summaryrefslogtreecommitdiff
path: root/mjit_worker.c
AgeCommit message (Collapse)Author
2020-03-14Use a human-readable funcname with --jit-debugTakashi Kokubun
for perf output like: Samples: 100K of event 'cycles:ppp', Event count (approx.): 1007750000 Children Self Command Shared Object Symbol + 81.58% 1.47% ruby ruby [.] rb_vm_exec + 81.06% 7.61% ruby ruby [.] vm_exec_core + 80.16% 0.00% ruby ruby [.] vm_sendish (inlined) + 75.03% 0.00% ruby ruby [.] mjit_exec (inlined) + 74.37% 0.00% ruby ruby [.] mjit_exec (inlined) + 73.42% 0.22% ruby _ruby_mjit_p11277u42.so [.] _mjit42_rack_method_override_rb_call + 73.25% 0.10% ruby _ruby_mjit_p11277u41.so [.] _mjit41_sinatra_show_exceptions_rb_call + 73.19% 0.22% ruby _ruby_mjit_p11277u44.so [.] _mjit44_rack_head_rb_call + 73.03% 0.15% ruby _ruby_mjit_p11277u45.so [.] _mjit45_sinatra_base_rb_call + 72.87% 0.26% ruby _ruby_mjit_p11277u49.so [.] _mjit49_rack_logger_rb_call + 70.56% 0.11% ruby _ruby_mjit_p11277u40.so [.] _mjit40_sinatra_base_rb_call + 68.70% 0.11% ruby _ruby_mjit_p11277u39.so [.] _mjit39_sinatra_base_rb_call + 68.39% 0.29% ruby _ruby_mjit_p11277u56.so [.] _mjit56_rack_protection_frame_options_rb_call + 67.89% 0.18% ruby _ruby_mjit_p11277u37.so [.] _mjit37_sinatra_base_rb_block_in_call + 67.04% 0.16% ruby _ruby_mjit_p11277u34.so [.] _mjit34_sinatra_base_rb_synchronize
2020-03-12Avoid referring to an old value of reallocTakashi Kokubun
OpenBSD RubyCI has failed with SEGV since 4bcd5981e80d3e1852c8723741a0069779464128. https://rubyci.org/logs/rubyci.s3.amazonaws.com/openbsd-current/ruby-master/log/20200312T223005Z.fail.html.gz This was because `status->cc_entries` could be stale after `realloc` call for inlined iseqs.
2020-03-10Capture inlined iseq's cc entries in root iseq'sTakashi Kokubun
jit_unit to avoid marking wrong cc entries when inlined iseq is compiled multiple times, resolving the TODO added by daf7c48d88. This obviates pseudo jit_unit in inlined iseq introduced by 7ec2359374 and fixes memory leak of the adhoc unit.
2020-03-04[win32] suppress false warning by MSVCNobuyoshi Nakada
2020-02-26Add missing free on cc_entriesTakashi Kokubun
Looks like an oversight in b9007b6c548f91e88fd3f2ffa23de740431fa969 and 7ec23593746c8ccabd6c005cc34dde77d564c6c9.
2020-02-26Internalize rb_mjit_unit definition againTakashi Kokubun
Fixed a TODO in b9007b6c548f91e88fd3f2ffa23de740431fa969
2020-02-25Explain the situation of inner cc_entries [ci skip]Takashi Kokubun
Add a little more details in 7ec23593746c8ccabd6c005cc34dde77d564c6c9
2020-02-25Note a situation around xmalloc vs free in MJIT [ci skip]Takashi Kokubun
shared by ko1
2020-02-25should be initialize jit_unit->cc_entries.Koichi Sasada
GC can invoke just after allocation of jit_unit->cc_entries so it should be zero-cleared.
2020-02-25prevent GC from mjit worker.Koichi Sasada
ALLOC_N() can causes GC. Sometimes `mjit_copy_job_handler()` can be called by mjit_worker thread which is not a Ruby thread, so we need to prevent GC in this function. This patch has some issues, but I introduce it to pass the tests.
2020-02-22Introduce disposable call-cache.Koichi Sasada
This patch contains several ideas: (1) Disposable inline method cache (IMC) for race-free inline method cache * Making call-cache (CC) as a RVALUE (GC target object) and allocate new CC on cache miss. * This technique allows race-free access from parallel processing elements like RCU. (2) Introduce per-Class method cache (pCMC) * Instead of fixed-size global method cache (GMC), pCMC allows flexible cache size. * Caching CCs reduces CC allocation and allow sharing CC's fast-path between same call-info (CI) call-sites. (3) Invalidate an inline method cache by invalidating corresponding method entries (MEs) * Instead of using class serials, we set "invalidated" flag for method entry itself to represent cache invalidation. * Compare with using class serials, the impact of method modification (add/overwrite/delete) is small. * Updating class serials invalidate all method caches of the class and sub-classes. * Proposed approach only invalidate the method cache of only one ME. See [Feature #16614] for more details. Notes: Merged: https://github.com/ruby/ruby/pull/2888
2019-12-01Allow specifying arbitrary MJIT flags by --jit-debugTakashi Kokubun
This is a secret feature for me. It's only for testing and any behavior with this flag override is unsupported. I needed this because I sometimes want to add debug options but do not want to disable optimizations, for using Linux perf.
2019-10-13Eliminate the possibility to leave freed ISeqTakashi Kokubun
in active_units Hoping to fix: http://ci.rvm.jp/results/trunk-mjit@silicon-docker/2311375
2019-10-13Delay the free until we stop referring to a unitTakashi Kokubun
`if (unit->iseq)` might have referred to a freed unit. Therefore this commit delays its free.
2019-10-13Remove the quick stop path after convert_unit_to_funcTakashi Kokubun
Now I'm not exactly sure why I needed to check `stop_worker_p` after `mjit_copy_cache_from_main_thread` of `convert_unit_to_func` in 4161674b2fbea6bdd01783ac5d3b39d88db22972. If it's for avoiding deadlock under `in_gc` condition, we should keep it. However, if it's not the case and it's just for retrying accidental compilation failure or just to avoid `MJIT_ATOMIC_SET` and `compact_all_jit_code`, I think this quick stop path is not mandatory. Because this path is somewhat problematic in my upcoming fix in mjit_worker, let me try to remove this first and see how CI goes.
2019-10-10mjit_worker.c: Add `-lm` to the C compiler in MJIT on AndroidYusuke Endoh
To avoid: cannot locate symbol "modf" referenced by .../_ruby_mjit_XXX.so"
2019-09-26RubyVM::MJIT.pause(wait: true) should waitTakashi Kokubun
for all compilations and compaction. Prior to this commit, the last-compiled code has not been used because MJIT worker is stopped before setting the code, and compaction has also been skipped. But it was not intentional and `wait: true` pause should wait until those two things by its feature.
2019-08-30Split CC_LINKER_ARGS from CC_DLDFLAGS_ARGSTakashi Kokubun
to drop MJIT_DLDFLAGS from compile_c_to_o for OpenBSD failure https://rubyci.org/logs/rubyci.s3.amazonaws.com/openbsd65/ruby-master/log/20190830T110008Z.fail.html.gz. 8c7f4e8f8b7f9e74195ea0eb51f39014fec342c0 did not work for i686-linux https://travis-ci.org/ruby/ruby/jobs/578808112.
2019-08-30Try dropping DLDFLAGS from compile_c_to_oTakashi Kokubun
I think this did not work for some environments, but this seems to fix OpenBSD RubyCI failure: https://rubyci.org/logs/rubyci.s3.amazonaws.com/openbsd65/ruby-master/log/20190830T110008Z.fail.html.gz Let me check RubyCI impact by this.
2019-08-28Stop passing CC_LIBS in compile_c_to_oTakashi Kokubun
To fix "gcc: -lgcc: linker input file unused because linking not done" in https://rubyci.org/logs/rubyci.s3.amazonaws.com/openbsd65/ruby-master/log/20190826T200009Z.log.html.gz Still I'm intentionally leaving CC_DLDFLAGS_ARGS because making compiler options different might result in compile/link failure. (Of course CC_DLDFLAGS_ARGS usually should not have compiler options, but it might include by bad configure setup) The same problem may exist in MJIT_LIBS used inside CC_LIBS as well, but I just ignored that case because it impacts only MinGW / cygwin, hoping that their users do not perform a wrong configure. Notes: s/might include/might be included/
2019-07-15Handle failure of opening a null deviceTakashi Kokubun
This issue is detected by Coverity Scan.
2019-07-15Simplify start_process by exploiting C99Takashi Kokubun
Having a block for mixing a declaration was confusing. Also I moved `dev_null` and `pid` to limit their scope.
2019-07-15Always call va_end in form_args()Takashi Kokubun
This issue is detected by Coverity Scan.
2019-07-14Simplify link_o_to_so argumentsTakashi Kokubun
by C99 compound literal for array and non-constant array initializer
2019-07-14Fix a wrong buffer size to avoid stack corruptionTakashi Kokubun
[Bug #15986]
2019-07-14Refactor mjit_worker command line with C99Takashi Kokubun
Mostly non-constatnt array initializer, and mixed declarations and code.
2019-05-30Do not use rb_iseq_path() while moving ISeq pointersTakashi Kokubun
in GC.compact. While `in_jit` is false, GC.compact is allowed to run and it may be moving ISeq-related pointers. So calling rb_iseq_path() when `in_jit` is true is illegal.
2019-05-29Never make a method call from MJIT workerTakashi Kokubun
by showing line number only when it's Fixnum. When it's not Fixnum, we need to call a method to know the line number.
2019-05-29Do not call FIX2INT while GC.compact may be runningTakashi Kokubun
because FIX2INT might crash by moving method entry pointer: http://ci.rvm.jp/results/trunk-mjit@silicon-docker/2063412
2019-05-15Rename mjit_gc_finish_hook to mjit_gc_exit_hookTakashi Kokubun
because @ko1 said "gc_finish" is confusing like a finish of entire GC process
2019-04-20fix strict compile errorstenderlove
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67625 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-14Avoid crash on logging GC-ed ISeqk0kubun
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67552 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-14Do not execute MJIT copy job when ISeq is GC-edk0kubun
I assumed that ISeq is never GC-ed by `in_jit` + `mjit_mark` on copy job ISeq, but unfortunately I found SEGV on `mjit_copy_job_handler` in which iseq->body was somehow Qnil. And it seems to be fixed by disabling the job when `mjit_free_iseq` is called for the ISeq. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67551 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-14Add debug counter for MJIT stale_unitsk0kubun
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67546 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-14Unify comment styles across MJIT sourcesk0kubun
I'm writing `//` comments in newer MJIT code after C99 enablement (because I write 1-line comments more often than multi-line comments and `//` requires fewer chars on 1-line) and then they are mixed with `/* */` now. For consistency and to avoid the conversion in future changes, let me finish the rewrite in MJIT-related code. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67533 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-14Recompile JIT-ed code without optimizationk0kubun
based on inline cache when JIT cancel happens by that. This feature was in the original MJIT implementation by Vladimir, but on merging MJIT to Ruby it was removed for simplification. This commit adds the functionality again for the following benchmark: https://github.com/benchmark-driver/misc/blob/52f05781f65467baf895bf6ba79d172c9b0826fd/concurrent-map/bench.rb (shown float is duration seconds. shorter is better) * Before ``` $ INHERIT=0 ruby -v bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) [x86_64-linux] -- 1.6507579649914987 $ INHERIT=0 ruby -v --jit bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) +JIT [x86_64-linux] -- 1.5091587850474752 $ INHERIT=1 ruby -v bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) [x86_64-linux] -- 1.6124781150138006 $ INHERIT=1 ruby --jit -v bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) +JIT [x86_64-linux] -- 1.7495657080435194 # <-- this ``` * After ``` $ INHERIT=0 ruby -v bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) [x86_64-linux] last_commit=Recompile JIT-ed code without optimization -- 1.653559010999743 $ INHERIT=0 ruby --jit -v bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) +JIT [x86_64-linux] last_commit=Recompile JIT-ed code without optimization -- 1.4738391840364784 $ INHERIT=1 ruby -v bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) [x86_64-linux] last_commit=Recompile JIT-ed code without optimization -- 1.645227018976584 $ INHERIT=1 ruby --jit -v bench.rb ruby 2.7.0dev (2019-04-13 trunk 67523) +JIT [x86_64-linux] last_commit=Recompile JIT-ed code without optimization -- 1.523708809982054 # <-- this ``` git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67530 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-10Adjusted stylesnobu
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67504 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-10Reverting all commits from r67479 to r67496 because of CI failureskazu
Because hard to specify commits related to r67479 only. So please commit again. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67499 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-04-10Adjusted stylesnobu
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67493 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-29mjit_worker.c: explicitly ignore the unused result of RB_DEBUG_COUNTER_INC_IFmame
By adding `(void)`. This makes Coverity Scan happy. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67383 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-29Add mjit_compile_failures debug counterk0kubun
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67382 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-29Prevent mjit length counters from being 0k0kubun
It seems that `debug_counter_show_results_at_exit` could be called earlier than `mjit_finish` sometimes. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67380 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-18Use alloca again instead of malloc and freek0kubun
by changing interface of `mjit_copy_cache_from_main_thread`. This is also fixing deadlock introduced by r67299. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67300 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-18Resurrect r67287 and r67288k0kubun
I noticed that r67287 was illegal because memory allocated by `alloca` was used after the stack is expired. So I just replaced that with `malloc` and `free` for now. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67299 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-18Revert "Eliminate mjit_copy_job_t reference from mjit_worker"k0kubun
This reverts commit ba51ae0109ee1d1fa7ca90b43da115ea68d7214a. CI is failing again... git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67297 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-18Eliminate mjit_copy_job_t reference from mjit_workerk0kubun
Take 2 of r67287. For some reasons, passing pointer of pointer on stack to a function and assigning an addresse to a pointer dereferenced from the pointer seems not working on mswin. So I achieved to return multiple values by returning struct instead of taking pointers. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67296 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-18Revert "Request inline cache values from mjit_compile"k0kubun
This reverts commit 4161674b2fbea6bdd01783ac5d3b39d88db22972. Revert "Eliminate mjit_copy_job_t reference from mjit_worker" This reverts commit d86a1aa045959dfbf5cd472eae0d043180259727. Reverting them because of CI failures git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67291 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-17Request inline cache values from mjit_compilek0kubun
rather than preparing beforehand. By having this change, implementing inlining by calling `mjit_copy_cache_from_main_thread` for inlined methods was made possible. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67288 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-17Eliminate mjit_copy_job_t reference from mjit_workerk0kubun
and make `copy_cache_from_main_thread` easier to use. For implementing inlining later, I'll use `copy_cache_from_main_thread` inside `mjit_compile`. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67287 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2019-03-17Drop rb_mjit_unit from mjit_copy_jobk0kubun
and guard iseq from GC by marking iseq in mjit_copy_job. This is a refactoring for implementing inlining later and should not be fixing or introducing any bugs. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@67286 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-13 18:20:43 +0000