We need a deopt point for the case when we fail to find an initial
element from which to begin the reduction step.
Bug: v8:7384
Change-Id: I5e476ddc433be690577677b018639c4c0c70809b
Reviewed-on: https://chromium-review.googlesource.com/906508
Reviewed-by: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51146}
This is the last piece of the TypedArray constructors that was still
written in JS.
Bug: v8:7102
Change-Id: I7c4dc867b09408caa4eec2873ea7185b6c61a525
Reviewed-on: https://chromium-review.googlesource.com/888751
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51122}
This creates a uniform PerformPromiseThen builtin, which performs the
operation with the same name from the spec, except that it expects the
handlers to be either undefined or callable already, since this is only
relevant for a single callsite (namely Promise.prototype.then).
Introduce a matching operator JSPerformPromiseThen into TurboFan, which
represents this operation and removes the additional checks in case of
Promise.prototype.then based on the information we can derived from the
receiver maps.
This yields a nice 20-25% improvement on Promise.prototype.then, as
illustrated by the following micro-benchmark
```js
const N = 1e7;
function inc(x) { return x + 1; }
function chain(promise) {
return promise.then(inc).then(value => {
if (value < N) chain(Promise.resolve(value));
});
}
console.time('total');
chain(Promise.resolve(0));
setTimeout(console.timeEnd.bind(console, 'total'));
```
which goes from around 1230ms to 930ms with this patch.
Bug: v8:7253
Change-Id: I5712a863acdbe7da3bb8e621887c7b952148c51a
Reviewed-on: https://chromium-review.googlesource.com/899064
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51071}
Instead of inlining NewPromiseCapability everywhere, we should have it
as a separate builtin that can be called. The motivation for doing so is
to reduce the overhead of the builtins and just share the slow-path
handling in a single place, plus this way we can also utilize the
NewPromiseCapability functionality from TurboFan optimized code
eventually.
Also remove the unused %new_promise_capability intrinsic and the builtin
backing it.
Bug: v8:7253
Change-Id: Ic7c916a11f12d063a645e2c25cc4120badee5640
Reviewed-on: https://chromium-review.googlesource.com/897804
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51036}
There's now only a single caller to InternalPromiseThen left,
which is the Promise.prototype.then implementation, so there's
no need to have a separate helper function.
Bug: v8:7253
Change-Id: I0e1ea674c942f735dd069137182232f34d16a729
Reviewed-on: https://chromium-review.googlesource.com/897762
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51031}
This implements the ideas outlined in the section "Microtask queue"
of the exploration document "Promise and async/await performance" (at
https://goo.gl/WHRar2), except that the microtask queue stays a linear
FixedArray for now, to avoid running into trouble with the parallel
scavenger. This way we can already save a significant amount of
allocations, thereby reducing the GC frequency quite a bit.
All items on the microtask queue are now proper structs that subclass
Microtask, i.e. we also wrap JSFunction and MicrotaskCallback jobs
into structs. We also consistently remember the context for every
microtask (except for MicrotaskCallback where we don't have a
context), and execute it later in exactly that context (as required
by the spec anyways for the Promise related jobs). Particularly
interesting is the PromiseReactionJobTask and its subclasses, since
they are designed to have the same size as the PromiseReaction. When
we resolve a JSPromise we just take the existing PromiseReaction
instances and morph them into PromiseFulfillReactionJobTask or
PromiseRejectReactionJobTask (depending whether you "Fulfill" or
"Reject"). That way the JSPromise class is now only 6 words instead
of 10 words.
Also the PromiseReaction and the reaction tasks can either carry a
JSPromise (for the fast native case) or a PromiseCapability (for the
generic case), which means we don't always pay the overhead of having
to also remember the "deferred resolve" and "deferred reject" handlers
that are only relevant for the generic case anyways.
It also fixes a spec violation where we called "then" before we actually
enqueued the PromiseResolveThenableJob, which is observably wrong.
Calling it later has the advantage that it should be fairly
straight-forward now to completely avoid it for native Promise
instances.
This seems to save around 10-20% on the various Promise benchmarks and
micro-benchmarks. We expect to gain even more as we're now able to
inline various operations into TurboFan optimized code easily.
Bug: v8:7253
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: I893d24ca5bb046974b4f5826a8f6dd22f1210b6a
Reviewed-on: https://chromium-review.googlesource.com/892819
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50980}
- Port TypedArray.prototype.slice to CSA
- Implement TypedArraySpeciesCreateByLength as a CSA
- Fix spec bugs: Throw if a source typed array is neutered
after creating a result typed array
Bug: v8:5929
Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
Change-Id: Ia7ce2239d37db6db172c00aa120ef51c31a14bac
Reviewed-on: https://chromium-review.googlesource.com/830991
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50952}
This adjusts the RunMicrotask logic to invoke CallHandlerInfo microtasks
from CSA land directly (via a runtime function call), instead of bailing
out to C++ for the rest of the microtask queue entries. Even in simple
micro-benchmarks there doesn't seem to be a huge performance difference.
In fact performance get's better when CallHandlerInfo and promises are
mixed, which makes sense, since calling from C++ to JS land is more
expensive than the other way around.
But just in case the runtime function call overhead ever becomes the
bottleneck we can introduce a direct C++ call and setup a handle scope
around it, much like a very simple version of CallApiFunctionStub.
This greatly simplifies the microtask handling and paves the way for
refactoring the queue to significant reduce the GC overhead associated
with promises currently.
Bug: v8:7253
Change-Id: I33adb62a6bada138674d324f36d4be894e27f3c9
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Reviewed-on: https://chromium-review.googlesource.com/890441
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50934}
This CL uses UTF16 encoding internally in the string iterator,
thereby saving a few shifts, ors, and ands.
Bug: v8:7270
Change-Id: I3ac9e0e8c4b64eb1d6c796597eb0b3413c5f516b
Reviewed-on: https://chromium-review.googlesource.com/887085
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Sigurd Schneider <sigurds@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50904}
which was dropped from the proposal:
https://github.com/tc39/proposal-bigint/pull/101
Bug: v8:6791
Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
Change-Id: I5a3e0024900b34a322a23bcbc4bfdbdffbf8b9d4
Reviewed-on: https://chromium-review.googlesource.com/885510
Commit-Queue: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50877}
- Remove TypedArray.prototype.subarray in js/typedarray.js
- Implement TypedArray.prototype.subarray as a CSA
- Implement TypedArraySpeciesCreateByArrayBuffer as a CSA
- Move a helper function for relative index from builtins-string-gec.cc
to code-stub-assembler.cc
- Move SpeciesConstructor from builtins-promise-gen.cc to
code-stub-assembler.cc
Bug: v8:7161, v8:5929
Change-Id: If3340476e16aa21659540eb4b24e3ead54e6a313
Reviewed-on: https://chromium-review.googlesource.com/830992
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50831}
This reverts commit 03e9d415c2.
Reason for revert: Correctness issues, see https://crbug.com/804159.
Bug: chromium:804159
Original change's description:
> Reland: Reimplement Object.entries/values as CSA to optimize performance.
>
> Add Object.entries/values builtins to debug-evaluate.cc whitelist macro.
> This fix revert commit of https://chromium-review.googlesource.com/c/v8/v8/+/859937
> Original is https://chromium-review.googlesource.com/c/v8/v8/+/810504
> >> Reimplements Object.entries/values as CSA to optimize performance. See more detail about https://bugs.chromium.org/p/v8/issues/ Issue 6804.
>
> This reverts commit 1b49f725ac.
>
> Bug: v8:6804
> Change-Id: I57e8b66e1c4ece2abb52e1630a97fbfd4070d810
> Reviewed-on: https://chromium-review.googlesource.com/860679
> Commit-Queue: Yang Guo <yangguo@chromium.org>
> Reviewed-by: Yang Guo <yangguo@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#50492}
TBR=yangguo@chromium.org,cbruni@chromium.org,jgruber@chromium.org,ishell@chromium.org,brn@b6n.ch
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: v8:6804
Change-Id: I39b1854ca7c2f57819ba377f84560356d3756bfb
Reviewed-on: https://chromium-review.googlesource.com/877886
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50746}
This avoids the call to the StringCharCodeAt builtin from
within TurboFan optimized code and instead emits a loop
that does the character load. This (together with previously
reverted CL to the JSCallReducer) almost completely recovers
the performance regression caused when we shipped TurboFan.
Without untrusted code mitigations the benchmark goes from
580ms to roughly 490ms, and with the patch to the JSCallReducer
the time goes down to 280ms, which is very close to what we
had with Crankshaft.
This also renames the LoadFromString helper method in the
EffectControlLinearizer to LoadFromSeqString to make it
clear what it does.
Bug: v8:7326
Change-Id: I6c77209ae01a3eacbd1e8fd40e4ad842eaf1999a
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Reviewed-on: https://chromium-review.googlesource.com/876102
Commit-Queue: Sigurd Schneider <sigurds@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50725}
Until now, String.prototype.{trimLeft,trimRight} were non-standard
language extensions, required for Web compatibility.
The proposal at https://github.com/tc39/proposal-string-left-right-trim
standardizes this functionality as String.prototype.{trimStart,trimEnd},
and defines String.prototype.{trimLeft,trimRight} as aliases for
backwards compatibility.
This patch implements that proposal behind the --harmony-string-trimming
flag.
Bug: v8:6530
Change-Id: Id21e624c12a79e6b782efb049a48901b9da7db71
Reviewed-on: https://chromium-review.googlesource.com/867044
Commit-Queue: Mathias Bynens <mathias@chromium.org>
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50717}
This reverts commit 93d67d20ee.
Reason for revert: several layout test failures:
https://build.chromium.org/p/client.v8.fyi/builders/V8-Blink%20Linux%2064/builds/21062
Original change's description:
> [turbofan] Inline StringCharCodeAt like Crankshaft did.
>
> This avoids the call to the StringCharCodeAt builtin from
> within TurboFan optimized code and instead emits a loop
> that does the character load. This (together with previously
> reverted CL to the JSCallReducer) almost completely recovers
> the performance regression caused when we shipped TurboFan.
>
> Without untrusted code mitigations the benchmark goes from
> 580ms to roughly 490ms, and with the patch to the JSCallReducer
> the time goes down to 280ms, which is very close to what we
> had with Crankshaft.
>
> This also renames the LoadFromString helper method in the
> EffectControlLinearizer to LoadFromSeqString to make it
> clear what it does.
>
> Bug: v8:7326
> Change-Id: Ibe0ec1847911a234f244bd8dcec6be18b241fda0
> Reviewed-on: https://chromium-review.googlesource.com/873376
> Reviewed-by: Yang Guo <yangguo@chromium.org>
> Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#50702}
TBR=yangguo@chromium.org,bmeurer@chromium.org
Change-Id: I6e909adba82adc46e269ab14426ee24caaca6ff9
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:7326
Reviewed-on: https://chromium-review.googlesource.com/875963
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50709}
This avoids the call to the StringCharCodeAt builtin from
within TurboFan optimized code and instead emits a loop
that does the character load. This (together with previously
reverted CL to the JSCallReducer) almost completely recovers
the performance regression caused when we shipped TurboFan.
Without untrusted code mitigations the benchmark goes from
580ms to roughly 490ms, and with the patch to the JSCallReducer
the time goes down to 280ms, which is very close to what we
had with Crankshaft.
This also renames the LoadFromString helper method in the
EffectControlLinearizer to LoadFromSeqString to make it
clear what it does.
Bug: v8:7326
Change-Id: Ibe0ec1847911a234f244bd8dcec6be18b241fda0
Reviewed-on: https://chromium-review.googlesource.com/873376
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50702}
This removes the Javascript version of Array.of in js/array.js and adds
a CodeStubAssembler version in src/builtins/builtins-array-gen.cc.
Mostly this change is for code-health reasons but it also gives
performance improvements for nearly all cases with the exception of
"transplanted" arrays. E.g.
function ArrayLike {}
ArrayLike.of = Array.of
ArrayLike.of(...) is now slower in the perf tests. Most of this change
can be attributed to using CallRuntime(kSetProperty,...) to set the
length. The JS version can do better due to inline caches trained on
the same datatype for 1000s of iterations, but this kind of workload is
unlikely.
Change-Id: I18e5b19b185257e9e0d553e1183b40ba4a5d3289
Reviewed-on: https://chromium-review.googlesource.com/863625
Commit-Queue: Dan Elphick <delphick@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50680}
This is needed to easily port the constructor dispatcher to CSA.
Bug: v8:7102
Change-Id: I9672416495940ca12088a2980a9ecc61364aef9d
Reviewed-on: https://chromium-review.googlesource.com/785630
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50671}
This reverts commit d30a8fa9b4.
Reason for revert: no-snap test failures here
https://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20nosnap%20-%20debug/builds/17068
You need to update the whitelist in src/debug/debug-evaluate.cc.
I'm a bit surprised this only happens in no-snap builds.
Original change's description:
> Reimplement Object.entries/values as CSA to optimize performance.
>
> This implementation based on runtime implementation.
>
> Bug: v8:6804
> Change-Id: Ib8bfcc4648e44a999789237effc0275c5e4d9936
> Reviewed-on: https://chromium-review.googlesource.com/810504
> Commit-Queue: Camillo Bruni <cbruni@chromium.org>
> Reviewed-by: Camillo Bruni <cbruni@chromium.org>
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#50477}
TBR=cbruni@chromium.org,jgruber@chromium.org,ishell@chromium.org,brn@b6n.ch
Change-Id: I1a0c8e3c054a57ca4d15f7a064ff4b28ca133b16
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:6804
Reviewed-on: https://chromium-review.googlesource.com/859937
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50478}
Inlining the EnqueueMicrotask logic into the various uses blows up the
snapshot size significantly. So instead of doing that we just turn the
operation into a dedicated builtin that we call from the various uses.
This still avoids the runtime function call overhead and maintains the
fast path without write barriers for the common case of the microtask
queue fitting into new space.
This also moves back the microtask helper CSA functions to the
specialized assembler.
Bug: v8:7253, chromium:799563
Change-Id: I2d24d0e5c01e442c5ad7f5d4373fbc6e94351ac5
Reviewed-on: https://chromium-review.googlesource.com/856618
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50461}
This changes the names of the builtins on the Promise.prototype to match
the general naming convention and make it easier to tell what the
builtin is about just by looking at its name.
Bug: v8:7253
Change-Id: I906cc6fc14a6b9d8ffe9c0dcfd94b50439c4cfcf
Reviewed-on: https://chromium-review.googlesource.com/850698
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50367}
... to ease its inlining to the bytecode handlers.
The new code organisation still don't produce unwanted frame creation
code on a fast path.
Bug: v8:7206, chromium:576312
Change-Id: Ib516ae0795ff1788b3a7e0bb521f72dfa68444f0
Reviewed-on: https://chromium-review.googlesource.com/833869
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50257}
This reduces reduces code size by 16 KB while keeping the fast path (in
which the given argument is already a smi) inlined and extracting
remaining logic to a stub call.
Change-Id: I531999c990519eef1247cc3785ad4b16164f7a5e
Reviewed-on: https://chromium-review.googlesource.com/833912
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50224}
This is a reland of ab38b03d1b
Original change's description:
> [builtins] Port Object.p.toLocaleString to CSA from JS
>
> - Added ObjectPrototypeToLocaleString TFJ
> - Remove v8natives.js
> - Move GetMethod and GetIterator into prologue.js
>
> TBR=adamk@chromium.org
>
> Bug: v8:6005
> Change-Id: I2b5b65892304e62bf64375458f8ffb9473b2c9b7
> Reviewed-on: https://chromium-review.googlesource.com/826479
> Reviewed-by: Peter Wong <peter.wm.wong@gmail.com>
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Commit-Queue: Peter Wong <peter.wm.wong@gmail.com>
> Cr-Commit-Position: refs/heads/master@{#50120}
Cq-Include-Trybots: master.tryserver.blink:linux_trusty_blink_rel;master.tryserver.chromium.linux:linux_chromium_rel_ng
TBR=adamk@chromium.org
Bug: v8:6005
Change-Id: Ie8c8810c5231e933e61ea8babe963e58bb6dcaed
Reviewed-on: https://chromium-review.googlesource.com/831156
Reviewed-by: Peter Wong <peter.wm.wong@gmail.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Peter Wong <peter.wm.wong@gmail.com>
Cr-Commit-Position: refs/heads/master@{#50218}
Support inlining Array.prototype.findIndex in Turbofan.
Depending on array size, quick benchmarks show a >2x
improvement: https://github.com/peterwmwong/v8-perf/blob/master/array-find-findIndex-tf/README.md
Bug: chromium:791045, v8:1956, v8:7165
Change-Id: I250554885f924c97b0072e09ee289713df5cbe63
Reviewed-on: https://chromium-review.googlesource.com/824382
Commit-Queue: Peter Wong <peter.wm.wong@gmail.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50133}
- Added ObjectPrototypeToLocaleString TFJ
- Remove v8natives.js
- Move GetMethod and GetIterator into prologue.js
TBR=adamk@chromium.org
Bug: v8:6005
Change-Id: I2b5b65892304e62bf64375458f8ffb9473b2c9b7
Reviewed-on: https://chromium-review.googlesource.com/826479
Reviewed-by: Peter Wong <peter.wm.wong@gmail.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Peter Wong <peter.wm.wong@gmail.com>
Cr-Commit-Position: refs/heads/master@{#50120}
This CL also removes LoadICProtoArray* builtins which are no longer necessary.
Bug: v8:7206, v8:5561
Change-Id: Ic5d9a3d4d21c4bd5e5e1cd110bd029ced157a000
Reviewed-on: https://chromium-review.googlesource.com/819252
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50104}
The dispatcher is responsible for handling stores to lexical environment
variables and for storing directly to the JSGlobalObject. In the latter
case the dispatcher also ensures that JSGlobalProxy is provided as
a receiver if a setter function has to be called.
Unlike StoreIC the calling convention for the StoreGlobalIC does not include
receiver.
Bug: v8:7206, chromium:576312, v8:5561
Change-Id: Ifa896c7b41bf440785b757c2272ec91211e79c98
Reviewed-on: https://chromium-review.googlesource.com/818965
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50081}
- Removes JS implementation and InnerArrayFind/InnerArrayFindIndex
- Adds TFJ, with TFS for slow continuation path
Some quick benchmarks show ~2x improvement for unoptimized code
and up to 16% improvement against optimized code (diminishes with
larger arrays as iterating dominates).
https://github.com/peterwmwong/v8-perf/blob/master/array-find-findIndex/README.md
Bug: chromium:791045, v8:1956, v8:5049, v8:7165
Change-Id: Ie16252ed495bbd91fe548b16d5ef6764de791a50
Reviewed-on: https://chromium-review.googlesource.com/804704
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49851}
- Implement RunMicrotasks in CSA to prevent a potentially large number
of jumps between C++ and JS code while consuming te queue. Appears to
provide a ~60% speedup in microtask-heavy code, which from limited
testing appears to scale linearly.
The code-stub microtask pump bails out to the old C++ microtask pump
if it encounters a CallHandlerInfo microtask, and remains in C++ for
the remainder of the queue (returning to the JS/stub implementation
after the bailed out queue is exhausted).
- Add a variation of JSEntryStub which enters the new RunMicrotasks code
stub.
- Add a new RunMicrotasks helper to Execution, which uses the
RunMicrotasks entry stub.
Bug:
Change-Id: I4667d4dd633d24455ea5d7cef239da0af1a7365e
Reviewed-on: https://chromium-review.googlesource.com/650486
Commit-Queue: Caitlin Potter <caitp@igalia.com>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49842}
This reduces the overhead of calling the builtin.
Quick measurements show >5x improvement. As the
typed array's size grows, iterating dominates
and the performance gap closes.
https://github.com/peterwmwong/v8-perf/blob/master/typedarray-findIndex/README.md
Bug: v8:5929
Change-Id: I27d67776c83cbe28f4f9f5ef479a7eeabf594654
Reviewed-on: https://chromium-review.googlesource.com/792394
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49720}
This reduces the overhead of calling the builtin.
Quick measurements show >5x improvement. As the
typed array's size grows, iterating dominates
and the performance gap closes.
https://github.com/peterwmwong/v8-perf/blob/master/typedarray-find/README.md
Bug: v8:5929
Change-Id: Ia74546bb46d446c6161c8956e350d4b5cdc1b328
Reviewed-on: https://chromium-review.googlesource.com/792454
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49706}
Prior to this change, the exponentiation operator was rewritten by the
parser to a call of the Math.pow builtin. However, Math.pow does not
accept BigInt arguments, while the exponentiation operator must accept
them.
This CL
- removes the parser's special treatment of ** and **=, treating them
like any other binary op instead.
- adds a TFC builtin Exponentiate that does the right thing for
all inputs.
- adds interpreter bytecodes Exp and ExpSmi whose handlers call the
Exponentiate builtin. For simplicity, they currently always collect
kAny feedback.
- adds a Turbofan operator JSExponentiate with a typed-lowering to
the existing NumberPow and a generic-lowering to the Exponentiate
builtin. There is currently no speculative lowering.
Note that exponentiation for BigInts is actually not implemented yet,
so we can't yet test it.
Bug: v8:6791
Change-Id: Id90914c9c3fce310ce01e715c09eaa9f294f4f8a
Reviewed-on: https://chromium-review.googlesource.com/785694
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Mythri Alle <mythria@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49696}
... in the same style as the previous CLs for negation and bitwise-not.
R=jarin@chromium.org
Bug: v8:6791
Change-Id: I0aa96a72421e90c8c82a39dd4264fdcf00967504
Reviewed-on: https://chromium-review.googlesource.com/779141
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49658}
- Fast path for same type source typed array
- Move previous CPP implementation into a runtime function "TypedArraySet"
- Remove parts covered by the TFJ
- Basic receiver, offset, source checks
- Handling of same type source typed array
Bug: v8:3590
Change-Id: I0f19d961424c30cc8bbcb8648b623e7e6dfa33f4
Reviewed-on: https://chromium-review.googlesource.com/786414
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49608}
This introduces a JSBitwiseNot operator and lowers it either to a
speculative xor with -1 (when we have Number feedback) or to
a stub call. The stub is also new.
Bug: v8:6791
Change-Id: I362e52de8a741dc5db044c406543878e407eb2ed
Reviewed-on: https://chromium-review.googlesource.com/778839
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49539}
This introduces a JSNegate operator and lowers it either to a
speculative multiplication with -1 (when we have Number feedback)
or to a stub call. The stub is also new.
R=jarin@chromium.org
Bug: v8:6791
Change-Id: I8e20333fe49cc6088d2d10777be982e42eed2412
Reviewed-on: https://chromium-review.googlesource.com/774718
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49538}
- Remove weak-collection.js
- Adds TFJ builtins for WeakSet and WeakMap constructors
- Unified helpers and constructor behavior into a BaseCollectionsAssembler
- Fast paths for...
- unmodified constructor function
- argument is a fast JS array
- entries are fast JS arrays, for Map/WeakMap
- no arguments passed
Quick benchmarks shows significant improvements (1.12x - 5.7x!) for ALL collection constructors (weak and non-weak):
https://github.com/peterwmwong/v8-perf/blob/master/weakcollection-constructor/README.md
More could be done for performance. Currently we always call out to JS to add entries, if we knew the prototype was unmodified, we could call the builtins directly.
Bug: v8:5049, v8:6604
Change-Id: Id7912c1eed5bcf512df7fd6238f04166a8a5937e
Reviewed-on: https://chromium-review.googlesource.com/760385
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49343}
Previously StaNamedProperty and StaKeyedProperty were in a weird state where
they claimed to not touch the accumulator, but actually did in case they were
deopted in the middle. A frame was added in the middle to overwrite the
accumulator again with the right value before returning from the setter, using
a lot of complexity in the deoptimizer.
This changes those instructions to be marked as writing to the accumulator
(e.g., the result of the setter), and uses to manually store and reload into
the accumulator the value being stored.
If we want to avoid the additional bytecodes, we could make sure that bytecodes
that claim to leave the accumulator alone don't deopt back to Advance/Dispatch
but LoadAccumulatorWithValue/Advance/Dispatch. That's in a way similar to what
happened before this CL, but I believe could be implemented much simpler.
Bug:
Change-Id: I4850a690ef5a30976701d0e050951faa46fd1c18
Reviewed-on: https://chromium-review.googlesource.com/753487
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49159}
- Add WeakMapPrototypeDelete and WeakSetPrototypeDelete TFJ builtins
- Fast paths when it's not necessary to shrink the table
- Add WeakCollectionDelete TFS
Some quick benchmarks shows 1.4x - 2.15x gains in performance.
https://github.com/peterwmwong/v8-perf/blob/master/weakcollection-delete/README.md
Bug: v8:5049, v8:6604
Change-Id: I14036df153f3a0242f9083d751658b868b16660a
Reviewed-on: https://chromium-review.googlesource.com/743864
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49076}
This is preparation for supporting OOB loads from strings, which
requires the KeyedLoadIC to track this information.
Bug: v8:6936, v8:7014
Change-Id: Ide132244ee523397dd418d21fe3377976f6633fd
Reviewed-on: https://chromium-review.googlesource.com/743481
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49026}
We now represent the SameValue operation explicitly in TurboFan and the
operation can thus participate in all kinds of optimizations. Especially
we get rid of the JSCall node in the general case, which blocks several
optimizations across the call. The general, baseline performance is now
always on par with StrictEqual.
Once the StrictEqual operator is also a simplified operator, we should
start unifying the type based optimizations in SimplifiedLowering.
In the micro-benchmark we go from
testStrictEqual: 1422 ms.
testObjectIs: 1520 ms.
testManualSameValue: 1759 ms.
to
testStrictEqual: 1426 ms.
testObjectIs: 1357 ms.
testManualSameValue: 1766 ms.
which gives the expected result.
Bug: v8:7007
Change-Id: I0de3ff6ff6209ab4c3edb69de6a16e387295a9c8
Reviewed-on: https://chromium-review.googlesource.com/741228
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48994}
... mainly by adapting Object::Compare and
CodeStubAssembler::RelationalComparison.
TBR=rmcilroy@chromium.org
Change-Id: I34448d45b4950b9318263c4a667aa9db7d77232d
Bug: v8:6791
Reviewed-on: https://chromium-review.googlesource.com/730730
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48873}
Previously, V8's slice was implemented in a combination of C++ and a
Javascript fallback. The disadvantage of this approach was that the
fast-path required a call through the CEntryStub, which introduced
considerable overhead for small arrays with fast elements kinds.
Now the implementation primarily uses the CSA to generate both the
full spec-complaint implementation as well as fast paths for argument
objects and arrays with fast elements kinds. The CSA implementation
uses a C++ implementation fallback in select situations where the the
complexity of a CSA implementation would be too great and the
CEntryStub overhead is not decisive (e.g. slices of dictionary
elements arrays).
Performance results on semi-random arrays with small number of
elements (old vs. new):
smi copy: 48.7 ms vs. 12 ms
smi slice: 43.5 ms 14.8 ms
object copy: 35.5 ms 7.7 ms
object slice: 38.7 ms 8.8 ms
dictionary slice: 2398.3 ms vs. 5.4 ms
fast sloppy arguments slice: 9.6 ms vs. 7.2 ms
slow sloppy arguments slice: 28.9 ms vs. 8.5 ms
As a bonus, the new implementation is fully spec-compliant and fixes
at least one existing bug.
The design document for Array.prototype builtin rework can be found
at https://goo.gl/wFHe2n
Bug: v8:1956,v8:6601,v8:6710,v8:6978
Change-Id: Ia0155bedcf39b4577605ff754f416c2af938efb7
Reviewed-on: https://chromium-review.googlesource.com/574710
Commit-Queue: Daniel Clifford <danno@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48853}
Support inlining of Array.prototype.filter in TurboFan.
Bug: v8:1956
Change-Id: If50e230d14461063d378c0591dc27dea43371afa
Reviewed-on: https://chromium-review.googlesource.com/733089
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48846}
In the special case of KeyedLoadIC, where the key that is passed in is a
Name that is always the same we only checked for identity in both the
stub and the TurboFan case, which works fine for symbols and internalized
strings, but doesn't really work with non-internalized strings, where
the identity check will fail, the runtime will internalize the string,
and the IC will then see the original internalized string again and not
progress in the feedback lattice. This leads to tricky deoptimization
loops in TurboFan and constantly missing ICs.
This adds fixes the stub to always try to internalize strings first
when the identity check fails and then doing the check again. If the
name is not found in the string table we miss, since in that case the
string cannot match the previously recorded feedback name (which is
always a unique name).
In TurboFan we represent this checks with new CheckEqualsSymbol and
CheckEqualsInternalizedString operators, which validate the previously
recorded feedback, and the CheckEqualsInternalizedString operator does
the attempt to internalize the input.
Bug: v8:6936, v8:6948, v8:6969
Change-Id: I3f3b4a587c67f00f7c4b60d239eb98a9626fe04a
Reviewed-on: https://chromium-review.googlesource.com/730224
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48784}
This reverts commit 9fd029ef25.
Due to a ClusterFuzz issue.
TBR=mstarzinger@chromium.org
Bug: chromium:776511
Change-Id: I0f30e2e8de97f3c437a1756c82e645828358ad91
Reviewed-on: https://chromium-review.googlesource.com/730006
Reviewed-by: Michael Stanton <mvstanton@chromium.org>
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48775}
Support inlining of Array.prototype.filter in TurboFan.
(relanding with fix for chromium:766635, visible in the
diff between patchsets 2 and 3)
Bug: v8:1956,chromium:766635
Change-Id: Ia50be6770602513e3d91d17e2b2ca9d3b0e8b42a
Reviewed-on: https://chromium-review.googlesource.com/721119
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48697}
Port the baseline version of Reflect.has to the CodeStubAssembler and
reuse the existing logic for HasProperty (i.e. the HasProperty builtin).
Also inline the Reflect.has builtin into TurboFan, by adding a check
on the target in front of a use of the JSHasProperty operator.
Technically this additional check is not necessary, because the
JSHasProperty operator already throws if the target is not a JSReceiver,
but the exception message is confusing then.
This improves the performance of the micro-benchmark from
reflectHasPresent: 337 ms.
reflectHasAbsent: 472 ms.
to
reflectHasPresent: 121 ms.
reflectHasAbsent: 216 ms.
which is a nice 2.8x improvement in the best case. It also improves the
chai test on the web-tooling-benchmark by around 1-2%, which is roughly
the expected win (since Reflect.has overall accounts for around 3-4%).
Bug: v8:5996, v8:6936, v8:6937
Change-Id: I856183229677a71c19936f06f2a4fc7a794a9a4a
Reviewed-on: https://chromium-review.googlesource.com/720959
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48608}
- Extract core StringPrototypeRepeat code into a TFS builtin (StringRepeat)
- Assumes arguments are a string and smi (no range checks)
- Add StringPrototypePadStart and StringPrototypePadEnd TFJ builtins
- Added StringPadAssembler to ensure common behavior
- Removed functionality from string.js
A quick benchmark shows significant performance gains for unoptimized
code (2.1x to 2.46x) and optimized code (1.03x - 1.56x).
https: //github.com/peterwmwong/v8-perf/blob/master/string-pad/README.md
Bug: v8:5049
Change-Id: I6e4fe99fb62a3edb3d6906fd4f78b3576b5b0d13
Reviewed-on: https://chromium-review.googlesource.com/720067
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48595}
This switches all deoptimization events to go through use one single
continuation builtin (i.e. {Builtins::kNotifyDeoptimized}) instead of
handling builtin continuation specially. Fewer moving pieces.
R=jarin@chromium.org
Change-Id: Ic8a2316fa2f5c8717b4d50d1a619b87a38011564
Reviewed-on: https://chromium-review.googlesource.com/712156
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48496}
- Expose fast paths for RegExpPrototypeMatchBody/RegExpPrototypeSearchBody as TFS builtins
- Add StringPrototypeMatch and StringPrototypeSearch TFJ builtins
- Add StringMatchSearchAssembler to ensure same search/match behavior
- Remove functionality from string.js
A quick benchmark shows gains of 20-30% for unoptimized code and 0-20% for optimized code.
https://github.com/peterwmwong/v8-perf/blob/master/string-search-match/README.md
Bug: v8:5049
Change-Id: I0fffee6e94e62ecae049c9e5798da52d67ae1823
Reviewed-on: https://chromium-review.googlesource.com/707824
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48452}
This introduces a ToNumeric conversion to the runtime and interpreter.
ToNumeric behaves like ToNumber, except that it also lets BigInts pass.
Bug: v8:6791
Change-Id: Idf9d0b5d283638459fe5893de41cc120356247a7
Reviewed-on: https://chromium-review.googlesource.com/707013
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48440}
Rename the MapLookupHashIndex builtin to FindOrderedHashMapEntry and
also rename the TurboFan operators LookupHashStorageIndex and
LookupSigned32HashStorageIndex to FindOrderedHashMapEntry and
FindOrderedHashMapEntryForInt32Key respectively. This way the naming is
more consistent and it's immediately obvious from the operator name that
this operator deals with OrderedHashMaps, which wasn't clear before.
Also fix the result of the operation to be either -1 or the index of
the entry relative to the hash table start (that is, no longer eagerly
add hash table start plus value offset to the entry index). This removes
this non-foldable integer additon from TurboFan code for both Map#get
and Map#has.
Drive-by-fix: Also provide more concrete types for the
FindOrderedHashMapEntry operators.
Bug: v8:5049
Change-Id: I418d107b806f3031a52a525cffc20456dc2342db
Reviewed-on: https://chromium-review.googlesource.com/707414
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48404}
We use the naming convention <Constructor>Prototype<Method> for builtins
that implement methods on a certain builtin constructors prototype. Fix
the collection builtins (Map and Set) to match this naming convention.
Bug: v8:5049
Change-Id: I8ced50c2ac9ebc8f4390bcbbc6aec426a0026813
Reviewed-on: https://chromium-review.googlesource.com/707318
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48403}
Further optimize the
new Array(n)
expressions by also inlining the backing store allocation into TurboFan
optimized code. Currently these backing store allocations cannot be
eliminated by escape analysis and can also not be optimized further by
the memory optimizer (i.e. no allocation folding and not write barrier
elimination) yet, but this can be done in follow-up CLs.
This yields another ~5% improvement on the ARES6 ML benchmark (steady
state).
Drive-by-fix: Add support for loops to the GraphAssembler. This was
necessary to implement the initialization loops in for the backing
store allocations in the EffectControlLinearizer.
Bug: v8:6399, v8:6901
Change-Id: I759d6802db01eb797e78c7d82d82caaee3463e16
Reviewed-on: https://chromium-review.googlesource.com/705934
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48352}
Make calls like
new Array(n)
new A(n)
(where A is a subclass of Array) inlinable into TurboFan. We do this by
speculatively checking that n is an unsigned integer that is not greater
than JSArray::kInitialMaxFastElementArray, and then lowering the backing
store allocation to a builtin call. The speculative optimization is
either protected by the AllocationSite for the Array constructor
invocation (if we have one), or by a newly introduced global protector
cell that is used for Array constructor invocations that don't have an
AllocationSite, i.e. the ones from Array#map, Array#filter, or from
subclasses of Array.
Next step will be to implement the backing store allocations inline in
TurboFan, but that requires Loop support in the GraphAssembler, so it's
done as a separate CL. This should further boost the performance.
This boosts the ARES6 ML benchmark by up to 8% on the steady state,
and also improves monomorphic Array#map calls by around 20-25% on the
initial setup.
Bug: v8:6399
Tbr: ulan@chromium.org
Change-Id: I7c8bdecf7c814ce52db6ee3051c3206a4f7d4bb6
Reviewed-on: https://chromium-review.googlesource.com/704639
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Michael Stanton <mvstanton@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48348}
The Object.is builtin provides an entry point to the abstract operation
SameValue, which properly distinguishes -0 and 0, and also identifies
NaNs. Most of the time you don't need these, but rather just regular
strict equality, but when you do, Object.is(o, -0) is the most readable
way to check for minus zero.
This is for example used in Node.js by formatNumber to properly print -0
for negative zero. However since the builtin thus far implemented as C++
builtin and TurboFan didn't know anything about it, Node.js considering
to go with a more performant, less readable version (which also makes
assumptions about the input value) in
https://github.com/nodejs/node/pull/15726
until the performance of Object.is will be on par (so hopefully we can
go back to Object.is in Node 9).
This CL ports the baseline implementation of Object.is to CSA, which
is pretty straight-forward since SameValue is already available in
CodeStubAssembler, and inlines a few interesting cases into TurboFan,
i.e. comparing same SSA node, and checking for -0 and NaN explicitly.
On the micro-benchmarks we go from
testNumberIsMinusZero: 1000 ms.
testObjectIsMinusZero: 929 ms.
testObjectIsNaN: 954 ms.
testObjectIsSame: 793 ms.
testStrictEqualSame: 104 ms.
to
testNumberIsMinusZero: 89 ms.
testObjectIsMinusZero: 88 ms.
testObjectIsNaN: 88 ms.
testObjectIsSame: 86 ms.
testStrictEqualSame: 105 ms.
which is a nice 10x to 11x improvement and brings Object.is on par with
strict equality for most cases.
Drive-by-fix: Also refactor and optimize the SameValue check in the
CodeStubAssembler to avoid code bloat (by not inlining StrictEqual
into every user of SameValue, and also avoiding useless checks).
Bug: v8:6882
Change-Id: Ibffd8c36511f219fcce0d89ed4e1073f5d6c6344
Reviewed-on: https://chromium-review.googlesource.com/700254
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48275}
The TypedArray.prototype[Symbol.toStringTag] getter is currently the best (and
as far as I can tell only definitely side-effect free) way to check whether an
arbitrary object is a TypedArray - either generally TypedArray or a specific
one like Uint8Array. Using the getter is thus emerging as the general pattern
to detect TypedArrays, even Node.js now adapted it starting with
https://github.com/nodejs/node/pull/15663
for the isTypedArray and isUint8Array type checks in lib/internal/util/types.js
now.
The getter returns either the string with the TypedArray subclass name
(i.e. "Uint8Array") or undefined if the receiver is not a TypedArray.
This can be implemented with a simple elements kind dispatch, instead of
checking the instance type and then loading the class name from the
constructor, which requires a loop walking up the transition tree. This
CL ports the builtin to CSA and TurboFan, and changes the logic to a
simple elements kind check. On the micro-benchmark mentioned in the
referenced bug, the time goes from
testIsArrayBufferView: 565 ms.
testIsTypedArray: 2403 ms.
testIsUint8Array: 3847 ms.
to
testIsArrayBufferView: 566 ms.
testIsTypedArray: 965 ms.
testIsUint8Array: 965 ms.
which presents an up to 4x improvement.
Bug: v8:6874
Change-Id: I9c330b4529d9631df2f052acf023c6a4fae69611
Reviewed-on: https://chromium-review.googlesource.com/695021
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48254}
There are only very few custom compiled IC handlers left that go in there, and for each compiled handler we only have 1 cache hit on top25; maximally saving 60ms over 33s. Additionally we'll migrate the remaining handlers to data-driven handlers anyway. Let's try to remove this code.
Bug:
Change-Id: Ib874cc498015046a3ff67c83ea8b10b3c4eb7d0f
Reviewed-on: https://chromium-review.googlesource.com/668409
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48201}
When inlining based on SharedFunctionInfo rather than based on concrete
JSFunction, we weren't able to properly optimize array, object and
regexp literals inside the inlinee, because we didn't know the concrete
FeedbackVector for the inlinee inside JSCreateLowering. This was because
JSCreateLowering wasn't properly updated after the literals moved to the
FeedbackVector. Now with this CL we also have the VectorSlotPair on the
literal creation operators, just like we do for property accesses and
calls, and are thus able to always access the appropriate FeedbackVector
and optimize the literal creation.
The impact is illustrated by the micro-benchmark on the tracking bug,
which goes from
createEmptyArrayLiteral: 1846 ms.
createShallowArrayLiteral: 1868 ms.
createShallowObjectLiteral: 2246 ms.
to
createEmptyArrayLiteral: 1175 ms.
createShallowArrayLiteral: 1187 ms.
createShallowObjectLiteral: 1195 ms.
with this CL, so up to 2x faster now.
Drive-by-fix: Also remove the unused CreateEmptyObjectLiteral builtin
and cleanup the names of the other builtins to be consistent with the
names of the TurboFan operators and Ignition bytecodes.
Bug: v8:6856
Change-Id: I453828d019b27c9aa1344edac0dd84e91a457097
Reviewed-on: https://chromium-review.googlesource.com/680656
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48140}
Along with BigInt.prototype. Their functions only have skeleton
implementations. The purpose of this change is to make it easier
to gradually increase test coverage (e.g. for toString(radix)).
Of course this is still behind the --harmony-bigint flag.
Bug: v8:6791
Change-Id: Ic307fd9165c56ac782fba18d648ce893daaa718f
Reviewed-on: https://chromium-review.googlesource.com/671209
Commit-Queue: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48094}
Support inlining of Array.prototype.filter in TurboFan.
Bug: v8:1956
Change-Id: Iba4d683aaa86c6104e8a1cf4d0f549a0c516576a
Reviewed-on: https://chromium-review.googlesource.com/657021
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48040}
The spec calls out to Promise.prototype.then and also passes around
the constructor of the receiver to Promise.prototype.finally.
Adds a new constructor slot to PromiseFinallyContext enum and this is
used to create a new promise in the thenFinally/catchFinally callbacks.
Created a new PromiseResolve TFS builtin refactored from
the existing PromiseResolve builtin. PromiseResolveWrapper
calls out to this TFS Builtin and is now exposed as Promise.resolve.
The thenFinally and catchFinally callbacks also call out to the
PromiseResolve TFS builtin.
Spec -- https://tc39.github.io/proposal-promise-finally/
Bug: v8:5967
Change-Id: I2ce89f14d3b149619d11e424b6e37062e466c4d5
Reviewed-on: https://chromium-review.googlesource.com/652026
Reviewed-by: Georg Neis <neis@chromium.org>
Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47898}
This speeds up the baseline performance of Object by 20%.
With this change, the callViaObject when run with --noopt
goes from 10718ms to 8577ms on the benchmark from:
http://benediktmeurer.de/2017/08/31/object-constructor-calls-in-webpack-bundles
Bug: v8:6772
Change-Id: Id0e54ba44204a1700885185ec360e1c56834fb73
Reviewed-on: https://chromium-review.googlesource.com/654900
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47888}
There are two main reasons to move DeserializeLazy to ASM:
1. We avoid complications around the distinction between Call/Construct
cases by making sure relevant registers (e.g. new_target) remain
unclobbered.
2. We can avoid the tail-call through CodeFactory::Call/Construct by
jumping directly to the deserialized code object.
Bug: v8:6624
Change-Id: Idef8fa73d804e16d510f62766c735d1891729b81
Reviewed-on: https://chromium-review.googlesource.com/652472
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47876}
JS runtime calls are always created with undefined recievers, so make the
bytecode behave similarly to CallUndefinedReciever such that we don't need
to push an explicit undefined register for the receiver for such calls.
Modifies the Async[Generator/Function]Await[Caught/Uncaught] runtime calls
to pass the generator in the first argument rather than the reciever since
these runtime calls were desugered in the bytecode generator and explicitly
passed the generator in the receiver.
Change-Id: I36c8087bb3b663dccd805bfdb1eea04eb6a73269
Reviewed-on: https://chromium-review.googlesource.com/654257
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47870}
Only the error cases of overwriting readonly properties need the
language_mode to decide whether to throw or be silent. Reading it
from the feedback vector's metadata (just like the C++ code in
ic.cc does) removes the need to duplicate each stub for each
language_mode ("StoreIC" + "StoreICStrict" etc.).
Change-Id: Ic0c67f9d40ca36c65e41b4f162b2ab70d155e549
Reviewed-on: https://chromium-review.googlesource.com/647373
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Commit-Queue: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47836}
Always return to the InterpreterEntryTrampoline rather than calling the
InterpreterExitTrampoline from the Return bytecode handler. This fixes a
regression which occured if we upset the call/return stack by skipping the
return to the InterpreterEntryTrampoline from the return bytecode handler.
BUG=chromium:759390,chromium:753705
Change-Id: Ib625654a4a5072ac6c8d8e9611d1b9c0bbced4ca
Reviewed-on: https://chromium-review.googlesource.com/649517
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47826}
This CL removes the weak-list of JS functions from the context
and all the code that iterares over it. This list was being used
mainly during deoptimization (for code unlinking) and during
garbage collection. Removing it will improve performance of
programs that create many closures and trigger many scavenge GC
cycles.
No extra work is required during garbage collection. However,
given that we no longer unlink code from JS functions during
deoptimization, we leave it as it is, and on its next activation
we check whether the mark_for_deoptimization bit of that code is
set, and if it is, than we unlink it and jump to lazy compiled
code. This check happens in the prologue of every code object.
We needed to change/remove the cctests that used to check
something on this list.
Working in x64, ia32, arm64, arm, mips64 and mips.
Bug: v8:6637
Change-Id: Ica99a12fd0351ae985e9a287918bf28caf6d2e24
TBR: mstarzinger@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/647596
Commit-Queue: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47808}
This reverts commit 84c2dfce43.
Reason for revert:
https://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20nosnap%20-%20debug/builds/14876
Original change's description:
> Remove weak-list of optimized JS functions.
>
> This CL removes the weak-list of JS functions from the context
> and all the code that iterares over it. This list was being used
> mainly during deoptimization (for code unlinking) and during
> garbage collection. Removing it will improve performance of
> programs that create many closures and trigger many scavenge GC
> cycles.
>
> No extra work is required during garbage collection. However,
> given that we no longer unlink code from JS functions during
> deoptimization, we leave it as it is, and on its next activation
> we check whether the mark_for_deoptimization bit of that code is
> set, and if it is, than we unlink it and jump to lazy compiled
> code. This check happens in the prologue of every code object.
>
> We needed to change/remove the cctests that used to check
> something on this list.
>
> Working in x64, ia32, arm64, arm, mips64 and mips.
>
> Bug: v8:6637
> Change-Id: I7f192652c8034b16a9ea71303fa8e78cda3c48f3
> Reviewed-on: https://chromium-review.googlesource.com/600427
> Commit-Queue: Juliana Patricia Vicente Franco <jupvfranco@google.com>
> Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
> Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
> Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#47790}
TBR=mstarzinger@chromium.org,jarin@chromium.org,leszeks@chromium.org,bmeurer@chromium.org,jupvfranco@google.com
Change-Id: Ia4f1a8acf6ca5cd5c74266437a03d854b3739af2
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:6637
Reviewed-on: https://chromium-review.googlesource.com/647540
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47792}
This CL removes the weak-list of JS functions from the context
and all the code that iterares over it. This list was being used
mainly during deoptimization (for code unlinking) and during
garbage collection. Removing it will improve performance of
programs that create many closures and trigger many scavenge GC
cycles.
No extra work is required during garbage collection. However,
given that we no longer unlink code from JS functions during
deoptimization, we leave it as it is, and on its next activation
we check whether the mark_for_deoptimization bit of that code is
set, and if it is, than we unlink it and jump to lazy compiled
code. This check happens in the prologue of every code object.
We needed to change/remove the cctests that used to check
something on this list.
Working in x64, ia32, arm64, arm, mips64 and mips.
Bug: v8:6637
Change-Id: I7f192652c8034b16a9ea71303fa8e78cda3c48f3
Reviewed-on: https://chromium-review.googlesource.com/600427
Commit-Queue: Juliana Patricia Vicente Franco <jupvfranco@google.com>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47790}
This CL adds support to optimize for..in in fast enum-cache mode to the
same degree that it was optimized in Crankshaft, without adding the same
deoptimization loop that Crankshaft had with missing enum cache indices.
That means code like
for (var k in o) {
var v = o[k];
// ...
}
and code like
for (var k in o) {
if (Object.prototype.hasOwnProperty.call(o, k)) {
var v = o[k];
// ...
}
}
which follows the https://eslint.org/docs/rules/guard-for-in linter
rule, can now utilize the enum cache indices if o has only fast
properties on the receiver, which speeds up the access o[k]
significantly and reduces the pollution of the global megamorphic
stub cache.
For example the micro-benchmark in the tracking bug v8:6702 now runs
faster than ever before:
forIn: 1516 ms.
forInHasOwnProperty: 1674 ms.
forInHasOwnPropertySafe: 1595 ms.
forInSum: 2051 ms.
forInSumSafe: 2215 ms.
Compared to numbers from V8 5.8 which is the last version running with
Crankshaft
forIn: 1641 ms.
forInHasOwnProperty: 1719 ms.
forInHasOwnPropertySafe: 1802 ms.
forInSum: 2226 ms.
forInSumSafe: 2409 ms.
and V8 6.0 which is the current stable version with TurboFan:
forIn: 1713 ms.
forInHasOwnProperty: 5417 ms.
forInHasOwnPropertySafe: 5324 ms.
forInSum: 7556 ms.
forInSumSafe: 11067 ms.
It also improves the throughput on the string-fasta benchmark by
around 7-10%, and there seems to be a ~5% improvement on the
Speedometer/React benchmark locally.
For this to work, the ForInPrepare bytecode was split into
ForInEnumerate and ForInPrepare, which is very similar to how it was
handled in Fullcodegen initially. In TurboFan we introduce a new
operator LoadFieldByIndex that does the dynamic property load.
This also removes the CheckMapValue operator again in favor of
just using LoadField, ReferenceEqual and CheckIf, which work
automatically with the EscapeAnalysis and the
BranchConditionElimination.
Bug: v8:6702
Change-Id: I91235413eea478ba77ace7bd14bb2f62e155dd9a
Reviewed-on: https://chromium-review.googlesource.com/645949
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47768}
This adds support for lowering {JSCreateArguments} within outermost
frames of type {CreateArgumentsType::kMappedArguments}. It will hence
enable escape analysis to work with such objects and allow for further
optimization.
This also adds a new {NewMappedArgumentsElements} simplfied operator.
Note that escape analysis support for this new operator will be done as
a follow-up.
R=tebbi@chromium.org
Change-Id: I0e2fac25c654f796433f57b116964053b6b68635
Reviewed-on: https://chromium-review.googlesource.com/641454
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47761}
This adds an initial implementation of the DeserializeLazy builtin and
runtime function, as well as --lazy-deserialization and
--trace-lazy-deserialization feature flags.
Since lazy deserialization itself isn't implemented yet, DeserializeLazy
simply replaces itself with the appropriate builtin. The builtin_id is
loaded from the SFI, and the builtin itself is loaded from the Builtins
table.
Bug: v8:6624
Change-Id: I4ef8c3030a8cda19a086b8e569a24d97213b5ed8
Reviewed-on: https://chromium-review.googlesource.com/643289
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47757}
This simplfies the code for lowering of {JSCreateArguments} nodes under
the assumption that deoptimization support is always enabled, and that
arguments objects are only materialized for JavaScript frames and not
for internal stub frames.
R=tebbi@chromium.org
Change-Id: I5f86ae0f0442a03b516904d737c5a0eac293b5b9
Reviewed-on: https://chromium-review.googlesource.com/640381
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47685}
Optimize the common pattern
for (var i in o) {
if (Object.prototype.hasOwnProperty.call(o, i)) {
// do something
}
}
which is part of the guard-for-in style in ESLint (see the documentation
at https://eslint.org/docs/rules/guard-for-in for details). This pattern
also shows up in React and Ember applications quite a lot (and is tested
by the appropriate Speedometer benchmarks, although not dominating those
benchmarks, since they spent a lot of time in non-TurboFan'ed code).
This improves the forInHasOwnProperty and forInHasOwnPropertySafe micro-
benchmarks in v8:6702, which look like this
function forInHasOwnProperty(o) {
var result = 0;
for (var i in o) {
if (o.hasOwnProperty(i)) {
result += 1;
}
}
return result;
}
function forInHasOwnPropertySafe(o) {
var result = 0;
for (var i in o) {
if (Object.prototype.hasOwnProperty.call(o, i)) {
result += 1;
}
}
return result;
}
by around 4x and allows for additional optimizations in the future, by
also elimiating the megamorphic load when accessing the enumerated
properties.
This changes the interpreter ForInNext bytecode to collect more precise
feedback about the for-in state, which now consists of three individual
states: UNINITIALIZED, MEGAMORPHIC and GENERIC. The MEGAMORPHIC state
means that the ForInNext has only seen objects with a usable enum cache
thus far, whereas GENERIC means that we have seen some slow-mode for..in
objects as well.
R=jarin@chromium.org
Bug: v8:6702
Change-Id: Ibcd75ea9b58c3b4f9219f11bc37eb04a2b985604
Reviewed-on: https://chromium-review.googlesource.com/636964
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47632}
This change adapts the Call bytecode handlers such that they don't require
a stack frame. It does this by modifying the call bytecode handler to
tail-call the Call or InterpreterPushArgsAndCall builtins. As a result, the
callee function will return to the InterpreterEntryTrampoline when it returns
(since this is the return address on the interpreter frame), which is
adapted to dispatch to the next bytecode handler. The return bytecode
handler is modified to tail-call a new InterpreterExitTramoline instead
of returning to the InterpreterEntryTrampoline.
Overall this significanlty reduces the amount of stack space required for
interpreter frames, increasing the maximum depth of recursive calls from
around 6000 to around 12,500 on x64.
BUG=chromium:753705
Change-Id: I23328e4cef878df3aca4db763b47d72a2cce664c
Reviewed-on: https://chromium-review.googlesource.com/634364
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47617}
The CPP builtins execute the same piece of code to prepare context before
jumping into CEntryStub. By creating new ASM builtin to execute that common
piece of code, ~7KB code size (tested on x64) of snapshot_blob.bin can be
reduced without any negative performance impact.
BUG=
Change-Id: I744369e8723dcd902b61dc50645db66bea884441
Reviewed-on: https://chromium-review.googlesource.com/595119
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47590}
This introduces a {DebugAbort} machine-level operator as well as the
corresponding {ArchDebugAbort} backend instruction. The goal of this is
to speed up snapshot generation due to cheaper "CSA-asserts".
R=jgruber@chromium.org
BUG=v8:6688
Bug: v8:6688
Change-Id: If45f7da0652d4bb920c51ab7a7c41f9670434bbb
Also-By: jgruber@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/628560
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47568}
- Convert S.p.includes builtin from CPP to TFJ
- Fast paths S.p.includes(str) and S.p.includes(str, smi)
- Add Runtime kStringIncludes
- Add StringIncludesIndexOfAssembler (Generate is based on
StringPrototypeIndexOf builtin)
- S.p.includes and S.p.indexOf both use StringIncludesIndexOfAssembler
Quick measurements show 3x improvement for S.p.includes(str).
More about the measurements: https://gist.github.com/peterwmwong/7a2a96f3171a52f16ca8125a089f38e7
Bug: v8:6680
Change-Id: I79cb8dbe2b79e6df15aa734e128eee25c7e6aaf5
Reviewed-on: https://chromium-review.googlesource.com/620150
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47546}
Code aging is no longer supported by any remaining compilers now
that full codegen has been removed. This CL removes all vestiges of
code aging.
BUG=v8:6409
Change-Id: I945ebcc20c7c55120550c8ee36188bfa042ea65e
Reviewed-on: https://chromium-review.googlesource.com/619153
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Marja Hölttä <marja@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47501}
The quite common empty object literal doesn't need an AllocationSite
since it starts off with the general ElementsKind. By using a separate
bytecode we can directly instantiate the empty object without jumping
to the runtime first.
Note: this experimentally disables pretenuring for empty object
literals. Depending on the outcome of our benchmarks pretenuring
will be enabled again or fully removed for empty object literals.
Bug: v8:6211
Change-Id: I2fee81cbefc70865fc436dbd3bc5fc8de04db91c
Reviewed-on: https://chromium-review.googlesource.com/577555
Commit-Queue: Camillo Bruni <cbruni@chromium.org>
Reviewed-by: Adam Klein <adamk@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47467}
This removes:
- CodeBreakIterator for FCG code.
- RelocModes for debug breaks.
- Code generator for debug break slots.
- GC support for debug break slots.
- Code flag to indicate code with debug break slots.
- Builtin type DBG.
- Mechanisms to replace FCG code in the debugger and LiveEdit.
- Runtime entry to the debugger from debug break slots.
R=bmeurer@chromium.org, rmcilroy@chromium.org, ulan@chromium.org
Bug: v8:6409
Change-Id: I5662c8800e3ef1b1584ad107bfe0aae26c9d8abb
Reviewed-on: https://chromium-review.googlesource.com/613263
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47364}
AsyncGenerators, when resumed with a "return" completion, Await the sent
value to provide consistency with syntactic return statements. This
moves the await to during AsyncGeneratorResumeNext, shrinking the number
of bytecodes.
There's a minor change to BytecodeGenerator which removes a
%_GeneratorClose() call, since it's inserted implicitly by the parser.
BUG=v8:5855
TBR=neis@chromium.org
Change-Id: I2965c610e5985ac24c713b481e62f6b97f96a3d8
Reviewed-on: https://chromium-review.googlesource.com/582218
Commit-Queue: Caitlin Potter <caitp@igalia.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47253}
A spec change (a0dfeba1a8) introduced a number of Await operations to the spec. In turn, this caused generated bytecode for async generators to grow drastically.
This commit moves the Await within AsyncGeneratorYield (https://tc39.github.io/proposal-async-iteration/#sec-asyncgeneratoryield step 5) into a new TFJ builtin, similar in structure to AsyncGeneratorAwait, but instead of resuming the generator on resolution of the Promise, the current generator request's Promise is fulfilled instead.
This results in a reduction in generated bytecode without losing any statically available information.
BUG=v8:5855
Change-Id: Ib5bcf06132d221beffdea30639a7b4437030143b
Reviewed-on: https://chromium-review.googlesource.com/582487
Commit-Queue: Caitlin Potter <caitp@igalia.com>
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47224}
This reverts commit 15ef03cbf3.
Reason for revert: Found the following bugs
Bug: chromium:752846, chromium:752712, chromium:752850
Original change's description:
> Reland "[builtins] Port getting property from Proxy to CSA"
>
> This reland is after fix in [heap] Delete wrong DCHECK.
> It includes moving ProxyGetProperty to its own stub to reduce
> binary size.
>
> This is a reland of 47a97aa53b
> Original change's description:
> > [builtins] Port getting property from Proxy to CSA
> >
> > Bug: v8:6559, v8:6557
> > Change-Id: If6c51f5483adb73ddd2495cede5d85e887a3c298
> > Reviewed-on: https://chromium-review.googlesource.com/589212
> > Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
> > Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
> > Commit-Queue: Maya Lekova <mslekova@google.com>
> > Cr-Commit-Position: refs/heads/master@{#47113}
>
> Bug: v8:6559, v8:6557
> Change-Id: I76acd97ba1acb62b7e7983db1741441d997050f0
> Reviewed-on: https://chromium-review.googlesource.com/600215
> Commit-Queue: Maya Lekova <mslekova@google.com>
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Reviewed-by: Franziska Hinkelmann <franzih@chromium.org>
> Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#47159}
TBR=jkummerow@chromium.org,mstarzinger@chromium.org,franzih@chromium.org,jgruber@chromium.org,ishell@chromium.org,bmeurer@chromium.org,mslekova@google.com
# Not skipping CQ checks because original CL landed > 1 day ago.
Change-Id: I51bef25a031b02cf4deab11282473acae57f1ed3
Reviewed-on: https://chromium-review.googlesource.com/603708
Commit-Queue: Maya Lekova <mslekova@google.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47200}
Drop the deprecated CallConstructStub and remove the use of CallICStub
from fullcodegen, since that feedback is unused completely every since
Crankshaft got removed, thus we can safely unlink all the CallIC stuff
from fullcodegen nowadays, and completely nuke the CallICStub and the
CallICTrampolineStub now (we can also transitively nuke the unused
CreateAllocationSiteStub and CreateWeakCellStub).
Instead the CallIC logic is integrated into Ignition now, and part of
the bytecode handlers for [[Call]] and [[Construct]]. There's still some
follow-up cleanup with the way the Array constructor feedback is
integrated, but that's way easier now.
Bug: v8:5517, v8:6399, v8:6409, v8:6679
Change-Id: I0a6c6046faceca9b1606577bc9e63d9295e44619
Reviewed-on: https://chromium-review.googlesource.com/603609
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47196}
As of https://chromium-review.googlesource.com/600968 the CallIC no
longer supports AllocationSite feedback for [[Call]], so we can drop
the TurboFan bits that deal with AllocationSites for JSCall nodes as
well. This further simplifies the handling of the Array constructor.
Drive-by-fix: Rename Builtins::kArrayCode to Builtins::kArrayConstructor
for sake of consistency.
Bug: v8:6399
Change-Id: I9e6a684fc00dd72e25f925db5f407c3f3f715873
Reviewed-on: https://chromium-review.googlesource.com/602354
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47194}
This reverts commit 6c541561ef.
Reason for revert:
https://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20nosnap/builds/17240
Original change's description:
> [ic] Properly integrate the CallIC into Ignition.
>
> Drop the deprecated CallConstructStub and remove the use of CallICStub
> from fullcodegen, since that feedback is unused completely every since
> Crankshaft got removed, thus we can safely unlink all the CallIC stuff
> from fullcodegen nowadays, and completely nuke the CallICStub and the
> CallICTrampolineStub now (we can also transitively nuke the unused
> CreateAllocationSiteStub and CreateWeakCellStub).
>
> Instead the CallIC logic is integrated into Ignition now, and part of
> the bytecode handlers for [[Call]] and [[Construct]]. There's still some
> follow-up cleanup with the way the Array constructor feedback is
> integrated, but that's way easier now.
>
> Bug: v8:5517, v8:6399, v8:6409, v8:6679
> Change-Id: Ia0efc6145ee64633757a6c3fd1879d4906ea2835
> Reviewed-on: https://chromium-review.googlesource.com/602134
> Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
> Reviewed-by: Yang Guo <yangguo@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#47192}
TBR=rmcilroy@chromium.org,yangguo@chromium.org,bmeurer@chromium.org
Change-Id: I416ce6646f62ceb4127b3acee43912ee0d701c23
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:5517, v8:6399, v8:6409, v8:6679
Reviewed-on: https://chromium-review.googlesource.com/603647
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47193}
Drop the deprecated CallConstructStub and remove the use of CallICStub
from fullcodegen, since that feedback is unused completely every since
Crankshaft got removed, thus we can safely unlink all the CallIC stuff
from fullcodegen nowadays, and completely nuke the CallICStub and the
CallICTrampolineStub now (we can also transitively nuke the unused
CreateAllocationSiteStub and CreateWeakCellStub).
Instead the CallIC logic is integrated into Ignition now, and part of
the bytecode handlers for [[Call]] and [[Construct]]. There's still some
follow-up cleanup with the way the Array constructor feedback is
integrated, but that's way easier now.
Bug: v8:5517, v8:6399, v8:6409, v8:6679
Change-Id: Ia0efc6145ee64633757a6c3fd1879d4906ea2835
Reviewed-on: https://chromium-review.googlesource.com/602134
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47192}
This reland is after fix in [heap] Delete wrong DCHECK.
It includes moving ProxyGetProperty to its own stub to reduce
binary size.
This is a reland of 47a97aa53b
Original change's description:
> [builtins] Port getting property from Proxy to CSA
>
> Bug: v8:6559, v8:6557
> Change-Id: If6c51f5483adb73ddd2495cede5d85e887a3c298
> Reviewed-on: https://chromium-review.googlesource.com/589212
> Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
> Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
> Commit-Queue: Maya Lekova <mslekova@google.com>
> Cr-Commit-Position: refs/heads/master@{#47113}
Bug: v8:6559, v8:6557
Change-Id: I76acd97ba1acb62b7e7983db1741441d997050f0
Reviewed-on: https://chromium-review.googlesource.com/600215
Commit-Queue: Maya Lekova <mslekova@google.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Franziska Hinkelmann <franzih@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47159}
Bug: 749486
The feature is off by default, and could be turned on via
`v8_enable_csa_write_barrier = true`. With this CL, only x64 uses this
feature
Change-Id: Ie024f08b7d796a4cc4d55285dc9fe796780f0e53
Reviewed-on: https://chromium-review.googlesource.com/588891
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Albert Mingkun Yang <albertnetymk@google.com>
Cr-Commit-Position: refs/heads/master@{#47122}
Until now, when generating a builtin, it can only embed builtins
(as call targets) that have already been generated. This is either
achieved by reordering the builtins list, or by loading the call
target at runtime from the builtins list (see
MacroAssembler::TailCallBuiltin).
This patch works around this issue by filling the builtins list
with dummy code objects, which are later replaced with the completed
actual builtins. In release mode, this adds around 3ms to 140ms we
previously needed to populate the builtins list.
Change-Id: I7d451b3c09a1db4b9e755548102a80c7f0dfada2
Reviewed-on: https://chromium-review.googlesource.com/586531
Commit-Queue: Yang Guo <yangguo@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47062}
Per https://github.com/tc39/proposal-async-iteration/pull/102/files:
AsyncGeneratorResolve no longer unwraps a value component. Instead, the value is
unwrapped before the builtin call via Await, allowing Promise rejections to
affect the generator control flow.
Thus, all `yield <expr>` implicitly become `yield await <expr>`.
Additionally, `return <expr>` becomes `return await <expr>`. Finally, when the
generator is resumed with `.return()`, the parameter passed to .return() is
awaited before generator execution properly continues).
BUG=v8:6187, v8:5855
R=littledan@chromium.org, neis@chromium.org, adamk@chromium.orgTBR=rmcilroy@chromium.org, neis@chromium.org
Cq-Include-Trybots: master.tryserver.v8:v8_linux_noi18n_rel_ng
Change-Id: Id7718028fd555481f9f4ca0dbecfa788e3057c48
Reviewed-on: https://chromium-review.googlesource.com/594500
Reviewed-by: Caitlin Potter <caitp@igalia.com>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Caitlin Potter <caitp@igalia.com>
Cr-Commit-Position: refs/heads/master@{#47058}
The @@toStringTag lookup in Object.prototype.toString causes quite a
lot of overhead and oftentimes dominates the builtin performance. These
lookups are almost always negative, especially for primitive values,
and Object.prototype.toString is often used to implement predicates
(like in Node core or in AngularJS), so having a way to skip the
negative lookup yields big performance gains.
This CL introduces a "MayHaveInterestingSymbols" bit on every map,
which says whether instances with this map may have an interesting
symbol. Currently only @@toStringTag is considered an interesting
symbol, but we can extend that in the future.
In the Object.prototype.toString we can use the interesting symbols
bit to do a quick check on the prototype chain to see if there are
any maps that might have the @@toStringTag, and if not, we can just
immediately return the result, which is very fast because it's derived
from the instance type. This also avoids the ToObject conversions for
primitive values, which is important, since this causes unnecessary
GC traffic and in for example AngularJS, strings are also often probed
via the Object.prototype.toString based predicates.
This boosts Speedometer/AngularJS by over 3% and Speedometer overall
by up to 1%. On the microbenchmark from the similar SpiderMonkey bug
(https://bugzilla.mozilla.org/show_bug.cgi?id=1369042), we go from
roughly 450ms to 70ms, which corresponds to a 6.5x improvement.
```
function f() {
var res = "";
var a = [1, 2, 3];
var toString = Object.prototype.toString;
var t = new Date;
for (var i = 0; i < 5000000; i++)
res = toString.call(a);
print(new Date - t);
return res;
}
f();
```
The design document at https://goo.gl/e8CruQ has some additional
data points.
TBR=ulan@chromium.org
Bug: v8:6654
Change-Id: I31932cf41ecddad079d294e2c322a852af0ed244
Reviewed-on: https://chromium-review.googlesource.com/593620
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47034}
This reverts commit 409f84c93b.
Reason for revert: Breaks nosnap debug:
https://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20nosnap%20-%20debug/builds/14288
Original change's description:
> [async-iteration] implement spec-change to `yield` in async generators
>
> Per https://github.com/tc39/proposal-async-iteration/pull/102/files:
>
> AsyncGeneratorResolve no longer unwraps a value component. Instead, the
> value is unwrapped before the builtin call via Await, allowing Promise
> rejections to affect the generator control flow.
>
> Thus, all `yield <expr>` implicitly become `yield await <expr>`.
>
> Additionally, `return <expr>` becomes `return await <expr>`. Finally, when
> the generator is resumed with `.return()`, the parameter passed to .return()
> is awaited before generator execution properly continues).
>
> BUG=v8:5855
> R=littledan@chromium.org, neis@chromium.org, adamk@chromium.org
>
> Cq-Include-Trybots: master.tryserver.v8:v8_linux_noi18n_rel_ng
> Change-Id: Ife084076c3ed434b5467e6aeba14082f8b410ad5
> Reviewed-on: https://chromium-review.googlesource.com/523844
> Commit-Queue: Caitlin Potter <caitp@igalia.com>
> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
> Reviewed-by: Georg Neis <neis@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#47011}
TBR=rmcilroy@chromium.org,adamk@chromium.org,yangguo@chromium.org,neis@chromium.org,littledan@chromium.org,gsathya@chromium.org,caitp@igalia.com
Change-Id: Ie6ad7e5410a3a89aab7a5dc68de36eb27b9354fe
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:5855
Cq-Include-Trybots: master.tryserver.v8:v8_linux_noi18n_rel_ng
Reviewed-on: https://chromium-review.googlesource.com/593952
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47013}
Per https://github.com/tc39/proposal-async-iteration/pull/102/files:
AsyncGeneratorResolve no longer unwraps a value component. Instead, the
value is unwrapped before the builtin call via Await, allowing Promise
rejections to affect the generator control flow.
Thus, all `yield <expr>` implicitly become `yield await <expr>`.
Additionally, `return <expr>` becomes `return await <expr>`. Finally, when
the generator is resumed with `.return()`, the parameter passed to .return()
is awaited before generator execution properly continues).
BUG=v8:5855
R=littledan@chromium.org, neis@chromium.org, adamk@chromium.org
Cq-Include-Trybots: master.tryserver.v8:v8_linux_noi18n_rel_ng
Change-Id: Ife084076c3ed434b5467e6aeba14082f8b410ad5
Reviewed-on: https://chromium-review.googlesource.com/523844
Commit-Queue: Caitlin Potter <caitp@igalia.com>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47011}
Empty Array literals are amongst the most commonly used literal types on our
top25 page list. Using a custom bytecode we can drop the boilerplate for empty
Array literals alltogether. However, we still need a proper AllocationSite to
track ElementsKind transitions.
Bug: v8:6211, chromium:746935
Change-Id: I891eaa778e4e81e138e483a65f04ae00ae30bd28
Reviewed-on: https://chromium-review.googlesource.com/580932
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Commit-Queue: Camillo Bruni <cbruni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46875}
