00001
	2	0.125 0.0333333333333333 3.375 12000. 1000010.
2	3	0 2 3.375
3	4	7 7 7

tests/.in/copyfile.txt

-              r07d867b
+              r22f94a4
 //
 // Author           : Peter A. Buhr
 // Created On       : Tue Jul 16 16:47:22 2019
+// Created On       : Fri Jun 19 13:44:05 2020
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Jul 17 18:04:44 2019
 // Update Count     : 26
+// Last Modified On : Fri Jun 19 17:58:03 2020
+// Update Count     : 4
 //
 #include <fstream.hfa>
 #include <stdlib.hfa>                                                                   // new/delete
+#include <exception.hfa>
 int main( int argc, char * argv[] ) {
+        ifstream * in  = &stdin;                                                        // default files
+        ofstream * out = &stdout;
+        ifstream in  = stdin;                                                           // copy default files
+        ofstream out = stdout;
         try {
                 choose ( argc ) {
                   case 2, 3:
                           in = new( (const char *)argv[1] );            // open input file first as output creates file
                           if ( argc == 3 ) out = new( (const char *)argv[2] ); // only open output if input opens as output created if nonexistent
                   case 1: ;                                     // use default files
+                        open( in, argv[1] );                                            // open input file first as output creates file
+                        if ( argc == 3 ) open( out, argv[2] );          // do not create output unless input opens
+                  case 1: ;                                                                             // use default files
                   default:
                           exit | "Usage [ input-file (default stdin) [ output-file (default stdout) ] ]";
+                        exit | "Usage" | argv[0] | "[ input-file (default stdin) [ output-file (default stdout) ] ]";
                 } // choose
+        } catch( Open_Failure * ex ; ex->istream == &in ) {
+                exit | "Unable to open input file" | argv[1];
+        } catch( Open_Failure * ex ; ex->ostream == &out ) {
+                close( in );                                                                    // optional
+                exit | "Unable to open output file" | argv[2];
+        } // try
+                char ch;
+                *out | nlOff;                                                                   // turn off auto newline
+                *in  | nlOn;                                                                    // turn on reading newline
+        out | nlOff;                                                                            // turn off auto newline
+        in  | nlOn;                                                                                     // turn on reading newline
+                for () {                                                                                // read all characters
+                        *in | ch;
+                  if ( eof( *in ) ) break;                                              // eof ?
+                        *out | ch;
+                } // for
+        } finally {
+                if ( in  != &stdin  ) delete( in );                             // close file, do not delete stdin!
+                if ( out != &stdout ) delete( out );                    // close file, do not delete stdout!
+        } // try
+        char ch;
+        for () {                                                                                        // read all characters
+                in | ch;
+          if ( eof( in ) ) break;                                                       // eof ?
+                out | ch;
+        } //for
 } // main

tests/.in/manipulatorsInput.txt

-              r07d867b
+              r22f94a4
 .5 3.5 3.456E+23.456E+2 -0x1.2p-3 3.5 0X1.23p3     3.5
 .5 3.5 3.456E+23.456E+2 -0x1.2p-3 3.5 0X1.23p3     3.5
+-25 42798
+1402432282 1505850196993244515
+394749758663249135511342
+12935154696204706112391834394
+423859149128410414395372834994551
+13889016598639747063234935497057631587
+170141183460469231731687303715884105727
+340282366920938463463374607431768211455
+-340282366920938463463374607431768211455
+340282366920938463463374607431768211455999
+1234567890123456789 -1234567890123456789

tests/Makefile.am

-              r07d867b
+              r22f94a4
 ACLOCAL_AMFLAGS  = -I automake
+include $(top_srcdir)/src/cfa.make
+include $(top_srcdir)/tools/build/cfa.make
+DEFAULT_INCLUDES = -I${abs_srcdir}
 debug=yes

tests/avltree/avl1.cfa

r07d867b	r22f94a4
24	24	tree(K, V) * create(K key, V value) {
25	25	// infinite loop trying to resolve ... t = malloc();
26		tree(K, V) * t = malloc(sizeof(tree(K,V)));
	26	tree(K, V) * t = ( tree(K, V) * ) malloc(sizeof(tree(K,V)));
27	27	(*t){ key, value };
28	28	return t;

tests/bitmanip3.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Tue Apr  7 21:22:59 2020
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Tue Apr 21 16:25:09 2020
 // Update Count     : 61
+// Last Modified On : Mon Aug 10 09:31:35 2020
+// Update Count     : 65
 //
 …
         printf( "signed char\n" );
         sc = 0;
         scr1 = floor2( sc, sc ), scr2 = floor2( sc + 2hh, sc ), scr3 = floor2( -sc - 2hh, sc );
         printf( "floor2(%hhd, %hhd) = %hhd, floor2(%hhd, %hhd) = %hhd, floor2(%hhd, %hhd) = %hhd\n", sc, sc, scr1, sc + 2hh, sc, scr2, -sc - 2hh, sc, scr3 );
+//      sc = 0;
+//      scr1 = floor2( sc, sc ), scr2 = floor2( sc + 2hh, sc ), scr3 = floor2( -sc - 2hh, sc );
+//      printf( "floor2(%hhd, %hhd) = %hhd, floor2(%hhd, %hhd) = %hhd, floor2(%hhd, %hhd) = %hhd\n", sc, sc, scr1, sc + 2hh, sc, scr2, -sc - 2hh, sc, scr3 );
         for ( sc = 1; sc != 0; sc <<= 1 ) {
                 scr1 = floor2( sc, sc ); scr2 = floor2( sc + 2hh, sc ); scr3 = floor2( -sc - 2hh, sc );
 …
         printf( "unsigned char\n" );
         uc = 0;
         ucr1 = floor2( uc, uc ), ucr2 = floor2( uc + 2hh, uc ), ucr3 = floor2( -uc - 2hh, uc );
         printf( "floor2(%hhu, %hhu) = %hhu, floor2(%hhu, %hhu) = %hhu, floor2(%hhu, %hhu) = %hhu\n", uc, uc, ucr1, uc + 2uhh, uc, ucr2, -uc - 2uhh, uc, ucr3 );
+//      uc = 0;
+//      ucr1 = floor2( uc, uc ), ucr2 = floor2( uc + 2hh, uc ), ucr3 = floor2( -uc - 2hh, uc );
+//      printf( "floor2(%hhu, %hhu) = %hhu, floor2(%hhu, %hhu) = %hhu, floor2(%hhu, %hhu) = %hhu\n", uc, uc, ucr1, uc + 2uhh, uc, ucr2, -uc - 2uhh, uc, ucr3 );
         for ( uc = 1; uc != 0; uc <<= 1 ) {
                 ucr1 = floor2( uc, uc ); ucr2 = floor2( uc + 2hh, uc ); ucr3 = floor2( -uc - 2hh, uc );
 …
         printf( "short int\n" );
         si = 0;
         sir1 = floor2( si, si ), sir2 = floor2( si + 2hh, si ), sir3 = floor2( -si - 2hh, si );
         printf( "floor2(%hd, %hd) = %hd, floor2(%hd, %hd) = %hd, floor2(%hd, %hd) = %hd\n", si, si, sir1, si + 2h, si, sir2, -si - 2h, si, sir3 );
+//      si = 0;
+//      sir1 = floor2( si, si ), sir2 = floor2( si + 2hh, si ), sir3 = floor2( -si - 2hh, si );
+//      printf( "floor2(%hd, %hd) = %hd, floor2(%hd, %hd) = %hd, floor2(%hd, %hd) = %hd\n", si, si, sir1, si + 2h, si, sir2, -si - 2h, si, sir3 );
         for ( si = 1; si != 0; si <<= 1 ) {
                 sir1 = floor2( si, si ); sir2 = floor2( si + 2hh, si ); sir3 = floor2( -si - 2hh, si );
 …
         printf( "unsigned short int\n" );
         usi = 0;
         usir1 = floor2( usi, usi ), usir2 = floor2( usi + 2hh, usi ), usir3 = floor2( -usi - 2hh, usi );
         printf( "floor2(%hu, %hu) = %hu, floor2(%hu, %hu) = %hu, floor2(%hu, %hu) = %hu\n", usi, usi, usir1, usi + 2uh, usi, usir2, -usi - 2uh, usi, usir3 );
+//      usi = 0;
+//      usir1 = floor2( usi, usi ), usir2 = floor2( usi + 2hh, usi ), usir3 = floor2( -usi - 2hh, usi );
+//      printf( "floor2(%hu, %hu) = %hu, floor2(%hu, %hu) = %hu, floor2(%hu, %hu) = %hu\n", usi, usi, usir1, usi + 2uh, usi, usir2, -usi - 2uh, usi, usir3 );
         for ( usi = 1; usi != 0; usi <<= 1 ) {
                 usir1 = floor2( usi, usi ); usir2 = floor2( usi + 2hh, usi ); usir3 = floor2( -usi - 2hh, usi );
 …
         printf( "int\n" );
         i = 0;
         ir1 = floor2( i, i ), ir2 = floor2( i + 2hh, i ), ir3 = floor2( -i - 2hh, i );
         printf( "floor2(%d, %d) = %d, floor2(%d, %d) = %d, floor2(%d, %d) = %d\n", i, i, ir1, i + 2h, i, ir2, -i - 2h, i, ir3 );
+//      i = 0;
+//      ir1 = floor2( i, i ), ir2 = floor2( i + 2hh, i ), ir3 = floor2( -i - 2hh, i );
+//      printf( "floor2(%d, %d) = %d, floor2(%d, %d) = %d, floor2(%d, %d) = %d\n", i, i, ir1, i + 2h, i, ir2, -i - 2h, i, ir3 );
         for ( i = 1; i != 0; i <<= 1 ) {
                 ir1 = floor2( i, i ); ir2 = floor2( i + 2hh, i ); ir3 = floor2( -i - 2hh, i );
 …
         printf( "unsigned int\n" );
         ui = 0;
         uir1 = floor2( ui, ui ), uir2 = floor2( ui + 2hh, ui ), uir3 = floor2( -ui - 2hh, ui );
         printf( "floor2(%u, %u) = %u, floor2(%u, %u) = %u, floor2(%u, %u) = %u\n", ui, ui, uir1, ui + 2h, ui, uir2, -ui - 2h, ui, uir3 );
+//      ui = 0;
+//      uir1 = floor2( ui, ui ), uir2 = floor2( ui + 2hh, ui ), uir3 = floor2( -ui - 2hh, ui );
+//      printf( "floor2(%u, %u) = %u, floor2(%u, %u) = %u, floor2(%u, %u) = %u\n", ui, ui, uir1, ui + 2h, ui, uir2, -ui - 2h, ui, uir3 );
         for ( ui = 1; ui != 0; ui <<= 1 ) {
                 uir1 = floor2( ui, ui ); uir2 = floor2( ui + 2hh, ui ); uir3 = floor2( -ui - 2hh, ui );
 …
         printf( "long int\n" );
         li = 0;
         lir1 = floor2( li, li ), lir2 = floor2( li + 2hh, li ), lir3 = floor2( -li - 2hh, li );
         printf( "floor2(%ld, %ld) = %ld, floor2(%ld, %ld) = %ld, floor2(%ld, %ld) = %ld\n", li, li, lir1, li + 2h, li, lir2, -li - 2h, li, lir3 );
+//      li = 0;
+//      lir1 = floor2( li, li ), lir2 = floor2( li + 2hh, li ), lir3 = floor2( -li - 2hh, li );
+//      printf( "floor2(%ld, %ld) = %ld, floor2(%ld, %ld) = %ld, floor2(%ld, %ld) = %ld\n", li, li, lir1, li + 2h, li, lir2, -li - 2h, li, lir3 );
         for ( li = 1; li != 0; li <<= 1 ) {
                 lir1 = floor2( li, li ); lir2 = floor2( li + 2hh, li ); lir3 = floor2( -li - 2hh, li );
 …
         printf( "unsigned long int\n" );
         uli = 0;
         ulir1 = floor2( uli, uli ), ulir2 = floor2( uli + 2hh, uli ), ulir3 = floor2( -uli - 2hh, uli );
         printf( "floor2(%lu, %lu) = %lu, floor2(%lu, %lu) = %lu, floor2(%lu, %lu) = %lu\n", uli, uli, ulir1, uli + 2h, uli, ulir2, -uli - 2h, uli, ulir3 );
+//      uli = 0;
+//      ulir1 = floor2( uli, uli ), ulir2 = floor2( uli + 2hh, uli ), ulir3 = floor2( -uli - 2hh, uli );
+//      printf( "floor2(%lu, %lu) = %lu, floor2(%lu, %lu) = %lu, floor2(%lu, %lu) = %lu\n", uli, uli, ulir1, uli + 2h, uli, ulir2, -uli - 2h, uli, ulir3 );
         for ( uli = 1; uli != 0; uli <<= 1 ) {
                 ulir1 = floor2( uli, uli ); ulir2 = floor2( uli + 2hh, uli ); ulir3 = floor2( -uli - 2hh, uli );
 …
         printf( "long long int\n" );
         lli = 0;
         llir1 = floor2( lli, lli ), llir2 = floor2( lli + 2hh, lli ), llir3 = floor2( -lli - 2hh, lli );
         printf( "floor2(%lld, %lld) = %lld, floor2(%lld, %lld) = %lld, floor2(%lld, %lld) = %lld\n", lli, lli, llir1, lli + 2h, lli, llir2, -lli - 2h, lli, llir3 );
+//      lli = 0;
+//      llir1 = floor2( lli, lli ), llir2 = floor2( lli + 2hh, lli ), llir3 = floor2( -lli - 2hh, lli );
+//      printf( "floor2(%lld, %lld) = %lld, floor2(%lld, %lld) = %lld, floor2(%lld, %lld) = %lld\n", lli, lli, llir1, lli + 2h, lli, llir2, -lli - 2h, lli, llir3 );
         for ( lli = 1; lli != 0; lli <<= 1 ) {
                 llir1 = floor2( lli, lli ); llir2 = floor2( lli + 2hh, lli ); llir3 = floor2( -lli - 2hh, lli );
 …
         printf( "unsigned long long int\n" );
         ulli = 0;
         ullir1 = floor2( ulli, ulli ), ullir2 = floor2( ulli + 2hh, ulli ), ullir3 = floor2( -ulli - 2hh, ulli );
         printf( "floor2(%llu, %llu) = %llu, floor2(%llu, %llu) = %llu, floor2(%llu, %llu) = %llu\n", ulli, ulli, ullir1, ulli + 2h, ulli, ullir2, -ulli - 2h, ulli, ullir3 );
+//      ulli = 0;
+//      ullir1 = floor2( ulli, ulli ), ullir2 = floor2( ulli + 2hh, ulli ), ullir3 = floor2( -ulli - 2hh, ulli );
+//      printf( "floor2(%llu, %llu) = %llu, floor2(%llu, %llu) = %llu, floor2(%llu, %llu) = %llu\n", ulli, ulli, ullir1, ulli + 2h, ulli, ullir2, -ulli - 2h, ulli, ullir3 );
         for ( ulli = 1; ulli != 0; ulli <<= 1 ) {
                 ullir1 = floor2( ulli, ulli ); ullir2 = floor2( ulli + 2hh, ulli ); ullir3 = floor2( -ulli - 2hh, ulli );
 …
         printf( "signed char\n" );
         sc = 0;
         scr1 = ceiling2( sc, sc ), scr2 = ceiling2( sc + 2hh, sc ), scr3 = ceiling2( -sc - 2hh, sc );
         printf( "ceiling2(%hhd, %hhd) = %hhd, ceiling2(%hhd, %hhd) = %hhd, ceiling2(%hhd, %hhd) = %hhd\n", sc, sc, scr1, sc + 2hh, sc, scr2, -sc - 2hh, sc, scr3 );
+//      sc = 0;
+//      scr1 = ceiling2( sc, sc ), scr2 = ceiling2( sc + 2hh, sc ), scr3 = ceiling2( -sc - 2hh, sc );
+//      printf( "ceiling2(%hhd, %hhd) = %hhd, ceiling2(%hhd, %hhd) = %hhd, ceiling2(%hhd, %hhd) = %hhd\n", sc, sc, scr1, sc + 2hh, sc, scr2, -sc - 2hh, sc, scr3 );
         for ( sc = 1; sc != 0; sc <<= 1 ) {
                 scr1 = ceiling2( sc, sc ); scr2 = ceiling2( sc + 2hh, sc ); scr3 = ceiling2( -sc - 2hh, sc );
 …
         printf( "unsigned char\n" );
         uc = 0;
         ucr1 = ceiling2( uc, uc ), ucr2 = ceiling2( uc + 2hh, uc ), ucr3 = ceiling2( -uc - 2hh, uc );
         printf( "ceiling2(%hhu, %hhu) = %hhu, ceiling2(%hhu, %hhu) = %hhu, ceiling2(%hhu, %hhu) = %hhu\n", uc, uc, ucr1, uc + 2uhh, uc, ucr2, -uc - 2uhh, uc, ucr3 );
+//      uc = 0;
+//      ucr1 = ceiling2( uc, uc ), ucr2 = ceiling2( uc + 2hh, uc ), ucr3 = ceiling2( -uc - 2hh, uc );
+//      printf( "ceiling2(%hhu, %hhu) = %hhu, ceiling2(%hhu, %hhu) = %hhu, ceiling2(%hhu, %hhu) = %hhu\n", uc, uc, ucr1, uc + 2uhh, uc, ucr2, -uc - 2uhh, uc, ucr3 );
         for ( uc = 1; uc != 0; uc <<= 1 ) {
                 ucr1 = ceiling2( uc, uc ); ucr2 = ceiling2( uc + 2hh, uc ); ucr3 = ceiling2( -uc - 2hh, uc );
 …
         printf( "short int\n" );
         si = 0;
         sir1 = ceiling2( si, si ), sir2 = ceiling2( si + 2hh, si ), sir3 = ceiling2( -si - 2hh, si );
         printf( "ceiling2(%hd, %hd) = %hd, ceiling2(%hd, %hd) = %hd, ceiling2(%hd, %hd) = %hd\n", si, si, sir1, si + 2h, si, sir2, -si - 2h, si, sir3 );
+//      si = 0;
+//      sir1 = ceiling2( si, si ), sir2 = ceiling2( si + 2hh, si ), sir3 = ceiling2( -si - 2hh, si );
+//      printf( "ceiling2(%hd, %hd) = %hd, ceiling2(%hd, %hd) = %hd, ceiling2(%hd, %hd) = %hd\n", si, si, sir1, si + 2h, si, sir2, -si - 2h, si, sir3 );
         for ( si = 1; si != 0; si <<= 1 ) {
                 sir1 = ceiling2( si, si ); sir2 = ceiling2( si + 2hh, si ); sir3 = ceiling2( -si - 2hh, si );
 …
         printf( "unsigned short int\n" );
         usi = 0;
         usir1 = ceiling2( usi, usi ), usir2 = ceiling2( usi + 2hh, usi ), usir3 = ceiling2( -usi - 2hh, usi );
         printf( "ceiling2(%hu, %hu) = %hu, ceiling2(%hu, %hu) = %hu, ceiling2(%hu, %hu) = %hu\n", usi, usi, usir1, usi + 2uh, usi, usir2, -usi - 2uh, usi, usir3 );
+//      usi = 0;
+//      usir1 = ceiling2( usi, usi ), usir2 = ceiling2( usi + 2hh, usi ), usir3 = ceiling2( -usi - 2hh, usi );
+//      printf( "ceiling2(%hu, %hu) = %hu, ceiling2(%hu, %hu) = %hu, ceiling2(%hu, %hu) = %hu\n", usi, usi, usir1, usi + 2uh, usi, usir2, -usi - 2uh, usi, usir3 );
         for ( usi = 1; usi != 0; usi <<= 1 ) {
                 usir1 = ceiling2( usi, usi ); usir2 = ceiling2( usi + 2hh, usi ); usir3 = ceiling2( -usi - 2hh, usi );
 …
         printf( "int\n" );
         i = 0;
         ir1 = ceiling2( i, i ), ir2 = ceiling2( i + 2hh, i ), ir3 = ceiling2( -i - 2hh, i );
         printf( "ceiling2(%d, %d) = %d, ceiling2(%d, %d) = %d, ceiling2(%d, %d) = %d\n", i, i, ir1, i + 2h, i, ir2, -i - 2h, i, ir3 );
+//      i = 0;
+//      ir1 = ceiling2( i, i ), ir2 = ceiling2( i + 2hh, i ), ir3 = ceiling2( -i - 2hh, i );
+//      printf( "ceiling2(%d, %d) = %d, ceiling2(%d, %d) = %d, ceiling2(%d, %d) = %d\n", i, i, ir1, i + 2h, i, ir2, -i - 2h, i, ir3 );
         for ( i = 1; i != 0; i <<= 1 ) {
                 ir1 = ceiling2( i, i ); ir2 = ceiling2( i + 2hh, i ); ir3 = ceiling2( -i - 2hh, i );
 …
         printf( "unsigned int\n" );
         ui = 0;
         uir1 = ceiling2( ui, ui ), uir2 = ceiling2( ui + 2hh, ui ), uir3 = ceiling2( -ui - 2hh, ui );
         printf( "ceiling2(%u, %u) = %u, ceiling2(%u, %u) = %u, ceiling2(%u, %u) = %u\n", ui, ui, uir1, ui + 2h, ui, uir2, -ui - 2h, ui, uir3 );
+//      ui = 0;
+//      uir1 = ceiling2( ui, ui ), uir2 = ceiling2( ui + 2hh, ui ), uir3 = ceiling2( -ui - 2hh, ui );
+//      printf( "ceiling2(%u, %u) = %u, ceiling2(%u, %u) = %u, ceiling2(%u, %u) = %u\n", ui, ui, uir1, ui + 2h, ui, uir2, -ui - 2h, ui, uir3 );
         for ( ui = 1; ui != 0; ui <<= 1 ) {
                 uir1 = ceiling2( ui, ui ); uir2 = ceiling2( ui + 2hh, ui ); uir3 = ceiling2( -ui - 2hh, ui );
 …
         printf( "long int\n" );
         li = 0;
         lir1 = ceiling2( li, li ), lir2 = ceiling2( li + 2hh, li ), lir3 = ceiling2( -li - 2hh, li );
         printf( "ceiling2(%ld, %ld) = %ld, ceiling2(%ld, %ld) = %ld, ceiling2(%ld, %ld) = %ld\n", li, li, lir1, li + 2h, li, lir2, -li - 2h, li, lir3 );
+//      li = 0;
+//      lir1 = ceiling2( li, li ), lir2 = ceiling2( li + 2hh, li ), lir3 = ceiling2( -li - 2hh, li );
+//      printf( "ceiling2(%ld, %ld) = %ld, ceiling2(%ld, %ld) = %ld, ceiling2(%ld, %ld) = %ld\n", li, li, lir1, li + 2h, li, lir2, -li - 2h, li, lir3 );
         for ( li = 1; li != 0; li <<= 1 ) {
                 lir1 = ceiling2( li, li ); lir2 = ceiling2( li + 2hh, li ); lir3 = ceiling2( -li - 2hh, li );
 …
         printf( "unsigned long int\n" );
         uli = 0;
         ulir1 = ceiling2( uli, uli ), ulir2 = ceiling2( uli + 2hh, uli ), ulir3 = ceiling2( -uli - 2hh, uli );
         printf( "ceiling2(%lu, %lu) = %lu, ceiling2(%lu, %lu) = %lu, ceiling2(%lu, %lu) = %lu\n", uli, uli, ulir1, uli + 2h, uli, ulir2, -uli - 2h, uli, ulir3 );
+//      uli = 0;
+//      ulir1 = ceiling2( uli, uli ), ulir2 = ceiling2( uli + 2hh, uli ), ulir3 = ceiling2( -uli - 2hh, uli );
+//      printf( "ceiling2(%lu, %lu) = %lu, ceiling2(%lu, %lu) = %lu, ceiling2(%lu, %lu) = %lu\n", uli, uli, ulir1, uli + 2h, uli, ulir2, -uli - 2h, uli, ulir3 );
         for ( uli = 1; uli != 0; uli <<= 1 ) {
                 ulir1 = ceiling2( uli, uli ); ulir2 = ceiling2( uli + 2hh, uli ); ulir3 = ceiling2( -uli - 2hh, uli );
 …
         printf( "long long int\n" );
         lli = 0;
         llir1 = ceiling2( lli, lli ), llir2 = ceiling2( lli + 2hh, lli ), llir3 = ceiling2( -lli - 2hh, lli );
         printf( "ceiling2(%lld, %lld) = %lld, ceiling2(%lld, %lld) = %lld, ceiling2(%lld, %lld) = %lld\n", lli, lli, llir1, lli + 2h, lli, llir2, -lli - 2h, lli, llir3 );
+//      lli = 0;
+//      llir1 = ceiling2( lli, lli ), llir2 = ceiling2( lli + 2hh, lli ), llir3 = ceiling2( -lli - 2hh, lli );
+//      printf( "ceiling2(%lld, %lld) = %lld, ceiling2(%lld, %lld) = %lld, ceiling2(%lld, %lld) = %lld\n", lli, lli, llir1, lli + 2h, lli, llir2, -lli - 2h, lli, llir3 );
         for ( lli = 1; lli != 0; lli <<= 1 ) {
                 llir1 = ceiling2( lli, lli ); llir2 = ceiling2( lli + 2hh, lli ); llir3 = ceiling2( -lli - 2hh, lli );
 …
         printf( "unsigned long long int\n" );
         ulli = 0;
         ullir1 = ceiling2( ulli, ulli ), ullir2 = ceiling2( ulli + 2hh, ulli ), ullir3 = ceiling2( -ulli - 2hh, ulli );
         printf( "ceiling2(%llu, %llu) = %llu, ceiling2(%llu, %llu) = %llu, ceiling2(%llu, %llu) = %llu\n", ulli, ulli, ullir1, ulli + 2h, ulli, ullir2, -ulli - 2h, ulli, ullir3 );
+//      ulli = 0;
+//      ullir1 = ceiling2( ulli, ulli ), ullir2 = ceiling2( ulli + 2hh, ulli ), ullir3 = ceiling2( -ulli - 2hh, ulli );
+//      printf( "ceiling2(%llu, %llu) = %llu, ceiling2(%llu, %llu) = %llu, ceiling2(%llu, %llu) = %llu\n", ulli, ulli, ullir1, ulli + 2h, ulli, ullir2, -ulli - 2h, ulli, ullir3 );
         for ( ulli = 1; ulli != 0; ulli <<= 1 ) {
                 ullir1 = ceiling2( ulli, ulli ); ullir2 = ceiling2( ulli + 2hh, ulli ); ullir3 = ceiling2( -ulli - 2hh, ulli );
                 printf( "ceiling2(%llu, %llu) = %llu, ceiling2(%llu, %llu) = %llu, ceiling2(%llu, %llu) = %llu\n", ulli, ulli, ullir1, ulli + 2h, ulli, ullir2, -ulli - 2h, ulli, ullir3 );
+        } // for
+        printf( "\n" );
+#endif // 0
+        //============================================================
+#if 1
+        sout | nl | "ceiling_div" | nl | nl;
+        printf( "signed char\n" );
+        for ( sc = 1; sc != 0; sc <<= 1 ) {
+                scr1 = ceiling_div( sc, sc ); scr2 = ceiling_div( sc + 2hh, sc ); scr3 = ceiling_div( -sc - 2hh, sc );
+                printf( "ceiling_div(%hhd, %hhd) = %hhd, ceiling_div(%hhd, %hhd) = %hhd, ceiling_div(%hhd, %hhd) = %hhd\n", sc, sc, scr1, sc + 2hh, sc, scr2, -sc - 2hh, sc, scr3 );
+        } // for
+        printf( "\n" );
+        printf( "unsigned char\n" );
+        for ( uc = 1; uc != 0; uc <<= 1 ) {
+                ucr1 = ceiling_div( uc, uc ); ucr2 = ceiling_div( uc + 2hh, uc ); ucr3 = ceiling_div( -uc - 2hh, uc );
+                printf( "ceiling_div(%hhu, %hhu) = %hhu, ceiling_div(%hhu, %hhu) = %hhu, ceiling_div(%hhu, %hhu) = %hhu\n", uc, uc, ucr1, uc + 2uhh, uc, ucr2, -uc - 2uhh, uc, ucr3 );
+        } // for
+        printf( "\n" );
+        printf( "short int\n" );
+        for ( si = 1; si != 0; si <<= 1 ) {
+                sir1 = ceiling_div( si, si ); sir2 = ceiling_div( si + 2hh, si ); sir3 = ceiling_div( -si - 2hh, si );
+                printf( "ceiling_div(%hd, %hd) = %hd, ceiling_div(%hd, %hd) = %hd, ceiling_div(%hd, %hd) = %hd\n", si, si, sir1, si + 2h, si, sir2, -si - 2h, si, sir3 );
+        } // for
+        printf( "\n" );
+        printf( "unsigned short int\n" );
+        for ( usi = 1; usi != 0; usi <<= 1 ) {
+                usir1 = ceiling_div( usi, usi ); usir2 = ceiling_div( usi + 2hh, usi ); usir3 = ceiling_div( -usi - 2hh, usi );
+                printf( "ceiling_div(%hu, %hu) = %hu, ceiling_div(%hu, %hu) = %hu, ceiling_div(%hu, %hu) = %hu\n", usi, usi, usir1, usi + 2uh, usi, usir2, -usi - 2uh, usi, usir3 );
+        } // for
+        printf( "\n" );
+        printf( "int\n" );
+        for ( i = 1; i != 0; i <<= 1 ) {
+                ir1 = ceiling_div( i, i ); ir2 = ceiling_div( i + 2hh, i ); ir3 = ceiling_div( -i - 2hh, i );
+                printf( "ceiling_div(%d, %d) = %d, ceiling_div(%d, %d) = %d, ceiling_div(%d, %d) = %d\n", i, i, ir1, i + 2h, i, ir2, -i - 2h, i, ir3 );
+        } // for
+        printf( "\n" );
+        printf( "unsigned int\n" );
+        for ( ui = 1; ui != 0; ui <<= 1 ) {
+                uir1 = ceiling_div( ui, ui ); uir2 = ceiling_div( ui + 2hh, ui ); uir3 = ceiling_div( -ui - 2hh, ui );
+                printf( "ceiling_div(%u, %u) = %u, ceiling_div(%u, %u) = %u, ceiling_div(%u, %u) = %u\n", ui, ui, uir1, ui + 2h, ui, uir2, -ui - 2h, ui, uir3 );
+        } // for
+        printf( "\n" );
+        printf( "long int\n" );
+        for ( li = 1; li != 0; li <<= 1 ) {
+                lir1 = ceiling_div( li, li ); lir2 = ceiling_div( li + 2hh, li ); lir3 = ceiling_div( -li - 2hh, li );
+                printf( "ceiling_div(%ld, %ld) = %ld, ceiling_div(%ld, %ld) = %ld, ceiling_div(%ld, %ld) = %ld\n", li, li, lir1, li + 2h, li, lir2, -li - 2h, li, lir3 );
+        } // for
+        printf( "\n" );
+        printf( "unsigned long int\n" );
+        for ( uli = 1; uli != 0; uli <<= 1 ) {
+                ulir1 = ceiling_div( uli, uli ); ulir2 = ceiling_div( uli + 2hh, uli ); ulir3 = ceiling_div( -uli - 2hh, uli );
+                printf( "ceiling_div(%lu, %lu) = %lu, ceiling_div(%lu, %lu) = %lu, ceiling_div(%lu, %lu) = %lu\n", uli, uli, ulir1, uli + 2h, uli, ulir2, -uli - 2h, uli, ulir3 );
+        } // for
+        printf( "\n" );
+        printf( "long long int\n" );
+        for ( lli = 1; lli != 0; lli <<= 1 ) {
+                llir1 = ceiling_div( lli, lli ); llir2 = ceiling_div( lli + 2hh, lli ); llir3 = ceiling_div( -lli - 2hh, lli );
+                printf( "ceiling_div(%lld, %lld) = %lld, ceiling_div(%lld, %lld) = %lld, ceiling_div(%lld, %lld) = %lld\n", lli, lli, llir1, lli + 2h, lli, llir2, -lli - 2h, lli, llir3 );
+        } // for
+        printf( "\n" );
+        printf( "unsigned long long int\n" );
+        for ( ulli = 1; ulli != 0; ulli <<= 1 ) {
+                ullir1 = ceiling_div( ulli, ulli ); ullir2 = ceiling_div( ulli + 2hh, ulli ); ullir3 = ceiling_div( -ulli - 2hh, ulli );
+                printf( "ceiling_div(%llu, %llu) = %llu, ceiling_div(%llu, %llu) = %llu, ceiling_div(%llu, %llu) = %llu\n", ulli, ulli, ullir1, ulli + 2h, ulli, ullir2, -ulli - 2h, ulli, ullir3 );
         } // for
         printf( "\n" );

tests/castError.cfa

-              r07d867b
+              r22f94a4
 //
+forall(otype T) struct S { T p; };
 int f;
+S(int) sint;
 void f() {
 …
         short int v;
 , v;           // implicit void cast
+        (S(char)) sint;
+}

tests/concurrent/examples/datingService.cfa

-              r07d867b
+              r22f94a4
                 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/signal/block.cfa

r07d867b	r22f94a4
82	82	if( !is_empty( cond ) ) {
83	83
84		$thread * next = front( cond );
	84	$thread * next = ( $thread * ) front( cond );
85	85
86	86	if( ! signal_block( cond ) ) {

tests/concurrent/signal/disjoint.cfa

-              r07d867b
+              r22f94a4
 #endif
+// This tests checks what happens when someone barges in the midle of the release
+// of a bulk of monitors.
 enum state_t { WAIT, SIGNAL, BARGE };
 monitor global_t {};
-global_t mut;
 monitor global_data_t;
 …
         int counter;
         state_t state;
+} data;
+};
+// Use a global struct because the order needs to match with Signaller thread
+struct {
+        global_t mut;
+        global_data_t data;
+} globals;
 condition cond;
 …
 void ?{}( global_data_t & this ) {
         this.counter == 0;
+        this.counter = 0;
         this.state = BARGE;
+}
 …
 thread Barger {};
+void ?{}( Barger & this ) {
+        ((thread&)this){ "Barger Thread" };
+}
 void main( Barger & this ) {
         while( !all_done ) {
                 barge( data );
+                barge( globals.data );
                 yield();
+        }
 …
 thread Waiter {};
+void ?{}( Waiter & this ) {
+        ((thread&)this){ "Waiter Thread" };
+}
 void main( Waiter & this ) {
         while( wait( mut, data ) ) { KICK_WATCHDOG; yield(); }
+        while( wait( globals.mut, globals.data ) ) { KICK_WATCHDOG; yield(); }
+}
 …
 void logic( global_t & mutex a ) {
         signal( cond, a, data );
+        signal( cond, a, globals.data );
         yield( random( 10 ) );
         //This is technically a mutual exclusion violation but the mutex monitor protects us
         bool running = TEST(data.counter < N) && data.counter > 0;
         if( data.state != SIGNAL && running ) {
                 sout | "ERROR Eager signal" | data.state;
+        bool running = TEST(globals.data.counter < N) && globals.data.counter > 0;
+        if( globals.data.state != SIGNAL && running ) {
+                sout | "ERROR Eager signal" | globals.data.state;
+        }
+}
 thread Signaller {};
+void ?{}( Signaller & this ) {
+        ((thread&)this){ "Signaller Thread" };
+}
 void main( Signaller & this ) {
         while( !all_done ) {
                 logic( mut );
+                logic( globals.mut );
                 yield();
+        }

tests/concurrent/waitfor/when.cfa

-              r07d867b
+              r22f94a4
 void arbiter( global_t & mutex this ) {
+        // There is a race at start where callers can get in before the arbiter.
+        // It doesn't really matter here so just restart the loop correctly and move on
+        this.last_call = 6;
         for( int i = 0; i < N; i++ ) {
                    when( this.last_call == 6 ) waitfor( call1 : this ) { if( this.last_call != 1) { serr | "Expected last_call to be 1 got" | this.last_call; } }

tests/copyfile.cfa

-              r07d867b
+              r22f94a4
 //
 // Author           : Peter A. Buhr
 // Created On       : Tue Jul 16 16:47:22 2019
+// Created On       : Fri Jun 19 13:44:05 2020
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Wed Jul 17 18:04:44 2019
 // Update Count     : 26
+// Last Modified On : Sun Jul  5 11:27:43 2020
+// Update Count     : 5
 //
 #include <fstream.hfa>
 #include <stdlib.hfa>                                                                   // new/delete
+#include <exception.hfa>
 int main( int argc, char * argv[] ) {
+        ifstream * in  = &stdin;                                                        // default files
+        ofstream * out = &stdout;
+        ifstream in  = stdin;                                                           // copy default files
+        ofstream out = stdout;
         try {
                 choose ( argc ) {
+                choose ( argc ) {                                                               // terminate if command-line errors
                   case 2, 3:
                           in = new( (const char *)argv[1] );            // open input file first as output creates file
                           if ( argc == 3 ) out = new( (const char *)argv[2] ); // only open output if input opens as output created if nonexistent
                   case 1: ;                                     // use default files
                   default:
                           exit | "Usage [ input-file (default stdin) [ output-file (default stdout) ] ]";
+                        open( in, argv[1] );                                            // open input file first as output creates file
+                        if ( argc == 3 ) open( out, argv[2] );          // do not create output unless input opens
+                  case 1: ;                                                                             // use default files
+                  default:                                                                              // wrong number of options
+                        exit | "Usage" | argv[0] | "[ input-file (default stdin) [ output-file (default stdout) ] ]";
                 } // choose
+        } catch( Open_Failure * ex ; ex->istream == &in ) {
+                exit | "Unable to open input file" | argv[1];
+        } catch( Open_Failure * ex ; ex->ostream == &out ) {
+                close( in );                                                                    // optional
+                exit | "Unable to open output file" | argv[2];
+        } // try
+                char ch;
+                *out | nlOff;                                                                   // turn off auto newline
+                *in  | nlOn;                                                                    // turn on reading newline
+        out | nlOff;                                                                            // turn off auto newline
+        in  | nlOn;                                                                                     // turn on reading newline
+                for () {                                                                                // read all characters
+                        *in | ch;
+                  if ( eof( *in ) ) break;                                              // eof ?
+                        *out | ch;
+                } // for
+        } finally {
+                if ( in  != &stdin  ) delete( in );                             // close file, do not delete stdin!
+                if ( out != &stdout ) delete( out );                    // close file, do not delete stdout!
+        } // try
+        char ch;
+        for () {                                                                                        // read all characters
+                in | ch;
+          if ( eof( in ) ) break;                                                       // eof ?
+                out | ch;
+        } //for
 } // main

tests/errors/.expect/completeType.x64.txt

-              r07d867b
+              r22f94a4
 ?=?: pointer to function
         ... with parameters
           reference to instance of type _104_0_T (not function type)
           instance of type _104_0_T (not function type)
+          reference to instance of type _109_0_T (not function type)
+          instance of type _109_0_T (not function type)
         ... returning
           _retval__operator_assign: instance of type _104_0_T (not function type)
+          _retval__operator_assign: instance of type _109_0_T (not function type)
           ... with attributes:
             Attribute with name: unused

tests/exceptions/.expect/resume.txt

-              r07d867b
+              r22f94a4
 end of try clause
 Exiting: simple try clause
+catch-all
 throwing child exception
 …
 inner catch
 outer catch
+throw
+rethrow
+handle

tests/exceptions/.expect/terminate.txt

-              r07d867b
+              r22f94a4
 simple catch
 Exiting: simple catch clause
+catch-all
 throwing child exception
 …
 inner catch
 outer catch
+throw
+rethrow
+handle

tests/exceptions/conditional.cfa

-              r07d867b
+              r22f94a4
 };
 void num_error_msg(num_error * this) {
+const char * num_error_msg(num_error * this) {
     if ( ! this->msg ) {
         static const char * base = "Num Error with code: X";
 …
         try {
                 throw &exc;
+                throw exc;
         } catch (num_error * error ; 3 == error->virtual_table->code( error )) {
                 caught_num_error(3, error);
 …
         try {
                 throwResume &exc;
+                throwResume exc;
         } catchResume (num_error * error ; 3 == error->virtual_table->code( error )) {
                 caught_num_error(3, error);

tests/exceptions/data-except.cfa

-              r07d867b
+              r22f94a4
         try {
                 throw &except;
+                throw except;
         } catch (paired * exc) {
                 printf("%s(%d, %d)\n", paired_msg(exc), exc->first, exc->second);
 …
         try {
                 throwResume &except;
+                throwResume except;
         } catchResume (paired * exc) {
                 printf("%s(%d, %d)\n", paired_msg(exc), exc->first, exc->second);

tests/exceptions/finally.cfa

-              r07d867b
+              r22f94a4
                 try {
                         printf("termination throw\n");
                         throw &exc;
+                        throw exc;
                 } finally {
                         loud_exit a = "termination inner finally";
 …
                 try {
                         printf("resumption throw\n");
                         throwResume &exc;
+                        throwResume exc;
                 } finally {
                         loud_exit a = "resumption inner finally";

tests/exceptions/interact.cfa

-              r07d867b
+              r22f94a4
         // Resume falls back to terminate.
         try {
                 throwResume &(star){};
+                throwResume (star){};
         } catch (star *) {
                 printf("caught as termination\n");
 …
         try {
                 loud_region a = "try block with resume throw";
                 throwResume &(star){};
+                throwResume (star){};
         } catch (star *) {
                 printf("caught as termination\n");
 …
         try {
                 try {
                         throw &(star){};
+                        throw (star){};
                 } catchResume (star *) {
                         printf("resume catch on terminate\n");
 …
         try {
                 try {
                         throwResume &(star){};
+                        throwResume (star){};
                 } catch (star *) {
                         printf("terminate catch on resume\n");
 …
                 try {
                         try {
                                 throw &(star){};
+                                throw (star){};
                         } catchResume (star *) {
                                 printf("inner resume catch (error)\n");
 …
                 } catch (star * error) {
                         printf("termination catch, will resume\n");
                         throwResume error;
+                        throwResume *error;
+                }
         } catchResume (star *) {
 …
                 try {
                         try {
                                 throwResume &(star){};
+                                throwResume (star){};
                         } catch (star *) {
                                 printf("inner termination catch\n");
 …
                 } catchResume (star * error) {
                         printf("resumption catch, will terminate\n");
                         throw error;
+                        throw *error;
+                }
         } catch (star *) {
 …
                                 try {
                                         printf("throwing resume moon\n");
                                         throwResume &(moon){};
+                                        throwResume (moon){};
                                 } catch (star *) {
                                         printf("termination catch\n");
+                                }
                                 printf("throwing resume star\n");
                                 throwResume &(star){};
+                                throwResume (star){};
                         } catchResume (star *) {
                                 printf("resumption star catch\n");
 …
                 } catchResume (moon *) {
                         printf("resumption moon catch, will terminate\n");
                         throw &(star){};
+                        throw (star){};
+                }
         } catchResume (star *) {

tests/exceptions/resume.cfa

-              r07d867b
+              r22f94a4
 TRIVIAL_EXCEPTION(moment_of, zen);
+void in_void(void);
 int main(int argc, char * argv[]) {
         // The simple throw catchResume test.
 …
                 loud_exit a = "simple try clause";
                 printf("simple throw\n");
                 throwResume &(zen){};
+                throwResume (zen){};
                 printf("end of try clause\n");
         } catchResume (zen * error) {
 …
         printf("\n");
+        // Throw catch-all test.
+        try {
+                throwResume (zen){};
+        } catchResume (exception_t * error) {
+                printf("catch-all\n");
+        }
+        printf("\n");
         // Catch a parent of the given exception.
         try {
                 printf("throwing child exception\n");
                 throwResume &(moment_of){};
+                throwResume (moment_of){};
         } catchResume (zen *) {
                 printf("inner parent match\n");
 …
         try {
                 try {
                         throwResume &(yin){};
+                        throwResume (yin){};
                 } catchResume (zen *) {
                         printf("caught yin as zen\n");
 …
                         loud_exit a = "rethrow inner try";
                         printf("rethrow inner try\n");
                         throwResume &(zen){};
+                        throwResume (zen){};
                 } catchResume (zen *) {
                         loud_exit a = "rethrowing catch clause";
 …
         try {
                 try {
                         throwResume &(yin){};
+                        throwResume (yin){};
                 } catchResume (yin *) {
                         printf("caught yin, will throw yang\n");
                         throwResume &(yang){};
+                        throwResume (yang){};
                 } catchResume (yang *) {
                         printf("caught exception from same try\n");
 …
                 try {
                         printf("throwing first exception\n");
                         throwResume &(yin){};
+                        throwResume (yin){};
                 } catchResume (yin *) {
                         printf("caught first exception\n");
                         try {
                                 printf("throwing second exception\n");
                                 throwResume &(yang){};
+                                throwResume (yang){};
                         } catchResume (yang *) {
                                 printf("caught second exception\n");
 …
         try {
                 try {
                         throwResume &(zen){};
                         throwResume &(zen){};
+                        throwResume (zen){};
+                        throwResume (zen){};
                 } catchResume (zen *) {
                         printf("inner catch\n");
+                }
                 throwResume &(zen){};
+                throwResume (zen){};
         } catchResume (zen *) {
                 printf("outer catch\n");
+        }
+        printf("\n");
+        in_void();
+}
+// Do a throw and rethrow in a void function.
+void in_void(void) {
+        try {
+                try {
+                        printf("throw\n");
+                        throwResume (zen){};
+                } catchResume (zen *) {
+                        printf("rethrow\n");
+                        throwResume;
+                }
+        } catchResume (zen *) {
+                printf("handle\n");
+        }
+}

tests/exceptions/terminate.cfa

-              r07d867b
+              r22f94a4
 TRIVIAL_EXCEPTION(moment_of, zen);
+void in_void(void);
 int main(int argc, char * argv[]) {
         // The simple throw catch test.
 …
                 loud_exit a = "simple try clause";
                 printf("simple throw\n");
                 throw &(zen){};
+                throw (zen){};
                 printf("end of try clause\n");
         } catch (zen * error) {
         loud_exit a = "simple catch clause";
+                loud_exit a = "simple catch clause";
                 printf("simple catch\n");
+        }
+        printf("\n");
+        // Throw catch-all test.
+        try {
+                throw (zen){};
+        } catch (exception_t * error) {
+                printf("catch-all\n");
+        }
         printf("\n");
 …
         try {
                 printf("throwing child exception\n");
                 throw &(moment_of){};
+                throw (moment_of){};
         } catch (zen *) {
                 printf("inner parent match\n");
 …
         try {
                 try {
                         throw &(yin){};
+                        throw (yin){};
                 } catch (zen *) {
                         printf("caught yin as zen\n");
 …
                         loud_exit a = "rethrow inner try";
                         printf("rethrow inner try\n");
                         throw &(zen){};
+                        throw (zen){};
                 } catch (zen *) {
                         loud_exit a = "rethrowing catch clause";
 …
         try {
                 try {
                         throw &(yin){};
+                        throw (yin){};
                 } catch (yin *) {
                         printf("caught yin, will throw yang\n");
                         throw &(yang){};
+                        throw (yang){};
                 } catch (yang *) {
                         printf("caught exception from same try\n");
 …
                 try {
                         printf("throwing first exception\n");
                         throw &(yin){};
+                        throw (yin){};
                 } catch (yin *) {
                         printf("caught first exception\n");
                         try {
                                 printf("throwing second exception\n");
                                 throw &(yang){};
+                                throw (yang){};
                         } catch (yang *) {
                                 printf("caught second exception\n");
 …
         try {
                 try {
                         throw &(zen){};
                         throw &(zen){};
+                        throw (zen){};
+                        throw (zen){};
                 } catch (zen *) {
                         printf("inner catch\n");
+                }
                 throw &(zen){};
+                throw (zen){};
         } catch (zen *) {
                 printf("outer catch\n");
+        }
+        printf("\n");
+        in_void();
+}
+// Do a throw and rethrow in a void function.
+void in_void(void) {
+        try {
+                try {
+                        printf("throw\n");
+                        throw (zen){};
+                } catch (zen *) {
+                        printf("rethrow\n");
+                        throw;
+                }
+        } catch (zen *) {
+                printf("handle\n");
+        }
+}

tests/heap.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Tue Nov  6 17:54:56 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Nov 24 12:34:51 2019
 // Update Count     : 28
+// Last Modified On : Tue Aug  4 06:36:17 2020
+// Update Count     : 56
 //
 …
                 size_t s = (i + 1) * 20;
                 char * area = (char *)malloc( s );
-                if ( area == 0p ) abort( "malloc/free out of memory" );
                 area[0] = '\345'; area[s - 1] = '\345';                 // fill first/last
                 area[malloc_usable_size( area ) - 1] = '\345';  // fill ultimate byte
 …
                 size_t s = i + 1;                                                               // +1 to make initialization simpler
                 locns[i] = (char *)malloc( s );
-                if ( locns[i] == 0p ) abort( "malloc/free out of memory" );
                 locns[i][0] = '\345'; locns[i][s - 1] = '\345'; // fill first/last
                 locns[i][malloc_usable_size( locns[i] ) - 1] = '\345'; // fill ultimate byte
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 char * area = (char *)malloc( s );
-                if ( area == 0p ) abort( "malloc/free out of memory" );
                 area[0] = '\345'; area[s - 1] = '\345';                 // fill first/last
                 area[malloc_usable_size( area ) - 1] = '\345';  // fill ultimate byte
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 locns[i] = (char *)malloc( s );
-                if ( locns[i] == 0p ) abort( "malloc/free out of memory" );
                 locns[i][0] = '\345'; locns[i][s - 1] = '\345'; // fill first/last
                 locns[i][malloc_usable_size( locns[i] ) - 1] = '\345'; // fill ultimate byte
 …
                 size_t s = (i + 1) * 20;
                 char * area = (char *)calloc( 5, s );
-                if ( area == 0p ) abort( "calloc/free out of memory" );
                 if ( area[0] != '\0' || area[s - 1] != '\0' ||
                          area[malloc_usable_size( area ) - 1] != '\0' ||
+                         area[malloc_size( area ) - 1] != '\0' ||
                          ! malloc_zero_fill( area ) ) abort( "calloc/free corrupt storage1" );
                 area[0] = '\345'; area[s - 1] = '\345';                 // fill first/last
 …
                 size_t s = i + 1;
                 locns[i] = (char *)calloc( 5, s );
-                if ( locns[i] == 0p ) abort( "calloc/free out of memory" );
                 if ( locns[i][0] != '\0' || locns[i][s - 1] != '\0' ||
                          locns[i][malloc_usable_size( locns[i] ) - 1] != '\0' ||
+                         locns[i][malloc_size( locns[i] ) - 1] != '\0' ||
                          ! malloc_zero_fill( locns[i] ) ) abort( "calloc/free corrupt storage2" );
                 locns[i][0] = '\345'; locns[i][s - 1] = '\345'; // fill first/last
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 char * area = (char *)calloc( 1, s );
-                if ( area == 0p ) abort( "calloc/free out of memory" );
                 if ( area[0] != '\0' || area[s - 1] != '\0' ) abort( "calloc/free corrupt storage4.1" );
                 if ( area[malloc_usable_size( area ) - 1] != '\0' ) abort( "calloc/free corrupt storage4.2" );
+                if ( area[malloc_size( area ) - 1] != '\0' ) abort( "calloc/free corrupt storage4.2" );
                 if ( ! malloc_zero_fill( area ) ) abort( "calloc/free corrupt storage4.3" );
                 area[0] = '\345'; area[s - 1] = '\345';                 // fill first/last
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 locns[i] = (char *)calloc( 1, s );
-                if ( locns[i] == 0p ) abort( "calloc/free out of memory" );
                 if ( locns[i][0] != '\0' || locns[i][s - 1] != '\0' ||
                          locns[i][malloc_usable_size( locns[i] ) - 1] != '\0' ||
+                         locns[i][malloc_size( locns[i] ) - 1] != '\0' ||
                          ! malloc_zero_fill( locns[i] ) ) abort( "calloc/free corrupt storage5" );
                 locns[i][0] = '\345'; locns[i][s - 1] = '\345'; // fill first/last
 …
                 for ( s; 1 ~ NoOfAllocs ) {                                             // allocation of size 0 can return null
                         char * area = (char *)memalign( a, s );
-                        if ( area == 0p ) abort( "memalign/free out of memory" );
                         //sout | i | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
 …
                         size_t s = i + default_mmap_start();            // cross over point
                         char * area = (char *)memalign( a, s );
-                        if ( area == 0p ) abort( "memalign/free out of memory" );
                         //sout | i | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
 …
                 // initial N byte allocation
                 char * area = (char *)calloc( 5, i );
-                if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
                 if ( area[0] != '\0' || area[i - 1] != '\0' ||
                          area[malloc_usable_size( area ) - 1] != '\0' ||
+                         area[malloc_size( area ) - 1] != '\0' ||
                          ! malloc_zero_fill( area ) ) abort( "calloc/realloc/free corrupt storage1" );
 …
                 for ( s; i ~ 256 * 1024 ~ 26 ) {                                // start at initial memory request
                         area = (char *)realloc( area, s );                      // attempt to reuse storage
-                        if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
                         if ( area[0] != '\0' || area[s - 1] != '\0' ||
                                  area[malloc_usable_size( area ) - 1] != '\0' ||
+                                 area[malloc_size( area ) - 1] != '\0' ||
                                  ! malloc_zero_fill( area ) ) abort( "calloc/realloc/free corrupt storage2" );
                 } // for
 …
                 size_t s = i + default_mmap_start();                    // cross over point
                 char * area = (char *)calloc( 1, s );
                 if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
+//              if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
                 if ( area[0] != '\0' || area[s - 1] != '\0' ||
+                         area[malloc_usable_size( area ) - 1] != '\0' ||
+                         ! malloc_zero_fill( area ) ) abort( "calloc/realloc/free corrupt storage1" );
+                         area[malloc_size( area ) - 1] != '\0' ||
+                         ! malloc_zero_fill( area ) ) //abort( "calloc/realloc/free corrupt storage3" );
+                        printf( "C %zd %d %d %d %d\n", s, area[0] != '\0', area[s - 1] != '\0', area[malloc_size( area ) - 1] != '\0', ! malloc_zero_fill( area ) );
                 // Do not start this loop index at 0 because realloc of 0 bytes frees the storage.
                 for ( r; i ~ 256 * 1024 ~ 26 ) {                                // start at initial memory request
                         area = (char *)realloc( area, r );                      // attempt to reuse storage
                         if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
+//                      if ( area == 0p ) abort( "calloc/realloc/free out of memory" );
                         if ( area[0] != '\0' || area[r - 1] != '\0' ||
                                  area[malloc_usable_size( area ) - 1] != '\0' ||
                                  ! malloc_zero_fill( area ) ) abort( "calloc/realloc/free corrupt storage2" );
+                                 area[malloc_size( area ) - 1] != '\0' ||
+                                 ! malloc_zero_fill( area ) ) abort( "calloc/realloc/free corrupt storage4" );
                 } // for
                 free( area );
 …
                 // initial N byte allocation
                 char * area = (char *)memalign( a, amount );    // aligned N-byte allocation
                 if ( area == 0p ) abort( "memalign/realloc/free out of memory" ); // no storage ?
+//              if ( area == 0p ) abort( "memalign/realloc/free out of memory" ); // no storage ?
                 //sout | alignments[a] | area;
                 if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
 …
                         if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "memalign/realloc/free corrupt storage" );
                         area = (char *)realloc( area, s );                      // attempt to reuse storage
-                        if ( area == 0p ) abort( "memalign/realloc/free out of memory" ); // no storage ?
                         //sout | i | area;
                         if ( (size_t)area % a != 0 ) {                          // check for initial alignment
 …
                 for ( s; 1 ~ limit ) {                                                  // allocation of size 0 can return null
                         char * area = (char *)cmemalign( a, 1, s );
-                        if ( area == 0p ) abort( "cmemalign/free out of memory" );
                         //sout | i | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
 …
                         } // if
                         if ( area[0] != '\0' || area[s - 1] != '\0' ||
                                  area[malloc_usable_size( area ) - 1] != '\0' ||
+                                 area[malloc_size( area ) - 1] != '\0' ||
                                  ! malloc_zero_fill( area ) ) abort( "cmemalign/free corrupt storage" );
                         area[0] = '\345'; area[s - 1] = '\345';         // fill first/last byte
 …
                 // initial N byte allocation
                 char * area = (char *)cmemalign( a, 1, amount ); // aligned N-byte allocation
-                if ( area == 0p ) abort( "cmemalign/realloc/free out of memory" ); // no storage ?
                 //sout | alignments[a] | area;
                 if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
 …
                 } // if
                 if ( area[0] != '\0' || area[amount - 1] != '\0' ||
                          area[malloc_usable_size( area ) - 1] != '\0' ||
+                         area[malloc_size( area ) - 1] != '\0' ||
                          ! malloc_zero_fill( area ) ) abort( "cmemalign/realloc/free corrupt storage1" );
                 area[0] = '\345'; area[amount - 2] = '\345';    // fill first/penultimate byte
 …
                         if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "cmemalign/realloc/free corrupt storage2" );
                         area = (char *)realloc( area, s );                      // attempt to reuse storage
-                        if ( area == 0p ) abort( "cmemalign/realloc/free out of memory" ); // no storage ?
                         //sout | i | area;
                         if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
                                 abort( "cmemalign/realloc/free bad alignment %p", area );
                         } // if
                         if ( area[s - 1] != '\0' || area[s - 1] != '\0' ||
                                  area[malloc_usable_size( area ) - 1] != '\0' ||
+                        if ( area[0] != '\345' || area[s - 1] != '\0' ||
+                                 area[malloc_size( area ) - 1] != '\0' ||
                                  ! malloc_zero_fill( area ) ) abort( "cmemalign/realloc/free corrupt storage3" );
                         area[s - 1] = '\345';                                           // fill last byte
 …
                 // initial N byte allocation
                 char * area = (char *)memalign( a, amount );    // aligned N-byte allocation
-                if ( area == 0p ) abort( "memalign/realloc with align/free out of memory" ); // no storage ?
                 //sout | alignments[a] | area | endl;
                 if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
 …
                         if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "memalign/realloc/free corrupt storage" );
                         area = (char *)realloc( area, a * 2, s );       // attempt to reuse storage
-                        if ( area == 0p ) abort( "memalign/realloc with align/free out of memory" ); // no storage ?
                         //sout | i | area | endl;
                         if ( (size_t)area % a * 2 != 0 ) {                      // check for initial alignment
 …
         for ( size_t a = libAlign() + libAlign(); a <= limit; a += a ) { // generate powers of 2
                 // initial N byte allocation
+                char *area = (char *)cmemalign( a, 1, amount ); // aligned N-byte allocation
+                if ( area == 0p ) abort( "cmemalign/realloc with align/free out of memory" ); // no storage ?
+                char * area = (char *)cmemalign( a, 1, amount ); // aligned N-byte allocation
                 //sout | alignments[a] | area | endl;
                 if ( (size_t)area % a != 0 || malloc_alignment( area ) != a ) { // check for initial alignment
 …
                 } // if
                 if ( area[0] != '\0' || area[amount - 1] != '\0' ||
                          area[malloc_usable_size( area ) - 1] != '\0' ||
+                         area[malloc_size( area ) - 1] != '\0' ||
                          ! malloc_zero_fill( area ) ) abort( "cmemalign/realloc with align/free corrupt storage1" );
                 area[0] = '\345'; area[amount - 2] = '\345';    // fill first/penultimate byte
 …
                         if ( area[0] != '\345' || area[s - 2] != '\345' ) abort( "cmemalign/realloc with align/free corrupt storage2" );
                         area = (char *)realloc( area, a * 2, s );       // attempt to reuse storage
-                        if ( area == 0p ) abort( "cmemalign/realloc with align/free out of memory" ); // no storage ?
                         //sout | i | area | endl;
                         if ( (size_t)area % a * 2 != 0 || malloc_alignment( area ) != a * 2 ) { // check for initial alignment
                                 abort( "cmemalign/realloc with align/free bad alignment %p %jd %jd", area, malloc_alignment( area ), a * 2 );
+                                abort( "cmemalign/realloc with align/free bad alignment %p %zd %zd", area, malloc_alignment( area ), a * 2 );
                         } // if
                         if ( area[s - 1] != '\0' || area[s - 1] != '\0' ||
                                  area[malloc_usable_size( area ) - 1] != '\0' ||
+                                 area[malloc_size( area ) - 1] != '\0' ||
                                  ! malloc_zero_fill( area ) ) abort( "cmemalign/realloc/free corrupt storage3" );
                         area[s - 1] = '\345';                                           // fill last byte

tests/io2.cfa

r07d867b	r22f94a4
121	121
122	122	[int, int, const char *, double] t3 = { 3, 4, "a", 7.2 };
123		sout \| [ 3, 4, ~~"a", 7.2 ];~~
	123	sout \| [ 3, 4, (const char*)"a", 7.2 ]; // workaround trac#207: the const cast should not be needed
124	124	sout \| t3;
125	125	sepSetTuple( sout, " " );

tests/list/.expect/dlist-insert-remove.txt

-              r07d867b
+              r22f94a4
 .7
+-
+~~~~~~~~~~ End removal tests on Headed List: First ~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Test 16-i.  Modifying Freds on MINE
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+==== fred by MINE before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by YOURS before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by MINE after
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+-
+==== fred by YOURS after
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by MINE after
+.7
+-
+.7
+-
+-
+-
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Test 16-ii.  Modifying Freds on YOURS
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+==== fred by MINE before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by YOURS before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by MINE after
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by YOURS after
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+-
+==== fred by YOURS after
+.7
+-
+.7
+-
+-
+-
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Test 16-iii.  Modifying Maries
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+==== mary before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== mary after
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+-
+==== mary after
+.7
+-
+.7
+-
+-
+-
+~~~~~~~~~~ End removal tests on Headed List: Last ~~~~~~~~~~
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Test 17-i.  Modifying Freds on MINE
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+==== fred by MINE before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by YOURS before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by MINE after
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+-
+==== fred by YOURS after
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by MINE after
+.7
+-
+.7
+-
+-
+-
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Test 17-ii.  Modifying Freds on YOURS
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+==== fred by MINE before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by YOURS before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by MINE after
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== fred by YOURS after
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+-
+==== fred by YOURS after
+.7
+-
+.7
+-
+-
+-
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Test 17-iii.  Modifying Maries
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+==== mary before
+.7
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+.7
+-
+==== mary after
+.7
+.7
+-
+.7
+-
+.7
+-
+.7
+.7
+-
+==== mary after
+.7
+-
+.7
+-
+-
+-

tests/list/dlist-insert-remove.cfa

-              r07d867b
+              r22f94a4
 ////////////////////////////////////////////////////////////
 //
+// Section 4f
+//
+// Test cases of pop_first, pop_last
+//
+// Example of call-side user code
+//
+////////////////////////////////////////////////////////////
+// These cases assume element removal at first-last is correct
+void test__pop_first__fred_mine() {
+        fred f1 = {1.7};
+        fred f2 = {2.7};
+        fred f3 = {3.7};
+        dlist(fred_in_mine, fred) flm;
+        insert_last(flm, f1);
+        insert_last(flm, f2);
+        insert_last(flm, f3);
+        dlist(fred_in_yours, fred) fly;
+        insert_last(fly, f1);
+        insert_last(fly, f2);
+        insert_last(fly, f3);
+        printMyFreddies(flm`first, flm`last, 1);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        printYourFreddies(fly`first, fly`last, 1);   // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        verify(validate(fly));
+        verify(validate(flm));
+        fred & popped = pop_first(flm);
+        verify(validate(fly));
+        verify(validate(flm));
+        printMyFreddies(flm`first, flm`last, 0);     // 2.7, 3.7;       2.7;  3.7;  3.7, 2.7      (modified)
+        printYourFreddies(fly`first, fly`last, 0);   // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7 (unmodified)
+        // observe f1 is now solo in mine; in yours, it was just traversed
+        printMyFreddies(f1, *0p, 0);    // 1.7; 1.7; ;
+        assert( &popped == & f1 );
+}
+void test__pop_first__fred_yours() {
+        fred f1 = {1.7};
+        fred f2 = {2.7};
+        fred f3 = {3.7};
+        dlist(fred_in_mine, fred) flm;
+        insert_last(flm, f1);
+        insert_last(flm, f2);
+        insert_last(flm, f3);
+        dlist(fred_in_yours, fred) fly;
+        insert_last(fly, f1);
+        insert_last(fly, f2);
+        insert_last(fly, f3);
+        printMyFreddies(flm`first, flm`last, 1);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        printYourFreddies(fly`first, fly`last, 1);   // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        verify(validate(fly));
+        verify(validate(flm));
+        fred & popped = pop_first(fly);
+        verify(validate(fly));
+        verify(validate(flm));
+        printMyFreddies(flm`first, flm`last, 0);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7 (unmodified)
+        printYourFreddies(fly`first, fly`last, 0);   // 2.7, 3.7;       2.7;  3.7;  3.7, 2.7      (modified)
+        // observe f1 is now solo in yours; in mine, it was just traversed
+        printYourFreddies(f1, *0p, 0);    // 1.7; 1.7; ;
+        assert( &popped == &f1 );
+}
+void test__pop_first__maries() {
+        mary m1 = {1.7};
+        mary m2 = {2.7};
+        mary m3 = {3.7};
+        dlist(mary, mary) ml;
+        insert_last(ml, m1);
+        insert_last(ml, m2);
+        insert_last(ml, m3);
+        printMariatheotokos(ml`first, ml`last, 1);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        verify(validate(ml));
+        mary & popped = pop_first(ml);
+        verify(validate(ml));
+        printMariatheotokos(ml`first, ml`last, 0);     // 2.7, 3.7;       2.7;  3.7;  3.7, 2.7      (modified)
+        // observe m1 is now solo
+        printMariatheotokos(m1, *0p, 0);               // 1.7; 1.7; ;
+        assert( &popped == &m1 );
+}
+void test__pop_last__fred_mine() {
+        fred f1 = {1.7};
+        fred f2 = {2.7};
+        fred f3 = {3.7};
+        dlist(fred_in_mine, fred) flm;
+        insert_last(flm, f1);
+        insert_last(flm, f2);
+        insert_last(flm, f3);
+        dlist(fred_in_yours, fred) fly;
+        insert_last(fly, f1);
+        insert_last(fly, f2);
+        insert_last(fly, f3);
+        printMyFreddies(flm`first, flm`last, 1);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        printYourFreddies(fly`first, fly`last, 1);   // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        verify(validate(fly));
+        verify(validate(flm));
+        fred & popped = pop_last(flm);
+        verify(validate(fly));
+        verify(validate(flm));
+        printMyFreddies(flm`first, flm`last, 0);     // 1.7, 2.7;       1.7;  2.7;  2.7, 1.7      (modified)
+        printYourFreddies(fly`first, fly`last, 0);   // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7 (unmodified)
+        // observe f3 is now solo in mine; in yours, it was just traversed
+        printMyFreddies(f3, *0p, 0);    // 3.7; 3.7; ;
+        assert( &popped == & f3 );
+}
+void test__pop_last__fred_yours() {
+        fred f1 = {1.7};
+        fred f2 = {2.7};
+        fred f3 = {3.7};
+        dlist(fred_in_mine, fred) flm;
+        insert_last(flm, f1);
+        insert_last(flm, f2);
+        insert_last(flm, f3);
+        dlist(fred_in_yours, fred) fly;
+        insert_last(fly, f1);
+        insert_last(fly, f2);
+        insert_last(fly, f3);
+        printMyFreddies(flm`first, flm`last, 1);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        printYourFreddies(fly`first, fly`last, 1);   // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        verify(validate(fly));
+        verify(validate(flm));
+        fred & popped = pop_last(fly);
+        verify(validate(fly));
+        verify(validate(flm));
+        printMyFreddies(flm`first, flm`last, 0);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7 (unmodified)
+        printYourFreddies(fly`first, fly`last, 0);   // 1.7, 2.7;       1.7;  2.7;  2.7, 1.7      (modified)
+        // observe f3 is now solo in yours; in mine, it was just traversed
+        printYourFreddies(f3, *0p, 0);    // 3.7; 3.7; ;
+        assert( &popped == & f3 );
+}
+void test__pop_last__maries() {
+        mary m1 = {1.7};
+        mary m2 = {2.7};
+        mary m3 = {3.7};
+        dlist(mary, mary) ml;
+        insert_last(ml, m1);
+        insert_last(ml, m2);
+        insert_last(ml, m3);
+        printMariatheotokos(ml`first, ml`last, 1);     // 1.7, 2.7, 3.7;  1.7;  3.7;  3.7, 2.7, 1.7
+        verify(validate(ml));
+        mary & popped = pop_last(ml);
+        verify(validate(ml));
+        printMariatheotokos(ml`first, ml`last, 0);     // 1.7, 1.7;       1.7;  2.7;  2.7, 1.7      (modified)
+        // observe m1 is now solo
+        printMariatheotokos(m3, *0p, 0);               // 3.7; 3.7; ;
+        assert( &popped == &m3 );
+}
+////////////////////////////////////////////////////////////
+//
 // Section 5
 //
 …
         test__remove_of_sole__mary();
+        sout | "";
+        sout | "~~~~~~~~~~ End removal tests on Headed List: First ~~~~~~~~~~";
+        sout | "";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        sout | "Test 16-i.  Modifying Freds on MINE";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        test__pop_first__fred_mine();
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        sout | "Test 16-ii.  Modifying Freds on YOURS";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        test__pop_first__fred_yours();
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        sout | "Test 16-iii.  Modifying Maries";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        test__pop_first__maries();
+        sout | "";
+        sout | "~~~~~~~~~~ End removal tests on Headed List: Last ~~~~~~~~~~";
+        sout | "";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        sout | "Test 17-i.  Modifying Freds on MINE";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        test__pop_last__fred_mine();
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        sout | "Test 17-ii.  Modifying Freds on YOURS";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        test__pop_last__fred_yours();
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        sout | "Test 17-iii.  Modifying Maries";
+        sout | "~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~";
+        test__pop_last__maries();
         return 0;
+}

tests/manipulatorsInput.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Sat Jun  8 17:58:54 2019
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Thu Jun 13 17:41:43 2019
 // Update Count     : 37
+// Last Modified On : Wed Jul 15 15:56:03 2020
+// Update Count     : 47
 //
 …
                 sin | ignore( wdi( 8, ldc ) );                  sout | ldc;
+        }
+#if defined( __SIZEOF_INT128__ )
+        {
+                int128 val;
+                for ( 15 ) {
+                        sin | val;
+                        sout | val;
+                }
+        }
+#endif // __SIZEOF_INT128__
 } // main

tests/manipulatorsOutput3.cfa

-              r07d867b
+              r22f94a4
 #include <fstream.hfa>
 int main() {
+    int128 x = 0xffff, y = 0x2;
+    x <<= 64;
+    x += 0xffff;
+    y <<= 64;
+    y += 0123;
+    y |= 0x8000000000000000;
+    x = -x;
+    sout | x;
+    sout | sign(x);
+    x = -x;
+    sout | sign(x);
+    sout | nl;
+#if 1
+    sout | bin(x);
+    sout | upcase(bin(x));
+    sout | nobase(bin(x));
+    sout | wd( 95, upcase(bin(x)) );
+    sout | wd( 95,90, upcase(bin(x)) );
+    sout | wd( 25,30, upcase(hex(x)) );
+    sout | nl;
+    printf( "%#.10o\n", 0123 );
+    sout | wd( 1,10, oct(0123) );
+    sout | oct(x);
+    sout | nobase(oct(x));
+    sout | wd( 45, oct(0123) );
+    sout | wd( 45,40, oct(0123) );
+    sout | wd( 40,45, oct(0123) );
+    sout | wd( 45, oct(x) );
+    sout | wd( 45,40, oct(x) );
+    sout | wd( 40,45, oct(x) );
+    sout | left(wd( 45, oct(0123) )) | 'X';
+    sout | left(wd( 45, oct(x) )) | 'X';
+    sout | left(wd( 45, oct(y) )) | 'X';
+    sout | left(wd( 45,40, oct(0123) )) | 'X';
+    sout | left(wd( 45,40, oct(x) )) | 'X';
+    sout | left(wd( 45,40, oct(y) )) | 'X';
+    sout | left(wd( 40,45, oct(0123) )) | 'X';
+    sout | left(wd( 40,45, oct(x) )) | 'X';
+    sout | left(wd( 40,45, oct(y) )) | 'X';
+    printf( "%#-1.10oX\n", 0123 );
+    sout | left(wd( 1,10, oct(0123) )) | 'X';
+    printf( "%#-40.10oX\n", 0123 );
+    sout | left(wd( 40,10, oct(0123) )) | 'X';
+    sout | left(wd( 40,10, oct(x) )) | 'X';
+    sout | left(wd( 40,10, oct(y) )) | 'X';
+    sout | left(wd( 10,40, oct(0123) )) | 'X';
+    sout | left(wd( 10,40, oct(x) )) | 'X';
+    sout | left(wd( 10,40, oct(y) )) | 'X';
+    y = 01234567;
+    sout | left(wd( 45, 49, oct(y) )) | 'X';
+    y = -y;
+    sout | wd(0, oct(y)) | 'Y';
+    sout | left(wd(0, oct(y))) | 'Y';
+    sout | nl;
+    sout | hex(x);
+    sout | upcase(hex(x));
+    sout | nobase(hex(x));
+    sout | wd( 45, upcase(hex(x)) );
+    sout | wd( 45,40, upcase(hex(x)) );
+    sout | wd( 45,49, upcase(hex(x)) );
+    sout | left(wd( 45, upcase(hex(x)) )) | 'X';
+    sout | left(wd( 45,40, upcase(hex(x)) )) | 'X';
+    sout | left(wd( 45,49, upcase(hex(x)) )) | 'X';
+    sout | nl;
+    int128 divisor = 0x4b3b4ca85a86c47a;
+    divisor <<= 16;
+    divisor += 0x98a224000000000;
+    // base 2
+    sout | "base 2";
+    sout | bin(divisor);
+    sout | upcase(bin(divisor));
+    sout | wd(38, upcase(bin(divisor)));
+    sout | wd(40, upcase(bin(divisor)));
+    sout | wd(40, 38, upcase(bin(divisor)));
+    sout | wd(40, 30, upcase(bin(divisor)));
+    sout | pad0(sign(wd(40, 38, upcase(bin(divisor)))));
+    sout | nl;
+    // oct
+    sout | "base 8";
+    sout | upcase(oct(divisor));
+    sout | wd(38, upcase(oct(divisor)));
+    sout | wd(40, upcase(oct(divisor)));
+    sout | wd(40, 38, upcase(oct(divisor)));
+    sout | wd(40, 30, upcase(oct(divisor)));
+    sout | pad0(sign(wd(40, 38, upcase(oct(divisor)))));
+    sout | nl;
+    // decimal
+    sout | "base 10";
+    sout | divisor;
+    sout | wd(2, divisor);
+    sout | wd(3, divisor);
+    sout | wd(10, divisor);
+    sout | wd(24, divisor);
+    sout | wd(38, divisor);
+    sout | wd(39, divisor);
+    sout | wd(40, divisor);
+    sout | wd(40, 30, divisor);
+    sout | wd(40, 38, divisor);
+    sout | wd(40, 40, divisor);
+    sout | pad0(wd(40, divisor));
+    sout | pad0(sign(wd(40,divisor)));
+    sout | nl;
+    // hex
+    sout | "base 16";
+    sout | upcase(hex(divisor));
+    sout | wd(38, upcase(hex(divisor)));
+    sout | wd(40, upcase(hex(divisor)));
+    sout | wd(40, 38, upcase(hex(divisor)));
+    sout | wd(40, 30, upcase(hex(divisor)));
+    sout | pad0(sign(wd(40, 38, upcase(hex(divisor)))));
+    sout | nl;
+    sout | bin(divisor);
+    sout | upcase(bin(divisor));
+    sout | oct(divisor);
+    sout | hex(divisor);
+    sout | upcase(hex(divisor));
+    sout | nobase(bin(divisor)) | nobase(oct(divisor)) | nobase(hex(divisor));
+    sout | sign(divisor);
+    sout | -divisor;
+    sout | sign(-divisor);
+    sout | wd(2, divisor);
+    sout | wd(3,10,divisor);
+    sout | left(wd(40,divisor)) | 'X';
+    sout | left(sign(wd(40, divisor))) | 'X';
+    sout | left(sign(wd(0,40, divisor))) | 'X';
+    printf( "%-+1.40dX\n", 123456789 );
+#endif // 0
+}
+        // int128 printing
+        int128 x = 0xffff, y = 0x2;
+        x <<= 64;
+        x += 0xffff;
+        y <<= 64;
+        y += 0123;
+        y |= 0x8000000000000000;
+        x = -x;
+        sout | x;
+        sout | sign(x);
+        x = -x;
+        sout | sign(x);
+        sout | nl;
+        int128 divisor = 0x4b3b4ca85a86c47a;
+        divisor <<= 16;
+        divisor += 0x98a224000000000;
+        // base 2
+        sout | "base 2";
+        sout | bin(x);
+        sout | upcase(bin(x));
+        sout | nobase(bin(x));
+        sout | wd( 95, upcase(bin(x)) );
+        sout | wd( 95,90, upcase(bin(x)) );
+        sout | wd( 25,30, upcase(hex(x)) );
+        sout | nl;
+        sout | bin(divisor);
+        sout | upcase(bin(divisor));
+        sout | wd(38, upcase(bin(divisor)));
+        sout | wd(40, upcase(bin(divisor)));
+        sout | wd(40, 38, upcase(bin(divisor)));
+        sout | wd(40, 30, upcase(bin(divisor)));
+        sout | pad0(sign(wd(40, 38, upcase(bin(divisor)))));
+        sout | nl;
+        // oct
+        sout | "base 8";
+        printf( "%#.10o\n", 0123 );
+        sout | wd( 1,10, oct(0123) );
+        sout | oct(x);
+        sout | nobase(oct(x));
+        sout | wd( 45, oct(0123) );
+        sout | wd( 45,40, oct(0123) );
+        sout | wd( 40,45, oct(0123) );
+        sout | wd( 45, oct(x) );
+        sout | wd( 45,40, oct(x) );
+        sout | wd( 40,45, oct(x) );
+        sout | left(wd( 45, oct(0123) )) | 'X';
+        sout | left(wd( 45, oct(x) )) | 'X';
+        sout | left(wd( 45, oct(y) )) | 'X';
+        sout | left(wd( 45,40, oct(0123) )) | 'X';
+        sout | left(wd( 45,40, oct(x) )) | 'X';
+        sout | left(wd( 45,40, oct(y) )) | 'X';
+        sout | left(wd( 40,45, oct(0123) )) | 'X';
+        sout | left(wd( 40,45, oct(x) )) | 'X';
+        sout | left(wd( 40,45, oct(y) )) | 'X';
+        printf( "%#-1.10oX\n", 0123 );
+        sout | left(wd( 1,10, oct(0123) )) | 'X';
+        printf( "%#-40.10oX\n", 0123 );
+        sout | left(wd( 40,10, oct(0123) )) | 'X';
+        sout | left(wd( 40,10, oct(x) )) | 'X';
+        sout | left(wd( 40,10, oct(y) )) | 'X';
+        sout | left(wd( 10,40, oct(0123) )) | 'X';
+        sout | left(wd( 10,40, oct(x) )) | 'X';
+        sout | left(wd( 10,40, oct(y) )) | 'X';
+        y = 01234567;
+        sout | left(wd( 45, 49, oct(y) )) | 'X';
+        y = -y;
+        sout | wd(0, oct(y)) | 'Y';
+        sout | left(wd(0, oct(y))) | 'Y';
+        sout | nl;
+        sout | upcase(oct(divisor));
+        sout | wd(38, upcase(oct(divisor)));
+        sout | wd(40, upcase(oct(divisor)));
+        sout | wd(40, 38, upcase(oct(divisor)));
+        sout | wd(40, 30, upcase(oct(divisor)));
+        sout | pad0(sign(wd(40, 38, upcase(oct(divisor)))));
+        sout | nl;
+        // decimal
+        sout | "base 10";
+        sout | divisor;
+        sout | wd(2, divisor);
+        sout | wd(3, divisor);
+        sout | wd(10, divisor);
+        sout | wd(24, divisor);
+        sout | wd(38, divisor);
+        sout | wd(39, divisor);
+        sout | wd(40, divisor);
+        sout | wd(40, 30, divisor);
+        sout | wd(40, 38, divisor);
+        sout | wd(40, 40, divisor);
+        sout | pad0(wd(40, divisor));
+        sout | pad0(sign(wd(40,divisor)));
+        sout | nl;
+        // hex
+        sout | "base 16";
+        sout | hex(x);
+        sout | upcase(hex(x));
+        sout | nobase(hex(x));
+        sout | wd( 45, upcase(hex(x)) );
+        sout | wd( 45,40, upcase(hex(x)) );
+        sout | wd( 45,49, upcase(hex(x)) );
+        sout | left(wd( 45, upcase(hex(x)) )) | 'X';
+        sout | left(wd( 45,40, upcase(hex(x)) )) | 'X';
+        sout | left(wd( 45,49, upcase(hex(x)) )) | 'X';
+        sout | nl;
+        sout | upcase(hex(divisor));
+        sout | wd(38, upcase(hex(divisor)));
+        sout | wd(40, upcase(hex(divisor)));
+        sout | wd(40, 38, upcase(hex(divisor)));
+        sout | wd(40, 30, upcase(hex(divisor)));
+        sout | pad0(sign(wd(40, 38, upcase(hex(divisor)))));
+        sout | nl;
+        // extras
+        sout | "extras";
+        sout | bin(divisor);
+        sout | upcase(bin(divisor));
+        sout | oct(divisor);
+        sout | hex(divisor);
+        sout | upcase(hex(divisor));
+        sout | nobase(bin(divisor)) | nobase(oct(divisor)) | nobase(hex(divisor));
+        sout | sign(divisor);
+        sout | -divisor;
+        sout | sign(-divisor);
+        sout | wd(2, divisor);
+        sout | wd(3,10,divisor);
+        sout | left(wd(40,divisor)) | 'X';
+        sout | left(sign(wd(40, divisor))) | 'X';
+        sout | left(sign(wd(0,40, divisor))) | 'X';
+        printf( "%-+1.40dX\n", 123456789 );
+        int128 i128;
+        unsigned int128 ui128;
+        i128 = -1000;
+        for ( 10 ) {
+                sout | left( sign( wd( 2, i128 ) ) ) | left( wd( 2, hex( i128 ) ) ) | left( wd( 2, oct( i128 ) ) );
+                sout | left( wd( 2, bin( i128 ) ) );
+                i128 += 1;
+        }
+        sout | nl;
+        sout | nl;
+        ui128 = 0x7fffffffffffffff;
+        ui128 <<= 64;
+        ui128 += 0xffffffffffffffff;
+        sout | left( wd( 160, i128 ) );
+        sout | left( sign( wd( 0, i128 ) ) );
+        sout | left( wd( 0, hex( i128 ) ) );
+        sout | left( wd( 0, oct( i128 ) ) );
+        sout | left( wd( 0, bin( i128 ) ) );
+        sout | left( sign( wd( 1, i128 ) ) );
+        sout | left( wd( 1, hex( i128 ) ) );
+        sout | left( wd( 1, oct( i128 ) ) );
+        sout | left( wd( 1, bin( i128 ) ) );
+        sout | left( sign( wd( 32, i128 ) ) );
+        sout | left( wd( 32, hex( i128 ) ) );
+        sout | left( wd( 32, oct( i128 ) ) );
+        sout | left( wd( 32, bin( i128 ) ) );
+        sout | left( sign( wd( 160, i128 ) ) );
+        sout | left( wd( 160, hex( i128 ) ) );
+        sout | left( wd( 160, oct( i128 ) ) );
+        sout | left( wd( 160, bin( i128 ) ) );
+        sout | left( sign( wd( 160, i128 ) ) );
+        sout | left( wd( 160, upcase(hex( i128 )) ) );
+        sout | left( wd( 160, upcase(oct( i128 ) )) );
+        sout | left( wd( 160, upcase(bin( i128 )) ) );
+        x = 1234;
+        x <<= 64;
+        x += 5678;
+        sout | x | 'X';
+        sout | wd(45, 20, oct(x)) | 'X';
+        sout | left(wd(45, 20, oct(x))) | 'X';
+        sout | wd(45, 21, oct(x)) | 'X';
+        sout | left(wd(45, 21, oct(x))) | 'X';
+        sout | wd(45, 22, oct(x)) | 'X';
+        sout | left(wd(45, 22, oct(x))) | 'X';
+        sout | wd(45, 36, oct(x)) | 'X';
+        sout | left(wd(45, 36, oct(x))) | 'X';
+        sout | wd(45, 46, oct(x)) | 'X';
+        sout | left(wd(45, 46, oct(x))) | 'X';
+        sout | left(wd(45, 20, oct(x))) | 'X';
+        sout | left(wd(45, 22, oct(x))) | 'X';
+        sout | left(wd(45, 24, oct(x))) | 'X';
+        sout | left(wd(45, 26, oct(x))) | 'X';
+        sout | left(wd(45, 28, oct(x))) | 'X';
+        y = -x;
+        sout | oct(y) | 'Y';
+        sout | left(wd(0, oct(y))) | 'Y';
+        sout | left(wd(20, oct(y))) | 'Y';
+        sout | left(wd(26, oct(y))) | 'Y';
+        sout | left(wd(40, oct(y))) | 'Y';
+        sout | left(wd(41, oct(y))) | 'Y';
+        sout | left(wd(45, oct(y))) | 'Y';
+        sout | left(wd(45, 49, oct(y))) | 'Y';
+        sout | left(wd(45, 4, oct(y))) | 'Y';
+        sout | left(wd( 45, oct(0123) )) | 'X';
+        sout | left(wd( 45, oct(x) )) | 'X';
+        sout | left(wd( 45,40, oct(0123) )) | 'X';
+        sout | left(wd( 45,40, oct(x) )) | 'X';
+        sout | left(wd( 40,45, oct(0123) )) | 'X';
+        sout | left(wd( 40,45, oct(x) )) | 'X';
+        sout | left(wd( 40,10, oct(0123) )) | 'X';
+        sout | left(wd( 40,10, oct(x) )) | 'X';
+        sout | left(wd( 40,10, oct(y) )) | 'X';
+        sout | left(wd( 10,40, oct(0123) )) | 'X';
+        sout | left(wd( 10,40, oct(x) )) | 'X';
+        sout | left(wd( 10,40, oct(y) )) | 'X';
+        x = 0xffff, y = 0x2;
+        int128 z = 0x_ffff_ffff_ffff_ffff;
+        x <<= 64;
+        x += 0xffff;
+        y <<= 64;
+        y += 0123;
+        y |= 0x8000000000000000;
+        for ( int128 i = 0; i < 8; i += 1 ) {
+                sout | nobase(oct(z + 0x8000000000000000 * i));
+        } // for
+        sout | bin(x);
+        sout | upcase(bin(x));
+        sout | nobase(bin(x));
+        sout | wd( 95, upcase(bin(x)) );
+        sout | wd( 95,90, upcase(bin(x)) );
+        sout | wd( 90,95, upcase(bin(x)) );
+        sout | left(bin(x));
+        sout | left(upcase(bin(x)));
+        sout | left(nobase(bin(x)));
+        sout | left(wd( 95, upcase(bin(x)) ));
+        sout | left(wd( 95,90, upcase(bin(x)) ));
+        sout | left(wd( 95,81, upcase(bin(x)) ));
+        sout | left(wd( 95,80, upcase(bin(x)) ));
+        sout | left(wd( 95,79, upcase(bin(x)) ));
+        sout | left(wd( 95,90, upcase(bin(0xffff)) ));
+        sout | left(wd( 68,64, upcase(bin(0xffff)) ));
+        sout | left(wd( 90,95, upcase(bin(x)) ));
+        printf( "%#30.25X\n", 0xffff );
+        sout | wd( 30,25, upcase(hex(0xffff)) );
+        printf( "%#25.30X\n", 0xffff );
+        sout | wd( 25,30, upcase(hex(0xffff)) );
+        sout | oct(y);
+        sout | wd( 45, oct(y) );
+        sout | left(wd( 45, oct(y) )) | 'X';
+        sout | left(wd( 40,10, oct(0123) )) | 'X';
+        sout | left(wd( 40,10, oct(x) )) | 'X';
+        sout | left(wd( 40,10, oct(y) )) | 'X';
+        sout | left(wd( 10,40, oct(0123) )) | 'X';
+        sout | left(wd( 10,40, oct(x) )) | 'X';
+        sout | left(wd( 10,40, oct(y) )) | 'X';
+        i128 = -10;
+        for ( 25 ) {
+                sout | left( sign( wd( 20, i128 ) ) ) | left( wd( 20, hex( i128 ) ) ) | left( wd( 20, oct( i128 ) ) );
+                sout | left( wd( 20, bin( i128 ) ) );
+                i128 += 1;
+        } // for
+        sout | nl;
+        i128 = 0x7fffffffffffffff;
+        i128 <<= 64;
+        i128 += 0xfffffffffffffffa;
+//      for ( 20 ) {
+        volatile int stop = 20;                                                         // gcc compiler bug
+        for ( int i = 0; i < stop; i += 1 ) {
+                sout | i128;
+                sout | left( sign( wd( 45, i128 ) ) ) | left( wd( 45, hex( i128 ) ) ) | left( wd( 45, oct( i128 ) ) );
+                sout | left( wd( 45, bin( i128 ) ) );
+                i128 += 1;
+        } // for
+        sout | nl;
+        ui128 = 0x7fffffffffffffff;
+        ui128 <<= 64;
+        ui128 += 0xfffffffffffffffa;
+        for ( 20 ) {
+                sout | ui128;
+                ui128 += 1;
+        }
+        sout | nl;
+        ui128 = 0xffffffffffffffff;
+        ui128 <<= 64;
+        ui128 += 0xfffffffffffffffa;
+        for ( 20 ) {
+                sout | ui128;
+                ui128 += 1;
+        }
+        // int128 constants (and printing)
+        int128 v = 0xffff_ffffffff_ffffffff_L128 + 0xffffffff_ffffffff_ffffffff_ffffffff_L128;
+        sout | hex(v);
+        v = 0xffff_ffffffff_ffffffff_L128 + 0xffffffff_ffffffff_ffffffff_ffffffff_L128;
+        sout | hex(v);
+        sout | nl;
+        sout | "binary";
+        sout | bin(v);
+        sout | bin(0b_11111111111111111111111111111111_L128);
+        sout | bin(0b_11111111111111111111111111111111_11111111111111111111111111111111_L128);
+        sout | bin(0b_11111111111111111111111111111111_11111111111111111111111111111111_11111111111111111111111111111111_L128);
+        sout | bin(0b_11111111111111111111111111111111_11111111111111111111111111111111_11111111111111111111111111111111_11111111111111111111111111111111_L128);
+        sout | hex(0b_10100010001101000101011001111000_L128);
+        sout | hex(0b_10100010001101000101011001111000_10100111011001010100001100100001_L128);
+        sout | hex(0b_10100010001101000101011001111000_10100111011001010100001100100001_11000010001101000101011001111000_L128);
+        sout | hex(0b_10100010001101000101011001111000_10100111011001010100001100100001_11000010001101000101011001111000_11010111010101010100001100100001_L128);
+        sout | nl;
+        sout | "octal";
+        sout | oct(v);
+        sout | oct(0_123456_L128u);
+        sout | oct(0_123456_65432_uL128);
+        sout | oct(0_123456_65432_34567_L128);
+        sout | oct(0_123456_65432_34567_76543_L128);
+        sout | oct(0_123456_65432_34567_76543_23456_L128);
+        sout | oct(0_123456_65432_34567_76543_23456_65432_L128);
+        sout | oct(0_123456_65432_34567_76543_23456_65432_34567_L128);
+        sout | oct(0_123456_65432_34567_76543_23456_65432_34567_76543_L128);
+        sout | oct(0_1111111111111111111L);
+        sout | oct(0_11111111111111111111L);
+        sout | oct(0_111111111111111111111L);
+        sout | nl;
+        sout | "decimal";
+        sout | v;
+        sout | 42_798_L128 | oct(42_798_L128);
+        sout | 1_402_432_282_L128 | oct(1_402_432_282_L128);
+        sout | 45_954_901_031_287_L128 | oct(45_954_901_031_287_L128);
+        sout | 1_505_850_196_993_244_515_L128 | oct(1_505_850_196_993_244_515_L128);
+        sout | 394_749_758_663_249_135_511_342_L128 | oct(394_749_758_663_249_135_511_342_L128);
+        sout | 12_935_154_696_204_706_112_391_834_394_L128 | oct(12_935_154_696_204_706_112_391_834_394_L128);
+        sout | 423_859_149_128_410_414_395_372_834_994_551_L128 | oct(423_859_149_128_410_414_395_372_834_994_551_L128);
+        sout | 13_889_016_598_639_747_063_234_935_497_057_631_587_L128 | oct(13_889_016_598_639_747_063_234_935_497_057_631_587_L128);
+        sout | 1234567890123456789_uL128;
+        sout | 1234567890123456789_L128u;
+        sout | 0x_7fffffff_ffffffff_ffffffff_ffffffff_L128;
+        sout | 0x_ffffffff_ffffffff_ffffffff_ffffffff_L128;
+        sout | 0x_80000000_00000000_00000000_00000000_L128;
+        unsigned int128 vv;
+        vv = 340282366920938463463374607431768211455_L128u;
+        sout | vv;
+        vv = 170141183460469231731687303715884105727_L128;
+        sout | vv;
+        sout | nl;
+        sout | "hexadecimal";
+        sout | hex(v);
+        sout | hex(0x_ffffffff_L128);
+        sout | hex(0x_ffffffff_ffffffff_L128);
+        sout | hex(0x_ffffffff_ffffffff_ffffffff_L128);
+        sout | hex(0xffffffff_ffffffff_ffffffff_ffffffff_L128);
+        sout | hex(0x_a2345678_L128);
+        sout | hex(0x_a2345678_b7654321_L128);
+        sout | hex(0x_a2345678_b7654321_c2345678_L128);
+        sout | hex(0x_a2345678_b7654321_c2345678_d7654321_L128);
+        sout | nl;
+} // main
+// Local Variables: //
+// tab-width: 4 //
+// compile-command: "cfa -Wall -Wextra manipulatorsOutput3.cfa" //
+// End: //

tests/pybin/test_run.py

-              r07d867b
+              r22f94a4
                         else :                                          text = "FAILED with code %d" % retcode
                 text += "    C%s - R%s" % (cls.fmtDur(duration[0]), cls.fmtDur(duration[1]))
+                text += "    C%s - R%s" % (fmtDur(duration[0]), fmtDur(duration[1]))
                 return text
-        @staticmethod
-        def fmtDur( duration ):
-                if duration :
-                        hours, rem = divmod(duration, 3600)
-                        minutes, rem = divmod(rem, 60)
-                        seconds, millis = divmod(rem, 1)
-                        return "%2d:%02d.%03d" % (minutes, seconds, millis * 1000)
-                return " n/a"

tests/pybin/tools.py

-              r07d867b
+              r22f94a4
                 while True:
                         yield i.next(max(expire - time.time(), 0))
+def fmtDur( duration ):
+        if duration :
+                hours, rem = divmod(duration, 3600)
+                minutes, rem = divmod(rem, 60)
+                seconds, millis = divmod(rem, 1)
+                return "%2d:%02d.%03d" % (minutes, seconds, millis * 1000)
+        return " n/a"

tests/searchsort.cfa

-              r07d867b
+              r22f94a4
         } // for
         sout | nl;
         for ( i; 0 ~ size ) {           // C version
+        for ( i; 0 ~ size ) {           // C version, returns void*
                 int key = size - i;
                 int * v = bsearch( &key, iarr, size, sizeof( iarr[0] ), comp );
+                int * v = ( int * ) bsearch( &key, iarr, size, sizeof( iarr[0] ), comp );
                 sout | key | ':' | *v | ", ";
         } // for

tests/test.py

-              r07d867b
+              r22f94a4
                 # for each build configurations, run the test
+                for arch, debug, install in itertools.product(settings.all_arch, settings.all_debug, settings.all_install):
+                        settings.arch    = arch
+                        settings.debug   = debug
+                        settings.install = install
+                        # filter out the tests for a different architecture
+                        # tests are the same across debug/install
+                        local_tests = settings.arch.filter( tests )
+                        options.jobs, forceJobs = job_count( options, local_tests )
+                        settings.update_make_cmd(forceJobs, options.jobs)
+                        # check the build configuration works
+                        settings.validate()
+                        # print configuration
+                        print('%s %i tests on %i cores (%s:%s)' % (
+                                'Regenerating' if settings.generating else 'Running',
+                                len(local_tests),
+                                options.jobs,
+                                settings.arch.string,
+                                settings.debug.string
+                        ))
+                        # otherwise run all tests and make sure to return the correct error code
+                        failed = run_tests(local_tests, options.jobs)
+                        if failed:
+                                result = 1
+                                if not settings.continue_:
+                                        break
+                with Timed() as total_dur:
+                        for arch, debug, install in itertools.product(settings.all_arch, settings.all_debug, settings.all_install):
+                                settings.arch    = arch
+                                settings.debug   = debug
+                                settings.install = install
+                                # filter out the tests for a different architecture
+                                # tests are the same across debug/install
+                                local_tests = settings.arch.filter( tests )
+                                options.jobs, forceJobs = job_count( options, local_tests )
+                                settings.update_make_cmd(forceJobs, options.jobs)
+                                # check the build configuration works
+                                settings.validate()
+                                # print configuration
+                                print('%s %i tests on %i cores (%s:%s)' % (
+                                        'Regenerating' if settings.generating else 'Running',
+                                        len(local_tests),
+                                        options.jobs,
+                                        settings.arch.string,
+                                        settings.debug.string
+                                ))
+                                # otherwise run all tests and make sure to return the correct error code
+                                failed = run_tests(local_tests, options.jobs)
+                                if failed:
+                                        result = 1
+                                        if not settings.continue_:
+                                                break
+                print('Tests took %s' % fmtDur( total_dur.duration ))
                 sys.exit( failed )

tests/time.cfa

-              r07d867b
+              r22f94a4
 // Created On       : Tue Mar 27 17:24:56 2018
 // Last Modified By : Peter A. Buhr
 // Last Modified On : Sun Jan  5 18:27:37 2020
 // Update Count     : 34
+// Last Modified On : Thu Jun 18 18:14:49 2020
+// Update Count     : 37
 //
 …
         Duration d1 = 3`h, d2 = 2`s, d3 = 3.375`s, d4 = 12`s, d5 = 1`s + 10_000`ns;
         sout | d1 | d2 | d3 | d4 | d5;
+        sout | d1`dd | d2`dm | d3`ds | d4`dms | d5`dus;
         d1 = 0;
         sout | d1 | d2 | d3;

tools/build/cfa.make

-              r07d867b
+              r22f94a4
+CFACOMPILE = $(CFACC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) $(AM_CFLAGS) $(CFLAGS)
+LTCFACOMPILE = $(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) \
+AM_T_CFA = $(am__t_CFA_@AM_T@)
+am__t_CFA_ =
+am__t_CFA_0 =
+am__t_CFA_1 = /usr/bin/time --quiet -f "$@ %E" # trailling space is necessary
+CFACOMPILE = $(AM_T_CFA)$(CFACC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(CFAFLAGS) $(AM_CFLAGS) $(CFLAGS)
+LTCFACOMPILE = $(AM_T_CFA)$(LIBTOOL) $(AM_V_lt) --tag=CC $(AM_LIBTOOLFLAGS) \
         $(LIBTOOLFLAGS) --mode=compile $(CFACC) $(DEFS) \
         $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CFAFLAGS) $(AM_CFLAGS) $(CFAFLAGS) $(CFLAGS)

tools/build/push2dist.sh

-              r07d867b
+              r22f94a4
 hash="$1"
+bwlim="$2"
 valid=$(distcc -j 2> /dev/null)
 # if test "${valid}" != 0
 …
 function push() {
         ssh ${host} "mkdir -p ~/.cfadistcc/${hash}/"
         rsync -a ${dV} ${files} ${host}:~/.cfadistcc/${hash}/.
+        rsync --bwlimit=${bwlim} -a ${dV} ${files} ${host}:~/.cfadistcc/${hash}/.
+}

tools/gdb/utils-gdb.py

r07d867b	r22f94a4
59	59	thread_ptr = gdb.lookup_type('struct $thread').pointer(),
60	60	int_ptr = gdb.lookup_type('int').pointer(),
61		thread_state = gdb.lookup_type('enum ~~coroutine_s~~tate'))
	61	thread_state = gdb.lookup_type('enum __Coroutine_State'))
62	62
63	63	def get_addr(addr):

Context Navigation

Legend:

Download in other formats: