- Aug 01, 2024
-
-
Haffner Immanuel authored
-
- Jul 22, 2024
-
-
v8-ci-autoroll-builder authored
Change-Id: I45a91f08a83f9cadff0d82b9090c2bc73fe93ed7 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5728997 Bot-Commit: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Cr-Commit-Position: refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63ed-refs/heads/main@{#95151}
-
Matthias Liedtke authored
Runtime::kWasmGenerateRandomModule only exists on non-official builds. Change-Id: I7579d28247a27d895ce350ba012ea7b27623c3e7 No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5724702 Auto-Submit: Matthias Liedtke <mliedtke@chromium.org> Reviewed-by:
Samuel Groß <saelo@chromium.org> Commit-Queue: Samuel Groß <saelo@chromium.org> Cr-Commit-Position: refs/heads/main@{#95151}
-
Samuel Groß authored
This is a reland of commit 121d624b More runtime functions are now opted out of (differential) fuzzing. Original change's description: > Make more runtime functions available to the fuzzers > > Previously, we had an explicit allow-list of runtime functions for > fuzzing, and every function that should be available to fuzzers needed > to be added to that. This approach is somewhat fragile, however, as it > is easy to forget adding new functions to that list. With this CL, all > runtime functions that are used for testing are automatically exposed to > fuzzers, but can be manually opted out, which is still necessary for > some functions for a number of reasons. > > This CL also makes a number of test functions fuzzer safe, as those are > now exposed to fuzzers. There will likely be a few more functions that > were missed and are not yet fully fuzzer-compatible. However, those > should quickly be identified by our fuzzers, at which point they can > either be made fuzzing-compatible or opted-out of fuzzing. > > Bug: 353685107 > Change-Id: Ibae038fee2926205bcbc91a8bcadcaec9a8242a1 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5720570 > Reviewed-by:
Matthias Liedtke <mliedtke@chromium.org> > Reviewed-by:
Michael Achenbach <machenbach@chromium.org> > Reviewed-by:
Toon Verwaest <verwaest@chromium.org> > Commit-Queue: Samuel Groß <saelo@chromium.org> > Cr-Commit-Position: refs/heads/main@{#95126} Bug: 353685107 Change-Id: Icff469bbee67bb65b8520ef57a42616615199916 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5728075 Commit-Queue: Samuel Groß <saelo@chromium.org> Reviewed-by:
-
Matthias Liedtke authored
Change-Id: I2a49b2d0978c167c1ee69fbd64a36b9042aae571 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5729412 Commit-Queue: Matthias Liedtke <mliedtke@chromium.org> Reviewed-by:
Jakob Kummerow <jkummerow@chromium.org> Auto-Submit: Matthias Liedtke <mliedtke@chromium.org> Commit-Queue: Jakob Kummerow <jkummerow@chromium.org> Cr-Commit-Position: refs/heads/main@{#95149}
-
Matthias Liedtke authored
In case of tail calls in the deoptimized frames the caller pc of the bottom most (outer most) function can end up in a different stack slot than in the optimized function as the number of parameter stack slots may differ. This means, the caller_pc_ needs to be stripped from the PAC and re-signed again prior to writing it to the output frame. Fixed: 353582136 Bug: 42204618 Change-Id: I1670c416d3af44af0507b82717ac1df4d3cc11ad Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5716734 Reviewed-by:
Stephen Röttger <sroettger@google.com> Cr-Commit-Position: refs/heads/main@{#95148}
-
Andreas Haas authored
Fast API calls from WebAssembly are based on the well-known imports optimization, which allows faster calls from WebAssembly when the import is known to be of a special kind. However, this optimization can only be used when all instances of a WebAssembly module use the same imports. This CL adds a test where a WebAssembly module gets instantiated with different fast API call targets to check that the well-known imports optimization indeed gets disabled. Bug: 41492790 Change-Id: I91ff611a6cbd26a025dd248cb83e2718b44fd956 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5720910 Reviewed-by:
Clemens Backes <clemensb@chromium.org> Commit-Queue: Andreas Haas <ahaas@chromium.org> Cr-Commit-Position: refs/heads/main@{#95147}
-
v8-ci-autoroll-builder authored
Rolling build: https://chromium.googlesource.com/chromium/src/build/+log/ad2f859..7a8285c Rolling buildtools: https://chromium.googlesource.com/chromium/src/buildtools/+log/adc2a53..3ef44a2 Rolling buildtools/reclient: re_client_version:0.148.0.41b09b51-gomaip..re_client_version:0.150.1.d9707319-gomaip Rolling third_party/catapult: https://chromium.googlesource.com/catapult/+log/29445d0..1699936 Rolling third_party/depot_tools: https://chromium.googlesource.com/chromium/tools/depot_tools/+log/b4102e4..cdcdd6e Rolling third_party/googletest/src: https://chromium.googlesource.com/external/github.com/google/googletest/+log/9ff2450..cee1ba1 Rolling third_party/markupsafe: https://chromium.googlesource.com/chromium/src/third_party/markupsafe/+log/e582d7f..6638e9b Rolling third_party/siso: git_revision:b41f9eaf44dfdaec51ab2d7089aaa4cb6d9e7f5e..git_revision:50a6db5dae3978d2d2e8dce29f6df024dde48d1b Rolling tools/clang: https://chromium.googlesource.com/chromium/src/tools/clang/+log/a5953b1..4dc76da Rolling tools/luci-go: git_revision:c9c95dda5f310610a3fc574af44faeed298b3938..git_revision:771ea9a614a104c71655f699ef82219a2a474817 Rolling tools/luci-go: git_revision:c9c95dda5f310610a3fc574af44faeed298b3938..git_revision:771ea9a614a104c71655f699ef82219a2a474817 Roll created at https://cr-buildbucket.appspot.com/build/8741743458899784145 Change-Id: I8848908bd3d7046ce9a7afb7304d2c6f59867871 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5728992 Commit-Queue: Michael Achenbach <machenbach@chromium.org> Bot-Commit: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Cr-Commit-Position: refs/heads/main@{#95146}
-
Victor Gomes authored
Fixed: 353877568 Change-Id: I31bba348ae0deff17ea1568d2791df51710cedc6 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5724700 Commit-Queue: Victor Gomes <victorgomes@chromium.org> Auto-Submit: Victor Gomes <victorgomes@chromium.org> Commit-Queue: Olivier Flückiger <olivf@chromium.org> Reviewed-by:
Olivier Flückiger <olivf@chromium.org> Cr-Commit-Position: refs/heads/main@{#95145}
-
Daniel Cheng authored
GCTracer::ResetForTesting() invokes the destructor and then uses placement new to reinitialize GCTracer to its default state. Unfortunately, the placement new call uses `heap_`, but `heap_` is no longer valid to access after invoking the destructor. The fix is to cache the value in a local and then pass the value of the local to placement new. Bug: 40222690 Bug: 353960901 Change-Id: I64bce1a995fd648bda167f743a0f9f1ed8b34215 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5727716 Reviewed-by:
Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/main@{#95144}
-
Pan, Tao authored
This CL can improve Speedometer3 by ~0.3% and reduce ~4% (2903 -> 2793) turbofan compilation when running Speedometer3. Bug: v8:14296 Change-Id: I134013f6a895785e410c027c64120a1fd1910ffa Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5679484 Reviewed-by:
Leszek Swirski <leszeks@chromium.org> Commit-Queue: Hao A Xu <hao.a.xu@intel.com> Cr-Commit-Position: refs/heads/main@{#95143}
-
Pan, Tao authored
ExpectedTransitionKey directly returns key if TransitionAccessor has only one member, and returns null if TransitionAccessor has multiple entries. This CL compare string content if TransitionAccessor has less than 8 entries. This CL can improve the below micro case by ~17%. for (var i = 0; i < 1e7; ++i) JSON.parse('{"title":0,"completed":1,"id":2}'); Change-Id: I11307c5e68940a65ade244714652df6f8d655fbb Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5672333 Commit-Queue: Hao A Xu <hao.a.xu@intel.com> Reviewed-by: -
Lu Yahan authored
Change-Id: If29e4190ccc3b8f320273d61ba9117d3f4fd3836 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5717465 Commit-Queue: Ji Qiu <qiuji@iscas.ac.cn> Reviewed-by:
Ji Qiu <qiuji@iscas.ac.cn> Cr-Commit-Position: refs/heads/main@{#95141}
-
- Jul 19, 2024
-
-
Michael Achenbach authored
This reverts commit 121d624b. Reason for revert: Causes many differential-fuzzing false positives in the current state. Original change's description: > Make more runtime functions available to the fuzzers > > Previously, we had an explicit allow-list of runtime functions for > fuzzing, and every function that should be available to fuzzers needed > to be added to that. This approach is somewhat fragile, however, as it > is easy to forget adding new functions to that list. With this CL, all > runtime functions that are used for testing are automatically exposed to > fuzzers, but can be manually opted out, which is still necessary for > some functions for a number of reasons. > > This CL also makes a number of test functions fuzzer safe, as those are > now exposed to fuzzers. There will likely be a few more functions that > were missed and are not yet fully fuzzer-compatible. However, those > should quickly be identified by our fuzzers, at which point they can > either be made fuzzing-compatible or opted-out of fuzzing. > > Bug: 353685107 > Change-Id: Ibae038fee2926205bcbc91a8bcadcaec9a8242a1 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5720570 > Reviewed-by:
-
Michael Achenbach authored
This reverts commit 59798082. Reason for revert: Need to revert previous CL. Original change's description: > Don't expose %SerializeDeserializeNow() to fuzzers > > The runtime function is not currently fuzzing-safe and will frequently > cause crashes if used by fuzzers. > > Bug: 353971258, 353685107 > Change-Id: I692bbd752fc86930e4ea014d2557247422b6ba9a > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5725353 > Auto-Submit: Samuel Groß <saelo@chromium.org> > Reviewed-by:
-
Milad Fa authored
Port e8f70212 Original Commit Message: The CL introduces F16x8Splat, F16x8ExtractLane and F16x8ReplaceLane opcodes together with software implementation in Liftoff. R=irezvov@chromium.org, joransiu@ca.ibm.com, junyan@redhat.com, midawson@redhat.com BUG= LOG=N Change-Id: I3532d89122e4d67034398c609ce3c17bc46da037 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5725501 Commit-Queue: Milad Farazmand <mfarazma@redhat.com> Reviewed-by:
Junliang Yan <junyan@redhat.com> Cr-Commit-Position: refs/heads/main@{#95138}
-
Ilya Rezvov authored
The CL introduces F16x8Splat, F16x8ExtractLane and F16x8ReplaceLane opcodes together with software implementation in Liftoff. Bug: 337998764 Change-Id: I628d25feba492afe02d4f4d325e4aaf824a98d53 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5722510 Commit-Queue: Ilya Rezvov <irezvov@chromium.org> Reviewed-by:
-
Samuel Groß authored
The runtime function is not currently fuzzing-safe and will frequently cause crashes if used by fuzzers. Bug: 353971258, 353685107 Change-Id: I692bbd752fc86930e4ea014d2557247422b6ba9a Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5725353 Auto-Submit: Samuel Groß <saelo@chromium.org> Reviewed-by:
-
Clemens Backes authored
Based on the preceeding CL (https://crrev.com/c/5718231) we can greatly simplify the LogCodesTask, relying on the CancellableTaskManager. Instead of using a bunch of atomic values and pointers, we now update all fields under the WasmEngine's mutex. R=mliedtke@chromium.org Bug: 42204620 Cq-Include-Trybots: luci.v8.try:v8_linux64_tsan_dbg Cq-Include-Trybots: luci.v8.try:v8_linux64_tsan_isolates_rel Cq-Include-Trybots: luci.v8.try:v8_linux64_tsan_rel Change-Id: I13fabfbf3f388274030f72e36f7dad1b6b7b3d18 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5718451 Reviewed-by:
-
Nico Hartmann authored
Bug: chromium:42210561 Change-Id: I534d1f3b066fcba1c50671cfbda8c3fbd6c813d1 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5725552 Commit-Queue: Nico Hartmann <nicohartmann@chromium.org> Reviewed-by:
Darius Mercadier <dmercadier@chromium.org> Cr-Commit-Position: refs/heads/main@{#95134}
-
Toon Verwaest authored
They explicitly aren't Key*, but can be treated similarly. Change-Id: Ib2c403425681a16014640bb3768cf60c85c4a287 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5724451 Reviewed-by:
-
Milad Fa authored
Followup CL on porting f2aaa9fd Change-Id: I2f36c60e214d2583622e9ad828045acd2b822ed3 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5717298 Reviewed-by:
-
Benedikt Meurer authored
This CL removes most of the overhead of the `ConsoleHelper` from the implementation of `console.timeStamp()` by defering work to until it's actually needed. Bug: 350443026 Change-Id: I26a2bd3f246b24eaad349e213b17baeeed23fa6b Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5724349 Auto-Submit: Benedikt Meurer <bmeurer@chromium.org> Reviewed-by:
Simon Zünd <szuend@chromium.org> Commit-Queue: Benedikt Meurer <bmeurer@chromium.org> Cr-Commit-Position: refs/heads/main@{#95131}
-
Samuel Groß authored
With this CL, FeedbackCells now own an entry in the JSDispatchTable. The entry is not yet populated, but it is already managed by the GC. In follow-up CLs, the entry handle will be populated to point to the current code for the functions sharing the FeedbackCell. Bug: 40931165, 42204201 Change-Id: I4f2cb6850b31ba7fc7af7460a3f8e9d5a0b9ed65 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5677828 Reviewed-by:
-
Michael Achenbach authored
This is a reland of commit 857c5502 All of the bugs that were found when fuzzing turboshaft-from-maglev the 1st time have been fixed. Time for round 2! Original change's description: > [test] Add turboshaft-from-maglev to fuzzing experiments > > No-Try: true > Bug: 42204525 > Change-Id: If7b5a42b2b5ba8b225cfea476e20210f44f11f85 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5679230 > Reviewed-by:
-
Darius Mercadier authored
Bug: chromium:42204525 Change-Id: I3ae60b1a730a904bc23faab784a7dc8fe360bfb8 Fixed: chromium:350770683 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5713609 Auto-Submit: Darius Mercadier <dmercadier@chromium.org> Commit-Queue: Darius Mercadier <dmercadier@chromium.org> Reviewed-by:
Nico Hartmann <nicohartmann@chromium.org> Cr-Commit-Position: refs/heads/main@{#95128}
-
Nico Hartmann authored
Bug: chromium:42210561 Change-Id: If3ae2b7abe391395474097d336cf4e131b330384 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5723348 Reviewed-by:
-
Samuel Groß authored
Previously, we had an explicit allow-list of runtime functions for fuzzing, and every function that should be available to fuzzers needed to be added to that. This approach is somewhat fragile, however, as it is easy to forget adding new functions to that list. With this CL, all runtime functions that are used for testing are automatically exposed to fuzzers, but can be manually opted out, which is still necessary for some functions for a number of reasons. This CL also makes a number of test functions fuzzer safe, as those are now exposed to fuzzers. There will likely be a few more functions that were missed and are not yet fully fuzzer-compatible. However, those should quickly be identified by our fuzzers, at which point they can either be made fuzzing-compatible or opted-out of fuzzing. Bug: 353685107 Change-Id: Ibae038fee2926205bcbc91a8bcadcaec9a8242a1 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5720570 Reviewed-by:
-
Eva Herencsarova authored
The regression test uncovers more issues, so skip if for now until they are also fixed. Bug: 354103254, 354104317, 353980376 No-Try: true No-Tree-Checks: true Change-Id: I6275dba336838c9c1e5c0057e76c91adf06bdbf9 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5724449 Reviewed-by:
Eva Herencsárová <evih@chromium.org> Commit-Queue: Clemens Backes <clemensb@chromium.org> Cr-Commit-Position: refs/heads/main@{#95125}
-
Clemens Backes authored
It leads to rare lock-order inversion problems if the foreground task runner calls task destructors while holding the task runner's lock. This currently happened when a new task was posted when the task runner was already terminated. After normal execution we would only call the destructor after releasing the lock anyway. This CL ensures that the mutex is never held when calling a task destructor. R=mlippautz@chromium.org Bug: 42204620 Change-Id: Ie895e5a605dff064bc5a4a90bdf1c7cdd4dc8cfe Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5718231 Reviewed-by:
-
Victor Gomes authored
This allows setting/storing fields in the same basic block without copying the VO. Bug: v8:7700 Change-Id: I9ef2d0e5de39a9fa235b8dc95b5144ef3cab75f5 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5701106 Auto-Submit: Victor Gomes <victorgomes@chromium.org> Reviewed-by:
-
Darius Mercadier authored
AnyUseMarkingProcessor iterates the graph forward, and can only remove the node it's currently visiting. This means that if a node n1 is only used in a node n2, which itself has no uses, then {n1} cannot be removed when it's visited because it has a use. When visiting {n2}, {n2} is removed and the use-count of {n1} is set to 0 (or -1?), but {n1} is not removed. Note that in the regular Maglev pipeline, such unused nodes are removed by LiveRangeAndNextUseProcessor, but this processor doesn't run in the Turboshaft pipeline, which is how we ended up reaching this DCHECK. Bug: chromium:42204525 Change-Id: Id00950fafeb530e48198af93fa6d639b61ee4026 Fixed: chromium:351283984 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5717696 Reviewed-by: -
Clemens Backes authored
Just a little cleanup before doing the actual change I want to do in the next CL. We have lots of "FooLocked()" methods in the code which assume that some mutex is locked. In the case of the {DefaultForegroundTaskRunner} we were passing a reference to the {MutexGuard} to those methods to further enforce that the mutex is held by the caller. This CL removes this and instead adds a DCHECK that the mutex cannot be locked again in the caller. Drive-by: Rename "lock_" to "mutex_". R=mlippautz@chromium.org Bug: 42204620 Change-Id: I3738db79d45875c975bd37398ef7e9052dc11896 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5717890 Commit-Queue: Clemens Backes <clemensb@chromium.org> Reviewed-by: -
Benedikt Meurer authored
The handling of the `console.time*()` functions by the embedder is now done through an alternative interface that passes `v8::String` directly and thereby saves unnecessary copies of string bytes. As of https://crrev.com/c/5721448 Chromium is migrated to the new interface and Node.js never implemented the interface in the first place, so we are good to remove the old interface now. Bug: 350443026, 41433391 Change-Id: I53a6b371f85b2150beaef0a28fc18206fa07ff17 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5724348 Commit-Queue: Yang Guo <yangguo@chromium.org> Reviewed-by:
Yang Guo <yangguo@chromium.org> Commit-Queue: Benedikt Meurer <bmeurer@chromium.org> Auto-Submit: Benedikt Meurer <bmeurer@chromium.org> Cr-Commit-Position: refs/heads/main@{#95120}
-
Lu Yahan authored
Port commit 06847fbb Change-Id: I3fb708c7e3a55a02443484f743ffa1c4f3c3c31c Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5716343 Reviewed-by:
-
Yang Guo authored
Some methods were not allowlisted because in the past, they were implemented as JS builtins and there were side-effects due to creating an i18n cache object. Some other methods were not allowlisted because they run regexps, which can affect RegExp.$1 etc – that has since then been worked around. For String.{match, matchAll, replace, replaceAll, split, search} we invalidate the protector cell to avoid the fast path that inlines calls to RegExp.prototype[Symbol.*]. Bug: 351769842 Change-Id: I2e5276b81d5b48488fc716d183a7ffa74a225604 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5720968 Commit-Queue: Simon Zünd <szuend@chromium.org> Auto-Submit: Yang Guo <yangguo@chromium.org> Commit-Queue: Yang Guo <yangguo@chromium.org> Reviewed-by: -
Lu Yahan authored
Port commit f2aaa9fd Change-Id: I036cd6a7bd9f9c60a8248dcc3b2f4a8e05caa181 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5716342 Reviewed-by:
-
- Jul 18, 2024
-
-
Seth Brenith authored
The Edge performance team requested this change as part of an investigation into memory-related system calls. Change-Id: Ia65a96149df3e5e3a427ff058dde260806ad28ce Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5667927 Commit-Queue: Seth Brenith <seth.brenith@microsoft.com> Reviewed-by:
-
Andreas Haas authored
When a parameter for a fast API call gets converted from float64 or float32 to int64 (see https://crrev.com/c/5710270 for an explanation), the float value may be out of int64 range. In that case, an out-of-range error should be thrown. This situation was ignored so far. With this CL, an error gets thrown by doing a regular API call instead of a fast API call, and let the regular call (actually the regular API call callback) take care of the range check. Bug: 41492790 Change-Id: I6c74d7fc97fa0d450fcc202db9ade2f555bec9df Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5712993 Commit-Queue: Andreas Haas <ahaas@chromium.org> Reviewed-by:
-
v8-ci-autoroll-builder authored
Rolling build: https://chromium.googlesource.com/chromium/src/build/+log/0e57d88..ad2f859 Rolling third_party/abseil-cpp: https://chromium.googlesource.com/chromium/src/third_party/abseil-cpp/+log/b7e091d..9d1552f Roll created at https://cr-buildbucket.appspot.com/build/8742060549809517889 Change-Id: I835b83e597de47761dfb60d53403a7f71a55f491 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5722348 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@{#95114}
-
