- 29 Jan, 2021 19 commits
-
-
Clemens Backes authored
This adds a regression test for https://crrev.com/c/2652488. The test reduces the available code space such that it would trigger an OOM condition if code is not garbage-collected. In order to guarantee garbage-collection in all configurations, an explicit interrupt check is added to the WasmDebugBreak runtime function. R=thibaudm@chromium.org Bug: chromium:1168564 Change-Id: I8fce7aa5128c9e3c9a7e2d2e7397c394fec7de85 Cq-Include-Trybots: luci.v8.try:v8_linux64_asan_rel_ng Cq-Include-Trybots: luci.v8.try:v8_linux64_tsan_isolates_rel_ng Cq-Include-Trybots: luci.v8.try:v8_linux64_tsan_rel_ng Cq-Include-Trybots: luci.v8.try:v8_mac64_asan_rel_ng Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2652490 Commit-Queue: Clemens Backes <clemensb@chromium.org> Reviewed-by:
Thibaud Michaud <thibaudm@chromium.org> Cr-Commit-Position: refs/heads/master@{#72435}
-
Manos Koukoutos authored
The latest wasm-gc spec introduces classification functions {ref.as, ref.is, br_on}_{func, data, i31}, to cast values typed as anyref and eqref. Bug: v8:7748 Change-Id: I39c288e4a4c96466e64e490f164ccc76b00011ee Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2655506 Commit-Queue: Manos Koukoutos <manoskouk@chromium.org> Reviewed-by:Jakob Kummerow <jkummerow@chromium.org> Cr-Commit-Position: refs/heads/master@{#72434}
-
Andreas Haas authored
... LiftoffStackSlots::Construct R=thibaudm@chromium.org Bug: chromium:1171788 Change-Id: Ifb8e20f4e81fe2c698fe1f51c0b833a6049f7558 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2659255Reviewed-by:
-
Mike Stanton authored
Actual FeedbackVector IC slots don't participate in the Ref class, since they are read during serialization as we process bytecode. So FeedbackVectorRef really only deals with the FeedbackCell array and the SharedFunctionInfo. These two types are already in the no serialized list, so it's pretty easy to move this class over there too. Bug: v8:7790 Change-Id: I51b7bf4c3404ae5bcfb16d29b5e719787ddd6b17 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656317 Commit-Queue: Michael Stanton <mvstanton@chromium.org> Reviewed-by:
Georg Neis <neis@chromium.org> Reviewed-by:
Nico Hartmann <nicohartmann@chromium.org> Cr-Commit-Position: refs/heads/master@{#72432}
-
Clemens Backes authored
We were doing the cache lookup using the original {wire_bytes_vec}, but then inserting the module with another copy stored in the {NativeModule}. This causes {NativeModuleCache::Key::operator<} to do a full {memcmp} on the two wire bytes copies. By using the same pointer consistently, we can avoid that costly {memcmp}. R=thibaudm@chromium.org Bug: v8:11164 Change-Id: I19ba1022f700d8da40671818ee2e2ebdbbddd5c0 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2658329Reviewed-by: -
Michael Lippautz authored
Adds getters for GC phases to be used by advanced embedders to ensure and check consistency conditions as needed. Bug: chromium:1056170 Change-Id: Ia0b219f838bf31f0edbfe40585b95bb5eafa734d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2658328Reviewed-by:
Ulan Degenbaev <ulan@chromium.org> Reviewed-by:
Omer Katz <omerkatz@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/master@{#72430}
-
Ulan Degenbaev authored
The new flag: - enables --concurrent-inlining, - disables --lazy-feedback-allocation, - sets the interrupt budget to 15K. Bug:v8:7790 Change-Id: Id236763f987a6b1309f0c43a2acd150c6e3723bb Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2652492 Commit-Queue: Ulan Degenbaev <ulan@chromium.org> Reviewed-by:
Michael Achenbach <machenbach@chromium.org> Reviewed-by:
Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#72429}
-
Thibaud Michaud authored
The delegate instruction is invalid in the following cases: - When the target is not a try block or the function block, - When the instruction is inside a catch handler of the target. R=clemensb@chromium.org Bug: v8:8091 Change-Id: Ic59e8314982166863ba2078e2b3b39e3ba488a74 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656318Reviewed-by:
Clemens Backes <clemensb@chromium.org> Commit-Queue: Thibaud Michaud <thibaudm@chromium.org> Cr-Commit-Position: refs/heads/master@{#72428}
-
Clemens Backes authored
Avoid constructing the frame summary (and a std::vector) just for getting the function index. Just get it from the code instead (where also the frame summary would get it from). R=jkummerow@chromium.org Bug: v8:11074 Change-Id: Ie9957e145d6b641fb211b03ef593d57afd310c91 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2653230Reviewed-by:
-
Omer Katz authored
Tolerance was set to 50 microseconds. This is not enough and causes flaky failures. Tolerance is needed because we record milliseconds and an empty scope can still take a few milliseconds to create and destroy. Increasing the tolerance to 500 microseconds should suffice and makes sense for the test. All expected values are in 10s of milliseconds, so half a millisecond tolerance should be reasonable. Bug: v8:11367 Change-Id: I2d284efe5176ec9938b6a378f4b5f1bda0e58781 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2659058 Commit-Queue: Omer Katz <omerkatz@chromium.org> Auto-Submit: Omer Katz <omerkatz@chromium.org> Reviewed-by:
Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/master@{#72426}
-
Mythri A authored
We use a heuristic that tiers up small functions at the first tick to optimize the small functions early. When tiering up from Turboprop it isn't important to tier up these functions quite early since they are already executing optimized code. Bug: v8:9684 Change-Id: Iaa647e0e03f0b4bf9cd0da7feb1e2d0e36004bc1 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656258Reviewed-by:
Sathya Gunasekaran <gsathya@chromium.org> Commit-Queue: Mythri Alle <mythria@chromium.org> Cr-Commit-Position: refs/heads/master@{#72425}
-
Thibaud Michaud authored
MiscField was still used in places where it should be replaced with the new AccessModeField, to avoid conflicts with the LaneSizeField. LaneSizeField is only used on arm64, but update other architectures too for consistency. Bug: v8:11098 Change-Id: Ib012ab7b1243f3bd586a8d824ca686d914f482bb Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2658322Reviewed-by:
-
Clemens Backes authored
Insertion into the map is expensive, hence avoid inserting every single code object. Instead, collect them in a {std::vector}, and only insert them when the {owned_code_} map is being used. By sorting the vector before inserting into the map, we can make most insertions constant time instead of logarithmic in the size of the map, by using the previous insert position as a hint for the next one. Drive-by: Remove an unneeded {WasmCodeRefScope}. R=thibaudm@chromium.org Bug: v8:11164 Change-Id: I3cc47f627eca40ea747d3e8388f93094650bbe19 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656259Reviewed-by: -
Marja Hölttä authored
Fix 1: Track Scope::needs_home_object and Scope::uses_super_property accurately. When "eval" is seen, figure out whether it can access "super" and if yes, set the corresponding home object as needed. Fix 2: The object literal scope shouldn't be entered for things inside spreads. Original: https://chromium-review.googlesource.com/c/v8/v8/+/2563275 Previous reland: https://chromium-review.googlesource.com/c/v8/v8/+/2637220 This saves memory (the home object doesn't need to be stored for each method, but only once per class) and hopefully makes the home object a constant in the optimized code. Detailed documentation of the changes: https://docs.google.com/document/d/1ZVXcoQdf9IdMsnRI9iyUjyq9NDoEyx9nA3XqMgwflMs/edit?usp=sharing Bug: v8:9237 Bug: chromium:1167918 Bug: chromium:1167981 Bug: chromium:1167988 Bug: chromium:1168055 Bug: chromium:1171195 Bug: chromium:1171600 Change-Id: I9686e0d90cd0c1128757eca440a88748897ee91e Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2655509 Commit-Queue: Marja Hölttä <marja@chromium.org> Reviewed-by:
Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#72422}
-
Marat Dukhan authored
i8x16.popcnt uses PSHUFB instruction, which is slow on the old Atom processors. Add an extra i8x16.popcnt implementation for those using HACKMEM-inspired divide-and-conquer algorithm. R=zhin@chromium.org, gdeepti@chromium.org Change-Id: I4e130428fea8c3cf3be1bd6da7308fc752b2132a Bug: v8:11002 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656858Reviewed-by:
Zhi An Ng <zhin@chromium.org> Reviewed-by:
Deepti Gandluri <gdeepti@chromium.org> Commit-Queue: Marat Dukhan <maratek@google.com> Cr-Commit-Position: refs/heads/master@{#72421}
-
v8-ci-autoroll-builder authored
Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/5e38be7..c083992 Rolling v8/buildtools: https://chromium.googlesource.com/chromium/src/buildtools/+log/e3db55b..9904915 Rolling v8/third_party/depot_tools: https://chromium.googlesource.com/chromium/tools/depot_tools/+log/9a18e37..558a304 Rolling v8/third_party/icu: https://chromium.googlesource.com/chromium/deps/icu/+log/2eefd9a..f097906 Rolling v8/tools/clang: https://chromium.googlesource.com/chromium/src/tools/clang/+log/0d2b891..cc8d00a TBR=machenbach@chromium.org,tmrts@chromium.org,v8-waterfall-sheriff@grotations.appspotmail.com Change-Id: Id6c419c461504e350c2b1f6bdf97943e583f7bc6 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2658835Reviewed-by:
v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Commit-Queue: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Cr-Commit-Position: refs/heads/master@{#72420}
-
Liu Yu authored
Port: 3965dcd5 Change-Id: I84064ebcffde490a44a341f0e65a3715746ca693 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2658655Reviewed-by:
Zhao Jiazhong <zhaojiazhong-hf@loongson.cn> Commit-Queue: Zhao Jiazhong <zhaojiazhong-hf@loongson.cn> Auto-Submit: Liu yu <liuyu@loongson.cn> Cr-Commit-Position: refs/heads/master@{#72419}
-
Frank Tang authored
Make sure there are space between day and hour in the pattern when we alternate date pattern for the change of hour cycle. Bug: chromium:1170305 Change-Id: I2714111dcdedeefafdb854d1684f301786273303 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2654002 Commit-Queue: Frank Tang <ftang@chromium.org> Reviewed-by:
Frank Tang <ftang@chromium.org> Reviewed-by:
-
Frank Tang authored
Bug: v8:11350 Change-Id: Ic34b40c4d88d6c2a0ac62bbebee4e2a95ebba826 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2648973 Commit-Queue: Frank Tang <ftang@chromium.org> Reviewed-by:
-
- 28 Jan, 2021 21 commits
-
-
Andrew Comminos authored
Ensure that we don't concurrently modify the StringsStorage map when getting a copy of a string. Bug: v8:11054 Change-Id: I6ad61838d7c5e8a6e9ff21aac04da8d353e41ad5 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2648821Reviewed-by:
Peter Marshall <petermarshall@chromium.org> Commit-Queue: Andrew Comminos <acomminos@fb.com> Cr-Commit-Position: refs/heads/master@{#72416}
-
Deepti Gandluri authored
PostMessage of an ArrayBuffer that is not detachable should result in a DataCloneError. Bug: chromium:1170176, chromium:961059 Change-Id: Ib89bbc10d2b58918067fd1a90365cad10a0db9ec Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2653810Reviewed-by:
Adam Klein <adamk@chromium.org> Reviewed-by:
Andreas Haas <ahaas@chromium.org> Commit-Queue: Deepti Gandluri <gdeepti@chromium.org> Cr-Commit-Position: refs/heads/master@{#72415}
-
Junliang Yan authored
Change-Id: I7f1d5e39033957410a8f3601100c7b7c5839271f Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2657475Reviewed-by:
Milad Fa <mfarazma@redhat.com> Commit-Queue: Junliang Yan <junyan@redhat.com> Cr-Commit-Position: refs/heads/master@{#72414}
-
Michael Lippautz authored
Allows for prohibiting GCs and will result in a crash in case a GC finalization event is triggered. Complements NoGarbageCollectionScope which ignores GC finalization events. Bug: chromium:1056170 Change-Id: Ie2a72a8675462b24692225af17c8f284318337ba Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656260Reviewed-by:
-
Michael Lippautz authored
Embedders forward the Value in TraceEphemeron as Member reference (as depicted in the API docs). Add TraceTrait<Member<T>> that forwards to TraceTrait<T> accordingly, supporting the intended use case. Bug: chromium:1056170 Change-Id: I3b247cb3553ae34d9ff5393aefeaec24068e78c2 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656255 Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Reviewed-by:
-
Marja Hölttä authored
The original commit implementing private accessor propertiers (*) claims it's not a thing, but it is. (*) https://chromium-review.googlesource.com/c/v8/v8/+/1695205/11/src/interpreter/bytecode-generator.cc#3959 Bug: v8:11360, v8:8330 Change-Id: If497f2b6a77dc28e4ade4ef78d901299f4e37593 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2652495Reviewed-by:
Joyee Cheung <joyee@igalia.com> Commit-Queue: Marja Hölttä <marja@chromium.org> Cr-Commit-Position: refs/heads/master@{#72411}
-
Clemens Backes authored
1) Wasm code is not associated with a Code object, hence WasmFrame::unchecked_code will always return a null object. Hence we can use the default implementation from TypedFrame and avoid the lookup on the heap which will always fail. 2) InternalFrame inherits from TypedFrame, hence can also reuse the unchecked_code implementation from TypedFrame. 3) Use "{}" instead of "Code()" to return "nothing". R=jkummerow@chromium.org Bug: v8:11074 Change-Id: I142d2f21c05bf87cafa5ba6e7f463510be6c70bb Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2653229Reviewed-by: -
Camillo Bruni authored
Fix filename creation when printing RCS to files. Change-Id: I04929366496ce2ac52d96540c2d41f753b293f24 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2650209Reviewed-by:
Mythri Alle <mythria@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#72409}
-
Manos Koukoutos authored
As per the latest wasm-gc spec, type checks are only allowed from subtypes of dataref and funcref. To cast from a more general type, specialized classification instructions need to be used (will come later). Bug: v8:7748 Change-Id: I29de48f445d652c5fc4e266d82e7d2e32cd7b6d3 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2649262 Commit-Queue: Manos Koukoutos <manoskouk@chromium.org> Reviewed-by:
-
Jakob Gruber authored
They've started failed, and no work is planned for the foreseeable future. Bug: v8:8888 Change-Id: I89dfa8f972a5bffa2bbb09c7a6ca56a0c4da9a02 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656316 Commit-Queue: Jakob Gruber <jgruber@chromium.org> Auto-Submit: Jakob Gruber <jgruber@chromium.org> Reviewed-by:
-
Milad Fa authored
Port 3965dcd5 Original Commit Message: This very slightly improves the performance of bytecode array visitors. R=verwaest@chromium.org, joransiu@ca.ibm.com, junyan@redhat.com, midawson@redhat.com BUG= LOG=N Change-Id: I7952fcacc72f4166c73885fd715950b319458209 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656455Reviewed-by:
Junliang Yan <junyan@redhat.com> Commit-Queue: Milad Fa <mfarazma@redhat.com> Cr-Commit-Position: refs/heads/master@{#72406}
-
Manos Koukoutos authored
As per the latest wasm-gc spec, all immediate arguments are removed from ref.cast and ref.test. Bug: v8:7748 Change-Id: I5839103276c4c5d51f3fa82a21cf2447cbb8ecaa Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2649261 Commit-Queue: Manos Koukoutos <manoskouk@chromium.org> Reviewed-by:
-
Camillo Bruni authored
ALmost all tools have migrated to .mjs modules. Bug: v8:10667 Change-Id: I95f7c4a31a721be3000c990bdac1c4eb0779b693 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2642460Reviewed-by:
Victor Gomes <victorgomes@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#72404}
-
Benedikt Meurer authored
This adds a [[WebAssemblyMemory]] internal property to ArrayBuffer and SharedArrayBuffer instances that are owned by WebAssembly.Memory objects. This allows the devtools-frontend to find the WebAssembly.Memory for any given ArrayBuffer, making it possible to properly support WebAssembly.memory.grow() eventually, but also showing a reasonable tab title. Before: https://imgur.com/hod9jPR.png After: https://imgur.com/v195VoC.png Bug: chromium:1171621, chromium:1171619, chromium:1166577 Change-Id: Ife22cabdfcf54ab30c234ea4ca86bfbb711ab2f1 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2653155 Commit-Queue: Benedikt Meurer <bmeurer@chromium.org> Reviewed-by:
Kim-Anh Tran <kimanh@chromium.org> Reviewed-by:
-
Benedikt Meurer authored
Previously we had the debugger / inspector support declared in debug-interface.h, but the implementation was sprinkled all across api.cc, which was quite messy. This moves the relevant macros and other bits into api-macros.h (with api-macros-undef.h to support jumbo builds), and moves the debugger interface implementation to src/debug/debug-interface.cc. Bug: chromium:1162229 Change-Id: If2698cba7bcc0b54b0f889220588ec214405848b Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2656256 Commit-Queue: Benedikt Meurer <bmeurer@chromium.org> Commit-Queue: Yang Guo <yangguo@chromium.org> Reviewed-by:
Yang Guo <yangguo@chromium.org> Auto-Submit: Benedikt Meurer <bmeurer@chromium.org> Cr-Commit-Position: refs/heads/master@{#72402}
-
Thibaud Michaud authored
This is a reland of a80d51d4 There were two main issues: - Liftoff did 32 bit addition to get the offset, which overflowed and created a valid address where an OOB was expected - The access mode and landing pad was missing for SIMD load and store instructions Also fixed the macros to disable trap handlers from simulator builds. Original change's description: > [wasm][mac][arm64] Enable OOB trap handler > > R=ahaas@chromium.org,mark@chromium.org,mseaborn@chromium.org > > Bug: v8:11098 > Change-Id: Ic4eb02a96805e49da71f301269567a6e0ac1b843 > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2519555 > Commit-Queue: Thibaud Michaud <thibaudm@chromium.org> > Reviewed-by: Zhi An Ng <zhin@chromium.org> > Reviewed-by: Igor Sheludko <ishell@chromium.org> > Reviewed-by: Andreas Haas <ahaas@chromium.org> > Cr-Commit-Position: refs/heads/master@{#72136} Bug: v8:11098 Change-Id: I2ebe434577f2cbe168f0bccf2936f69ca211e3c1 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2637223Reviewed-by:
-
Dominik Inführ authored
10ms is most likely too short, this also forces us to rename those metrics. Also rename the corresponding tracing events. Bug: v8:10315 Change-Id: I42fb1572b150f15cf12c2b02444f015fae349344 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2652494Reviewed-by:
-
Manos Koukoutos authored
As per latest wasm-gc spec, the data heap type is a subtype of eq and a supertype of all array and struct types. The heap type expected for arrays and structs when interacting with JS changes from eq to data. Bug: v8:7748 Change-Id: Idd1670b9e47acc95c098559e674c629ea44ca49d Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2649044 Commit-Queue: Manos Koukoutos <manoskouk@chromium.org> Reviewed-by:
-
Chong Gu authored
Needed since default manifest assignment will be put into the test() template in https://chromium-review.googlesource.com/c/chromium/src/+/2596114 Bug: chromium:1132878 Change-Id: I2475280c590f67cc88bcf6e877cab347079d734c Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2650765 Commit-Queue: Wez <wez@chromium.org> Reviewed-by:
Wez <wez@chromium.org> Reviewed-by:
-
Clemens Backes authored
The "noexcept" annotation is part of the type since C++17, so we cannot just change v8.h to add the annotation everywhere. So instead ignore missing V8_NOEXCEPT annotations in api.cc R=marja@chromium.org CC=jdoerrie@chromium.org No-Try: true Bug: v8:8616 Change-Id: I8f68a6d20653ee8c0e3d995cf585b36cbc15c5f0 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2655446 Commit-Queue: Clemens Backes <clemensb@chromium.org> Reviewed-by:
Marja Hölttä <marja@chromium.org> Reviewed-by:
-
Michael Lippautz authored
This allows construction and destruction of empty Persistent and friends, which simplifiest the use for embedders. Bug: chromium:1056170 Change-Id: I4286639aa5d50f9f98654b859de10bb80cbada21 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2655505 Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Reviewed-by:
-
