Changeset 9e1eabc
- Timestamp:
- Nov 17, 2017, 2:41:55 PM (7 years ago)
- Branches:
- ADT, aaron-thesis, arm-eh, ast-experimental, cleanup-dtors, deferred_resn, demangler, enum, forall-pointer-decay, jacob/cs343-translation, jenkins-sandbox, master, new-ast, new-ast-unique-expr, new-env, no_list, persistent-indexer, pthread-emulation, qualifiedEnum, resolv-new, with_gc
- Children:
- 48fa824, c38ae92
- Parents:
- 65d6de4 (diff), 5f91d650 (diff)
Note: this is a merge changeset, the changes displayed below correspond to the merge itself.
Use the(diff)
links above to see all the changes relative to each parent. - Files:
-
- 3 added
- 2 deleted
- 31 edited
Legend:
- Unmodified
- Added
- Removed
-
Jenkinsfile
r65d6de4 r9e1eabc 15 15 arch_name = '' 16 16 architecture = '' 17 17 18 18 do_alltests = false 19 19 do_benchmark = false … … 183 183 sh 'make clean > /dev/null' 184 184 sh 'make > /dev/null 2>&1' 185 } 185 } 186 186 catch (Exception caughtError) { 187 187 err = caughtError //rethrow error later … … 257 257 def build() { 258 258 build_stage('Build') { 259 259 260 260 def install_dir = pwd tmp: true 261 261 262 262 //Configure the conpilation (Output is not relevant) 263 263 //Use the current directory as the installation target so nothing … … 290 290 if( !do_benchmark ) return 291 291 292 //Write the commit id to Benchmark293 writeFile file: 'bench.csv', text:'data=' + gitRefNewValue + ',' + arch_name + ','294 295 292 //Append bench results 296 sh 'make -C src/benchmark --no-print-directory csv-data >> bench.csv'293 sh 'make -C src/benchmark --no-print-directory jenkins githash=' + gitRefNewValue + ' arch=' + arch_name + ' | tee bench.json' 297 294 } 298 295 } … … 327 324 328 325 //Then publish the results 329 sh 'curl - -silent --data @bench.csvhttp://plg2:8082/jenkins/publish > /dev/null || true'326 sh 'curl -H "Content-Type: application/json" --silent --data @bench.json http://plg2:8082/jenkins/publish > /dev/null || true' 330 327 } 331 328 } -
doc/proposals/concurrency/Makefile
r65d6de4 r9e1eabc 32 32 PICTURES = ${addprefix build/, ${addsuffix .pstex, \ 33 33 system \ 34 monitor_structs \ 34 35 }} 35 36 … … 83 84 dvips $< -o $@ 84 85 85 build/${basename ${DOCUMENT}}.dvi : Makefile ${GRAPHS} ${PROGRAMS} ${PICTURES} ${FIGURES} ${SOURCES} ${basename ${DOCUMENT}}.tex ../../LaTeXmacros/common.tex ../../LaTeXmacros/indexstyle 86 build/${basename ${DOCUMENT}}.dvi : Makefile ${GRAPHS} ${PROGRAMS} ${PICTURES} ${FIGURES} ${SOURCES} ${basename ${DOCUMENT}}.tex ../../LaTeXmacros/common.tex ../../LaTeXmacros/indexstyle annex/local.bib 86 87 87 88 @ if [ ! -r ${basename $@}.ind ] ; then touch ${basename $@}.ind ; fi # Conditionally create an empty *.ind (index) file for inclusion until makeindex is run. … … 94 95 @ -${BibTeX} ${basename $@} 95 96 @ echo "Glossary" 96 makeglossaries -q -s ${basename $@}.ist ${basename $@} # Make index from *.aux entries and input index at end of document97 @ makeglossaries -q -s ${basename $@}.ist ${basename $@} # Make index from *.aux entries and input index at end of document 97 98 @ echo ".dvi generation" 98 99 @ -build/bump_ver.sh -
doc/proposals/concurrency/annex/local.bib
r65d6de4 r9e1eabc 52 52 year = 2017 53 53 } 54 55 @manual{Cpp-Transactions, 56 keywords = {C++, Transactional Memory}, 57 title = {Technical Specification for C++ Extensions for Transactional Memory}, 58 organization= {International Standard ISO/IEC TS 19841:2015 }, 59 publisher = {American National Standards Institute}, 60 address = {http://www.iso.org}, 61 year = 2015, 62 } 63 64 @article{BankTransfer, 65 keywords = {Bank Transfer}, 66 title = {Bank Account Transfer Problem}, 67 publisher = {Wiki Wiki Web}, 68 address = {http://wiki.c2.com}, 69 year = 2010 70 } 71 72 @misc{2FTwoHardThings, 73 keywords = {Hard Problem}, 74 title = {TwoHardThings}, 75 author = {Martin Fowler}, 76 address = {https://martinfowler.com/bliki/TwoHardThings.html}, 77 year = 2009 78 } 79 80 @article{IntrusiveData, 81 title = {Intrusive Data Structures}, 82 author = {Jiri Soukup}, 83 journal = {CppReport}, 84 year = 1998, 85 month = May, 86 volume = {10/No5.}, 87 page = 22 88 } 89 90 @misc{affinityLinux, 91 title = "{Linux man page - sched\_setaffinity(2)}" 92 } 93 94 @misc{affinityWindows, 95 title = "{Windows (vs.85) - SetThreadAffinityMask function}" 96 } 97 98 @misc{affinityFreebsd, 99 title = "{FreeBSD General Commands Manual - CPUSET(1)}" 100 } 101 102 @misc{affinityNetbsd, 103 title = "{NetBSD Library Functions Manual - AFFINITY(3)}" 104 } 105 106 @misc{affinityMacosx, 107 title = "{Affinity API Release Notes for OS X v10.5}" 108 } -
doc/proposals/concurrency/figures/int_monitor.fig
r65d6de4 r9e1eabc 8 8 -2 9 9 1200 2 10 5 1 0 1 0 7 50 -1 -1 0.000 0 1 0 0 600.000 2625.000 600 2325 300 2625 600 2925 11 6 3225 4500 7425 4800 12 1 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3375 4650 80 80 3375 4650 3455 4730 13 1 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4725 4650 105 105 4725 4650 4830 4755 14 1 3 0 1 -1 -1 0 0 4 0.000 1 0.0000 6225 4650 105 105 6225 4650 6330 4755 15 4 0 -1 0 0 0 12 0.0000 2 135 1035 4950 4725 blocked task\001 16 4 0 -1 0 0 0 12 0.0000 2 135 870 3525 4725 active task\001 17 4 0 -1 0 0 0 12 0.0000 2 180 930 6450 4725 routine ptrs\001 10 5 1 0 1 0 7 50 -1 -1 0.000 0 1 0 0 675.000 2700.000 675 2400 375 2700 675 3000 11 6 4533 2866 4655 3129 12 5 1 0 1 0 7 50 -1 -1 0.000 0 1 0 0 4657.017 2997.000 4655 2873 4533 2997 4655 3121 13 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 14 4655 2866 4655 3129 18 15 -6 19 6 8445 1695 8655 1905 20 1 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 8550 1800 105 105 8550 1800 8655 1905 21 4 1 -1 0 0 0 10 0.0000 2 75 75 8550 1860 a\001 16 6 4725 2866 4847 3129 17 5 1 0 1 0 7 50 -1 -1 0.000 0 1 0 0 4849.017 2997.000 4847 2873 4725 2997 4847 3121 18 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 19 4847 2866 4847 3129 22 20 -6 23 6 8445 1395 8655 1605 24 1 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 8550 1500 105 105 8550 1500 8655 1605 25 4 1 -1 0 0 0 10 0.0000 2 105 90 8550 1560 b\001 21 6 4911 2866 5033 3129 22 5 1 0 1 0 7 50 -1 -1 0.000 0 1 0 0 5035.017 2997.000 5033 2873 4911 2997 5033 3121 23 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 24 5033 2866 5033 3129 26 25 -6 27 6 3945 1695 4155 1905 28 1 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4050 1800 105 105 4050 1800 4155 1905 29 4 1 -1 0 0 0 10 0.0000 2 75 75 4050 1860 a\001 26 6 9027 2866 9149 3129 27 5 1 0 1 0 7 50 -1 -1 0.000 0 0 0 0 9024.983 2997.000 9027 2873 9149 2997 9027 3121 28 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 29 9027 2866 9027 3129 30 30 -6 31 6 3945 1395 4155 1605 32 1 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4050 1500 105 105 4050 1500 4155 1605 33 4 1 -1 0 0 0 10 0.0000 2 105 90 4050 1560 b\001 31 6 9253 2866 9375 3129 32 5 1 0 1 0 7 50 -1 -1 0.000 0 0 0 0 9250.983 2997.000 9253 2873 9375 2997 9253 3121 33 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 34 9253 2866 9253 3129 35 -6 36 6 9478 2866 9600 3129 37 5 1 0 1 0 7 50 -1 -1 0.000 0 0 0 0 9475.983 2997.000 9478 2873 9600 2997 9478 3121 38 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 39 9478 2866 9478 3129 34 40 -6 35 41 1 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 7650 3675 80 80 7650 3675 7730 3755 36 42 1 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3150 3675 80 80 3150 3675 3230 3755 43 1 3 0 1 0 7 50 -1 -1 0.000 1 0.0000 4047 1793 125 125 4047 1793 3929 1752 44 1 3 0 1 0 7 50 -1 -1 0.000 1 0.0000 4050 1500 125 125 4050 1500 3932 1459 45 1 3 0 1 0 7 50 -1 -1 0.000 1 0.0000 8550 1500 125 125 8550 1500 8432 1459 46 1 3 0 1 0 7 50 -1 -1 0.000 1 0.0000 8550 1800 125 125 8550 1800 8432 1759 47 1 3 0 1 0 7 50 -1 -1 0.000 1 0.0000 1200 2850 125 125 1200 2850 1082 2809 48 1 3 0 1 0 7 50 -1 -1 0.000 1 0.0000 900 2850 125 125 900 2850 782 2809 49 1 3 0 1 -1 -1 0 0 4 0.000 1 0.0000 6225 4650 105 105 6225 4650 6330 4755 50 1 3 0 1 -1 -1 0 0 20 0.000 1 0.0000 3150 4650 80 80 3150 4650 3230 4730 51 1 3 0 1 -1 -1 0 0 -1 0.000 1 0.0000 4575 4650 105 105 4575 4650 4680 4755 37 52 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 38 53 3900 1950 4200 2100 … … 62 77 3000 4050 3300 4200 63 78 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 64 6 00 2925 1350 292579 675 3000 1425 3000 65 80 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 66 6 00 2325 1350 232581 675 2400 1425 2400 67 82 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 68 1 350 2625 1425 285083 1425 2700 1500 2925 69 84 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 70 1 350 2325 1275 255085 1425 2400 1350 2625 71 86 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 72 600 2625 1350 2625 73 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 74 1350 2775 1275 2645 1125 2645 1050 2775 1125 2905 1275 2905 75 1350 2775 76 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 77 975 2775 900 2645 750 2645 675 2775 750 2905 900 2905 78 975 2775 79 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 80 4800 3000 4725 2870 4575 2870 4500 3000 4575 3130 4725 3130 81 4800 3000 82 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 83 5100 3000 5025 2870 4875 2870 4800 3000 4875 3130 5025 3130 84 5100 3000 85 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 86 9300 3000 9225 2870 9075 2870 9000 3000 9075 3130 9225 3130 87 9300 3000 88 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 89 9600 3000 9525 2870 9375 2870 9300 3000 9375 3130 9525 3130 90 9600 3000 91 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 92 675 2775 975 2775 93 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 94 1050 2775 1350 2775 95 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 96 4875 4950 4800 4820 4650 4820 4575 4950 4650 5080 4800 5080 97 4875 4950 98 2 1 0 1 0 7 50 -1 -1 0.000 0 0 -1 0 0 2 99 4575 4950 4875 4950 100 2 3 0 1 0 7 50 -1 -1 0.000 0 0 0 0 0 7 101 3525 4970 3450 4840 3300 4840 3225 4970 3300 5100 3450 5100 102 3525 4970 87 675 2700 1425 2700 103 88 4 1 -1 0 0 0 12 0.0000 2 135 315 2850 4275 exit\001 104 89 4 1 -1 0 0 0 12 0.0000 2 135 315 7350 4275 exit\001 … … 121 106 4 1 -1 0 0 0 12 0.0000 2 135 495 4050 1275 queue\001 122 107 4 1 -1 0 0 0 12 0.0000 2 165 420 4050 1050 entry\001 123 4 0 0 50 -1 0 11 0.0000 2 120 705 450 2250 Condition\001 124 4 0 0 50 -1 0 11 0.0000 2 165 630 3600 5025 signalled\001 125 4 0 0 50 -1 0 11 0.0000 2 165 525 4950 5025 waiting\001 108 4 0 0 50 -1 0 11 0.0000 2 120 705 600 2325 Condition\001 109 4 0 -1 0 0 0 12 0.0000 2 180 930 6450 4725 routine ptrs\001 110 4 0 -1 0 0 0 12 0.0000 2 135 1050 3300 4725 active thread\001 111 4 0 -1 0 0 0 12 0.0000 2 135 1215 4725 4725 blocked thread\001 -
doc/proposals/concurrency/text/basics.tex
r65d6de4 r9e1eabc 11 11 Execution with a single thread and multiple stacks where the thread is self-scheduling deterministically across the stacks is called coroutining. Execution with a single and multiple stacks but where the thread is scheduled by an oracle (non-deterministic from the thread perspective) across the stacks is called concurrency. 12 12 13 Therefore, a minimal concurrency system can be achieved by creating coroutines, which instead of context switching among each other, always ask an oracle where to context switch next. While coroutines can execute on the caller's stack-frame, stackfull coroutines allow full generality and are sufficient as the basis for concurrency. The aforementioned oracle is a scheduler and the whole system now follows a cooperative threading-model \cit. The oracle/scheduler can either be a stackless or stackfull entity and correspondingly require one or two context switches to run a different coroutine. In any case, a subset of concurrency related challenges start to appear. For the complete set of concurrency challenges to occur, the only feature missing is preemption.14 15 A scheduler introduces order of execution uncertainty, while preemption introduces uncertainty about where context-switches occur. Mutual-exclusion and synchronisation are ways of limiting non-determinism in a concurrent system. Now it is important to understand that uncertainty is desireable; uncertainty can be used by runtime systems to significantly increase performance and is often the basis of giving a user the illusion that tasks are running in parallel. Optimal performance in concurrent applications is often obtained by having as much non-determinism as correctness allows \cit.13 Therefore, a minimal concurrency system can be achieved by creating coroutines, which instead of context switching among each other, always ask an oracle where to context switch next. While coroutines can execute on the caller's stack-frame, stackfull coroutines allow full generality and are sufficient as the basis for concurrency. The aforementioned oracle is a scheduler and the whole system now follows a cooperative threading-model (a.k.a non-preemptive scheduling). The oracle/scheduler can either be a stackless or stackfull entity and correspondingly require one or two context switches to run a different coroutine. In any case, a subset of concurrency related challenges start to appear. For the complete set of concurrency challenges to occur, the only feature missing is preemption. 14 15 A scheduler introduces order of execution uncertainty, while preemption introduces uncertainty about where context-switches occur. Mutual-exclusion and synchronisation are ways of limiting non-determinism in a concurrent system. Now it is important to understand that uncertainty is desireable; uncertainty can be used by runtime systems to significantly increase performance and is often the basis of giving a user the illusion that tasks are running in parallel. Optimal performance in concurrent applications is often obtained by having as much non-determinism as correctness allows. 16 16 17 17 \section{\protect\CFA 's Thread Building Blocks} … … 307 307 \subsection{Alternative: Lamda Objects} 308 308 309 For coroutines as for threads, many implementations are based on routine pointers or function objects\cit . For example, Boost implements coroutines in terms of four functor object types:309 For coroutines as for threads, many implementations are based on routine pointers or function objects\cite{Butenhof97, ANSI14:C++, MS:VisualC++, BoostCoroutines15}. For example, Boost implements coroutines in terms of four functor object types: 310 310 \begin{cfacode} 311 311 asymmetric_coroutine<>::pull_type -
doc/proposals/concurrency/text/concurrency.tex
r65d6de4 r9e1eabc 8 8 Approaches based on shared memory are more closely related to non-concurrent paradigms since they often rely on basic constructs like routine calls and shared objects. At the lowest level, concurrent paradigms are implemented as atomic operations and locks. Many such mechanisms have been proposed, including semaphores~\cite{Dijkstra68b} and path expressions~\cite{Campbell74}. However, for productivity reasons it is desireable to have a higher-level construct be the core concurrency paradigm~\cite{HPP:Study}. 9 9 10 An approach that is worth mentioning because it is gaining in popularity is transactionnal memory~\cite{Dice10}[Check citation]. While this approach is even pursued by system languages like \CC\cit , the performance and feature set is currently too restrictive to be the main concurrency paradigm for systems language, which is why it was rejected as the core paradigm for concurrency in \CFA.10 An approach that is worth mentioning because it is gaining in popularity is transactionnal memory~\cite{Dice10}[Check citation]. While this approach is even pursued by system languages like \CC\cite{Cpp-Transactions}, the performance and feature set is currently too restrictive to be the main concurrency paradigm for systems language, which is why it was rejected as the core paradigm for concurrency in \CFA. 11 11 12 12 One of the most natural, elegant, and efficient mechanisms for synchronization and communication, especially for shared-memory systems, is the \emph{monitor}. Monitors were first proposed by Brinch Hansen~\cite{Hansen73} and later described and extended by C.A.R.~Hoare~\cite{Hoare74}. Many programming languages---e.g., Concurrent Pascal~\cite{ConcurrentPascal}, Mesa~\cite{Mesa}, Modula~\cite{Modula-2}, Turing~\cite{Turing:old}, Modula-3~\cite{Modula-3}, NeWS~\cite{NeWS}, Emerald~\cite{Emerald}, \uC~\cite{Buhr92a} and Java~\cite{Java}---provide monitors as explicit language constructs. In addition, operating-system kernels and device drivers have a monitor-like structure, although they often use lower-level primitives such as semaphores or locks to simulate monitors. For these reasons, this project proposes monitors as the core concurrency-construct. … … 139 139 The \gls{multi-acq} monitor lock allows a monitor lock to be acquired by both \code{bar} or \code{baz} and acquired again in \code{foo}. In the calls to \code{bar} and \code{baz} the monitors are acquired in opposite order. 140 140 141 However, such use leads to the lock acquiring order problem. In the example above, the user uses implicit ordering in the case of function \code{foo} but explicit ordering in the case of \code{bar} and \code{baz}. This subtle mistake means that calling these routines concurrently may lead to deadlock and is therefore undefined behavior. As shown\cit , solving this problem requires:141 However, such use leads to the lock acquiring order problem. In the example above, the user uses implicit ordering in the case of function \code{foo} but explicit ordering in the case of \code{bar} and \code{baz}. This subtle mistake means that calling these routines concurrently may lead to deadlock and is therefore undefined behavior. As shown\cite{Lister77}, solving this problem requires: 142 142 \begin{enumerate} 143 143 \item Dynamically tracking of the monitor-call order. 144 144 \item Implement rollback semantics. 145 145 \end{enumerate} 146 While the first requirement is already a significant constraint on the system, implementing a general rollback semantics in a C-like language is prohibitively complex \cit. In \CFA, users simply need to be carefull when acquiring multiple monitors at the same time or only use \gls{bulk-acq} of all the monitors. While \CFA provides only a partial solution, many system provide no solution and the \CFA partial solution handles many useful cases.146 While the first requirement is already a significant constraint on the system, implementing a general rollback semantics in a C-like language is still prohibitively complex \cite{Dice10}. In \CFA, users simply need to be carefull when acquiring multiple monitors at the same time or only use \gls{bulk-acq} of all the monitors. While \CFA provides only a partial solution, many system provide no solution and the \CFA partial solution handles many useful cases. 147 147 148 148 For example, \gls{multi-acq} and \gls{bulk-acq} can be used together in interesting ways: … … 157 157 } 158 158 \end{cfacode} 159 This example shows a trivial solution to the bank-account transfer-problem\cit . Without \gls{multi-acq} and \gls{bulk-acq}, the solution to this problem is much more involved and requires carefull engineering.159 This example shows a trivial solution to the bank-account transfer-problem\cite{BankTransfer}. Without \gls{multi-acq} and \gls{bulk-acq}, the solution to this problem is much more involved and requires carefull engineering. 160 160 161 161 \subsection{\code{mutex} statement} \label{mutex-stmt} 162 162 163 The call semantics discussed aboved have one software engineering issue, only a named routine can acquire the mutual-exclusion of a set of monitor. \CFA offers the \code{mutex} statement to workaround the need for unnecessary names, avoiding a major software engineering problem\cit . Listing \ref{lst:mutex-stmt} shows an example of the \code{mutex} statement, which introduces a new scope in which the mutual-exclusion of a set of monitor is acquired. Beyond naming, the \code{mutex} statement has no semantic difference from a routine call with \code{mutex} parameters.163 The call semantics discussed aboved have one software engineering issue, only a named routine can acquire the mutual-exclusion of a set of monitor. \CFA offers the \code{mutex} statement to workaround the need for unnecessary names, avoiding a major software engineering problem\cite{2FTwoHardThings}. Listing \ref{lst:mutex-stmt} shows an example of the \code{mutex} statement, which introduces a new scope in which the mutual-exclusion of a set of monitor is acquired. Beyond naming, the \code{mutex} statement has no semantic difference from a routine call with \code{mutex} parameters. 164 164 165 165 \begin{figure} … … 232 232 % ====================================================================== 233 233 % ====================================================================== 234 In addition to mutual exclusion, the monitors at the core of \CFA's concurrency can also be used to achieve synchronisation. With monitors, this capability is generally achieved with internal or external scheduling as in \cit. Since internal scheduling within a single monitor is mostly a solved problem, this thesis concentrates on extending internal scheduling to multiple monitors. Indeed, like the \gls{bulk-acq} semantics, internal scheduling extends to multiple monitors in a way that is natural to the user but requires additional complexity on the implementation side.234 In addition to mutual exclusion, the monitors at the core of \CFA's concurrency can also be used to achieve synchronisation. With monitors, this capability is generally achieved with internal or external scheduling as in \cite{Hoare74}. Since internal scheduling within a single monitor is mostly a solved problem, this thesis concentrates on extending internal scheduling to multiple monitors. Indeed, like the \gls{bulk-acq} semantics, internal scheduling extends to multiple monitors in a way that is natural to the user but requires additional complexity on the implementation side. 235 235 236 236 First, here is a simple example of such a technique: … … 305 305 This version uses \gls{bulk-acq} (denoted using the {\sf\&} symbol), but the presence of multiple monitors does not add a particularly new meaning. Synchronization happens between the two threads in exactly the same way and order. The only difference is that mutual exclusion covers more monitors. On the implementation side, handling multiple monitors does add a degree of complexity as the next few examples demonstrate. 306 306 307 While deadlock issues can occur when nesting monitors, these issues are only a symptom of the fact that locks, and by extension monitors, are not perfectly composable. For monitors, a well known deadlock problem is the Nested Monitor Problem \cit, which occurs when a \code{wait} is made by a thread that holds more than one monitor. For example, the following pseudo-code runs into the nested-monitor problem :307 While deadlock issues can occur when nesting monitors, these issues are only a symptom of the fact that locks, and by extension monitors, are not perfectly composable. For monitors, a well known deadlock problem is the Nested Monitor Problem \cite{Lister77}, which occurs when a \code{wait} is made by a thread that holds more than one monitor. For example, the following pseudo-code runs into the nested-monitor problem : 308 308 \begin{multicols}{2} 309 309 \begin{pseudo} … … 771 771 For the first two conditions, it is easy to implement a check that can evaluate the condition in a few instruction. However, a fast check for \pscode{monitor accepts me} is much harder to implement depending on the constraints put on the monitors. Indeed, monitors are often expressed as an entry queue and some acceptor queue as in the following figure: 772 772 773 \begin{figure}[H] 773 774 \begin{center} 774 775 {\resizebox{0.4\textwidth}{!}{\input{monitor}}} 775 776 \end{center} 777 \label{fig:monitor} 778 \end{figure} 776 779 777 780 There are other alternatives to these pictures, but in the case of this picture, implementing a fast accept check is relatively easy. Restricted to a fixed number of mutex members, N, the accept check reduces to updating a bitmask when the acceptor queue changes, a check that executes in a single instruction even with a fairly large number (e.g., 128) of mutex members. This technique cannot be used in \CFA because it relies on the fact that the monitor type enumerates (declares) all the acceptable routines. For OO languages this does not compromise much since monitors already have an exhaustive list of member routines. However, for \CFA this is not the case; routines can be added to a type anywhere after its declaration. It is important to note that the bitmask approach does not actually require an exhaustive list of routines, but it requires a dense unique ordering of routines with an upper-bound and that ordering must be consistent across translation units. -
doc/proposals/concurrency/text/future.tex
r65d6de4 r9e1eabc 6 6 7 7 \section{Flexible Scheduling} \label{futur:sched} 8 8 An important part of concurrency is scheduling. Different scheduling algorithm can affact peformance (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 tweaking options to users, allowing the scheduler to be adjusted the to requirements of the workload. However, in order to be truly flexible, it would be interesting to allow users to add arbitrary data and arbirary scheduling algorithms to the scheduler. For example, a web server could attach Type-of-Service information to threads and have a ``ToS aware'' scheduling algorithm tailored to this specific web server. This path of flexible schedulers will be explored for \CFA. 9 9 10 10 \section{Non-Blocking IO} \label{futur:nbio} 11 11 While most of the parallelism tools 12 However, many modern workloads are not bound on computation but on IO operations, an common case being webservers and XaaS (anything as a service). These type of workloads often require significant engineering around amortising costs of blocking IO operations. While improving throughtput of these operations is outside what \CFA can do as a language, it can help users to make better use of the CPU time otherwise spent waiting on IO operations. The current trend is to use asynchronous programming using tools like callbacks and/or futurs and promises\cit. However, while these are valid solutions, they lead to code that is harder to read and maintain because it is much less linear 13 14 12 However, many modern workloads are not bound on computation but on IO operations, an common case being webservers and XaaS (anything as a service). These type of workloads often require significant engineering around amortising costs of blocking IO operations. While improving throughtput of these operations is outside what \CFA can do as a language, it can help users to make better use of the CPU time otherwise spent waiting on IO operations. The current trend is to use asynchronous programming using tools like callbacks and/or futurs and promises\cite. However, while these are valid solutions, they lead to code that is harder to read and maintain because it is much less linear 15 13 16 14 \section{Other concurrency tools} \label{futur:tools} 17 15 While monitors offer a flexible and powerful concurent core for \CFA, other concurrency tools are also necessary for a complete multi-paradigm concurrency package. Example of such tools can include simple locks and condition variables, futures and promises\cite{promises}, and executors. These additional features are useful when monitors offer a level of abstraction which is indaquate for certain tasks. 18 16 19 17 \section{Implicit threading} \label{futur:implcit} 20 Simpler applications can benefit greatly from having implicit parallelism. That is, parallelism that does not rely on the user to write concurrency. This type of parallelism can be achieved both at the language level and at the library level. The cannonical example of implcit parallelism is parallel for loops, which are the simplest example of a divide and conquer algorithm\cit . Listing \ref{lst:parfor} shows three different code examples that accomplish pointwise sums of large arrays. Note that none of these example explicitly declare any concurrency or parallelism objects.18 Simpler applications can benefit greatly from having implicit parallelism. That is, parallelism that does not rely on the user to write concurrency. This type of parallelism can be achieved both at the language level and at the library level. The cannonical example of implcit parallelism is parallel for loops, which are the simplest example of a divide and conquer algorithm\cite{uC++book}. Listing \ref{lst:parfor} shows three different code examples that accomplish pointwise sums of large arrays. Note that none of these example explicitly declare any concurrency or parallelism objects. 21 19 22 20 \begin{figure} … … 103 101 \end{figure} 104 102 105 Implicit parallelism is a general solution and therefore is 106 107 \section{Multiple Paradigms} \label{futur:paradigms} 103 Implicit parallelism is a general solution and therefore has its limitations. However, it is a quick and simple approach to parallelism which may very well be sufficient for smaller applications and reduces the amount of boiler-plate that is needed to start benefiting from parallelism in modern CPUs. 108 104 109 105 110 \section{Transactions} \label{futur:transaction}111 Concurrency and parallelism is still a very active field that strongly benefits from hardware advances. As such certain features that aren't necessarily mature enough in their current state could become relevant in the lifetime of \CFA. -
doc/proposals/concurrency/text/internals.tex
r65d6de4 r9e1eabc 1 1 2 2 \chapter{Behind the scene} 3 There are several challenges specific to \CFA when implementing concurrency. These challenges are direct results of \gls{bulk-acq} and loose object definitions. These two constraints are to root cause of most design decisions in the implementation. Furthermore, to avoid the head-aches of dynamically allocating memory in a concurrent environment, the internal-scheduling design is (almost) entirely free of mallocs and other dynamic memory allocation scheme. This is to avoid the chicken and egg problem \cite{Chicken} of having a memory allocator that relies on the threading system and a threading system that relies on the runtime. This extra goal, means that memory management is a constant concern in the design of the system.4 5 The main memory concern for concurrency is queues. All blocking operations are made by parking threads onto queues. The se queues need to be intrinsic\cit to avoid the need memory allocation. This entails that all the fields needed to keep track of all needed information. Since many conconcurrency operations can use an unbound amount of memory (depending on \gls{bulk-acq}) statically defining information in the intrusive fields of threads is insufficient. The only variable sized container that does not require memory allocation is the callstack, which is heavily used in the implementation of internal scheduling. Particularly the GCC extension variable length arrays which isused extensively.3 There are several challenges specific to \CFA when implementing concurrency. These challenges are a direct result of \gls{bulk-acq} and loose object-definitions. These two constraints are the root cause of most design decisions in the implementation. Furthermore, to avoid contention from dynamically allocating memory in a concurrent environment, the internal-scheduling design is (almost) entirely free of mallocs. This is to avoid the chicken and egg problem \cite{Chicken} of having a memory allocator that relies on the threading system and a threading system that relies on the runtime. This extra goal, means that memory management is a constant concern in the design of the system. 4 5 The main memory concern for concurrency is queues. All blocking operations are made by parking threads onto queues. The queue design needs to be intrusive\cite{IntrusiveData} to avoid the need for memory allocation, which entails that all the nodes need specific fields to keep track of all needed information. Since many concurrency operations can use an unbound amount of memory (depending on \gls{bulk-acq}), statically defining information in the intrusive fields of threads is insufficient. The only variable sized container that does not require memory allocation is the callstack, which is heavily used in the implementation of internal scheduling. Particularly variable length arrays, which are used extensively. 6 6 7 7 Since stack allocation is based around scope, the first step of the implementation is to identify the scopes that are available to store the information, and which of these can have a variable length. The threads and the condition both allow a fixed amount of memory to be stored, while mutex-routines and the actual blocking call allow for an unbound amount (though the later is preferable in terms of performance). 8 8 9 Note that since the major contributions of this thesis are extending monitor semantics to \gls{bulk-acq} and loose object definitions, any challenges that are not resulting of these characteristiques of \CFA are consi red as problems which have already been solved and therefore will not bediscussed further.9 Note that since the major contributions of this thesis are extending monitor semantics to \gls{bulk-acq} and loose object definitions, any challenges that are not resulting of these characteristiques of \CFA are considered as solved problems and therefore not discussed further. 10 10 11 11 % ====================================================================== … … 15 15 % ====================================================================== 16 16 17 The first step towards the monitor implementation is simple mutex-routines using monitors. In the single monitor case, this is done using the entry/exit procedure highlighted in listing \ref{lst:entry1}. This entry/exit procedure does n't actually have to be extended to support multiple monitors, indeed it is sufficient to enter/leave monitors one-by-one as long as the order is correct to prevent deadlocks\cit. In \CFA, ordering of monitor relies on memory ordering, this is sufficient because all objects are guaranteed to have distinct non-overlaping memory layouts and mutual-exclusion for a monitor is only defined for its lifetime, meaning that destroying a monitor while it is acquired is undefined behavior. When a mutex call is made, the concerned monitors are agregated into an variable-length pointer array and sorted based on pointer values. This array is concerved during the entire duration of the mutual-exclusion and it's ordering reused extensively.17 The first step towards the monitor implementation is simple mutex-routines using monitors. In the single monitor case, this is done using the entry/exit procedure highlighted in listing \ref{lst:entry1}. This entry/exit procedure does not actually have to be extended to support multiple monitors, indeed it is sufficient to enter/leave monitors one-by-one as long as the order is correct to prevent deadlocks\cite{Havender68}. In \CFA, ordering of monitor relies on memory ordering, this is sufficient because all objects are guaranteed to have distinct non-overlaping memory layouts and mutual-exclusion for a monitor is only defined for its lifetime, meaning that destroying a monitor while it is acquired is undefined behavior. When a mutex call is made, the concerned monitors are agregated into a variable-length pointer array and sorted based on pointer values. This array presists for the entire duration of the mutual-exclusion and its ordering reused extensively. 18 18 \begin{figure} 19 19 \begin{multicols}{2} … … 96 96 \end{tabular} 97 97 \end{center} 98 \caption{Call site vs entry-point locking for mutex calls}98 \caption{Call-site vs entry-point locking for mutex calls} 99 99 \label{fig:locking-site} 100 100 \end{figure} 101 101 102 Note the \code{mutex} keyword relies on the type system, which means that in cases where a generic monitor routine is actually desired, writing amutex routine is possible with the proper trait, for example:102 Note the \code{mutex} keyword relies on the type system, which means that in cases where a generic monitor routine is desired, writing the mutex routine is possible with the proper trait, for example: 103 103 \begin{cfacode} 104 //Incorrect: T is not amonitor104 //Incorrect: T may not be monitor 105 105 forall(dtype T) 106 106 void foo(T * mutex t); … … 111 111 \end{cfacode} 112 112 113 Both entry-point and callsite locking are valid implementations. The current \CFA implementations uses entry-point locking because it seems to require less work if done using \gls{raii}, effectively transferring the burden of implementation to object construction/destruction. The same could be said of callsite locking, the difference being that the later does not necessarily have an existing scope that matches exactly the scope of the mutual exclusion, i.e.: the function body.113 Both entry-point and callsite locking are feasible implementations. The current \CFA implementations uses entry-point locking because it requires less work when using \gls{raii}, effectively transferring the burden of implementation to object construction/destruction. The same could be said of callsite locking, the difference being that the later does not necessarily have an existing scope that matches exactly the scope of the mutual exclusion, i.e.: the function body. Furthermore, entry-point locking requires less code generation since any useful routine is called at least as often as it is define, there can be only one entry-point but many callsites. 114 114 115 115 % ====================================================================== … … 119 119 % ====================================================================== 120 120 121 Figure \ref{fig:system1} shows a high-level picture if the \CFA runtime system in regards to concurrency. 121 Figure \ref{fig:system1} shows a high-level picture if the \CFA runtime system in regards to concurrency. Each component of the picture is explained in details in the fllowing sections. 122 122 123 123 \begin{figure} … … 130 130 131 131 \subsection{Context Switching} 132 As mentionned in section \ref{coroutine}, coroutines are a stepping stone for implementing threading. This is because they share the same mechanism for context-switching between different stacks. To improve performance and simplicity, context-switching is implemented using the following assumption: all context-switches happen inside a specific function call. This assumption s means that the basic recipe for context-switch is only to copy all callee-saved registers unto the stack and then switch the stack registers with the ones of the target coroutine/thread. Note that instruction pointer can be left untouched since the context-switch always inside the same function. In the case of coroutines, that is the entire story. Threads however do not simply context-switch between each other directly. The context-switch to processors which is where the scheduling happens. This method is called a 2-step context-switch and has the advantage of having a clear distinction between user code and the "kernel" where scheduling and other system operation happen. Obiously, this has the cost of doubling the context-switch cost frombecause threads must context-switch to an intermediate stack. However, the performance of the 2-step context-switch is still superior to a \code{pthread_yield}(see section \ref{results}). additionally, for users in need for optimal performance, it is important to note that having a 2-step context-switch as the default does not prevent \CFA from offering a 1-step context-switch to use manually (or as part of monitors). This option is not currently present in \CFA but the changes required to add it are strictly additive.132 As mentionned in section \ref{coroutine}, coroutines are a stepping stone for implementing threading. This is because they share the same mechanism for context-switching between different stacks. To improve performance and simplicity, context-switching is implemented using the following assumption: all context-switches happen inside a specific function call. This assumption means that the context-switch only has to copy the callee-saved registers onto the stack and then switch the stack registers with the ones of the target coroutine/thread. Note that the instruction pointer can be left untouched since the context-switch is always inside the same function. Threads however do not context-switch between each other directly. They context-switch to the scheduler. This method is called a 2-step context-switch and has the advantage of having a clear distinction between user code and the kernel where scheduling and other system operation happen. Obiously, this has the cost of doubling the context-switch cost because threads must context-switch to an intermediate stack. However, the performance of the 2-step context-switch is still superior to a \code{pthread_yield}(see section \ref{results}). additionally, for users in need for optimal performance, it is important to note that having a 2-step context-switch as the default does not prevent \CFA from offering a 1-step context-switch to use manually (or as part of monitors). This option is not currently present in \CFA but the changes required to add it are strictly additive. 133 133 134 134 \subsection{Processors} 135 Parallelism in \CFA are built around using processors to specify how much parallelism is desired. \CFA processors are object wrappers around kernel threads, specifically pthreads in the current implementation of \CFA. Indeed, any parallelism must go through operatiing system librairies. However, \gls{cfathread} are still the main source of concurrency, processors are simply the underlying source of parallelism. Indeed, processor kernel threads simply fetch a user-level thread from the scheduler and run, they are effectively executers for user-threads. The main benefit of this approach is that it offers a well defined boundary between kernel code and user-code, for examplekernel thread quiescing, scheduling and interrupt handling. Processors internally use coroutines to take advantage of the existing context-switching semantics.135 Parallelism in \CFA is built around using processors to specify how much parallelism is desired. \CFA processors are object wrappers around kernel threads, specifically pthreads in the current implementation of \CFA. Indeed, any parallelism must go through operating-system librairies. However, \glspl{uthread} are still the main source of concurrency, processors are simply the underlying source of parallelism. Indeed, processor \glspl{kthread} simply fetch a \glspl{uthread} from the scheduler and run, they are effectively executers for user-threads. The main benefit of this approach is that it offers a well defined boundary between kernel code and user code, for example, kernel thread quiescing, scheduling and interrupt handling. Processors internally use coroutines to take advantage of the existing context-switching semantics. 136 136 137 137 \subsection{Stack management} 138 138 One of the challenges of this system is to reduce the footprint as much as possible. Specifically, all pthreads created also have a stack created with them, which should be used as much as possible. Normally, coroutines also create there own stack to run on, however, in the case of the coroutines used for processors, these coroutines run directly on the kernel thread stack, effectively stealing the processor stack. The exception to this rule is the Main Processor, i.e. the initial kernel thread that is given to any program. In order to respect user expectations, the stack of the initial kernel thread, the main stack of the program, is used by the main user thread rather than the main processor. 139 139 140 \subsection{Preemption} 141 Finally, an important aspect for any complete threading system is preemption. As mentionned in chapter \ref{basics}, preemption introduces an extra degree of uncer etainty, which enables users to have multiple threads interleave transparrently between eachother, rather than having to cooperate between thread for proper scheduling and CPU distribution. Indeed, preemption is desireable because it adds a degree of isolation between tasks. In a fully cooperative system, any thread that runs into a long loop can starve other threads, while in a preemptive system starvation can still occur but it does not rely on every thread having to yield or block on a regular basis, which reduces significantlyprogrammer burden. Obviously, preemption is not optimal for every workload, however any preemptive system can become a cooperative system by making the time-slices extremely large. Which is why \CFA uses a preemptive threading system.142 143 Preemption in \CFA is based on kernel timers which are used to run a discreet event simulation. Every processor keeps track of the current time and registers an expiration time with the preemption system. When the preemption system receives a change in preemption it sorts these expiration times in a list and sets a kernel timer for the closest one, effectiveling stepping between preemption events on each signals sent by the timer. These timers use the linux signal {\tt SIGALRM}, which is delivered to the process. This is important because when delivering signals to a process, the kernel documentation states that the signal can be delivered to any kernel thread for which the signal isn't blocki.e. :140 \subsection{Preemption} \label{preemption} 141 Finally, an important aspect for any complete threading system is preemption. As mentionned in chapter \ref{basics}, preemption introduces an extra degree of uncertainty, which enables users to have multiple threads interleave transparently, rather than having to cooperate among threads for proper scheduling and CPU distribution. Indeed, preemption is desireable because it adds a degree of isolation among threads. In a fully cooperative system, any thread that runs into a long loop can starve other threads, while in a preemptive system starvation can still occur but it does not rely on every thread having to yield or block on a regular basis, which reduces significantly a programmer burden. Obviously, preemption is not optimal for every workload, however any preemptive system can become a cooperative system by making the time-slices extremely large. Which is why \CFA uses a preemptive threading system. 142 143 Preemption in \CFA is based on kernel timers, which are used to run a discrete-event simulation. Every processor keeps track of the current time and registers an expiration time with the preemption system. When the preemption system receives a change in preemption, it sorts these expiration times in a list and sets a kernel timer for the closest one, effectively stepping between preemption events on each signals sent by the timer. These timers use the linux signal {\tt SIGALRM}, which is delivered to the process rather than the kernel-thread. This results in an implementation problem,because when delivering signals to a process, the kernel documentation states that the signal can be delivered to any kernel thread for which the signal is not blocked i.e. : 144 144 \begin{quote} 145 145 A process-directed signal may be delivered to any one of the threads that does not currently have the signal blocked. If more than one of the threads has the signal unblocked, then the kernel chooses an arbitrary thread to which to deliver the signal. … … 148 148 For the sake of simplicity and in order to prevent the case of having two threads receiving alarms simultaneously, \CFA programs block the {\tt SIGALRM} signal on every thread except one. Now because of how involontary context-switches are handled, the kernel thread handling {\tt SIGALRM} cannot also be a processor thread. 149 149 150 Involontary context-switching is done by sending {\tt SIGUSER1} to the corresponding processor and having the thread yield from inside the signal handler. Effectively context-switch away from the signal-handler back to the kernel and the signal-handler frame will be unwound when the thread is scheduled again. This means that a signal-handler can start on one kernel thread and terminate on a second kernel thread (but the same user thread). It is important to note that signal-handlers save and restore signal masks because user-thread migration can cause signal mask to migrate from one kernel thread to another. This is only a problem if all kernel threads among which a user thread can migrate differ in terms of signal masks. However, since the kernel thread hanlding preemption requires a different signal mask, executing user threads on the kernel alarm thread can cause deadlocks. For this reason, the alarm thread is on a tight loop around a system call to \code{sigwait} or more specifically \code{sigwaitinfo}, requiring very little CPU time for preemption. One final detail about the alarm thread is how to wake it when additional communication is required (e.g. on thread termination). This is also done using {\tt SIGALRM}, but sent throught the \code{pthread_sigqueue}. Indeed, \code{sigwait} can differentiate signals sent from \code{pthread_sigqueue} from signals sent from alarms or the kernel. 151 152 \subsection{Scheduler} \footnote{ I'm not sure what to write here, is this section even needed. } 153 Finally, an aspect that was not mentionned yet is the scheduling algorithm. Currently, the \CFA scheduler uses a single ready queue for all processors. Will this is not the highest performance algorithm, it has the significant advantage of being robust to heterogenous workloads. This is a very simple scheduling approach but is sufficient to for the context of this thesis. 154 155 What to do here? 156 157 However, when 158 As will be mentionned \ref{futur:sched} it needs to be updated when clusters will be 159 160 clusters 161 162 163 164 Among the most pressing updates to the \CFA 165 uses single queue 166 in future should move to multiple queues with workstealing 167 general purpouse means robust > fast 168 worksharing can higher standard deviation in performance 169 150 Involuntary context-switching is done by sending signal {\tt SIGUSER1} to the corresponding processor and having the thread yield from inside the signal handler. Effectively context-switching away from the signal-handler back to the kernel and the signal-handler frame is eventually unwound when the thread is scheduled again. This approach means that a signal-handler can start on one kernel thread and terminate on a second kernel thread (but the same user thread). It is important to note that signal-handlers save and restore signal masks because user-thread migration can cause signal mask to migrate from one kernel thread to another. This behaviour is only a problem if all kernel threads among which a user thread can migrate differ in terms of signal masks\footnote{Sadly, official POSIX documentation is silent on what distiguishes ``async-signal-safe'' functions from other functions}. However, since the kernel thread hanlding preemption requires a different signal mask, executing user threads on the kernel alarm thread can cause deadlocks. For this reason, the alarm thread is on a tight loop around a system call to \code{sigwaitinfo}, requiring very little CPU time for preemption. One final detail about the alarm thread is how to wake it when additional communication is required (e.g., on thread termination). This unblocking is also done using {\tt SIGALRM}, but sent throught the \code{pthread_sigqueue}. Indeed, \code{sigwait} can differentiate signals sent from \code{pthread_sigqueue} from signals sent from alarms or the kernel. 151 152 \subsection{Scheduler} 153 Finally, an aspect that was not mentionned yet is the scheduling algorithm. Currently, the \CFA scheduler uses a single ready queue for all processors, which is the simplest approach to scheduling. Further discussion on scheduling is present in section \label{futur:sched}. 170 154 171 155 % ====================================================================== … … 174 158 % ====================================================================== 175 159 % ====================================================================== 176 To ease the understanding of monitors, like many other concepts, they are generelly represented graphically. While non-scheduled monitors are simple enough for a graphical representation to be useful, internal scheduling is complex enough to justify a visual representation. The following figure is the traditionnal illustration of a monitor : 177 160 The following figure is the traditional illustration of a monitor (repeated from page~\pageref{fig:monitor} for convenience) : 161 162 \begin{figure}[H] 178 163 \begin{center} 179 164 {\resizebox{0.4\textwidth}{!}{\input{monitor}}} 180 165 \end{center} 181 182 This picture has several components, the two most important being the entry-queue and the AS-stack. The entry-queue is a (almost) FIFO list where threads waiting to enter are parked, while the AS-stack is a FILO list used for threads that have been signaled or otherwise marked as running next. For \CFA, the previous picture does not have support for blocking multiple monitors on a single condition. To support \gls{bulk-acq} two changes to this picture are required. First, it doesn't make sense to tie the condition to a single monitor since blocking two monitors as one would require arbitrarily picking a monitor to hold the condition. Secondly, the object waiting on the conditions and AS-stack cannot simply contain the waiting thread since a single thread can potentially wait on multiple monitors. As mentionned in section \ref{intsched}, the handling in multiple monitors is done by partially passing, which entails that each concerned monitor needs to have a node object. However, for waiting on the condition, since all threads need to wait together, a single object needs to be queued in the condition. Moving out the condition and updating the node types yields : 183 166 \caption{Traditional illustration of a monitor} 167 \label{fig:monitor} 168 \end{figure} 169 170 This picture has several components, the two most important being the entry-queue and the AS-stack. The entry-queue is an (almost) FIFO list where threads waiting to enter are parked, while the acceptor-signalor (AS) stack is a FILO list used for threads that have been signalled or otherwise marked as running next. 171 172 For \CFA, this picture does not have support for blocking multiple monitors on a single condition. To support \gls{bulk-acq} two changes to this picture are required. First, it is non longer helpful to attach the condition to a single monitor. Secondly, the thread waiting on the conditions has to be seperated multiple monitors, which yields : 173 174 \begin{figure}[H] 184 175 \begin{center} 185 176 {\resizebox{0.8\textwidth}{!}{\input{int_monitor}}} 186 177 \end{center} 187 188 This picture and the proper entry and leave algorithms is the fundamental implementation of internal scheduling (see listing \ref{lst:entry2}). 178 \caption{Illustration of \CFA monitor} 179 \label{fig:monitor_cfa} 180 \end{figure} 181 182 This picture and the proper entry and leave algorithms is the fundamental implementation of internal scheduling (see listing \ref{lst:entry2}). Note that when threads are moved from the condition to the AS-stack, it splits the thread into to pieces. The thread is woken up when all the pieces have moved from the AS-stacks to the active thread seat. In this picture, the threads are split into halves but this is only because there are two monitors in this picture. For a specific signaling operation every monitor needs a piece of thread on its AS-stack. 189 183 190 184 \begin{figure}[b] … … 219 213 \end{figure} 220 214 221 Some important things to notice about the exit routine. The solution discussed in \ref{intsched} can be seen in the exit routine of listing \ref{lst:entry2}. Basically, the solution boils down to having a seperate data structure for the condition queue and the AS-stack, and unconditionally transferring ownership of the monitors but only unblocking the thread when the last monitor has transferred ownership. This solution is deadlock safe as well as preventing any potential barging. 222 223 The data structure used for the AS-stack are reused extensively for external scheduling, but in the case of internal scheduling, the data is allocated using variable-length arrays on the callstack of the \code{wait} and \code{signal_block} routines. 215 Some important things to notice about the exit routine. The solution discussed in \ref{intsched} can be seen in the exit routine of listing \ref{lst:entry2}. Basically, the solution boils down to having a seperate data structure for the condition queue and the AS-stack, and unconditionally transferring ownership of the monitors but only unblocking the thread when the last monitor has transferred ownership. This solution is deadlock safe as well as preventing any potential barging. The data structure used for the AS-stack are reused extensively for external scheduling, but in the case of internal scheduling, the data is allocated using variable-length arrays on the callstack of the \code{wait} and \code{signal_block} routines. 216 217 \begin{figure}[H] 218 \begin{center} 219 {\resizebox{0.8\textwidth}{!}{\input{monitor_structs.pstex_t}}} 220 \end{center} 221 \caption{Data structures involved in internal/external scheduling} 222 \label{fig:structs} 223 \end{figure} 224 225 Figure \ref{fig:structs} shows a high level representation of these data-structures. The main idea behind them is that, while figure \ref{fig:monitor_cfa} is a nice illustration in theory, in practice breaking a threads into multiple pieces to put unto intrusive stacks does not make sense. The \code{condition node} is the data structure that is queued into a condition variable and, when signaled, the condition queue is popped and each \code{condition criterion} are moved to the AS-stack. Once all the criterion have be popped from their respective AS-stacks, the thread is woken-up, which is what is shown in listing \ref{lst:entry2}. 224 226 225 227 % ====================================================================== … … 228 230 % ====================================================================== 229 231 % ====================================================================== 230 Similarly to internal scheduling, external scheduling for multiple monitors relies on the idea that entry-queues are no longer specific to a single monitor, as mentionned in section \ref{extsched}. This means that some kind of entry-queues must be used that is aware of both monitors and which holds threads that are currently waiting to enter the critical section. This challenge is solved for internal scheduling by having the entry-queues in conditions no longer be tied to a monitor, effectively allowing conditions to be moved outside of monitors. However, in the case of external scheduling, acceptable routines must be aware of the entry queues, which means they must be stored inside at least one of the monitors that will be acquired. This in turn adds the requirement that a systematic algorithm of disambiguating which monitor holds the relevant queue regardless of user ordering. The proposed algorithm is to fall back on monitor lock ordering and specify that the monitor that is acquired first is the one with the relevant entry queue. This assumes that the lock acquiring order is static for the lifetime of all concerned objects but that is a reasonable constraint. 231 232 This algorithm choice has two consequences, the entry queue of the highest priority monitor is no longer a true FIFO queue and the queue of the lowest priority monitor is both required and probably unused. The queue can no longer be a FIFO queue because instead of simply containing the waiting threads in order of arrival, they also contain a set of monitors. Therefore, another thread whos set contains the same highest priority monitor but different lower priority monitors may arrive first but enter the critical section after a thread with the correct pairing. Secondly, since it is not known at compile time which monitor will be the lowest priority monitor, every monitor needs to have the correct queues even though it is probable that some queues will go unused for the entire duration of the program, for example if a monitor is only used in a pair. 232 Similarly to internal scheduling, external scheduling for multiple monitors relies on the idea that waiting-thread queues are no longer specific to a single monitor, as mentionned in section \ref{extsched}. For internal scheduling, these queues are part of condition variables which are still unique for a given scheduling operation (e.g., no single statment uses multiple conditions). However, in the case of external scheduling, there is no equivalent object which is associated with \code{waitfor} statements. This absence means the queues holding the waiting threads must be stored inside at least one of the monitors that is acquired. The monitors being the only objects that have sufficient lifetime and are available on both sides of the \code{waitfor} statment. This requires an algorithm to choose which monitor holds the relevant queue. It is also important that said algorithm be independent of the order in which users list parameters. The proposed algorithm is to fall back on monitor lock ordering and specify that the monitor that is acquired first is the one with the relevant wainting queue. This assumes that the lock acquiring order is static for the lifetime of all concerned objects but that is a reasonable constraint. 233 234 This algorithm choice has two consequences : 235 \begin{itemize} 236 \item The queue of the highest priority monitor is no longer a true FIFO queue because threads can be moved to the front of the queue. These queues need to contain a set of monitors for each of the waiting threads. Therefore, another thread whose set contains the same highest priority monitor but different lower priority monitors may arrive first but enter the critical section after a thread with the correct pairing. 237 \item The queue of the lowest priority monitor is both required and potentially unused. Indeed, since it is not known at compile time which monitor will be the lowest priority monitor, every monitor needs to have the correct queues even though it is possible that some queues will go unused for the entire duration of the program, for example if a monitor is only used in a specific pair. 238 \end{itemize} 233 239 234 240 Therefore, the following modifications need to be made to support external scheduling : 235 241 \begin{itemize} 236 \item The threads waiting on the entry-queue need to keep track of which routine is trying to enter, and using which set of monitors. The \code{mutex} routine already has all the required information on it 's stack so the thread only needs to keep a pointer to that information.242 \item The threads waiting on the entry-queue need to keep track of which routine is trying to enter, and using which set of monitors. The \code{mutex} routine already has all the required information on its stack so the thread only needs to keep a pointer to that information. 237 243 \item The monitors need to keep a mask of acceptable routines. This mask contains for each acceptable routine, a routine pointer and an array of monitors to go with it. It also needs storage to keep track of which routine was accepted. Since this information is not specific to any monitor, the monitors actually contain a pointer to an integer on the stack of the waiting thread. Note that the complete mask can be pushed to any owned monitors, regardless of \code{when} statements, the \code{waitfor} statement is used in a context where the thread already has full ownership of (at least) every concerned monitor and therefore monitors will refuse all calls no matter what. 238 244 \item The entry/exit routine need to be updated as shown in listing \ref{lst:entry3}. 239 245 \end{itemize} 240 246 247 \subsection{External scheduling - destructors} 241 248 Finally, to support the ordering inversion of destructors, the code generation needs to be modified to use a special entry routine. This routine is needed because of the storage requirements of the call order inversion. Indeed, when waiting for the destructors, storage is need for the waiting context and the lifetime of said storage needs to outlive the waiting operation it is needed for. For regular \code{waitfor} statements, the callstack of the routine itself matches this requirement but it is no longer the case when waiting for the destructor since it is pushed on to the AS-stack for later. The waitfor semantics can then be adjusted correspondingly, as seen in listing \ref{lst:entry-dtor} 242 249 … … 250 257 continue 251 258 elif matches waitfor mask 252 push waiterto AS-stack259 push criterions to AS-stack 253 260 continue 254 261 else … … 265 272 if all monitors ready 266 273 wake-up thread 274 endif 275 endif 267 276 268 277 if entry queue not empty 269 278 wake-up thread 279 endif 270 280 \end{pseudo} 271 281 \end{multicols} … … 295 305 Waitfor 296 306 \begin{pseudo} 297 lock all monitors298 307 if matching thread is already there 299 308 if found destructor … … 303 312 push self to AS-stack 304 313 baton pass 314 endif 305 315 return 306 316 endif 307 317 if non-blocking 308 318 Unlock all monitors 309 319 Return 320 endif 310 321 311 322 push self to AS-stack -
doc/proposals/concurrency/text/parallelism.tex
r65d6de4 r9e1eabc 15 15 Examples of languages that support \glspl{uthread} are Erlang~\cite{Erlang} and \uC~\cite{uC++book}. 16 16 17 \subsection{Fibers : user-level threads without preemption} 17 \subsection{Fibers : user-level threads without preemption} \label{fibers} 18 18 A popular varient of \glspl{uthread} is what is often refered to as \glspl{fiber}. However, \glspl{fiber} do not present meaningful semantical differences with \glspl{uthread}. The significant difference between \glspl{uthread} and \glspl{fiber} is the lack of \gls{preemption} in the later one. Advocates of \glspl{fiber} list their high performance and ease of implementation as majors strenghts of \glspl{fiber} but the performance difference between \glspl{uthread} and \glspl{fiber} is controversial, and the ease of implementation, while true, is a weak argument in the context of language design. Therefore this proposal largely ignores fibers. 19 19 … … 33 33 34 34 \subsection{Future Work: Machine setup}\label{machine} 35 While this was not done in the context of this thesis, another important aspect of clusters is affinity. While many common desktop and laptop PCs have homogeneous CPUs, other devices often have more heteregenous setups. For example, system using \acrshort{numa} configurations may benefit from users being able to tie clusters and /or kernel threads to certains CPU cores. OS support for CPU affinity is now common \cit,which means it is both possible and desirable for \CFA to offer an abstraction mechanism for portable CPU affinity.35 While this was not done in the context of this thesis, another important aspect of clusters is affinity. While many common desktop and laptop PCs have homogeneous CPUs, other devices often have more heteregenous setups. For example, system using \acrshort{numa} configurations may benefit from users being able to tie clusters and\/or kernel threads to certains CPU cores. OS support for CPU affinity is now common \cite{affinityLinux, affinityWindows, affinityFreebsd, affinityNetbsd, affinityMacosx} which means it is both possible and desirable for \CFA to offer an abstraction mechanism for portable CPU affinity. 36 36 37 \subsection{Paradigms}\label{cfaparadigms}38 Given these building blocks, it is possible to reproduce all three of the popular paradigms. Indeed, \glspl{uthread} is the default paradigm in \CFA. However, disabling \gls{preemption} on the \gls{cfacluster} means \glspl{cfathread} effectively become \glspl{fiber}. Since several \glspl{cfacluster} with different scheduling policy can coexist in the same application, this allows \glspl{fiber} and \glspl{uthread} to coexist in the runtime of an application. Finally, it is possible to build executors for thread pools from \glspl{uthread} or \glspl{fiber}.37 % \subsection{Paradigms}\label{cfaparadigms} 38 % Given these building blocks, it is possible to reproduce all three of the popular paradigms. Indeed, \glspl{uthread} is the default paradigm in \CFA. However, disabling \gls{preemption} on the \gls{cfacluster} means \glspl{cfathread} effectively become \glspl{fiber}. Since several \glspl{cfacluster} with different scheduling policy can coexist in the same application, this allows \glspl{fiber} and \glspl{uthread} to coexist in the runtime of an application. Finally, it is possible to build executors for thread pools from \glspl{uthread} or \glspl{fiber}. -
doc/proposals/concurrency/text/results.tex
r65d6de4 r9e1eabc 1 1 % ====================================================================== 2 2 % ====================================================================== 3 \chapter{Performance results} 3 \chapter{Performance results} \label{results} 4 4 % ====================================================================== 5 5 % ====================================================================== 6 7 6 \section{Machine setup} 8 9 \begin{figure} 7 Table \ref{tab:machine} shows the characteristiques of the machine used to run the benchmarks. All tests where made on this machine. 8 \begin{figure}[H] 10 9 \begin{center} 11 10 \begin{tabular}{| l | r | l | r |} … … 37 36 38 37 \section{Micro benchmarks} 38 All benchmarks are run using the same harness to produce the results, seen as the \code{BENCH()} macro in the following examples. This macro uses the following logic to benchmark the code : 39 \begin{pseudo} 40 #define BENCH(run, result) 41 gettime(); 42 run; 43 gettime(); 44 result = (after - before) / N; 45 \end{pseudo} 46 The method used to get time is \code{clock_gettime(CLOCK_THREAD_CPUTIME_ID);}. Each benchmark is using many interations of a simple call to measure the cost of the call. The specific number of interation dependes on the specific benchmark. 47 48 \subsection{Context-switching} 49 The first interesting benchmark is to measure how long context-switches take. The simplest approach to do this is to yield on a thread, which executes a 2-step context switch. In order to make the comparison fair, coroutines also execute a 2-step context-switch, which is a resume/suspend cycle instead of a yield. Listing \ref{lst:ctx-switch} shows the code for coroutines and threads. All omitted tests are functionally identical to one of these tests. The results can be shown in table \ref{tab:ctx-switch}. 50 \begin{figure} 51 \begin{multicols}{2} 52 \CFA Coroutines 53 \begin{cfacode} 54 coroutine GreatSuspender {}; 55 void main(GreatSuspender& this) { 56 while(true) { suspend(); } 57 } 58 int main() { 59 GreatSuspender s; 60 resume(s); 61 BENCH( 62 for(size_t i=0; i<n; i++) { 63 resume(s); 64 }, 65 result 66 ) 67 printf("%llu\n", result); 68 } 69 \end{cfacode} 70 \columnbreak 71 \CFA Threads 72 \begin{cfacode} 73 74 75 76 77 int main() { 78 79 80 BENCH( 81 for(size_t i=0; i<n; i++) { 82 yield(); 83 }, 84 result 85 ) 86 printf("%llu\n", result); 87 } 88 \end{cfacode} 89 \end{multicols} 90 \caption{\CFA benchmark code used to measure context-switches for coroutines and threads.} 91 \label{lst:ctx-switch} 92 \end{figure} 39 93 40 94 \begin{figure} … … 54 108 \caption{Context Switch comparaison. All numbers are in nanoseconds(\si{\nano\second})} 55 109 \label{tab:ctx-switch} 110 \end{figure} 111 112 \subsection{Mutual-exclusion} 113 The next interesting benchmark is to measure the overhead to enter/leave a critical-section. For monitors, the simplest appraoch is to measure how long it takes enter and leave a monitor routine. Listing \ref{lst:mutex} shows the code for \CFA. To put the results in context, the cost of entering a non-inline function and the cost of acquiring and releasing a pthread mutex lock are also mesured. The results can be shown in table \ref{tab:mutex}. 114 115 \begin{figure} 116 \begin{cfacode} 117 monitor M {}; 118 void __attribute__((noinline)) call( M & mutex m /*, m2, m3, m4*/ ) {} 119 120 int main() { 121 M m/*, m2, m3, m4*/; 122 BENCH( 123 for(size_t i=0; i<n; i++) { 124 call(m/*, m2, m3, m4*/); 125 }, 126 result 127 ) 128 printf("%llu\n", result); 129 } 130 \end{cfacode} 131 \caption{\CFA benchmark code used to measure mutex routines.} 132 \label{lst:mutex} 56 133 \end{figure} 57 134 … … 75 152 \end{figure} 76 153 154 \subsection{Internal scheduling} 155 The Internal scheduling benchmark measures the cost of waiting on and signaling a condition variable. Listing \ref{lst:int-sched} shows the code for \CFA. The results can be shown in table \ref{tab:int-sched}. As with all other benchmarks, all omitted tests are functionally identical to one of these tests. 156 157 \begin{figure} 158 \begin{cfacode} 159 volatile int go = 0; 160 condition c; 161 monitor M {}; 162 M m1; 163 164 void __attribute__((noinline)) do_call( M & mutex a1 ) { signal(c); } 165 166 thread T {}; 167 void ^?{}( T & mutex this ) {} 168 void main( T & this ) { 169 while(go == 0) { yield(); } 170 while(go == 1) { do_call(m1); } 171 } 172 int __attribute__((noinline)) do_wait( M & mutex a1 ) { 173 go = 1; 174 BENCH( 175 for(size_t i=0; i<n; i++) { 176 wait(c); 177 }, 178 result 179 ) 180 printf("%llu\n", result); 181 go = 0; 182 return 0; 183 } 184 int main() { 185 T t; 186 return do_wait(m1); 187 } 188 \end{cfacode} 189 \caption{Benchmark code for internal scheduling} 190 \label{lst:int-sched} 191 \end{figure} 192 77 193 \begin{figure} 78 194 \begin{center} … … 92 208 \end{figure} 93 209 210 \subsection{External scheduling} 211 The Internal scheduling benchmark measures the cost of the \code{waitfor} statement (\code{_Accept} in \uC). Listing \ref{lst:ext-sched} shows the code for \CFA. The results can be shown in table \ref{tab:ext-sched}. As with all other benchmarks, all omitted tests are functionally identical to one of these tests. 212 213 \begin{figure} 214 \begin{cfacode} 215 volatile int go = 0; 216 monitor M {}; 217 M m1; 218 thread T {}; 219 220 void __attribute__((noinline)) do_call( M & mutex a1 ) {} 221 222 void ^?{}( T & mutex this ) {} 223 void main( T & this ) { 224 while(go == 0) { yield(); } 225 while(go == 1) { do_call(m1); } 226 } 227 int __attribute__((noinline)) do_wait( M & mutex a1 ) { 228 go = 1; 229 BENCH( 230 for(size_t i=0; i<n; i++) { 231 waitfor(call, a1); 232 }, 233 result 234 ) 235 printf("%llu\n", result); 236 go = 0; 237 return 0; 238 } 239 int main() { 240 T t; 241 return do_wait(m1); 242 } 243 \end{cfacode} 244 \caption{Benchmark code for external scheduling} 245 \label{lst:ext-sched} 246 \end{figure} 247 94 248 \begin{figure} 95 249 \begin{center} … … 109 263 \end{figure} 110 264 111 \begin{figure} 112 \begin{center} 113 \begin{tabular}{| l | S[table-format=5.2,table-number-alignment=right] | S[table-format=5.2,table-number-alignment=right] | S[table-format=5.2,table-number-alignment=right] |} 114 \cline{2-4} 115 \multicolumn{1}{c |}{} & \multicolumn{1}{c |}{ Median } &\multicolumn{1}{c |}{ Average } & \multicolumn{1}{c |}{ Standard Deviation} \\ 116 \hline 117 Pthreads & 26974.5 & 26977 & 124.12 \\ 118 \CFA Coroutines & 5 & 5 & 0 \\ 119 \CFA Threads & 1122.5 & 1109.86 & 36.54 \\ 120 \uC Coroutines & 106 & 107.04 & 1.61 \\ 121 \uC Threads & 525.5 & 533.04 & 11.14 \\ 265 \subsection{Object creation} 266 Finaly, the last benchmark measured is the cost of creation for concurrent objects. Listing \ref{lst:creation} shows the code for pthreads and \CFA threads. The results can be shown in table \ref{tab:creation}. As with all other benchmarks, all omitted tests are functionally identical to one of these tests. The only note here is that the callstacks of \CFA coroutines are lazily created, therefore without priming the coroutine, the creation cost is very low. 267 268 \begin{figure} 269 \begin{multicols}{2} 270 pthread 271 \begin{cfacode} 272 int main() { 273 BENCH( 274 for(size_t i=0; i<n; i++) { 275 pthread_t thread; 276 if(pthread_create( 277 &thread, 278 NULL, 279 foo, 280 NULL 281 ) < 0) { 282 perror( "failure" ); 283 return 1; 284 } 285 286 if(pthread_join( 287 thread, 288 NULL 289 ) < 0) { 290 perror( "failure" ); 291 return 1; 292 } 293 }, 294 result 295 ) 296 printf("%llu\n", result); 297 } 298 \end{cfacode} 299 \columnbreak 300 \CFA Threads 301 \begin{cfacode} 302 int main() { 303 BENCH( 304 for(size_t i=0; i<n; i++) { 305 MyThread m; 306 }, 307 result 308 ) 309 310 printf("%llu\n", result); 311 } 312 \end{cfacode} 313 \end{multicols} 314 \caption{Bechmark code for pthreads and \CFA to measure object creation} 315 \label{lst:creation} 316 \end{figure} 317 318 \begin{figure} 319 \begin{center} 320 \begin{tabular}{| l | S[table-format=5.2,table-number-alignment=right] | S[table-format=5.2,table-number-alignment=right] | S[table-format=5.2,table-number-alignment=right] |} 321 \cline{2-4} 322 \multicolumn{1}{c |}{} & \multicolumn{1}{c |}{ Median } &\multicolumn{1}{c |}{ Average } & \multicolumn{1}{c |}{ Standard Deviation} \\ 323 \hline 324 Pthreads & 26974.5 & 26977 & 124.12 \\ 325 \CFA Coroutines Lazy & 5 & 5 & 0 \\ 326 \CFA Coroutines Eager & 335.0 & 357.67 & 34.2 \\ 327 \CFA Threads & 1122.5 & 1109.86 & 36.54 \\ 328 \uC Coroutines & 106 & 107.04 & 1.61 \\ 329 \uC Threads & 525.5 & 533.04 & 11.14 \\ 122 330 \hline 123 331 \end{tabular} -
doc/proposals/concurrency/text/together.tex
r65d6de4 r9e1eabc 7 7 8 8 \section{Threads as monitors} 9 As it was subtely alluded in section \ref{threads}, \code{threads} in \CFA are in fact monitors . This means that all the monitorsfeatures are available when using threads. For example, here is a very simple two thread pipeline that could be used for a simulator of a game engine :9 As it was subtely alluded in section \ref{threads}, \code{threads} in \CFA are in fact monitors, which means that all monitor features are available when using threads. For example, here is a very simple two thread pipeline that could be used for a simulator of a game engine : 10 10 \begin{cfacode} 11 11 // Visualization declaration … … 72 72 } 73 73 } 74 75 // Call destructor for simulator once simulator finishes 76 // Call destructor for renderer to signify shutdown 74 77 \end{cfacode} 75 78 76 79 \section{Fibers \& Threads} 80 As mentionned in section \ref{preemption}, \CFA uses preemptive threads by default but can use fibers on demand. Currently, using fibers is done by adding the following line of code to the program~: 81 \begin{cfacode} 82 unsigned int default_preemption() { 83 return 0; 84 } 85 \end{cfacode} 86 This function is called by the kernel to fetch the default preemption rate, where 0 signifies an infinite time-slice i.e. no preemption. However, once clusters are fully implemented, it will be possible to create fibers and uthreads in on the same system : 87 \begin{figure} 88 \begin{cfacode} 89 //Cluster forward declaration 90 struct cluster; 91 92 //Processor forward declaration 93 struct processor; 94 95 //Construct clusters with a preemption rate 96 void ?{}(cluster& this, unsigned int rate); 97 //Construct processor and add it to cluster 98 void ?{}(processor& this, cluster& cluster); 99 //Construct thread and schedule it on cluster 100 void ?{}(thread& this, cluster& cluster); 101 102 //Declare two clusters 103 cluster thread_cluster = { 10`ms }; //Preempt every 10 ms 104 cluster fibers_cluster = { 0 }; //Never preempt 105 106 //Construct 4 processors 107 processor processors[4] = { 108 //2 for the thread cluster 109 thread_cluster; 110 thread_cluster; 111 //2 for the fibers cluster 112 fibers_cluster; 113 fibers_cluster; 114 }; 115 116 //Declares thread 117 thread UThread {}; 118 void ?{}(UThread& this) { 119 //Construct underlying thread to automatically 120 //be scheduled on the thread cluster 121 (this){ thread_cluster } 122 } 123 124 void main(UThread & this); 125 126 //Declares fibers 127 thread Fiber {}; 128 void ?{}(Fiber& this) { 129 //Construct underlying thread to automatically 130 //be scheduled on the fiber cluster 131 (this.__thread){ fibers_cluster } 132 } 133 134 void main(Fiber & this); 135 \end{cfacode} 136 \end{figure} -
doc/proposals/concurrency/version
r65d6de4 r9e1eabc 1 0.11. 471 0.11.129 -
src/Common/Debug.h
r65d6de4 r9e1eabc 24 24 #include "SynTree/Declaration.h" 25 25 26 /// debug codegen a translation unit 27 static inline void debugCodeGen( const std::list< Declaration * > & translationUnit, const std::string & label ) { 28 std::list< Declaration * > decls; 26 #define DEBUG 29 27 30 filter( translationUnit.begin(), translationUnit.end(), back_inserter( decls ), []( Declaration * decl ) { 31 return ! LinkageSpec::isBuiltin( decl->get_linkage() ); 32 }); 28 namespace Debug { 29 /// debug codegen a translation unit 30 static inline void codeGen( __attribute__((unused)) const std::list< Declaration * > & translationUnit, __attribute__((unused)) const std::string & label ) { 31 #ifdef DEBUG 32 std::list< Declaration * > decls; 33 33 34 std::cerr << "======" << label << "======" << std::endl; 35 CodeGen::generate( decls, std::cerr, false, true ); 36 } // dump 34 filter( translationUnit.begin(), translationUnit.end(), back_inserter( decls ), []( Declaration * decl ) { 35 return ! LinkageSpec::isBuiltin( decl->get_linkage() ); 36 }); 37 38 std::cerr << "======" << label << "======" << std::endl; 39 CodeGen::generate( decls, std::cerr, false, true ); 40 #endif 41 } // dump 42 43 static inline void treeDump( __attribute__((unused)) const std::list< Declaration * > & translationUnit, __attribute__((unused)) const std::string & label ) { 44 #ifdef DEBUG 45 std::list< Declaration * > decls; 46 47 filter( translationUnit.begin(), translationUnit.end(), back_inserter( decls ), []( Declaration * decl ) { 48 return ! LinkageSpec::isBuiltin( decl->get_linkage() ); 49 }); 50 51 std::cerr << "======" << label << "======" << std::endl; 52 printAll( decls, std::cerr ); 53 #endif 54 } // dump 55 } 37 56 38 57 // Local Variables: // -
src/ResolvExpr/AlternativeFinder.cc
r65d6de4 r9e1eabc 581 581 std::vector<unsigned> tupleEls; /// Number of elements in current tuple element(s) 582 582 583 ArgPack(const TypeEnvironment& env, const AssertionSet& need, const AssertionSet& have, 583 ArgPack(const TypeEnvironment& env, const AssertionSet& need, const AssertionSet& have, 584 584 const OpenVarSet& openVars) 585 585 : actuals(), env(env), need(need), have(have), openVars(openVars), nextArg(0), 586 586 expls(), nextExpl(0), tupleEls() {} 587 587 588 588 /// Starts a new tuple expression 589 589 void beginTuple() { … … 620 620 621 621 /// Instantiates an argument to match a formal, returns false if no results left 622 bool instantiateArgument( Type* formalType, Initializer* initializer, 623 const std::vector< AlternativeFinder >& args, 624 std::vector<ArgPack>& results, std::vector<ArgPack>& nextResults, 622 bool instantiateArgument( Type* formalType, Initializer* initializer, 623 const std::vector< AlternativeFinder >& args, 624 std::vector<ArgPack>& results, std::vector<ArgPack>& nextResults, 625 625 const SymTab::Indexer& indexer ) { 626 626 if ( TupleType* tupleType = dynamic_cast<TupleType*>( formalType ) ) { … … 629 629 for ( Type* type : *tupleType ) { 630 630 // xxx - dropping initializer changes behaviour from previous, but seems correct 631 if ( ! instantiateArgument( type, nullptr, args, results, nextResults, indexer ) ) 631 if ( ! instantiateArgument( type, nullptr, args, results, nextResults, indexer ) ) 632 632 return false; 633 633 } … … 658 658 Type* argType = result.actuals.back().expr->get_result(); 659 659 if ( result.tupleEls.back() == 1 && Tuples::isTtype( argType ) ) { 660 // the case where a ttype value is passed directly is special, e.g. for 660 // the case where a ttype value is passed directly is special, e.g. for 661 661 // argument forwarding purposes 662 662 // xxx - what if passing multiple arguments, last of which is ttype? 663 // xxx - what would happen if unify was changed so that unifying tuple 663 // xxx - what would happen if unify was changed so that unifying tuple 664 664 // types flattened both before unifying lists? then pass in TupleType 665 665 // (ttype) below. … … 671 671 } 672 672 // check unification for ttype before adding to final 673 if ( unify( ttype, argType, result.env, result.need, result.have, 673 if ( unify( ttype, argType, result.env, result.need, result.have, 674 674 result.openVars, indexer ) ) { 675 675 finalResults.push_back( std::move(result) ); … … 684 684 aResult.env.addActual( actual.env, aResult.openVars ); 685 685 Cost cost = actual.cost; 686 686 687 687 // explode argument 688 688 std::vector<Alternative> exploded; 689 689 Tuples::explode( actual, indexer, back_inserter( exploded ) ); 690 690 691 691 // add exploded argument to tuple 692 692 for ( Alternative& aActual : exploded ) { … … 706 706 return ! results.empty(); 707 707 } 708 708 709 709 // iterate each current subresult 710 710 for ( unsigned iResult = 0; iResult < results.size(); ++iResult ) { … … 724 724 std::cerr << std::endl; 725 725 ) 726 727 if ( unify( formalType, actualType, result.env, result.need, result.have, 726 727 if ( unify( formalType, actualType, result.env, result.need, result.have, 728 728 result.openVars, indexer ) ) { 729 729 ++result.nextExpl; … … 736 736 if ( ConstantExpr* cnstExpr = getDefaultValue( initializer ) ) { 737 737 if ( Constant* cnst = dynamic_cast<Constant*>( cnstExpr->get_constant() ) ) { 738 if ( unify( formalType, cnst->get_type(), result.env, result.need, 738 if ( unify( formalType, cnst->get_type(), result.env, result.need, 739 739 result.have, result.openVars, indexer ) ) { 740 740 nextResults.push_back( std::move(result.withArg( cnstExpr )) ); … … 791 791 results.swap( nextResults ); 792 792 nextResults.clear(); 793 793 794 794 return ! results.empty(); 795 } 795 } 796 796 797 797 template<typename OutputIterator> 798 void AlternativeFinder::makeFunctionAlternatives( const Alternative &func, 799 FunctionType *funcType, const std::vector< AlternativeFinder > &args, 798 void AlternativeFinder::makeFunctionAlternatives( const Alternative &func, 799 FunctionType *funcType, const std::vector< AlternativeFinder > &args, 800 800 OutputIterator out ) { 801 801 OpenVarSet funcOpenVars; … … 803 803 TypeEnvironment funcEnv( func.env ); 804 804 makeUnifiableVars( funcType, funcOpenVars, funcNeed ); 805 // add all type variables as open variables now so that those not used in the parameter 805 // add all type variables as open variables now so that those not used in the parameter 806 806 // list are still considered open. 807 807 funcEnv.add( funcType->get_forall() ); 808 808 809 809 if ( targetType && ! targetType->isVoid() && ! funcType->get_returnVals().empty() ) { 810 810 // attempt to narrow based on expected target type 811 811 Type * returnType = funcType->get_returnVals().front()->get_type(); 812 if ( ! unify( returnType, targetType, funcEnv, funcNeed, funcHave, funcOpenVars, 812 if ( ! unify( returnType, targetType, funcEnv, funcNeed, funcHave, funcOpenVars, 813 813 indexer ) ) { 814 814 // unification failed, don't pursue this function alternative … … 822 822 for ( DeclarationWithType* formal : funcType->get_parameters() ) { 823 823 ObjectDecl* obj = strict_dynamic_cast< ObjectDecl* >( formal ); 824 if ( ! instantiateArgument( 824 if ( ! instantiateArgument( 825 825 obj->get_type(), obj->get_init(), args, results, nextResults, indexer ) ) 826 826 return; … … 904 904 905 905 std::vector< AlternativeFinder > argAlternatives; 906 findSubExprs( untypedExpr->begin_args(), untypedExpr->end_args(), 906 findSubExprs( untypedExpr->begin_args(), untypedExpr->end_args(), 907 907 back_inserter( argAlternatives ) ); 908 908 … … 934 934 Alternative newFunc( *func ); 935 935 referenceToRvalueConversion( newFunc.expr ); 936 makeFunctionAlternatives( newFunc, function, argAlternatives, 936 makeFunctionAlternatives( newFunc, function, argAlternatives, 937 937 std::back_inserter( candidates ) ); 938 938 } … … 943 943 Alternative newFunc( *func ); 944 944 referenceToRvalueConversion( newFunc.expr ); 945 makeFunctionAlternatives( newFunc, function, argAlternatives, 945 makeFunctionAlternatives( newFunc, function, argAlternatives, 946 946 std::back_inserter( candidates ) ); 947 947 } // if 948 948 } // if 949 } 949 } 950 950 } catch ( SemanticError &e ) { 951 951 errors.append( e ); … … 962 962 try { 963 963 // check if type is a pointer to function 964 if ( PointerType* pointer = dynamic_cast<PointerType*>( 964 if ( PointerType* pointer = dynamic_cast<PointerType*>( 965 965 funcOp->expr->get_result()->stripReferences() ) ) { 966 if ( FunctionType* function = 966 if ( FunctionType* function = 967 967 dynamic_cast<FunctionType*>( pointer->get_base() ) ) { 968 968 Alternative newFunc( *funcOp ); 969 969 referenceToRvalueConversion( newFunc.expr ); 970 makeFunctionAlternatives( newFunc, function, argAlternatives, 970 makeFunctionAlternatives( newFunc, function, argAlternatives, 971 971 std::back_inserter( candidates ) ); 972 972 } … … 1007 1007 candidates.splice( candidates.end(), alternatives ); 1008 1008 1009 findMinCost( candidates.begin(), candidates.end(), std::back_inserter( alternatives ) ); 1009 // use a new list so that alternatives are not examined by addAnonConversions twice. 1010 AltList winners; 1011 findMinCost( candidates.begin(), candidates.end(), std::back_inserter( winners ) ); 1010 1012 1011 1013 // function may return struct or union value, in which case we need to add alternatives for implicit 1012 1014 // conversions to each of the anonymous members, must happen after findMinCost since anon conversions 1013 1015 // are never the cheapest expression 1014 for ( const Alternative & alt : alternatives ) {1016 for ( const Alternative & alt : winners ) { 1015 1017 addAnonConversions( alt ); 1016 1018 } 1019 alternatives.splice( alternatives.begin(), winners ); 1017 1020 1018 1021 if ( alternatives.empty() && targetType && ! targetType->isVoid() ) { -
src/ResolvExpr/RenameVars.cc
r65d6de4 r9e1eabc 29 29 RenameVars global_renamer; 30 30 31 RenameVars::RenameVars() : level( 0 ) {31 RenameVars::RenameVars() : level( 0 ), resetCount( 0 ) { 32 32 mapStack.push_front( std::map< std::string, std::string >() ); 33 33 } … … 35 35 void RenameVars::reset() { 36 36 level = 0; 37 resetCount++; 37 38 } 38 39 … … 130 131 for ( Type::ForallList::iterator i = type->get_forall().begin(); i != type->get_forall().end(); ++i ) { 131 132 std::ostringstream output; 132 output << "_" << level << "_" << (*i)->get_name();133 output << "_" << resetCount << "_" << level << "_" << (*i)->get_name(); 133 134 std::string newname( output.str() ); 134 135 mapStack.front()[ (*i)->get_name() ] = newname; -
src/ResolvExpr/RenameVars.h
r65d6de4 r9e1eabc 48 48 void typeBefore( Type *type ); 49 49 void typeAfter( Type *type ); 50 int level ;50 int level, resetCount; 51 51 std::list< std::map< std::string, std::string > > mapStack; 52 52 }; -
src/benchmark/Makefile.am
r65d6de4 r9e1eabc 23 23 STATS = ${TOOLSDIR}stat.py 24 24 repeats = 30 25 TIME_FORMAT = "%E" 26 PRINT_FORMAT = %20s: #Comments needed for spacing 25 27 26 28 .NOTPARALLEL: … … 29 31 30 32 all : ctxswitch$(EXEEXT) mutex$(EXEEXT) signal$(EXEEXT) waitfor$(EXEEXT) creation$(EXEEXT) 31 32 bench$(EXEEXT) :33 @for ccflags in "-debug" "-nodebug"; do \34 echo ${CC} ${AM_CFLAGS} ${CFLAGS} ${ccflags} @CFA_FLAGS@ -lrt bench.c;\35 ${CC} ${AM_CFLAGS} ${CFLAGS} $${ccflags} -lrt bench.c;\36 ./a.out ; \37 done ; \38 rm -f ./a.out ;39 40 csv-data$(EXEEXT):41 @${CC} ${AM_CFLAGS} ${CFLAGS} ${ccflags} @CFA_FLAGS@ -nodebug -lrt -quiet -DN=50000000 csv-data.c42 @./a.out43 @rm -f ./a.out44 45 ## =========================================================================================================46 ctxswitch$(EXEEXT): \47 ctxswitch-pthread.run \48 ctxswitch-cfa_coroutine.run \49 ctxswitch-cfa_thread.run \50 ctxswitch-upp_coroutine.run \51 ctxswitch-upp_thread.run52 53 ctxswitch-cfa_coroutine$(EXEEXT):54 ${CC} ctxswitch/cfa_cor.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}55 56 ctxswitch-cfa_thread$(EXEEXT):57 ${CC} ctxswitch/cfa_thrd.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}58 59 ctxswitch-upp_coroutine$(EXEEXT):60 u++ ctxswitch/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}61 62 ctxswitch-upp_thread$(EXEEXT):63 u++ ctxswitch/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}64 65 ctxswitch-pthread$(EXEEXT):66 @BACKEND_CC@ ctxswitch/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}67 68 ## =========================================================================================================69 mutex$(EXEEXT) :\70 mutex-function.run \71 mutex-pthread_lock.run \72 mutex-upp.run \73 mutex-cfa1.run \74 mutex-cfa2.run \75 mutex-cfa4.run76 77 mutex-function$(EXEEXT):78 @BACKEND_CC@ mutex/function.c -DBENCH_N=500000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}79 80 mutex-pthread_lock$(EXEEXT):81 @BACKEND_CC@ mutex/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}82 83 mutex-upp$(EXEEXT):84 u++ mutex/upp.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}85 86 mutex-cfa1$(EXEEXT):87 ${CC} mutex/cfa1.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}88 89 mutex-cfa2$(EXEEXT):90 ${CC} mutex/cfa2.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}91 92 mutex-cfa4$(EXEEXT):93 ${CC} mutex/cfa4.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}94 95 ## =========================================================================================================96 signal$(EXEEXT) :\97 signal-upp.run \98 signal-cfa1.run \99 signal-cfa2.run \100 signal-cfa4.run101 102 signal-upp$(EXEEXT):103 u++ schedint/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}104 105 signal-cfa1$(EXEEXT):106 ${CC} schedint/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}107 108 signal-cfa2$(EXEEXT):109 ${CC} schedint/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}110 111 signal-cfa4$(EXEEXT):112 ${CC} schedint/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}113 114 ## =========================================================================================================115 waitfor$(EXEEXT) :\116 waitfor-upp.run \117 waitfor-cfa1.run \118 waitfor-cfa2.run \119 waitfor-cfa4.run120 121 waitfor-upp$(EXEEXT):122 u++ schedext/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}123 124 waitfor-cfa1$(EXEEXT):125 ${CC} schedext/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}126 127 waitfor-cfa2$(EXEEXT):128 ${CC} schedext/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}129 130 waitfor-cfa4$(EXEEXT):131 ${CC} schedext/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}132 133 ## =========================================================================================================134 creation$(EXEEXT) :\135 creation-pthread.run \136 creation-cfa_coroutine.run \137 creation-cfa_coroutine_eager.run \138 creation-cfa_thread.run \139 creation-upp_coroutine.run \140 creation-upp_thread.run141 142 creation-cfa_coroutine$(EXEEXT):143 ${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}144 145 creation-cfa_coroutine_eager$(EXEEXT):146 ${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} -DEAGER147 148 creation-cfa_thread$(EXEEXT):149 ${CC} creation/cfa_thrd.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}150 151 creation-upp_coroutine$(EXEEXT):152 u++ creation/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}153 154 creation-upp_thread$(EXEEXT):155 u++ creation/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}156 157 creation-pthread$(EXEEXT):158 @BACKEND_CC@ creation/pthreads.c -DBENCH_N=250000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}159 160 ## =========================================================================================================161 33 162 34 %.run : %$(EXEEXT) ${REPEAT} … … 169 41 @rm -f a.out .result.log 170 42 43 %.runquiet : 44 @+make $(basename $@) 45 @./a.out 46 @rm -f a.out 47 48 %.make : 49 @printf "${PRINT_FORMAT}" $(basename $(subst compile-,,$@)) 50 @+/usr/bin/time -f ${TIME_FORMAT} make $(basename $@) 2>&1 51 171 52 ${REPEAT} : 172 53 @+make -C ${TOOLSDIR} repeat 54 55 ## ========================================================================================================= 56 57 jenkins$(EXEEXT): 58 @echo "{" 59 @echo -e '\t"githash": "'${githash}'",' 60 @echo -e '\t"arch": "' ${arch} '",' 61 @echo -e '\t"compile": {' 62 @+make compile TIME_FORMAT='%e,' PRINT_FORMAT='\t\t\"%s\" :' 63 @echo -e '\t\t"dummy" : {}' 64 @echo -e '\t},' 65 @echo -e '\t"ctxswitch": {' 66 @echo -en '\t\t"coroutine":' 67 @+make ctxswitch-cfa_coroutine.runquiet 68 @echo -en '\t\t,"thread":' 69 @+make ctxswitch-cfa_thread.runquiet 70 @echo -e '\t},' 71 @echo -e '\t"mutex": [' 72 @echo -en '\t\t' 73 @+make mutex-cfa1.runquiet 74 @echo -en '\t\t,' 75 @+make mutex-cfa2.runquiet 76 @echo -e '\t],' 77 @echo -e '\t"scheduling": [' 78 @echo -en '\t\t' 79 @+make signal-cfa1.runquiet 80 @echo -en '\t\t,' 81 @+make signal-cfa2.runquiet 82 @echo -en '\t\t,' 83 @+make waitfor-cfa1.runquiet 84 @echo -en '\t\t,' 85 @+make waitfor-cfa2.runquiet 86 @echo -e '\n\t],' 87 @echo -e '\t"epoch": ' $(shell date +%s) 88 @echo "}" 89 90 ## ========================================================================================================= 91 ctxswitch$(EXEEXT): \ 92 ctxswitch-pthread.run \ 93 ctxswitch-cfa_coroutine.run \ 94 ctxswitch-cfa_thread.run \ 95 ctxswitch-upp_coroutine.run \ 96 ctxswitch-upp_thread.run 97 98 ctxswitch-cfa_coroutine$(EXEEXT): 99 @${CC} ctxswitch/cfa_cor.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 100 101 ctxswitch-cfa_thread$(EXEEXT): 102 @${CC} ctxswitch/cfa_thrd.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 103 104 ctxswitch-upp_coroutine$(EXEEXT): 105 @u++ ctxswitch/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 106 107 ctxswitch-upp_thread$(EXEEXT): 108 @u++ ctxswitch/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 109 110 ctxswitch-pthread$(EXEEXT): 111 @@BACKEND_CC@ ctxswitch/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 112 113 ## ========================================================================================================= 114 mutex$(EXEEXT) :\ 115 mutex-function.run \ 116 mutex-pthread_lock.run \ 117 mutex-upp.run \ 118 mutex-cfa1.run \ 119 mutex-cfa2.run \ 120 mutex-cfa4.run 121 122 mutex-function$(EXEEXT): 123 @@BACKEND_CC@ mutex/function.c -DBENCH_N=500000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 124 125 mutex-pthread_lock$(EXEEXT): 126 @@BACKEND_CC@ mutex/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 127 128 mutex-upp$(EXEEXT): 129 @u++ mutex/upp.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 130 131 mutex-cfa1$(EXEEXT): 132 @${CC} mutex/cfa1.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 133 134 mutex-cfa2$(EXEEXT): 135 @${CC} mutex/cfa2.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 136 137 mutex-cfa4$(EXEEXT): 138 @${CC} mutex/cfa4.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 139 140 ## ========================================================================================================= 141 signal$(EXEEXT) :\ 142 signal-upp.run \ 143 signal-cfa1.run \ 144 signal-cfa2.run \ 145 signal-cfa4.run 146 147 signal-upp$(EXEEXT): 148 @u++ schedint/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 149 150 signal-cfa1$(EXEEXT): 151 @${CC} schedint/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 152 153 signal-cfa2$(EXEEXT): 154 @${CC} schedint/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 155 156 signal-cfa4$(EXEEXT): 157 @${CC} schedint/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 158 159 ## ========================================================================================================= 160 waitfor$(EXEEXT) :\ 161 waitfor-upp.run \ 162 waitfor-cfa1.run \ 163 waitfor-cfa2.run \ 164 waitfor-cfa4.run 165 166 waitfor-upp$(EXEEXT): 167 @u++ schedext/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 168 169 waitfor-cfa1$(EXEEXT): 170 @${CC} schedext/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 171 172 waitfor-cfa2$(EXEEXT): 173 @${CC} schedext/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 174 175 waitfor-cfa4$(EXEEXT): 176 @${CC} schedext/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 177 178 ## ========================================================================================================= 179 creation$(EXEEXT) :\ 180 creation-pthread.run \ 181 creation-cfa_coroutine.run \ 182 creation-cfa_coroutine_eager.run \ 183 creation-cfa_thread.run \ 184 creation-upp_coroutine.run \ 185 creation-upp_thread.run 186 187 creation-cfa_coroutine$(EXEEXT): 188 @${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 189 190 creation-cfa_coroutine_eager$(EXEEXT): 191 @${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} -DEAGER 192 193 creation-cfa_thread$(EXEEXT): 194 @${CC} creation/cfa_thrd.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 195 196 creation-upp_coroutine$(EXEEXT): 197 @u++ creation/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 198 199 creation-upp_thread$(EXEEXT): 200 @u++ creation/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 201 202 creation-pthread$(EXEEXT): 203 @@BACKEND_CC@ creation/pthreads.c -DBENCH_N=250000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 204 205 ## ========================================================================================================= 206 207 compile$(EXEEXT) :\ 208 compile-array.make \ 209 compile-attributes.make \ 210 compile-empty.make \ 211 compile-expression.make \ 212 compile-io.make \ 213 compile-monitor.make \ 214 compile-operators.make \ 215 compile-typeof.make 216 217 218 compile-array$(EXEEXT): 219 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/array.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 220 221 compile-attributes$(EXEEXT): 222 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/attributes.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 223 224 compile-empty$(EXEEXT): 225 @${CC} -nodebug -quiet -fsyntax-only -w compile/empty.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 226 227 compile-expression$(EXEEXT): 228 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/expression.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 229 230 compile-io$(EXEEXT): 231 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/io.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 232 233 compile-monitor$(EXEEXT): 234 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/monitor.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 235 236 compile-operators$(EXEEXT): 237 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/operators.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 238 239 compile-thread$(EXEEXT): 240 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/thread.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 241 242 compile-typeof$(EXEEXT): 243 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/typeof.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 244 -
src/benchmark/Makefile.in
r65d6de4 r9e1eabc 124 124 esac 125 125 am__tagged_files = $(HEADERS) $(SOURCES) $(TAGS_FILES) $(LISP) 126 am__DIST_COMMON = $(srcdir)/Makefile.in 126 am__DIST_COMMON = $(srcdir)/Makefile.in compile 127 127 DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) 128 128 ACLOCAL = @ACLOCAL@ … … 253 253 STATS = ${TOOLSDIR}stat.py 254 254 repeats = 30 255 TIME_FORMAT = "%E" 256 PRINT_FORMAT = %20s: #Comments needed for spacing 255 257 all: all-am 256 258 … … 446 448 all : ctxswitch$(EXEEXT) mutex$(EXEEXT) signal$(EXEEXT) waitfor$(EXEEXT) creation$(EXEEXT) 447 449 448 bench$(EXEEXT) :449 @for ccflags in "-debug" "-nodebug"; do \450 echo ${CC} ${AM_CFLAGS} ${CFLAGS} ${ccflags} @CFA_FLAGS@ -lrt bench.c;\451 ${CC} ${AM_CFLAGS} ${CFLAGS} $${ccflags} -lrt bench.c;\452 ./a.out ; \453 done ; \454 rm -f ./a.out ;455 456 csv-data$(EXEEXT):457 @${CC} ${AM_CFLAGS} ${CFLAGS} ${ccflags} @CFA_FLAGS@ -nodebug -lrt -quiet -DN=50000000 csv-data.c458 @./a.out459 @rm -f ./a.out460 461 ctxswitch$(EXEEXT): \462 ctxswitch-pthread.run \463 ctxswitch-cfa_coroutine.run \464 ctxswitch-cfa_thread.run \465 ctxswitch-upp_coroutine.run \466 ctxswitch-upp_thread.run467 468 ctxswitch-cfa_coroutine$(EXEEXT):469 ${CC} ctxswitch/cfa_cor.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}470 471 ctxswitch-cfa_thread$(EXEEXT):472 ${CC} ctxswitch/cfa_thrd.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}473 474 ctxswitch-upp_coroutine$(EXEEXT):475 u++ ctxswitch/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}476 477 ctxswitch-upp_thread$(EXEEXT):478 u++ ctxswitch/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}479 480 ctxswitch-pthread$(EXEEXT):481 @BACKEND_CC@ ctxswitch/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}482 483 mutex$(EXEEXT) :\484 mutex-function.run \485 mutex-pthread_lock.run \486 mutex-upp.run \487 mutex-cfa1.run \488 mutex-cfa2.run \489 mutex-cfa4.run490 491 mutex-function$(EXEEXT):492 @BACKEND_CC@ mutex/function.c -DBENCH_N=500000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}493 494 mutex-pthread_lock$(EXEEXT):495 @BACKEND_CC@ mutex/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}496 497 mutex-upp$(EXEEXT):498 u++ mutex/upp.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}499 500 mutex-cfa1$(EXEEXT):501 ${CC} mutex/cfa1.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}502 503 mutex-cfa2$(EXEEXT):504 ${CC} mutex/cfa2.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}505 506 mutex-cfa4$(EXEEXT):507 ${CC} mutex/cfa4.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}508 509 signal$(EXEEXT) :\510 signal-upp.run \511 signal-cfa1.run \512 signal-cfa2.run \513 signal-cfa4.run514 515 signal-upp$(EXEEXT):516 u++ schedint/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}517 518 signal-cfa1$(EXEEXT):519 ${CC} schedint/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}520 521 signal-cfa2$(EXEEXT):522 ${CC} schedint/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}523 524 signal-cfa4$(EXEEXT):525 ${CC} schedint/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}526 527 waitfor$(EXEEXT) :\528 waitfor-upp.run \529 waitfor-cfa1.run \530 waitfor-cfa2.run \531 waitfor-cfa4.run532 533 waitfor-upp$(EXEEXT):534 u++ schedext/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}535 536 waitfor-cfa1$(EXEEXT):537 ${CC} schedext/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}538 539 waitfor-cfa2$(EXEEXT):540 ${CC} schedext/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}541 542 waitfor-cfa4$(EXEEXT):543 ${CC} schedext/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}544 545 creation$(EXEEXT) :\546 creation-pthread.run \547 creation-cfa_coroutine.run \548 creation-cfa_coroutine_eager.run \549 creation-cfa_thread.run \550 creation-upp_coroutine.run \551 creation-upp_thread.run552 553 creation-cfa_coroutine$(EXEEXT):554 ${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}555 556 creation-cfa_coroutine_eager$(EXEEXT):557 ${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} -DEAGER558 559 creation-cfa_thread$(EXEEXT):560 ${CC} creation/cfa_thrd.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags}561 562 creation-upp_coroutine$(EXEEXT):563 u++ creation/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}564 565 creation-upp_thread$(EXEEXT):566 u++ creation/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags}567 568 creation-pthread$(EXEEXT):569 @BACKEND_CC@ creation/pthreads.c -DBENCH_N=250000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags}570 571 450 %.run : %$(EXEEXT) ${REPEAT} 572 451 @rm -f .result.log … … 578 457 @rm -f a.out .result.log 579 458 459 %.runquiet : 460 @+make $(basename $@) 461 @./a.out 462 @rm -f a.out 463 464 %.make : 465 @printf "${PRINT_FORMAT}" $(basename $(subst compile-,,$@)) 466 @+/usr/bin/time -f ${TIME_FORMAT} make $(basename $@) 2>&1 467 580 468 ${REPEAT} : 581 469 @+make -C ${TOOLSDIR} repeat 470 471 jenkins$(EXEEXT): 472 @echo "{" 473 @echo -e '\t"githash": "'${githash}'",' 474 @echo -e '\t"arch": "' ${arch} '",' 475 @echo -e '\t"compile": {' 476 @+make compile TIME_FORMAT='%e,' PRINT_FORMAT='\t\t\"%s\" :' 477 @echo -e '\t\t"dummy" : {}' 478 @echo -e '\t},' 479 @echo -e '\t"ctxswitch": {' 480 @echo -en '\t\t"coroutine":' 481 @+make ctxswitch-cfa_coroutine.runquiet 482 @echo -en '\t\t,"thread":' 483 @+make ctxswitch-cfa_thread.runquiet 484 @echo -e '\t},' 485 @echo -e '\t"mutex": [' 486 @echo -en '\t\t' 487 @+make mutex-cfa1.runquiet 488 @echo -en '\t\t,' 489 @+make mutex-cfa2.runquiet 490 @echo -e '\t],' 491 @echo -e '\t"scheduling": [' 492 @echo -en '\t\t' 493 @+make signal-cfa1.runquiet 494 @echo -en '\t\t,' 495 @+make signal-cfa2.runquiet 496 @echo -en '\t\t,' 497 @+make waitfor-cfa1.runquiet 498 @echo -en '\t\t,' 499 @+make waitfor-cfa2.runquiet 500 @echo -e '\n\t],' 501 @echo -e '\t"epoch": ' $(shell date +%s) 502 @echo "}" 503 504 ctxswitch$(EXEEXT): \ 505 ctxswitch-pthread.run \ 506 ctxswitch-cfa_coroutine.run \ 507 ctxswitch-cfa_thread.run \ 508 ctxswitch-upp_coroutine.run \ 509 ctxswitch-upp_thread.run 510 511 ctxswitch-cfa_coroutine$(EXEEXT): 512 @${CC} ctxswitch/cfa_cor.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 513 514 ctxswitch-cfa_thread$(EXEEXT): 515 @${CC} ctxswitch/cfa_thrd.c -DBENCH_N=50000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 516 517 ctxswitch-upp_coroutine$(EXEEXT): 518 @u++ ctxswitch/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 519 520 ctxswitch-upp_thread$(EXEEXT): 521 @u++ ctxswitch/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 522 523 ctxswitch-pthread$(EXEEXT): 524 @@BACKEND_CC@ ctxswitch/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 525 526 mutex$(EXEEXT) :\ 527 mutex-function.run \ 528 mutex-pthread_lock.run \ 529 mutex-upp.run \ 530 mutex-cfa1.run \ 531 mutex-cfa2.run \ 532 mutex-cfa4.run 533 534 mutex-function$(EXEEXT): 535 @@BACKEND_CC@ mutex/function.c -DBENCH_N=500000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 536 537 mutex-pthread_lock$(EXEEXT): 538 @@BACKEND_CC@ mutex/pthreads.c -DBENCH_N=50000000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 539 540 mutex-upp$(EXEEXT): 541 @u++ mutex/upp.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 542 543 mutex-cfa1$(EXEEXT): 544 @${CC} mutex/cfa1.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 545 546 mutex-cfa2$(EXEEXT): 547 @${CC} mutex/cfa2.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 548 549 mutex-cfa4$(EXEEXT): 550 @${CC} mutex/cfa4.c -DBENCH_N=5000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 551 552 signal$(EXEEXT) :\ 553 signal-upp.run \ 554 signal-cfa1.run \ 555 signal-cfa2.run \ 556 signal-cfa4.run 557 558 signal-upp$(EXEEXT): 559 @u++ schedint/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 560 561 signal-cfa1$(EXEEXT): 562 @${CC} schedint/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 563 564 signal-cfa2$(EXEEXT): 565 @${CC} schedint/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 566 567 signal-cfa4$(EXEEXT): 568 @${CC} schedint/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 569 570 waitfor$(EXEEXT) :\ 571 waitfor-upp.run \ 572 waitfor-cfa1.run \ 573 waitfor-cfa2.run \ 574 waitfor-cfa4.run 575 576 waitfor-upp$(EXEEXT): 577 @u++ schedext/upp.cc -DBENCH_N=5000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 578 579 waitfor-cfa1$(EXEEXT): 580 @${CC} schedext/cfa1.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 581 582 waitfor-cfa2$(EXEEXT): 583 @${CC} schedext/cfa2.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 584 585 waitfor-cfa4$(EXEEXT): 586 @${CC} schedext/cfa4.c -DBENCH_N=500000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 587 588 creation$(EXEEXT) :\ 589 creation-pthread.run \ 590 creation-cfa_coroutine.run \ 591 creation-cfa_coroutine_eager.run \ 592 creation-cfa_thread.run \ 593 creation-upp_coroutine.run \ 594 creation-upp_thread.run 595 596 creation-cfa_coroutine$(EXEEXT): 597 @${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 598 599 creation-cfa_coroutine_eager$(EXEEXT): 600 @${CC} creation/cfa_cor.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} -DEAGER 601 602 creation-cfa_thread$(EXEEXT): 603 @${CC} creation/cfa_thrd.c -DBENCH_N=10000000 -I. -nodebug -lrt -quiet @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 604 605 creation-upp_coroutine$(EXEEXT): 606 @u++ creation/upp_cor.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 607 608 creation-upp_thread$(EXEEXT): 609 @u++ creation/upp_thrd.cc -DBENCH_N=50000000 -I. -nodebug -lrt -quiet ${AM_CFLAGS} ${CFLAGS} ${ccflags} 610 611 creation-pthread$(EXEEXT): 612 @@BACKEND_CC@ creation/pthreads.c -DBENCH_N=250000 -I. -lrt -pthread ${AM_CFLAGS} ${CFLAGS} ${ccflags} 613 614 compile$(EXEEXT) :\ 615 compile-array.make \ 616 compile-attributes.make \ 617 compile-empty.make \ 618 compile-expression.make \ 619 compile-io.make \ 620 compile-monitor.make \ 621 compile-operators.make \ 622 compile-typeof.make 623 624 compile-array$(EXEEXT): 625 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/array.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 626 627 compile-attributes$(EXEEXT): 628 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/attributes.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 629 630 compile-empty$(EXEEXT): 631 @${CC} -nodebug -quiet -fsyntax-only -w compile/empty.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 632 633 compile-expression$(EXEEXT): 634 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/expression.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 635 636 compile-io$(EXEEXT): 637 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/io.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 638 639 compile-monitor$(EXEEXT): 640 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/monitor.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 641 642 compile-operators$(EXEEXT): 643 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/operators.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 644 645 compile-thread$(EXEEXT): 646 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/thread.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 647 648 compile-typeof$(EXEEXT): 649 @${CC} -nodebug -quiet -fsyntax-only -w ../tests/typeof.c @CFA_FLAGS@ ${AM_CFLAGS} ${CFLAGS} ${ccflags} 582 650 583 651 # Tell versions [3.59,3.63) of GNU make to not export all variables. -
src/main.cc
r65d6de4 r9e1eabc 206 206 FILE * extras = fopen( libcfap | treep ? "../prelude/extras.cf" : CFA_LIBDIR "/extras.cf", "r" ); 207 207 assertf( extras, "cannot open extras.cf\n" ); 208 parse( extras, LinkageSpec:: C );208 parse( extras, LinkageSpec::BuiltinC ); 209 209 210 210 if ( ! libcfap ) { -
src/tests/.expect/32/KRfunctions.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 signed int __f0__Fi_iPCii__1(signed int __a__i_1, const signed int *__b__PCi_1, signed int __c__i_1){ 8 2 __attribute__ ((unused)) signed int ___retval_f0__i_1; -
src/tests/.expect/32/attributes.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 signed int __la__Fi___1(){ 8 2 __attribute__ ((unused)) signed int ___retval_la__i_1; -
src/tests/.expect/32/declarationSpecifier.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 volatile const signed short int __x1__CVs_1; 8 2 static volatile const signed short int __x2__CVs_1; … … 701 695 } 702 696 static inline int invoke_main(int argc, char* argv[], char* envp[]) { (void)argc; (void)argv; (void)envp; return __main__Fi_iPPCc__1(argc, argv); } 703 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);704 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);705 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);706 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));707 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);708 extern signed int printf(const char *__restrict __format, ...);709 697 static inline signed int invoke_main(signed int argc, char **argv, char **envp); 710 698 signed int main(signed int __argc__i_1, char **__argv__PPc_1, char **__envp__PPc_1){ -
src/tests/.expect/32/extension.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 __extension__ signed int __a__i_1; 8 2 __extension__ signed int __b__i_1; -
src/tests/.expect/32/gccExtensions.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 extern signed int __x__i_1 asm ( "xx" ); 8 2 signed int __main__Fi_iPPCc__1(signed int __argc__i_1, const char **__argv__PPCc_1){ … … 174 168 } 175 169 static inline int invoke_main(int argc, char* argv[], char* envp[]) { (void)argc; (void)argv; (void)envp; return __main__Fi_iPPCc__1(argc, argv); } 176 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);177 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);178 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);179 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));180 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);181 extern signed int printf(const char *__restrict __format, ...);182 170 static inline signed int invoke_main(signed int argc, char **argv, char **envp); 183 171 signed int main(signed int __argc__i_1, char **__argv__PPc_1, char **__envp__PPc_1){ -
src/tests/.expect/32/literals.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 void __for_each__A0_2_0_0____operator_assign__PFd0_Rd0d0____constructor__PF_Rd0____constructor__PF_Rd0d0____destructor__PF_Rd0____operator_assign__PFd1_Rd1d1____constructor__PF_Rd1____constructor__PF_Rd1d1____destructor__PF_Rd1____operator_preincr__PFd0_Rd0____operator_predecr__PFd0_Rd0____operator_equal__PFi_d0d0____operator_notequal__PFi_d0d0____operator_deref__PFRd1_d0__F_d0d0PF_d1___1(__attribute__ ((unused)) void (*_adapterF_9telt_type__P)(void (*__anonymous_object0)(), void *__anonymous_object1), __attribute__ ((unused)) void *(*_adapterFP9telt_type_14titerator_type_M_P)(void (*__anonymous_object2)(), void *__anonymous_object3), __attribute__ ((unused)) signed int (*_adapterFi_14titerator_type14titerator_type_M_PP)(void (*__anonymous_object4)(), void *__anonymous_object5, void *__anonymous_object6), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type_P_M)(void (*__anonymous_object7)(), __attribute__ ((unused)) void *___retval__operator_preincr__14titerator_type_1, void *__anonymous_object8), __attribute__ ((unused)) void (*_adapterF_P9telt_type9telt_type__MP)(void (*__anonymous_object9)(), void *__anonymous_object10, void *__anonymous_object11), __attribute__ ((unused)) void (*_adapterF9telt_type_P9telt_type9telt_type_P_MP)(void (*__anonymous_object12)(), __attribute__ ((unused)) void *___retval__operator_assign__9telt_type_1, void *__anonymous_object13, void *__anonymous_object14), __attribute__ ((unused)) void (*_adapterF_P14titerator_type14titerator_type__MP)(void (*__anonymous_object15)(), void *__anonymous_object16, void *__anonymous_object17), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type14titerator_type_P_MP)(void (*__anonymous_object18)(), __attribute__ ((unused)) void *___retval__operator_assign__14titerator_type_1, void *__anonymous_object19, void *__anonymous_object20), __attribute__ ((unused)) unsigned long int _sizeof_14titerator_type, __attribute__ ((unused)) unsigned long int _alignof_14titerator_type, __attribute__ ((unused)) unsigned long int _sizeof_9telt_type, __attribute__ ((unused)) unsigned long int _alignof_9telt_type, __attribute__ ((unused)) void *(*___operator_assign__PF14titerator_type_R14titerator_type14titerator_type__1)(void *__anonymous_object21, void *__anonymous_object22), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type__1)(void *__anonymous_object23), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type14titerator_type__1)(void *__anonymous_object24, void *__anonymous_object25), __attribute__ ((unused)) void (*___destructor__PF_R14titerator_type__1)(void *__anonymous_object26), __attribute__ ((unused)) void *(*___operator_assign__PF9telt_type_R9telt_type9telt_type__1)(void *__anonymous_object27, void *__anonymous_object28), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type__1)(void *__anonymous_object29), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type9telt_type__1)(void *__anonymous_object30, void *__anonymous_object31), __attribute__ ((unused)) void (*___destructor__PF_R9telt_type__1)(void *__anonymous_object32), __attribute__ ((unused)) void *(*___operator_preincr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object33), __attribute__ ((unused)) void *(*___operator_predecr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object34), __attribute__ ((unused)) signed int (*___operator_equal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object35, void *__anonymous_object36), __attribute__ ((unused)) signed int (*___operator_notequal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object37, void *__anonymous_object38), __attribute__ ((unused)) void *(*___operator_deref__PFR9telt_type_14titerator_type__1)(void *__anonymous_object39), void *__begin__14titerator_type_1, void *__end__14titerator_type_1, void (*__func__PF_9telt_type__1)(void *__anonymous_object40)); 8 2 void __for_each_reverse__A0_2_0_0____operator_assign__PFd0_Rd0d0____constructor__PF_Rd0____constructor__PF_Rd0d0____destructor__PF_Rd0____operator_assign__PFd1_Rd1d1____constructor__PF_Rd1____constructor__PF_Rd1d1____destructor__PF_Rd1____operator_preincr__PFd0_Rd0____operator_predecr__PFd0_Rd0____operator_equal__PFi_d0d0____operator_notequal__PFi_d0d0____operator_deref__PFRd1_d0__F_d0d0PF_d1___1(__attribute__ ((unused)) void (*_adapterF_9telt_type__P)(void (*__anonymous_object41)(), void *__anonymous_object42), __attribute__ ((unused)) void *(*_adapterFP9telt_type_14titerator_type_M_P)(void (*__anonymous_object43)(), void *__anonymous_object44), __attribute__ ((unused)) signed int (*_adapterFi_14titerator_type14titerator_type_M_PP)(void (*__anonymous_object45)(), void *__anonymous_object46, void *__anonymous_object47), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type_P_M)(void (*__anonymous_object48)(), __attribute__ ((unused)) void *___retval__operator_preincr__14titerator_type_1, void *__anonymous_object49), __attribute__ ((unused)) void (*_adapterF_P9telt_type9telt_type__MP)(void (*__anonymous_object50)(), void *__anonymous_object51, void *__anonymous_object52), __attribute__ ((unused)) void (*_adapterF9telt_type_P9telt_type9telt_type_P_MP)(void (*__anonymous_object53)(), __attribute__ ((unused)) void *___retval__operator_assign__9telt_type_1, void *__anonymous_object54, void *__anonymous_object55), __attribute__ ((unused)) void (*_adapterF_P14titerator_type14titerator_type__MP)(void (*__anonymous_object56)(), void *__anonymous_object57, void *__anonymous_object58), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type14titerator_type_P_MP)(void (*__anonymous_object59)(), __attribute__ ((unused)) void *___retval__operator_assign__14titerator_type_1, void *__anonymous_object60, void *__anonymous_object61), __attribute__ ((unused)) unsigned long int _sizeof_14titerator_type, __attribute__ ((unused)) unsigned long int _alignof_14titerator_type, __attribute__ ((unused)) unsigned long int _sizeof_9telt_type, __attribute__ ((unused)) unsigned long int _alignof_9telt_type, __attribute__ ((unused)) void *(*___operator_assign__PF14titerator_type_R14titerator_type14titerator_type__1)(void *__anonymous_object62, void *__anonymous_object63), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type__1)(void *__anonymous_object64), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type14titerator_type__1)(void *__anonymous_object65, void *__anonymous_object66), __attribute__ ((unused)) void (*___destructor__PF_R14titerator_type__1)(void *__anonymous_object67), __attribute__ ((unused)) void *(*___operator_assign__PF9telt_type_R9telt_type9telt_type__1)(void *__anonymous_object68, void *__anonymous_object69), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type__1)(void *__anonymous_object70), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type9telt_type__1)(void *__anonymous_object71, void *__anonymous_object72), __attribute__ ((unused)) void (*___destructor__PF_R9telt_type__1)(void *__anonymous_object73), __attribute__ ((unused)) void *(*___operator_preincr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object74), __attribute__ ((unused)) void *(*___operator_predecr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object75), __attribute__ ((unused)) signed int (*___operator_equal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object76, void *__anonymous_object77), __attribute__ ((unused)) signed int (*___operator_notequal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object78, void *__anonymous_object79), __attribute__ ((unused)) void *(*___operator_deref__PFR9telt_type_14titerator_type__1)(void *__anonymous_object80), void *__begin__14titerator_type_1, void *__end__14titerator_type_1, void (*__func__PF_9telt_type__1)(void *__anonymous_object81)); … … 1377 1371 } 1378 1372 static inline int invoke_main(int argc, char* argv[], char* envp[]) { (void)argc; (void)argv; (void)envp; return __main__Fi___1(); } 1379 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned int __size);1380 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);1381 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);1382 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));1383 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);1384 extern signed int printf(const char *__restrict __format, ...);1385 1373 static inline signed int invoke_main(signed int argc, char **argv, char **envp); 1386 1374 signed int main(signed int __argc__i_1, char **__argv__PPc_1, char **__envp__PPc_1){ -
src/tests/.expect/64/KRfunctions.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 signed int __f0__Fi_iPCii__1(signed int __a__i_1, const signed int *__b__PCi_1, signed int __c__i_1){ 8 2 __attribute__ ((unused)) signed int ___retval_f0__i_1; -
src/tests/.expect/64/attributes.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 signed int __la__Fi___1(){ 8 2 __attribute__ ((unused)) signed int ___retval_la__i_1; -
src/tests/.expect/64/declarationSpecifier.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 volatile const signed short int __x1__CVs_1; 8 2 static volatile const signed short int __x2__CVs_1; … … 701 695 } 702 696 static inline int invoke_main(int argc, char* argv[], char* envp[]) { (void)argc; (void)argv; (void)envp; return __main__Fi_iPPCc__1(argc, argv); } 703 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);704 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);705 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);706 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));707 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);708 extern signed int printf(const char *__restrict __format, ...);709 697 static inline signed int invoke_main(signed int argc, char **argv, char **envp); 710 698 signed int main(signed int __argc__i_1, char **__argv__PPc_1, char **__envp__PPc_1){ -
src/tests/.expect/64/extension.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 __extension__ signed int __a__i_1; 8 2 __extension__ signed int __b__i_1; -
src/tests/.expect/64/gccExtensions.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 extern signed int __x__i_1 asm ( "xx" ); 8 2 signed int __main__Fi_iPPCc__1(signed int __argc__i_1, const char **__argv__PPCc_1){ … … 174 168 } 175 169 static inline int invoke_main(int argc, char* argv[], char* envp[]) { (void)argc; (void)argv; (void)envp; return __main__Fi_iPPCc__1(argc, argv); } 176 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);177 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);178 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);179 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));180 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);181 extern signed int printf(const char *__restrict __format, ...);182 170 static inline signed int invoke_main(signed int argc, char **argv, char **envp); 183 171 signed int main(signed int __argc__i_1, char **__argv__PPc_1, char **__envp__PPc_1){ -
src/tests/.expect/64/literals.txt
r65d6de4 r9e1eabc 1 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);2 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);3 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);4 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));5 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);6 extern signed int printf(const char *__restrict __format, ...);7 1 void __for_each__A0_2_0_0____operator_assign__PFd0_Rd0d0____constructor__PF_Rd0____constructor__PF_Rd0d0____destructor__PF_Rd0____operator_assign__PFd1_Rd1d1____constructor__PF_Rd1____constructor__PF_Rd1d1____destructor__PF_Rd1____operator_preincr__PFd0_Rd0____operator_predecr__PFd0_Rd0____operator_equal__PFi_d0d0____operator_notequal__PFi_d0d0____operator_deref__PFRd1_d0__F_d0d0PF_d1___1(__attribute__ ((unused)) void (*_adapterF_9telt_type__P)(void (*__anonymous_object0)(), void *__anonymous_object1), __attribute__ ((unused)) void *(*_adapterFP9telt_type_14titerator_type_M_P)(void (*__anonymous_object2)(), void *__anonymous_object3), __attribute__ ((unused)) signed int (*_adapterFi_14titerator_type14titerator_type_M_PP)(void (*__anonymous_object4)(), void *__anonymous_object5, void *__anonymous_object6), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type_P_M)(void (*__anonymous_object7)(), __attribute__ ((unused)) void *___retval__operator_preincr__14titerator_type_1, void *__anonymous_object8), __attribute__ ((unused)) void (*_adapterF_P9telt_type9telt_type__MP)(void (*__anonymous_object9)(), void *__anonymous_object10, void *__anonymous_object11), __attribute__ ((unused)) void (*_adapterF9telt_type_P9telt_type9telt_type_P_MP)(void (*__anonymous_object12)(), __attribute__ ((unused)) void *___retval__operator_assign__9telt_type_1, void *__anonymous_object13, void *__anonymous_object14), __attribute__ ((unused)) void (*_adapterF_P14titerator_type14titerator_type__MP)(void (*__anonymous_object15)(), void *__anonymous_object16, void *__anonymous_object17), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type14titerator_type_P_MP)(void (*__anonymous_object18)(), __attribute__ ((unused)) void *___retval__operator_assign__14titerator_type_1, void *__anonymous_object19, void *__anonymous_object20), __attribute__ ((unused)) unsigned long int _sizeof_14titerator_type, __attribute__ ((unused)) unsigned long int _alignof_14titerator_type, __attribute__ ((unused)) unsigned long int _sizeof_9telt_type, __attribute__ ((unused)) unsigned long int _alignof_9telt_type, __attribute__ ((unused)) void *(*___operator_assign__PF14titerator_type_R14titerator_type14titerator_type__1)(void *__anonymous_object21, void *__anonymous_object22), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type__1)(void *__anonymous_object23), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type14titerator_type__1)(void *__anonymous_object24, void *__anonymous_object25), __attribute__ ((unused)) void (*___destructor__PF_R14titerator_type__1)(void *__anonymous_object26), __attribute__ ((unused)) void *(*___operator_assign__PF9telt_type_R9telt_type9telt_type__1)(void *__anonymous_object27, void *__anonymous_object28), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type__1)(void *__anonymous_object29), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type9telt_type__1)(void *__anonymous_object30, void *__anonymous_object31), __attribute__ ((unused)) void (*___destructor__PF_R9telt_type__1)(void *__anonymous_object32), __attribute__ ((unused)) void *(*___operator_preincr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object33), __attribute__ ((unused)) void *(*___operator_predecr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object34), __attribute__ ((unused)) signed int (*___operator_equal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object35, void *__anonymous_object36), __attribute__ ((unused)) signed int (*___operator_notequal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object37, void *__anonymous_object38), __attribute__ ((unused)) void *(*___operator_deref__PFR9telt_type_14titerator_type__1)(void *__anonymous_object39), void *__begin__14titerator_type_1, void *__end__14titerator_type_1, void (*__func__PF_9telt_type__1)(void *__anonymous_object40)); 8 2 void __for_each_reverse__A0_2_0_0____operator_assign__PFd0_Rd0d0____constructor__PF_Rd0____constructor__PF_Rd0d0____destructor__PF_Rd0____operator_assign__PFd1_Rd1d1____constructor__PF_Rd1____constructor__PF_Rd1d1____destructor__PF_Rd1____operator_preincr__PFd0_Rd0____operator_predecr__PFd0_Rd0____operator_equal__PFi_d0d0____operator_notequal__PFi_d0d0____operator_deref__PFRd1_d0__F_d0d0PF_d1___1(__attribute__ ((unused)) void (*_adapterF_9telt_type__P)(void (*__anonymous_object41)(), void *__anonymous_object42), __attribute__ ((unused)) void *(*_adapterFP9telt_type_14titerator_type_M_P)(void (*__anonymous_object43)(), void *__anonymous_object44), __attribute__ ((unused)) signed int (*_adapterFi_14titerator_type14titerator_type_M_PP)(void (*__anonymous_object45)(), void *__anonymous_object46, void *__anonymous_object47), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type_P_M)(void (*__anonymous_object48)(), __attribute__ ((unused)) void *___retval__operator_preincr__14titerator_type_1, void *__anonymous_object49), __attribute__ ((unused)) void (*_adapterF_P9telt_type9telt_type__MP)(void (*__anonymous_object50)(), void *__anonymous_object51, void *__anonymous_object52), __attribute__ ((unused)) void (*_adapterF9telt_type_P9telt_type9telt_type_P_MP)(void (*__anonymous_object53)(), __attribute__ ((unused)) void *___retval__operator_assign__9telt_type_1, void *__anonymous_object54, void *__anonymous_object55), __attribute__ ((unused)) void (*_adapterF_P14titerator_type14titerator_type__MP)(void (*__anonymous_object56)(), void *__anonymous_object57, void *__anonymous_object58), __attribute__ ((unused)) void (*_adapterF14titerator_type_P14titerator_type14titerator_type_P_MP)(void (*__anonymous_object59)(), __attribute__ ((unused)) void *___retval__operator_assign__14titerator_type_1, void *__anonymous_object60, void *__anonymous_object61), __attribute__ ((unused)) unsigned long int _sizeof_14titerator_type, __attribute__ ((unused)) unsigned long int _alignof_14titerator_type, __attribute__ ((unused)) unsigned long int _sizeof_9telt_type, __attribute__ ((unused)) unsigned long int _alignof_9telt_type, __attribute__ ((unused)) void *(*___operator_assign__PF14titerator_type_R14titerator_type14titerator_type__1)(void *__anonymous_object62, void *__anonymous_object63), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type__1)(void *__anonymous_object64), __attribute__ ((unused)) void (*___constructor__PF_R14titerator_type14titerator_type__1)(void *__anonymous_object65, void *__anonymous_object66), __attribute__ ((unused)) void (*___destructor__PF_R14titerator_type__1)(void *__anonymous_object67), __attribute__ ((unused)) void *(*___operator_assign__PF9telt_type_R9telt_type9telt_type__1)(void *__anonymous_object68, void *__anonymous_object69), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type__1)(void *__anonymous_object70), __attribute__ ((unused)) void (*___constructor__PF_R9telt_type9telt_type__1)(void *__anonymous_object71, void *__anonymous_object72), __attribute__ ((unused)) void (*___destructor__PF_R9telt_type__1)(void *__anonymous_object73), __attribute__ ((unused)) void *(*___operator_preincr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object74), __attribute__ ((unused)) void *(*___operator_predecr__PF14titerator_type_R14titerator_type__1)(void *__anonymous_object75), __attribute__ ((unused)) signed int (*___operator_equal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object76, void *__anonymous_object77), __attribute__ ((unused)) signed int (*___operator_notequal__PFi_14titerator_type14titerator_type__1)(void *__anonymous_object78, void *__anonymous_object79), __attribute__ ((unused)) void *(*___operator_deref__PFR9telt_type_14titerator_type__1)(void *__anonymous_object80), void *__begin__14titerator_type_1, void *__end__14titerator_type_1, void (*__func__PF_9telt_type__1)(void *__anonymous_object81)); … … 1377 1371 } 1378 1372 static inline int invoke_main(int argc, char* argv[], char* envp[]) { (void)argc; (void)argv; (void)envp; return __main__Fi___1(); } 1379 __attribute__ ((__nothrow__,__leaf__,__malloc__)) extern void *malloc(unsigned long int __size);1380 __attribute__ ((__nothrow__,__leaf__)) extern void free(void *__ptr);1381 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void abort(void);1382 __attribute__ ((__nothrow__,__leaf__,__nonnull__(1))) extern signed int atexit(void (*__func)(void));1383 __attribute__ ((__nothrow__,__leaf__,__noreturn__)) extern void exit(signed int __status);1384 extern signed int printf(const char *__restrict __format, ...);1385 1373 static inline signed int invoke_main(signed int argc, char **argv, char **envp); 1386 1374 signed int main(signed int __argc__i_1, char **__argv__PPc_1, char **__envp__PPc_1){
Note: See TracChangeset
for help on using the changeset viewer.