Previously we would directly take the result from a fulfilled native
promise bypassing the microtask queue. This is observably different
from the spec.
Note: Our variant of the bluebird benchmark is heavily favored towards
fulfilled native promises because we don't use setTimeout (unlike the
original benchmark). I suspect this pattern doesn't appear often in
the wild so it's fine to take this hit for now.
PSA for Perf sheriffs: this is going to tank some benchmarks.
Bug: chromium:800651, v8:5691, v8:6007
Change-Id: Ic273bf2195529424b0d87359d28d5267060d5252
Reviewed-on: https://chromium-review.googlesource.com/895416
Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51010}
We still avoid the "then" lookup using the current fast-path
mega-guard in the baseline case, but in TurboFan we simply
constant-fold the "then" lookup in the JSCallReducer. So all
further optimizations on Promise#then in TurboFan will automatically
apply to Promise#catch as well.
Bug: v8:7253
Change-Id: Idf7252157375a0ae3a91c7a3b42c30c5f367c0a8
Reviewed-on: https://chromium-review.googlesource.com/895446
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Michael Stanton <mvstanton@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51008}
A given JSPromise can either be in pending state, and accumulates
reactions, or in settled state, where all reactions are scheduled
as microtasks, and it carries a result. So we can use a single field
on the JSPromise instance to hold both the result and the reactions
and that field is interpreted differently depending on the status of
the JSPromise.
Bug: v8:7253
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: I19a7d499c88f452f0d35979ab95deb110021cde9
Reviewed-on: https://chromium-review.googlesource.com/895528
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51004}
This allows base::nullopt to be used instead of base::Optional<Foo>()
and implicit conversion to Optional<T> from T.
Also added NOLINT comments to the implicit constructors.
Change-Id: I4c688045685e2a50e0c0d38a959570f77454ec61
Reviewed-on: https://chromium-review.googlesource.com/893268
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Commit-Queue: Dan Elphick <delphick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51003}
This prevents unnecessary switching back and forth between internal and
public API boundaries. It is also a step towards making all WebAssembly
internals completely independent of "scheduled exception" values.
R=ahaas@chromium.org
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: I46e8cb4fad3d255d9bd20b9c343901a03a25426c
Reviewed-on: https://chromium-review.googlesource.com/895742
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51002}
Implements the saturating opcode i64.trunc_s:sat/f32.
Also does some refactoring of the i32 saturating opcodes use a simplier
solution (calling a single method to handle all i32 values).
Also refactors code so that the remaining i64 saturating conversions
should be easy to add to the wasm compiler.
Bug: v8:7226
Change-Id: I031aca1e059b4baa989a56ecbc16941f591ff9b3
Reviewed-on: https://chromium-review.googlesource.com/887333
Commit-Queue: Karl Schimpf <kschimpf@chromium.org>
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#51001}
This downloads the mips tool chain into tools/mips_toolchain via
runhooks if the flag download_mips_toolchain is specified for
gclient.
The currently uploaded version is 2015/1/7.
Bug: chromium:632390
Change-Id: Ied396d1a980db962773a96e9dadd3f282c7541c0
Reviewed-on: https://chromium-review.googlesource.com/892864
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50999}
The FeedbackNexus classes initially were one-to-one with IC classes,
but over time this got out of date. We also found Nexus' useful, so
we made more classes even for cases that weren't ICs.
The inheritence and polymorphism became confusing and led to
duplication. Better, to just talk about a (single) FeedbackNexus.
Bug: v8:7344
Change-Id: I509dc9657895d56c3859de6e6589695cdff9e73e
Reviewed-on: https://chromium-review.googlesource.com/890452
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Franziska Hinkelmann <franzih@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50997}
The hash avoids assigning all CHECK failures to the same clusterfuzz
report.
Bug: chromium:805970
Change-Id: Ia52da335ea86fbc7cc924dd81a893722a6d3d92e
Reviewed-on: https://chromium-review.googlesource.com/894323
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50992}
The fuzzer found a couple of cases that exploited comments of the
form:
function test() {
const re = /*.../;
const str = '...*/...';
let result;
try { result = re.exec(str); } catch (e) { /* ... */ }
}
Note that the first line does not contain a regexp literal, it starts
a comment instead. The second line terminates the comment.
This fixes detection of such cases by initializing `result` to null.
TBR=yangguo@chromium.org
Bug: chromium:805970
Change-Id: I5d46db9892e2b4e71cdc2907cebf07a2e33b7a0e
Reviewed-on: https://chromium-review.googlesource.com/894403
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50991}
In order to support i64 values on 32 bit platforms, we extend the
{LiftoffRegister} class to support storing pairs of GP registers on
those platforms. On 64 bit platforms, this refactoring should cause no
slowdown, as the compiler can statically determine that
{LiftoffRegister::is_pair()} is always false.
R=titzer@chromium.org
Bug: v8:6600
Change-Id: Ie9966d599271b8779959b1809ab4e129b68b80d7
Reviewed-on: https://chromium-review.googlesource.com/890261
Reviewed-by: Ben Titzer <titzer@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50986}
For mips, if 'mips_arch_variant=="r6"' and if 'mips_use_msa' flag is set
to 'true', then test-run-wasm-simd tests won't be skipped for mips. It
will also force 'MIPS_SIMD' bit in CpuFeatures to be set.
ARM processors are assumed to support SIMD.
Change-Id: Iea668b97ef995ca4949ddbf2ffc734aad89d3aa3
Reviewed-on: https://chromium-review.googlesource.com/868430
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Ivica Bogosavljevic <ivica.bogosavljevic@mips.com>
Commit-Queue: Ivica Bogosavljevic <ivica.bogosavljevic@mips.com>
Cr-Commit-Position: refs/heads/master@{#50981}
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}
Register t3 is not allocatable, so register a7 is used instead.
Change-Id: Ic198f038a10b4c891b417656d639477aedf6b703
Reviewed-on: https://chromium-review.googlesource.com/891823
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Ivica Bogosavljevic <ivica.bogosavljevic@mips.com>
Commit-Queue: Ivica Bogosavljevic <ivica.bogosavljevic@mips.com>
Cr-Commit-Position: refs/heads/master@{#50979}
This is a reland of 0db74d4974.
Original change's description:
> [test] Random seed processor
>
> 1. --total-timeout-sec now available for ./run-tests.py. It can be
> useful with infinite seed stressing
> 2. random seed dropped from the context. Now JSON progress indicator
> gets it from the list of command args.
>
> Bug: v8:6917
> Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
> Change-Id: I73e535bc8face9b913c696b8d5e3a246fa231004
> Reviewed-on: https://chromium-review.googlesource.com/888524
> Commit-Queue: Michał Majewski <majeski@google.com>
> Reviewed-by: Michael Achenbach <machenbach@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#50964}
Bug: v8:6917
Change-Id: I1ea376a4abffce5ab65f4834ea7e6d6011765ffa
Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
Reviewed-on: https://chromium-review.googlesource.com/894204
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michał Majewski <majeski@google.com>
Cr-Commit-Position: refs/heads/master@{#50978}
Async generators didn't correctly handle the situation where one calls
.return on a suspended-at-start async generator and passes a
promise-like object whose awaiting causes a new request to the
generator.
Bug: chromium:805729
Change-Id: I4da13ab5bd97f8c2a2c5373242a2d5e2ab0f7f10
Reviewed-on: https://chromium-review.googlesource.com/891231
Reviewed-by: Caitlin Potter <caitp@igalia.com>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50974}
Bug=chromium:806582
Change-Id: I0d541903dfd1622ae6d4a2628c41dc28704680e6
Reviewed-on: https://chromium-review.googlesource.com/891626
Reviewed-by: Ben Smith <binji@chromium.org>
Commit-Queue: Malcolm White <malcolmwhite@google.com>
Cr-Commit-Position: refs/heads/master@{#50972}
Port 830e39abae
Original Commit Message:
Wide suspends have a "wide" (or "extra-wide") bytecode at their offset,
rather than the suspend itself, so they were failing the return check.
R=leszeks@chromium.org, joransiu@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=
LOG=N
Change-Id: I0c95b9fd34df7232ae07fd1e508f40cd139e9734
Reviewed-on: https://chromium-review.googlesource.com/894303
Reviewed-by: Joran Siu <joransiu@ca.ibm.com>
Commit-Queue: Junliang Yan <jyan@ca.ibm.com>
Cr-Commit-Position: refs/heads/master@{#50971}
This reverts commit 0db74d4974.
Reason for revert: https://chromium-swarm.appspot.com/task?id=3b609f9976bac610&refresh=10&show_raw=1
Original change's description:
> [test] Random seed processor
>
> 1. --total-timeout-sec now available for ./run-tests.py. It can be
> useful with infinite seed stressing
> 2. random seed dropped from the context. Now JSON progress indicator
> gets it from the list of command args.
>
> Bug: v8:6917
> Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
> Change-Id: I73e535bc8face9b913c696b8d5e3a246fa231004
> Reviewed-on: https://chromium-review.googlesource.com/888524
> Commit-Queue: Michał Majewski <majeski@google.com>
> Reviewed-by: Michael Achenbach <machenbach@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#50964}
TBR=machenbach@chromium.org,sergiyb@chromium.org,majeski@google.com
Change-Id: I2d96ea328cda2d09b01ff455e47c77d567fafe00
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:6917
Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
Reviewed-on: https://chromium-review.googlesource.com/894522
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50970}
This is a reland of 76195d9e08.
It was reverted because the new parallel tasks (with higher number
of workers) hang on client.v8.ports bots. Since each test task steals
the worker thread it's assigned but only processes one item before
waiting for completion by others: I think the problem is that there
aren't enough workers in client.v8.ports' config. There aren't any
try bots for this config... reduce the tests to use 4 tasks and
hope for the best (i.e. a 4 core machine that uses "num cores")...
Original change's description:
> Smoother distribution of worker assignment in parallel task array.
>
> This is a merge of https://chromium-review.googlesource.com/c/v8/v8/+/888704
> and https://chromium-review.googlesource.com/c/v8/v8/+/887084
>
> Which implements the fix in CL 887084 correctly in a world where
> there can be more tasks_ than items_ (crbug.com/806237).
>
> Bug: chromium:805932
> Change-Id: I05401be4fdce442644a8973281a9d88bd959b271
> Reviewed-on: https://chromium-review.googlesource.com/892883
> Commit-Queue: Gabriel Charette <gab@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#50956}
Reverted-on: https://chromium-review.googlesource.com/893462
Bug: chromium:805932
Change-Id: I4d0bda3b9f52e9160e613a8f34a95e48b814bb9e
Reviewed-on: https://chromium-review.googlesource.com/893362
Reviewed-by: Hannes Payer <hpayer@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50967}
The way the code generator's AssembleMove and AssembleSwap methods are written
makes it easy to forget which sort of move is being implemented when looking at
a sequence of instructions. This patch is an attempt to address this by
rewriting those methods using switch/case instead of a string of if/else.
To do this, introduce new utility functions to detect what type of move to
perform given a pair of InstructionOperands.
Bug:
Change-Id: I32b146c86409e595b7b59a66bf43220899024fdd
Reviewed-on: https://chromium-review.googlesource.com/749201
Commit-Queue: Pierre Langlois <pierre.langlois@arm.com>
Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50966}
This reverts commit 76195d9e08.
Reason for revert: New parallel tests timeout on the waterfall (I think because it's configured to use less worker threads and TaskProcessingOneItem is currently designed to steal a worker but only process one item...).
Original change's description:
> Smoother distribution of worker assignment in parallel task array.
>
> This is a merge of https://chromium-review.googlesource.com/c/v8/v8/+/888704
> and https://chromium-review.googlesource.com/c/v8/v8/+/887084
>
> Which implements the fix in CL 887084 correctly in a world where
> there can be more tasks_ than items_ (crbug.com/806237).
>
> Bug: chromium:805932
> Change-Id: I05401be4fdce442644a8973281a9d88bd959b271
> Reviewed-on: https://chromium-review.googlesource.com/892883
> Commit-Queue: Gabriel Charette <gab@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#50956}
TBR=gab@chromium.org,hpayer@chromium.org,mlippautz@chromium.org
Change-Id: Icf52eb3afeb9467557c1e0db6922d590466943f0
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: chromium:805932
Reviewed-on: https://chromium-review.googlesource.com/893462
Reviewed-by: Hannes Payer <hpayer@chromium.org>
Commit-Queue: Gabriel Charette <gab@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50965}
1. --total-timeout-sec now available for ./run-tests.py. It can be
useful with infinite seed stressing
2. random seed dropped from the context. Now JSON progress indicator
gets it from the list of command args.
Bug: v8:6917
Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
Change-Id: I73e535bc8face9b913c696b8d5e3a246fa231004
Reviewed-on: https://chromium-review.googlesource.com/888524
Commit-Queue: Michał Majewski <majeski@google.com>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50964}
This adds support for set_global and get_global.
R=ahaas@chromium.org
Change-Id: I08bfa3c23080f473616970e9894cfb6e55a4f76d
Reviewed-on: https://chromium-review.googlesource.com/890744
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#50963}
