- 25 Oct, 2018 17 commits
-
-
Toon Verwaest authored
Bug: v8:8363, v8:7926 Change-Id: Ib4e39c9b23cf0733c189480ba8fbe1df27c85395 Reviewed-on: https://chromium-review.googlesource.com/c/1299084Reviewed-by:
Igor Sheludko <ishell@chromium.org> Commit-Queue: Toon Verwaest <verwaest@chromium.org> Cr-Commit-Position: refs/heads/master@{#56986}
-
Toon Verwaest authored
Since tokens turn illegal automatically upon failure, we can delay checking in places where we'll return upon illegal token anyway. We still need checks before uses of a possible invalid expression (nullptr in parser), so keep those checks for now; possibly pushing a previously higher check to right before the dereference. Bug: v8:8363 ,v8:7926 Change-Id: Ia402fc83c15f4c3258a335614b9213f8211a3c53 Reviewed-on: https://chromium-review.googlesource.com/c/1299081Reviewed-by:
-
Jakob Gruber authored
This is a reland of a31a6230 Original change's description: > [ia32] Remove poisoning logic on ia32 > > Poisoning has been disabled by default on ia32 a while ago. This CL > removes its logic from ia32 code generation, which will let us move > towards fuller (and unconditional) root register support. > > Bug: chromium:860429, v8:8254 > Change-Id: I8f672cf48a6ffc7bf21e7794c1b7463d7f8b9594 > Reviewed-on: https://chromium-review.googlesource.com/c/1296131 > Commit-Queue: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Michael Starzinger <mstarzinger@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56978} Tbr: mstarzinger@chromium.org,jarin@chromium.org Bug: chromium:860429, v8:8254 Change-Id: Ia65ac57fdc6b9a0f59cc64455d6a000005e9be3b Reviewed-on: https://chromium-review.googlesource.com/c/1299080Reviewed-by:
Jakob Gruber <jgruber@chromium.org> Commit-Queue: Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#56984}
-
Clemens Hammacher authored
This CL extracts some functionality out of the {PrepareAndStartCompile} step, in order to reuse that from the {AsyncStreamingProcessor}. We currently schedule a {PrepareAndStartCompile} task to get the same effect, and rely on the internal implementation to do the right thing. R=ahaas@chromium.org Bug: v8:8238 Change-Id: I43135fe488a5f72c09307ac955381c69b7987ec1 Reviewed-on: https://chromium-review.googlesource.com/c/1297321Reviewed-by:Andreas Haas <ahaas@chromium.org> Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Cr-Commit-Position: refs/heads/master@{#56983}
-
Jakob Gruber authored
PrepareElementsForSort must return a number less than or equal the array length. Bug: chromium:897512, v8:7382 Change-Id: If5f9c4d052e623ab9f3300b8534603abbee859fa Reviewed-on: https://chromium-review.googlesource.com/c/1297958 Commit-Queue: Jakob Gruber <jgruber@chromium.org> Reviewed-by:
Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#56982}
-
Georg Neis authored
Bug: v8:7790 Change-Id: I842615412368b37a1be4a5dfff7ba37f781f7906 Reviewed-on: https://chromium-review.googlesource.com/c/1297965Reviewed-by:
Maya Lekova <mslekova@chromium.org> Reviewed-by:
Jaroslav Sevcik <jarin@chromium.org> Commit-Queue: Georg Neis <neis@chromium.org> Cr-Commit-Position: refs/heads/master@{#56981}
-
Michael Achenbach authored
This reverts commit a31a6230. Reason for revert: https://ci.chromium.org/p/v8/builders/luci.v8.ci/V8%20Linux%20-%20nosnap%20-%20debug/21332 Original change's description: > [ia32] Remove poisoning logic on ia32 > > Poisoning has been disabled by default on ia32 a while ago. This CL > removes its logic from ia32 code generation, which will let us move > towards fuller (and unconditional) root register support. > > Bug: chromium:860429, v8:8254 > Change-Id: I8f672cf48a6ffc7bf21e7794c1b7463d7f8b9594 > Reviewed-on: https://chromium-review.googlesource.com/c/1296131 > Commit-Queue: Jakob Gruber <jgruber@chromium.org> > Reviewed-by: Michael Starzinger <mstarzinger@chromium.org> > Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56978} TBR=mstarzinger@chromium.org,jarin@chromium.org,jgruber@chromium.org Change-Id: I305e9e1719fb4b3f8ef267c232723db9b52966e9 No-Presubmit: true No-Tree-Checks: true No-Try: true Bug: chromium:860429, v8:8254 Reviewed-on: https://chromium-review.googlesource.com/c/1299015Reviewed-by:
Michael Achenbach <machenbach@chromium.org> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#56980}
-
Ross McIlroy authored
Resets the isolate's context to nullptr in debug builds during bytecode finalization to ensure that we don't rely on the native context during context independent unoptimized compilation. BUG=chromium:898076, v8:8041 Change-Id: Ifaa5006a7a3d31d7fbd535ebb63f8889c75526c4 Reviewed-on: https://chromium-review.googlesource.com/c/1297961 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by:
Leszek Swirski <leszeks@chromium.org> Cr-Commit-Position: refs/heads/master@{#56979}
-
Jakob Gruber authored
Poisoning has been disabled by default on ia32 a while ago. This CL removes its logic from ia32 code generation, which will let us move towards fuller (and unconditional) root register support. Bug: chromium:860429, v8:8254 Change-Id: I8f672cf48a6ffc7bf21e7794c1b7463d7f8b9594 Reviewed-on: https://chromium-review.googlesource.com/c/1296131 Commit-Queue: Jakob Gruber <jgruber@chromium.org> Reviewed-by:
Michael Starzinger <mstarzinger@chromium.org> Reviewed-by:
-
Clemens Hammacher authored
The flag was only there to tell whether the {AsyncCompileJob} needs to be kept alive. We already have this information in all the other fields of the {AsyncCompileJob}, thus remove it. R=ahaas@chromium.org Bug: v8:8238 Change-Id: I8d1d76ba8d622d1816c240e7a824ecf31c3b1ce5 Reviewed-on: https://chromium-review.googlesource.com/c/1297957Reviewed-by: -
Peter Marshall authored
This is a reland of c92a1dda Original change's description: > [cpu-profiler] Fix a bug which caused a pure virtual function call > > We need to remove each Sampler from the SamplerManager before we call > the Sampler destructor. This is because the signal handler can interrupt > the destructor, and call DoSampler(), which calls sampler->SampleStack() > on the sampler being destructed, causing general unhappiness and > "Pure virtual function called!" crashes. > > Bug: v8:8346, v8:5193 > Change-Id: Iaa595a196eab33fb1af31584e9a68fd1ce0a18f6 > Reviewed-on: https://chromium-review.googlesource.com/c/1293949 > Commit-Queue: Peter Marshall <petermarshall@chromium.org> > Reviewed-by: Alexei Filippov <alph@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56882} TBR=yangguo@chromium.org Bug: v8:8346, v8:5193 Change-Id: I9878f65c868ff1aed6f3a587cba688c4241bad8c Reviewed-on: https://chromium-review.googlesource.com/c/1298893Reviewed-by:
Peter Marshall <petermarshall@chromium.org> Commit-Queue: Peter Marshall <petermarshall@chromium.org> Cr-Commit-Position: refs/heads/master@{#56976}
-
Sergiy Byelozyorov authored
R=machenbach@chromium.org Bug: chromium:865541 Change-Id: I98f21c278099bd2c90b4a1ff9b7dddb74d263e34 Reviewed-on: https://chromium-review.googlesource.com/c/1144923 Commit-Queue: Sergiy Byelozyorov <sergiyb@chromium.org> Reviewed-by:
-
Toon Verwaest authored
It's unnecessary to buffer in the zone, and using a deque is more expensive than an std::vector as a stack since we reuse areas very frequently. The top-of-stack that the State keeps track of is now simply tracked in the state, with a scope_depth_ counter to figure out if the fni_ is "open" (has an active state). Change-Id: I29ad3db7520340b8fe035feed400178bd50785bc Reviewed-on: https://chromium-review.googlesource.com/c/1298894Reviewed-by:
-
Igor Sheludko authored
Bug: v8:8182 Change-Id: I4dadd9cab071ecd4314c370be5f444e36acb708e Reviewed-on: https://chromium-review.googlesource.com/c/1297317Reviewed-by:
-
Florian Sattler authored
Replace the explicit ok tracing by setting the scanner to fail, allowing us to return automatically. RETURN_IF_PARSE_ERROR is now used instead of CHECK_OK to verify if the parser failed. In a follow-up CL we'll merge RETURN_IF_PARSE_ERROR after Expect* into an EXPECT* macro. We'll keep (for now) RETURN_IF_PARSE_ERROR that guard uses of possible NullExpression (e.g., impl()->IsIdentifier(...)). All other RETURN_IF* will be removed. Uses after failure can likely later be fixed too by introducing a FailureExpression. Bug: v8:8363 ,v8:7926 Change-Id: I9896449eb9be476c453da4417a0bfd17c169ff38 Reviewed-on: https://chromium-review.googlesource.com/c/1294649 Commit-Queue: Florian Sattler <sattlerf@google.com> Reviewed-by:
Marja Hölttä <marja@chromium.org> Reviewed-by:
Toon Verwaest <verwaest@chromium.org> Cr-Commit-Position: refs/heads/master@{#56972}
-
Michael Achenbach authored
Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/6455acf..5a371bc Rolling v8/test/wasm-js/data: https://chromium.googlesource.com/external/github.com/WebAssembly/spec/+log/7e3c46a..b0e7838 Rolling v8/third_party/catapult: https://chromium.googlesource.com/catapult/+log/b026043..ed6fe0f Rolling v8/third_party/depot_tools: https://chromium.googlesource.com/chromium/tools/depot_tools/+log/98f1e59..2b71832 Rolling v8/third_party/fuchsia-sdk: https://chromium.googlesource.com/chromium/src/third_party/fuchsia-sdk/+log/bac0433..29de0c2 Rolling v8/tools/clang: https://chromium.googlesource.com/chromium/src/tools/clang/+log/a245b95..3041f30 Rolling v8/tools/luci-go: https://chromium.googlesource.com/chromium/src/tools/luci-go/+log/445d7c4..86c09e8 TBR=machenbach@chromium.org,hablich@chromium.org,sergiyb@chromium.org Change-Id: Ieb48aa2ff019267e3be8d3f70e60829d545d34c3 Reviewed-on: https://chromium-review.googlesource.com/c/1298391Reviewed-by:
v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#56971}
-
Kai Ninomiya authored
Check for is_javascript in Debug::AllFramesOnStackAreBlackboxed, instead of assuming all frames are javascript frames. Fix is thanks to dgozman: https://crbug.com/896093#c3 Bug: chromium:896093 Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng Change-Id: I6b735d904c3fd036d4589d65f10673b81b8f326c Reviewed-on: https://chromium-review.googlesource.com/c/1295552Reviewed-by:
Yang Guo <yangguo@chromium.org> Reviewed-by:
Aleksey Kozyatinskiy <kozyatinskiy@chromium.org> Commit-Queue: Kai Ninomiya <kainino@chromium.org> Cr-Commit-Position: refs/heads/master@{#56970}
-
- 24 Oct, 2018 23 commits
-
-
Junliang Yan authored
R=joransiu@ca.ibm.com Change-Id: I154b3cac455d44a84f7d6363758841310010d749 Reviewed-on: https://chromium-review.googlesource.com/c/1296683 Commit-Queue: Junliang Yan <jyan@ca.ibm.com> Reviewed-by:
Joran Siu <joransiu@ca.ibm.com> Cr-Commit-Position: refs/heads/master@{#56969}
-
Junliang Yan authored
R=joransiu@ca.ibm.com Change-Id: Ic9a6a9055e3f05ef64723d995228673ee48eb994 Reviewed-on: https://chromium-review.googlesource.com/c/1296682Reviewed-by:
-
Peter Marshall authored
This would allow SP to be a nonsense value, which would confuse the CPU profiler when the signal handler read the SP before it was restored. Bug: v8:8355 Change-Id: If108c8cf00467904fe0f4f26fddc2e3122afcd93 Reviewed-on: https://chromium-review.googlesource.com/c/1298032Reviewed-by:
Martyn Capewell <martyn.capewell@arm.com> Reviewed-by:
Alexei Filippov <alph@chromium.org> Commit-Queue: Peter Marshall <petermarshall@chromium.org> Cr-Commit-Position: refs/heads/master@{#56967}
-
Toon Verwaest authored
This allows us to use one underlying buffer rather than 3, and allows memory to be freed as we're growing the vector. Change-Id: I45c178a31e8f6d3ee44d3319ce8bca2db2460d33 Reviewed-on: https://chromium-review.googlesource.com/c/1297328Reviewed-by:
-
Tom Tan authored
This is a reland of fcbb023b Original change's description: > Add Windows ARM64 ABI support to V8 > > This change added Windows ARM64 ABI support, major things are: > 1. Excluding x18 register from any usage because it is reserved as > platform register. Preserve alignment after the change. > 2. Fix the assumption of LP64 in arm64 backend. Windows ARM64 is > still LLP64. > 3. Stack guard page probe for large allocation on stack. > > Reference: > Windows ARM64 ABI: > https://docs.microsoft.com/en-us/cpp/build/arm64-windows-abi-conventions?view=vs-2017 > > Bug: chromium:893460 > Change-Id: I325884ac8dab719154a0047141e18a9fcb8dff7e > Reviewed-on: https://chromium-review.googlesource.com/c/1285129 > Commit-Queue: Michael Achenbach <machenbach@chromium.org> > Reviewed-by: Andreas Haas <ahaas@chromium.org> > Reviewed-by: Michael Lippautz <mlippautz@chromium.org> > Reviewed-by: Benedikt Meurer <bmeurer@chromium.org> > Reviewed-by: Ulan Degenbaev <ulan@chromium.org> > Cr-Commit-Position: refs/heads/master@{#56881} CQ_INCLUDE_TRYBOTS=luci.chromium.try:android_arm64_dbg_recipe TBR=mlippautz@chromium.org Bug: chromium:893460 Change-Id: Icc45fd091c33f7df805842a70236b79b14756f52 Reviewed-on: https://chromium-review.googlesource.com/c/1297300 Commit-Queue: Michael Achenbach <machenbach@chromium.org> Reviewed-by:
Ulan Degenbaev <ulan@chromium.org> Reviewed-by:
-
Jakob Kummerow authored
This CL gives a first look at the new way to represent tagged object pointers in C++. It adds infrastructure in Handles and the garbage collector to deal with the new object type, and ports a first class to the new world. Design overview: https://goo.gl/Ph4CGz Bug: v8:3770 Change-Id: I3e37fbf399612f95540cb386710a595069fb9d55 Reviewed-on: https://chromium-review.googlesource.com/c/1292673Reviewed-by:
-
Frank Tang authored
Add position to the return of %SegmentIterator%.next() which newly added to the spec in https://github.com/tc39/proposal-intl-segmenter/pull/42 Bug: v8:8305 Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng Change-Id: I8de7102acb670a6c529ab3e35601c78a8dc7703c Reviewed-on: https://chromium-review.googlesource.com/c/1278636Reviewed-by:
Sathya Gunasekaran <gsathya@chromium.org> Reviewed-by:
Daniel Ehrenberg <littledan@chromium.org> Commit-Queue: Frank Tang <ftang@chromium.org> Cr-Commit-Position: refs/heads/master@{#56963}
-
Georg Neis authored
Bug: v8:7790 Change-Id: I14bac46ef7457ea142f79f96fc5a2018d429dcc8 Reviewed-on: https://chromium-review.googlesource.com/c/1297323 Commit-Queue: Georg Neis <neis@chromium.org> Reviewed-by:
-
Frank Tang authored
Remove TODO that is already done Uncomment two working tests. Bug: v8:5751 Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng Change-Id: Icb28d84e92812996c2928e90961d75508ba4c401 Reviewed-on: https://chromium-review.googlesource.com/c/1296933Reviewed-by:
-
Dan Elphick authored
Creates the hash_seed byte array in RO_SPACE and moves the root from STRONG_MUTABLE_IMMOVABLE_ROOT_LIST to STRONG_READ_ONLY_ROOT_LIST. Bug: v8:8191 Change-Id: I3b044fbb3e51eb5d21ac2e68a54076623865b9d2 Reviewed-on: https://chromium-review.googlesource.com/c/1297959 Commit-Queue: Dan Elphick <delphick@chromium.org> Reviewed-by:
-
Aseem Garg authored
R=clemensh@chromium.org,kozyatinskiy@chromium.org Cq-Include-Trybots: luci.chromium.try:linux_chromium_headless_rel;master.tryserver.blink:linux_trusty_blink_rel Change-Id: Ic6c7e2eaa4463d945d00eb1e1123d7d1731b34db Reviewed-on: https://chromium-review.googlesource.com/c/1297671Reviewed-by:
Clemens Hammacher <clemensh@chromium.org> Reviewed-by:
-
Ross McIlroy authored
BuildClassBoilerplate accessed the native context to get the class_function_descriptors. Baseline compilation should be native context independent, so we shouldn't access the native context at all. As it happens, class_function_descriptors wasn't used so can just be removed. BUG=chromium:898076, v8:8041 Change-Id: If9c0edf3dfde68c76ea87820f9d4b080aac6d60e Reviewed-on: https://chromium-review.googlesource.com/c/1298033Reviewed-by:
-
Caitlin Potter authored
There are several core changes in this stub: 1) add a version of KeyedStoreGenericGenerator::SetPropertyInLiteral() which supports indexed properties directly, witthout KeyedStore 2) add a code stub for SetPropertyInLiteral which uses the version supporting indexed properties 3) Use the code stub in CloneObjectIC, rather than using the smaller special-cased version which does not handle Names. Item 1) involves a refactoring which adds a nice way to reuse code in KeyedStoreGenericAssembler, which allows deleting a bunch of copy/pasted code. This makes it easy to reuse the index handling in KeyedStoreGeneric() without adding adding a bunch more duplicated handling. Because of this, I consider this to be somewhat of a cleanup, though if the copied code is preferred, I'm happy to revert to that. Item 2) is needed for Object.fromEntries(), as it's better to not require falling back to the slow path if a key happens to be an Smi --- but this is also optional. Item 3) benefits the codebase by allowing Object.fromEntries() to use this fast path without calling into the runtime, and without duplicating code which is also used by CloneObjectIC. I am skeptical that this should affect performance significantly. I've run ObjectLiteralSpread tests, and the mean of scores over 100 runs is somewhat surprising: CloneObjectIC --- the only user of this code, has an increased average score, while the polyfill cases score slightly worse --- However, the overall changes are small and likely flukes. The complete processed test output is below: ``` // Mean of 100 runs of each benchmark Babel-ObjectLiteralSpread: -----+---------------------------+---------------------------+------- | With patch | Without patch | diff Mean | 11530.87 | 12142.92 | -5.04% -----+---------------------------+---------------------------+------- BabelAndOverwrite-ObjectLiteralSpread: -----+---------------------------+---------------------------+------- | With patch | Without patch | diff Mean | 10881.41 | 11260.81 | -3.37% -----+---------------------------+---------------------------+------- ObjectAssign-ObjectLiteralSpread: -----+---------------------------+---------------------------+------- | With patch | Without patch | diff Mean | 6188.92 | 6358.55 | -2.67% -----+---------------------------+---------------------------+------- ObjectAssignAndOverwrite-ObjectLiteralSpread: -----+---------------------------+---------------------------+------- | With patch | Without patch | diff Mean | 6112.80 | 6275.54 | -1.61% -----+---------------------------+---------------------------+------- ObjectSpread-ObjectLiteralSpread: -----+---------------------------+---------------------------+------- | With patch | Without patch | diff Mean | 51942.93 | 50713.17 | +3.46% -----+---------------------------+---------------------------+------- ObjectSpreadAndOverwrite-ObjectLiteralSpread: -----+---------------------------+---------------------------+------- | With patch | Without patch | diff Mean | 51375.23 | 50833.29 | +2.09% -----+---------------------------+---------------------------+------- ``` BUG=v8:8238, v8:8021 R=ishell@chromium.org, jkummerow@chromium.org Change-Id: I43e102fc461ffd389b5d6810a73f86e5012d7dee Reviewed-on: https://chromium-review.googlesource.com/c/1277751 Commit-Queue: Caitlin Potter <caitp@igalia.com> Reviewed-by:Jakob Kummerow <jkummerow@chromium.org> Cr-Commit-Position: refs/heads/master@{#56957}
-
Toon Verwaest authored
Bug: v8:8365 Change-Id: Ie938073551bf1af6fb59ac1c395e7fabbcfdebd7 Reviewed-on: https://chromium-review.googlesource.com/c/1298034Reviewed-by:
-
Michael Starzinger authored
This adds back another register to the record write stub to have one additional register on top of the parameter register as allocation general purpose register. It has only been recently reduced to just four registers due to embedded builtins. This is needed to be able to tail call a record write stub. R=ulan@chromium.org CC=jgruber@chromium.org BUG=v8:8341 Change-Id: Id16f9e96d611a871fbe1180581eaf14275a7332e Reviewed-on: https://chromium-review.googlesource.com/c/1297955Reviewed-by:
-
Sathya Gunasekaran authored
Change-Id: I8ce540dcd1dd5384f96d1c47c9784fdfb0933c1e Reviewed-on: https://chromium-review.googlesource.com/c/1298029Reviewed-by:
Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org> Cr-Commit-Position: refs/heads/master@{#56954}
-
Lei Zhang authored
GN should understand action outputs, so the header generated by the run_torque action do not need to be separately listed in another source_set. Change-Id: I309e8c012eb0a0597a247806d36658c1d6e5d97b Reviewed-on: https://chromium-review.googlesource.com/c/1297680Reviewed-by:
Tobias Tebbi <tebbi@chromium.org> Commit-Queue: Lei Zhang <thestig@chromium.org> Cr-Commit-Position: refs/heads/master@{#56953}
-
Hai Dang authored
Bug: v8:7980 Change-Id: Ic4c72b02c196b296105a6ddf9c3af9fb699ef8c5 Reviewed-on: https://chromium-review.googlesource.com/c/1297327Reviewed-by:
Georg Neis <neis@chromium.org> Commit-Queue: Hai Dang <dhai@google.com> Cr-Commit-Position: refs/heads/master@{#56952}
-
Michael Starzinger authored
R=ulan@chromium.org BUG=v8:8238 Change-Id: Idf6b3d4035b392dd1b20ff3e4cbdb60cdaada054 Reviewed-on: https://chromium-review.googlesource.com/c/1297325Reviewed-by:
-
Jaroslav Sevcik authored
Bug: v8:5495, v8:8361 Change-Id: I7a03c7a4897b15112b978d232754076ad8753c4e Reviewed-on: https://chromium-review.googlesource.com/c/1297311Reviewed-by:
-
Sergiy Byelozyorov authored
R=machenbach@chromium.org No-Try: true No-Tree-Checks: true Bug: chromium:892433 Change-Id: Id323739be44ea55d73c712059520d7f5e684c97e Reviewed-on: https://chromium-review.googlesource.com/c/1280304Reviewed-by:
-
Benedikt Meurer authored
This changes the ReceiverOrOddball feedback on JSStrictEqual to ReceiverOrNullOrUndefined feedback, which can also safely be consumed by JSEqual (we cannot generally accept any oddball here since booleans trigger implicit conversions, unfortunately). Thus we replace the previously introduced CheckReceiverOrOddball with CheckReceiverOrNullOrUndefined, and drop CheckOddball, since we will no longer collect Oddball feedback separately. TurboFan will then turn a JSEqual[ReceiverOrNullOrUndefined] into a sequence like this: ``` left = CheckReceiverOrNullOrUndefined(left); right = CheckReceiverOrNullOrUndefined(right); result = if ObjectIsUndetectable(left) then ObjectIsUndetectable(right) else ReferenceEqual(left, right); ``` This significantly improves the peak performance of abstract equality with Receiver, Null or Undefined inputs. On the test case outlined in http://crbug.com/v8/8356 we go from naive: 2946 ms. tenary: 2134 ms. to naive: 2230 ms. tenary: 2250 ms. which corresponds to a 25% improvement on the abstract equality case. For regular code this will probably yield more performance, since we get rid of the JSEqual operator, which might have arbitrary side effects and thus blocks all kinds of TurboFan optimizations. The JSStrictEqual case is slightly slower now, since it has to rule out booleans as well (even though that's not strictly necessary, but consistency is key here). This way developers can safely use `a == b` instead of doing a dance like `a == null ? b == null : a === b` (which is what dart2js does right now) when both `a` and `b` are known to be Receiver, Null or Undefined. The abstract equality is not only faster to parse than the tenary, but also generates a shorter bytecode sequence. In the test case referenced in http://crbug.com/v8/8356 the bytecode for `naive` is ``` StackCheck Ldar a1 TestEqual a0, [0] JumpIfFalse [5] LdaSmi [1] Return LdaSmi [2] Return ``` which is 14 bytes, whereas the `tenary` function generates ``` StackCheck Ldar a0 TestUndetectable JumpIfFalse [7] Ldar a1 TestUndetectable Jump [7] Ldar a1 TestEqualStrict a0, [0] JumpIfToBooleanFalse [5] LdaSmi [1] Return LdaSmi [2] Return ``` which is 24 bytes. So the `naive` version is 40% smaller and requires fewer bytecode dispatches. Bug: chromium:898455, v8:8356 Change-Id: If3961b2518b4438700706b3bd6071d546305e233 Reviewed-on: https://chromium-review.googlesource.com/c/1297315Reviewed-by: -
Michael Achenbach authored
Also use low experiment percentage on CQ, since the builder's output is unused. NOTRY=true TBR=sergiyb@chromium.org Bug: chromium:830557 Change-Id: Id024ab16e2944ec5e94b0209672ed6b77ae322a8 Reviewed-on: https://chromium-review.googlesource.com/c/1296466Reviewed-by:
Sergiy Byelozyorov <sergiyb@chromium.org> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#56947}
-
