aboutsummaryrefslogtreecommitdiffstats
path: root/mjit.c
Commit message (Collapse)AuthorAgeFilesLines
* MJIT: Mark MJIT's objects from mjit_markTakashi Kokubun2022-12-081-2/+5
| | | | | It looks like the current way of marking objects was breaking eightbitraptor's upcoming VWA changes and this seems to fix it.
* MJIT: Deal with a TODO with assertionTakashi Kokubun2022-12-071-1/+2
|
* MJIT: Fix an assertion broken with --mjit-wait enabledTakashi Kokubun2022-12-061-1/+1
|
* MJIT: Refactor mjit_waitTakashi Kokubun2022-12-061-39/+12
|
* MJIT: Remove obsoleted MJIT countersTakashi Kokubun2022-12-061-32/+0
|
* MJIT: Use xfree for unitsTakashi Kokubun2022-12-061-3/+3
| | | | now that we use ZALLOC_N for allocating units
* MJIT: Explain why we have MJIT_CFLAGS_PIPE [ci skip]Takashi Kokubun2022-12-061-2/+2
|
* MJIT: Just use ZALLOC_N in create_unitTakashi Kokubun2022-12-061-9/+1
| | | | We no longer use an MJIT worker thread, so there's no need to avoid GC.
* MJIT: Remove an unused argument and unused countersTakashi Kokubun2022-12-061-7/+4
| | | | I plan to rebuild MJIT metrics later, not using debug counters.
* MJIT: Refactor the jit_func enum for MJITTakashi Kokubun2022-12-061-14/+14
| | | | | | All values should have a MJIT_ prefix. We could address the warning for the end mark if we just define the macro for the check next to the enum. It even simplifies some code for checking the enum.
* MJIT: Merge mjit_unit.h into mjit_c.hTakashi Kokubun2022-11-291-1/+1
| | | | The distinction doesn't make much difference today.
* MJIT: Rename mjit_compiler.c to mjit_c.cTakashi Kokubun2022-11-291-2/+20
| | | | It's no longer about the compiler logic itself.
* MJIT: Cache an ISeq pointer instanceTakashi Kokubun2022-11-281-1/+5
| | | | to obviate one rb_funcall. This also removes rb_ptr as refactoring.
* MJIT: Make it parsable by Solargraph (#6827)Takashi Kokubun2022-11-281-2/+3
| | | | | | | * Revert "Revert "MJIT: Make it parsable by Solargraph"" This reverts commit 8e18761da1932df88bfb6505acbda4740e1b2930. * Call rb_gc_register_mark_object
* Revert "MJIT: Make it parsable by Solargraph"Takashi Kokubun2022-11-281-2/+2
| | | | | | | | | | | This reverts commit ccd8dd6ad395bbd9f5290e0fcb7929e5e5d36767. Revert "MJIT: Fix miniruby with MJIT_FORCE_ENABLE" This reverts commit b033775ed9d9226ba73c1d4a197e55ba89575142. GitHub Actions is failing. I ran out of time today to investigate it. will try it again tomorrow.
* MJIT: Make it parsable by SolargraphTakashi Kokubun2022-11-271-2/+2
|
* MJIT: Avoid hanging on mjit_wait with a JIT failureTakashi Kokubun2022-11-271-1/+6
|
* Synchronously run the MJIT compiler in the parentTakashi Kokubun2022-11-271-43/+50
| | | | | | | | | | | process. Completely isolating the MJIT compilation process complicates a lot of things for ensuring consistency before and after the fork. Just running this synchronously makes things a lot easier, for example the race condition of capture_cc_entries could be fixed by this patch alone. Hopefully, the bottleneck is the C compiler and not this Ruby code. Also, this change doesn't negatively impact MJIT's final numbers on yjit-bench while "1st itr" is degraded for sure.
* MJIT: Change default --mjit-max-cache back to 100Takashi Kokubun2022-11-251-1/+1
| | | | | | | | | | | | | | | | | These days we benchmark MJIT using yjit-bench. The warmup duration in yjit-bench is very short, so compiling many methods comes at a cost even while it's actually optimal for MJIT to compile everything / tens of thousands of methods once it reaches the peak performance. yjit-bench doesn't necessarily represent the peak performance on production. It measures the performance of Ruby 30~60s after boot. If your JIT takes more than 1 minute to warm up, there's no way for the JIT to make the numbers good on yjit-bench. Until we make MJIT's compilation much faster, we don't afford compiling 10,000 methods on yjit-bench. This change alone makes MJIT's benchmark number on railsbench 2x better :p
* Use `rb_sprintf` instead of deprecated `sprintf`Nobuyoshi Nakada2022-11-251-7/+7
|
* Rename --mjit-min-calls to --mjit-call-threshold (#6731)Takashi Kokubun2022-11-141-13/+13
| | | for consistency with YJIT
* s/mjit_func_t/jit_func_t/Takashi Kokubun2022-11-131-9/+9
|
* Allow passing a Rust closure to rb_iseq_callback (#6575)Takashi Kokubun2022-10-181-2/+2
|
* Make mjit_cont sharable with YJIT (#6556)Takashi Kokubun2022-10-171-99/+4
| | | | | | | * Make mjit_cont sharable with YJIT * Update dependencies * Update YJIT binding
* Rework vm_core to use `int first_lineno` struct member.Samuel Williams2022-09-261-16/+10
|
* Fix typoTakashi Kokubun2022-09-071-1/+1
| | | | I meant the other one. Otherwise this option doesn't make sense.
* Adjust pch_status for --mjit=pauseTakashi Kokubun2022-09-061-1/+2
| | | | to let mjit_add_iseq_to_process work
* Skip early mjit_add_iseq_to_process on --mjit=pauseTakashi Kokubun2022-09-061-1/+1
|
* Skip calling check_unit_queue with mjit_opts.customTakashi Kokubun2022-09-061-0/+1
| | | | fixing the behavior of b726c06e7eeafff52e368179dbe79a11b1aff975
* Allow hooking a different method after --mjit=pauseTakashi Kokubun2022-09-061-9/+36
| | | | | The interface is similar to RubyVM::ISeq.translate; it's used if defined. Same as --mjit=pause, this is an undocumented feature for MJIT experiments.
* Cache RubyVM::MJIT constantsTakashi Kokubun2022-09-061-0/+7
| | | | for performance
* Add an option to lazily boot MJIT for experimentsTakashi Kokubun2022-09-061-5/+18
| | | | | | You may use `RUBYOPT=--mjit=pause irb` to play with RubyVM::MJIT::C, control the boot timing of MJIT, or customize the implementation while paused. It's an undocumented feature for such experiments.
* Do not fork the process on --mjit-waitTakashi Kokubun2022-09-061-20/+22
| | | | | fork is for parallel compilation, but --mjit-wait cancels it. It's more useful to not fork it for binding.irb, debugging, etc.
* Deal with MJIT_FORCE_ENABLE minirubyTakashi Kokubun2022-09-051-0/+9
| | | | http://ci.rvm.jp/results/trunk-mjit@phosphorus-docker/4225469
* Ruby MJIT (#6028)Takashi Kokubun2022-09-041-6/+6
|
* Call appropriate hooks on MJIT's forkTakashi Kokubun2022-09-041-18/+4
| | | | This takes care of signal_self_pipe and other things.
* Prefer stdbool for MJIT optionsTakashi Kokubun2022-09-031-4/+4
| | | | same motivation as d6f21b308bcff03e82f8b3dbf11a852ce111b3b3
* Use `rb_fork` to suppress deprecated-declarations warningsNobuyoshi Nakada2022-08-211-2/+3
|
* Drop mswin support of MJIT (#6265)Takashi Kokubun2022-08-201-326/+8
| | | | | | | The current MJIT relies on SIGCHLD and fork(2) to be performant, and it's something mswin can't offer. You could run Linux MJIT on WSL instead. [Misc #18968]
* Execute MJIT in a forked Ruby process (#6264)Takashi Kokubun2022-08-201-23/+62
| | | [Misc #18968]
* Add `-bundle_loader` to mjit compilation args on macOSYuta Saito2022-08-041-2/+10
|
* Adjust styles [ci skip]Nobuyoshi Nakada2022-07-271-1/+2
|
* MJIT: Share rb_mjit_unit through mjit_unit.hTakashi Kokubun2022-07-141-30/+1
| | | | mjit_compile.c should be able to access this more easily.
* MJIT: Merge mjit_worker.c back to mjit.c (#6138)Takashi Kokubun2022-07-141-5/+1264
| | | | | Since #6006, we no longer avoid executing GC on mjit_worker.c and thus there's no need to carefully change how we write code whether you're in mjit.c or mjit_worker.c anymore.
* MJIT: s/assert/VM_ASSERT/Takashi Kokubun2022-07-111-2/+2
| | | | | | include/ruby/assert.h sets NDEBUG by default. As MJIT CI uses -DVM_CHECK_MODE, using only VM_ASSERT might be more helpful.
* MJIT: Consider compaction on CC failureTakashi Kokubun2022-06-161-1/+3
|
* MJIT: Handle compaction units on stop_workerTakashi Kokubun2022-06-151-6/+13
|
* MJIT: Get rid of obsoleted compiling_iseqsTakashi Kokubun2022-06-151-33/+3
|
* MJIT: Deal with mjit_free_iseq on mjit_notify_waitpidTakashi Kokubun2022-06-151-7/+9
|
* MJIT: Fix test_jit_failure for mswinTakashi Kokubun2022-06-151-1/+1
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-22 07:23:52 +0000