- 14 Apr, 2022 15 commits
-
-
Camillo Bruni authored
- Start moving profiler scripts to tools/profiling - Add linux-perf-d8.py wrapper script that runs `perf record` and `perf inject` - Improve waiting for the d8/chrome process and allow for early termination if --timeout is provided - Allow fractional seconds for --timeout - Delete run-perf.sh and provide equivalent functionality in linux-perf-d8.py Change-Id: Iac1d6cf895aa7159a9bbb387aca7632df27a0ca3 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585951Reviewed-by:
Leszek Swirski <leszeks@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/main@{#79991}
-
Camillo Bruni authored
- Rename CodeEventDispatcher to LogEventDispatcher - Use std::vector instead of std::unordered_set, dispatching speed is more important than addition/removal of listeners - Changing the LogEventDispatcher code to be more code-search friendly - Use a raw pointer for the LogEventDispatcher instance on the isolate it's a single-owned entity Bug: v8:12795 Change-Id: I139f05431519c18cba33d1506467be918f52658c Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582125Reviewed-by:
Victor Gomes <victorgomes@chromium.org> Reviewed-by:
Jakob Linke <jgruber@chromium.org> Reviewed-by:
Jakob Kummerow <jkummerow@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/main@{#79990}
-
Clemens Backes authored
This allows to turn two field in the LiftoffAssembler into constants. R=thibaudm@chromium.org Bug: v8:12425 Change-Id: Ie39ca73d6bb704b42bd449eed984f426e69deeb5 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585956Reviewed-by:
Thibaud Michaud <thibaudm@chromium.org> Commit-Queue: Clemens Backes <clemensb@chromium.org> Cr-Commit-Position: refs/heads/main@{#79989}
-
Michael Achenbach authored
No-Try: true Bug: v8:11428 Change-Id: Ia06f5a99f851325103ad1860bf44a7313ea541f7 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3581723Reviewed-by:
Almothana Athamneh <almuthanna@chromium.org> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/main@{#79988}
-
Leszek Swirski authored
We don't actually ever need the MaglevCompiler instance. Bug: v8:7700 Change-Id: I876353310cf34971b72b08d2113d87caaa255e13 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585957 Auto-Submit: Leszek Swirski <leszeks@chromium.org> Reviewed-by:
-
V8 Autoroll authored
Change-Id: Idb40d6a544cebbdd127eb3c8d71f11adf6ac6a97 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3586227Reviewed-by:
Lutz Vahl <vahl@chromium.org> Reviewed-by:
Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/main@{#79986}
-
Tobias Tebbi authored
Change-Id: I0dbb0e800908d02a783393c3e329d306316b03c7 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585949 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Tobias Tebbi <tebbi@chromium.org> Cr-Commit-Position: refs/heads/main@{#79985}
-
Tobias Tebbi authored
Bug: chromium:1315901 Change-Id: I99ed1562356676f54e69a832c8e862c1cf74fb07 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585948Reviewed-by:
Samuel Groß <saelo@chromium.org> Commit-Queue: Tobias Tebbi <tebbi@chromium.org> Cr-Commit-Position: refs/heads/main@{#79984}
-
Jakob Gruber authored
Bug: v8:12161 Change-Id: Idb393836d787f348462c263c7f61e1a8b02b6729 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582390 Auto-Submit: Jakob Linke <jgruber@chromium.org> Reviewed-by:
-
jameslahm authored
..., this.x(), this.?x and this?.x(). For common case like these, it's not necessary to store the source position for the ThisExpression. And we could use the cached kNoSourcePosition ThisExpression for these. Bug: v8:10914 Change-Id: I02e2bc1633f6da036535d7a76bdabeac0d22f4d9 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585490Reviewed-by:
-
Jakob Gruber authored
.. instead of through urgency. Whenever we target a specific JumpLoop, the install target should be used instead of the generic urgency mechanism. Bug: v8:12161 Change-Id: I2e5cb604ce237bcc788dfc4288150881510bf6af Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3578800Reviewed-by:
-
Jakob Gruber authored
This is a reland of commit a4216b7b Original change's description: > [osr] Extract extended OSR checks to BaselineOnStackReplacement builtin > > .. to reduce Sparkplug code size. > > Bug: v8:12161 > Change-Id: I4029a75dfa37f716c285ce27153c077a0a82a341 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3576119 > Reviewed-by: Leszek Swirski <leszeks@chromium.org> > Commit-Queue: Jakob Linke <jgruber@chromium.org> > Cr-Commit-Position: refs/heads/main@{#79962} Bug: v8:12161 Change-Id: I69afd0832d7ca447b5481651ef47ebaa8d023ded Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585943 Auto-Submit: Jakob Linke <jgruber@chromium.org> Reviewed-by:
-
Lu Yahan authored
Change-Id: I95a5768af19174275e2828136ce3ff772886e84f Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585499Reviewed-by:
ji qiu <qiuji@iscas.ac.cn> Commit-Queue: Yahan Lu <yahan@iscas.ac.cn> Auto-Submit: Yahan Lu <yahan@iscas.ac.cn> Cr-Commit-Position: refs/heads/main@{#79979}
-
Lu Yahan authored
Update the size to kTierupBudgetOffset. Change-Id: Ibe241211ef67148fae3a4a9eed271f9293ca4801 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585492Reviewed-by:
-
v8-ci-autoroll-builder authored
Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/b37c340..67d9897 Rolling v8/buildtools/linux64: git_revision:ae110f8b525009255ba1f9ae96982176d3bfad3d..git_revision:0cbe341c1a28037ee32d21b589030a7df0b9fdab Rolling v8/buildtools/third_party/libunwind/trunk: https://chromium.googlesource.com/external/github.com/llvm/llvm-project/libunwind/+log/c39fea8..bbcbce9 Rolling v8/third_party/catapult: https://chromium.googlesource.com/catapult/+log/3a1ae18..90316ac Rolling v8/third_party/depot_tools: https://chromium.googlesource.com/chromium/tools/depot_tools/+log/3b97fa8..3c6f0eb Rolling v8/tools/clang: https://chromium.googlesource.com/chromium/src/tools/clang/+log/b5e2f7c..9199dc2 R=v8-waterfall-sheriff@grotations.appspotmail.com,mtv-sf-v8-sheriff@grotations.appspotmail.com Change-Id: I0b4e855b5c28652f425db99a109bfcc26ad9dca3 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3586225 Commit-Queue: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Bot-Commit: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Cr-Commit-Position: refs/heads/main@{#79977}
-
- 13 Apr, 2022 25 commits
-
-
Milad Fa authored
CL also adds TODO items to implement this functionality on disassembler and the simulator. The R bit which enables PC-relative capabilities when set is not being used during codegen at the moment. Change-Id: I0927d9be37a50beecf9b6a052e378dca7604513b Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584535Reviewed-by:
Junliang Yan <junyan@redhat.com> Commit-Queue: Milad Farazmand <mfarazma@redhat.com> Cr-Commit-Position: refs/heads/main@{#79976}
-
Corentin Pescheloche authored
With the switch to primitive ids make sure no profiler can be returned with the same id. Bug: chromium:1297283 Change-Id: I9cf944e9a472ea45679feb0f30137dba95a32ca6 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582786 Auto-Submit: Corentin Pescheloche <cpescheloche@fb.com> Reviewed-by:
Camillo Bruni <cbruni@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/main@{#79975}
-
Anton Bikineev authored
Bug: chromium:1029379 Change-Id: I34e988c0e14441afe821a35505c4ce3f1eee8fc4 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584118Reviewed-by:
Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Anton Bikineev <bikineev@chromium.org> Cr-Commit-Position: refs/heads/main@{#79974}
-
Camillo Bruni authored
CacheInitialJSArrayMaps was called in the middle of JSFunction::SetPrototype even though this only happens during bootstrapping given that Array.prototype os non-configurable and non-writable. Changes: - Rename CacheInitialArrayMaps to InitializeJSArrayMaps - Add more explicit checks in InitializeJSArrayMaps to link back to the Context indices for better code searching Change-Id: Iad6d20e3d67d715bfd6429037c75ac35ab7f399f Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3571889Reviewed-by:
Shu-yu Guo <syg@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/main@{#79973}
-
Junliang Yan authored
Change-Id: I205f055af95bd92a84f0a2bc5bbe4d2633e4a8cc Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584530Reviewed-by:
Milad Farazmand <mfarazma@redhat.com> Commit-Queue: Junliang Yan <junyan@redhat.com> Cr-Commit-Position: refs/heads/main@{#79972}
-
Shu-yu Guo authored
This is a reland of commit 38767476 Changes since revert: None, original failure was caused by another CL Original change's description: > [heap] Skip weak heap objects when sharedness doesn't match > > During marking, shared objects should not be marked when not > marking the shared heap (i.e. when not doing a shared GC). > Doing so is unsafe, as marking can race with sweeper threads > sweeping the shared heap. This CL adds the missing check on > weak object marking. > > Bug: v8:12687 > Change-Id: I1e0b8ba6b09bbcf665e5ff0f6242ed88f543c1fa > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3583610 > Reviewed-by: Dominik Inführ <dinfuehr@chromium.org> > Commit-Queue: Shu-yu Guo <syg@chromium.org> > Cr-Commit-Position: refs/heads/main@{#79964} Bug: v8:12687 Change-Id: Ia24ff5daa4d37daf73391c0e577b086611c5e496 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584600 Owners-Override: Shu-yu Guo <syg@chromium.org> Reviewed-by:
-
Milad Fa authored
This CL adds prefixed load scalar and floating point instructions to the assembler and uses it during code generation if the processor supports it. They have also been added to the disassembler and the simulator. Drive-by: emit_prefix is now making sure a trampoline doesn't get emitted in between prefix and suffix. Assembler needs to block it using BlockTrampolinePoolScope. Drive-by: more CHECKs are added to the assembler to make sure immediate values have the correct length. Change-Id: Ic5e38926c65364bc97aba0b7638dde5f365ec706 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582479Reviewed-by:
-
Junliang Yan authored
drive-by: Add code comment macros Change-Id: I1bded0f0e49f9e590cb7a8a5d86e26e8d9d5eea6 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585530Reviewed-by:
-
Tobias Tebbi authored
Bug: chromium:1315901 Change-Id: Ic44bfcae32aba202ba25c5f59fe579214a444584 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584117Reviewed-by:
-
Junliang Yan authored
Change-Id: Ic65f0cccaba6cea2d441540b57689e2c68af3125 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3585526Reviewed-by:
-
Tobias Tebbi authored
This reverts commit a4216b7b. Reason for revert: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux%20-%20arm64%20-%20sim%20-%20MSAN/43174/overview Original change's description: > [osr] Extract extended OSR checks to BaselineOnStackReplacement builtin > > .. to reduce Sparkplug code size. > > Bug: v8:12161 > Change-Id: I4029a75dfa37f716c285ce27153c077a0a82a341 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3576119 > Reviewed-by: Leszek Swirski <leszeks@chromium.org> > Commit-Queue: Jakob Linke <jgruber@chromium.org> > Cr-Commit-Position: refs/heads/main@{#79962} Bug: v8:12161 Change-Id: I382609d0b8cd951a3df5c9c834fe7071eb90faa5 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584121 Commit-Queue: Tobias Tebbi <tebbi@chromium.org> Owners-Override: Tobias Tebbi <tebbi@chromium.org> Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Cr-Commit-Position: refs/heads/main@{#79966}
-
Tobias Tebbi authored
This reverts commit 38767476. Reason for revert: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux%20-%20arm64%20-%20sim%20-%20MSAN/43174/overview Original change's description: > [heap] Skip weak heap objects when sharedness doesn't match > > During marking, shared objects should not be marked when not > marking the shared heap (i.e. when not doing a shared GC). > Doing so is unsafe, as marking can race with sweeper threads > sweeping the shared heap. This CL adds the missing check on > weak object marking. > > Bug: v8:12687 > Change-Id: I1e0b8ba6b09bbcf665e5ff0f6242ed88f543c1fa > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3583610 > Reviewed-by: Dominik Inführ <dinfuehr@chromium.org> > Commit-Queue: Shu-yu Guo <syg@chromium.org> > Cr-Commit-Position: refs/heads/main@{#79964} Bug: v8:12687 Change-Id: If068d31f23a5280667dea1085b3076bb2107d07a No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584119 Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com> Commit-Queue: Tobias Tebbi <tebbi@google.com> Owners-Override: Tobias Tebbi <tebbi@google.com> Cr-Commit-Position: refs/heads/main@{#79965}
-
Shu-yu Guo authored
During marking, shared objects should not be marked when not marking the shared heap (i.e. when not doing a shared GC). Doing so is unsafe, as marking can race with sweeper threads sweeping the shared heap. This CL adds the missing check on weak object marking. Bug: v8:12687 Change-Id: I1e0b8ba6b09bbcf665e5ff0f6242ed88f543c1fa Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3583610Reviewed-by:
Dominik Inführ <dinfuehr@chromium.org> Commit-Queue: Shu-yu Guo <syg@chromium.org> Cr-Commit-Position: refs/heads/main@{#79964}
-
Junliang Yan authored
Change-Id: I8bae34db4d1aa02516dcf4610f77d6e8edaa5775 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3581490Reviewed-by:
-
Jakob Gruber authored
.. to reduce Sparkplug code size. Bug: v8:12161 Change-Id: I4029a75dfa37f716c285ce27153c077a0a82a341 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3576119Reviewed-by:
-
Shu-yu Guo authored
Computing the length for variable-length TAs is a lot of code and was regressing microbenchmarks. Bug: v8:11111 Change-Id: Ia7c3c92bfb43938068aaf539b290f6a30b049c18 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3583898Reviewed-by:
Marja Hölttä <marja@chromium.org> Commit-Queue: Shu-yu Guo <syg@chromium.org> Cr-Commit-Position: refs/heads/main@{#79961}
-
Anton Bikineev authored
This reduces card granularity from 4096 to 512 bytes with the goal to improve write barrier filtering. Bug: chromium:1029379 Change-Id: I22e2a9c61ef4c36c3db65404370213d0a8048e08 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582393Reviewed-by:
-
Marja Hölttä authored
For a while, we shipped a version which writes version 13 data with JSArrayBufferView flags, and then fixed version 13 to not include the flags. This CL adds a compatibility mode for parsing the the version 13 data which includes the flags, since it still occurs in the wild. Bug: chromium:1314833,chromium:1284506 Change-Id: I96cc432c8574a40b11ec0037394feb1853515760 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3583982Reviewed-by:
-
Michael Lippautz authored
Bug: v8:12781 Change-Id: I4cf23035a7679efdb4830335b75d8616abfbbe08 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582385Reviewed-by:
-
Nikolaos Papaspyrou authored
Tracer scopes are used in numerous places in src/heap to track time spent during various phases of the garbage collection. Usually, they are introduced with the TRACE_GC* family of macros, taking the scope identifier as a parameter. At most call sites, the scope identifier is known at compile time. This CL inlines the constructor and destructor of GCTracer::Scope, in order to enable the C++ compiler to properly optimize the introduction of such scopes when the scope identifier is known at compile time, using constant propagation. This is expected to have a performance impact for short-lived and frequently used scopes, e.g., in incremental marking and sweeping steps. Change-Id: I6f1a2954a437de1fa6dab5e464c20c952d84ffd4 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3581774Reviewed-by:
Omer Katz <omerkatz@chromium.org> Reviewed-by:
-
Jakob Kummerow authored
This makes the installation sequence of WebAssembly.Tag slightly shorter, slightly faster, slightly cleaner in corner-case semantics, and slightly better documented. To allow testing this code, Isolate::InstallConditionalFeatures is exposed as d8.test.installConditionalFeatures(). Fixed: chromium:1314616 Change-Id: I44285e398b8797e0e7d2d8c782cecec3ba68a503 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582382 Commit-Queue: Jakob Kummerow <jkummerow@chromium.org> Reviewed-by:
Clemens Backes <clemensb@chromium.org> Reviewed-by:
Toon Verwaest <verwaest@chromium.org> Cr-Commit-Position: refs/heads/main@{#79956}
-
Omer Katz authored
NewSpace and PagedSpace both inherit from SpaceWithLinearArena and implement allocation logic on top of it. The parts of the allocation path that deal specifically with the linear allocation area are equivalent (only minor syntactic differences between them). This CL refactors the allocation from a linear allocation area out of NewSpace and PagedSpace and moves it to SpaceWithLinearArea. This eliminates code duplication and keeps everything generally still working the same. This is done as part of an effort to create a stable NewSpace interface to allow introducing an alternative paged new space. Bug: v8:12612 Change-Id: Ie24345a2d51f6e67ebe8a1d67e586038f7aec8de Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3578547Reviewed-by:
-
Dominik Inführ authored
Change-Id: I11ffc3197e78a15e38978724da07e10b62857b3b Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582392 Auto-Submit: Dominik Inführ <dinfuehr@chromium.org> Reviewed-by:
-
Jakob Gruber authored
The code generator accesses the heap even without --code-comments set: remove the related condition from the UnparkedScopeIfNeeded. Fixed: v8:12794 Change-Id: I0099f22a9382373c4f75538615fbf431c4d71283 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3582389 Commit-Queue: Jakob Linke <jgruber@chromium.org> Auto-Submit: Jakob Linke <jgruber@chromium.org> Reviewed-by:
-
Leszek Swirski authored
Unconditional eager deopts from lack of feedback (née soft deopts) mean that the remainder of the basic block is dead. Avoid emitting this code by fast forwarding the iterator until the next merge. The EagerDeopt node becomes a Deopt control node which terminates its own block (this is to avoid spurious control flow after the EagerDeopt, or weirdness with liveness). A concept of "merging dead blocks" has to be introduced so that the successors of the killed block still have the right number of predecessors. Bug: v8:7700 Change-Id: Id9c442c3b18d3f394dc2411604d0c8503d6aaae2 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3578647Reviewed-by:
-
