aboutsummaryrefslogtreecommitdiffstats
path: root/rjit.c
Commit message (Collapse)AuthorAgeFilesLines
* Revisions for #10198Takashi Kokubun2024-03-121-6/+6
| | | | This fixes some inconsistencies introduced by that PR.
* Do not include a backtick in error messages and backtracesYusuke Endoh2024-02-151-1/+1
| | | | [Feature #16495]
* RJIT: Rewind $! after compilationTakashi Kokubun2023-12-221-1/+5
|
* RJIT: Make --rjit-trace actually workTakashi Kokubun2023-12-211-9/+25
|
* RJIT: Add --rjit-trace to allow TracePoint during JITTakashi Kokubun2023-12-211-9/+14
|
* RJIT: Avoid incrementing vm_insns_countTakashi Kokubun2023-12-211-0/+4
| | | | during compilation
* RJIT: Rename pause/resume to disable/enableTakashi Kokubun2023-12-211-4/+3
| | | | | | like YJIT. They don't work in the same way yet, but it's nice to make the naming consistent first so that we will not need to rename them later.
* RJIT: Share rb_vm_insns_count for vm_insns_countTakashi Kokubun2023-12-181-13/+2
|
* add `flags` to `rb_postponed_job_preregister`Koichi Sasada2023-12-101-1/+1
| | | | for future extensions.
* Change the semantics of rb_postponed_job_registerKJ Tsanaktsidis2023-12-101-1/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Our current implementation of rb_postponed_job_register suffers from some safety issues that can lead to interpreter crashes (see bug #1991). Essentially, the issue is that jobs can be called with the wrong arguments. We made two attempts to fix this whilst keeping the promised semantics, but: * The first one involved masking/unmasking when flushing jobs, which was believed to be too expensive * The second one involved a lock-free, multi-producer, single-consumer ringbuffer, which was too complex The critical insight behind this third solution is that essentially the only user of these APIs are a) internal, or b) profiling gems. For a), none of the usages actually require variable data; they will work just fine with the preregistration interface. For b), generally profiling gems only call a single callback with a single piece of data (which is actually usually just zero) for the life of the program. The ringbuffer is complex because it needs to support multi-word inserts of job & data (which can't be atomic); but nobody actually even needs that functionality, really. So, this comit: * Introduces a pre-registration API for jobs, with a GVL-requiring rb_postponed_job_prereigster, which returns a handle which can be used with an async-signal-safe rb_postponed_job_trigger. * Deprecates rb_postponed_job_register (and re-implements it on top of the preregister function for compatability) * Moves all the internal usages of postponed job register pre-registration
* RJIT: Tweak the default call thresholdTakashi Kokubun2023-04-261-1/+1
| | | | This number works better on yjit-bench (ruby-lsp).
* RJIT: Add --rjit-verify-ctx optionTakashi Kokubun2023-04-041-7/+10
|
* RJIT: Support entry with different PCsTakashi Kokubun2023-04-021-0/+20
|
* RJIT: Prefix rjit_options with rb_Takashi Kokubun2023-03-181-3/+3
|
* RJIT: Support --rjit-stats on release build as wellTakashi Kokubun2023-03-171-2/+2
|
* RJIT: Implement --rjit-trace-exitsTakashi Kokubun2023-03-121-1/+26
|
* RJIT: Introduce --rjit-exec-mem-sizeTakashi Kokubun2023-03-101-114/+18
|
* RJIT: Resurrect --rjit-pause and RJIT.resumeTakashi Kokubun2023-03-091-1/+1
|
* RJIT: Make functions in rjit_c.c staticTakashi Kokubun2023-03-081-0/+4
| | | | They don't need to be global.
* RJIT: Reorganize rjit.cTakashi Kokubun2023-03-081-54/+45
| | | | Reordering functions for readability, adding stats whenever possible.
* RJIT: Stop allowing leaked globals rjit_*Takashi Kokubun2023-03-081-43/+43
|
* RJIT: Get rid of verbose loggingTakashi Kokubun2023-03-081-26/+2
|
* RJIT: Clean up the declaration messTakashi Kokubun2023-03-081-5/+1
|
* Remove obsoleted functions in rjit.cTakashi Kokubun2023-03-071-93/+3
|
* Update documentation about RJITTakashi Kokubun2023-03-071-1/+3
|
* Fix a typoTakashi Kokubun2023-03-071-4/+4
|
* Allow enabling YJIT and RJIT independently (#7474)Takashi Kokubun2023-03-071-2/+106
| | | | | We used to require MJIT is supported when YJIT is supported. However, now that RJIT dropped some platforms that YJIT supports, it no longer makes sense. We should be able to enable only YJIT, and vice versa.
* Actually stop supporting those optionsTakashi Kokubun2023-03-071-23/+2
|
* Update options available in RJITTakashi Kokubun2023-03-071-7/+4
|
* s/mjit/rjit/Takashi Kokubun2023-03-061-131/+130
|
* s/MJIT/RJIT/Takashi Kokubun2023-03-061-67/+67
|
* Rename MJIT filenames to RJITTakashi Kokubun2023-03-061-0/+567
generated by cgit v1.2.3 (git 2.39.1) at 2024-09-22 04:27:57 +0000