- Clear flags to avoid the quite expensive query for whether this page
is to be swept.
- Use a vector instead of a list as we always expect a small number of
pages to go through the pool and we want to avoid memory management on
this path.
BUG=
Change-Id: If3c0ad480b8e4f3ccf5a0ef43200c5269822245d
Reviewed-on: https://chromium-review.googlesource.com/443248
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Hannes Payer <hpayer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43230}
This wraps float arrays with a proxy to make raw buffer use slow paths avoiding different NAN patterns.
This also mocks out large typed arrays when passing the lenth as third constructor parameter.
BUG=chromium:691287,chromium:690898
NOTRY=true
Change-Id: Ic4295b0d8690e5209aceeda9ed93efdd580194c0
Reviewed-on: https://chromium-review.googlesource.com/441624
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43229}
Some of these functions are invoked by BytecodeGenerator due to parser
desugarings, and moving the context indices cause
BytecodeExpectationsPrinter to render them as something
useful/meaningful.
BUG=
R=jgruber@chromium.org, adamk@chromium.org
Review-Url: https://codereview.chromium.org/2695323002
Cr-Commit-Position: refs/heads/master@{#43225}
When --harmony-async-iteration is enabled, it is now possible to
use the for-await-of loop, which uses the Async Iteration protocol
rather than the ordinary ES6 Iteration protocol.
the Async-from-Sync Iterator object is not implemented in this CL,
and so for-await-of loops will abort execution if the iterated object
does not have a Symbol.asyncIterator() method. Async-from-Sync
Iterators are implemented seperately in https://codereview.chromium.org/2645313003/
BUG=v8:5855, v8:4483
R=neis@chromium.org, littledan@chromium.org, adamk@chromium.org
Review-Url: https://codereview.chromium.org/2637403008
Cr-Commit-Position: refs/heads/master@{#43224}
The mips64 implementation always ended up in the slowpath due to some
loads that were the wrong width, so that is also fixed here.
BUG=v8:5974
Change-Id: Ie448a1fab5b7fca87597c5a1bf75443864e30c28
Reviewed-on: https://chromium-review.googlesource.com/443247
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43222}
These experimental natives previously only installed functions to the
appropriate parent. In this patch, the exports container is retained
so that the bootstrapper may install the functions instead. This
change is intended to reduce startup time. SharedArrayBuffer retains
some experimental natives exported from JS; this may be addressed
in a follow-on patch. The patch includes some minor cleanup of the
bootstrap process by removing "experimental exports", which was unused.
R=yangguo@chromium.org
BUG=v8:5880
CQ_INCLUDE_TRYBOTS=master.tryserver.v8:v8_linux_noi18n_rel_ng
Review-Url: https://codereview.chromium.org/2683083003
Cr-Commit-Position: refs/heads/master@{#43221}
- Different places used is_simple to mean different things; renamed one.
- No need to do Scope::SetHasNoSimpleParameters multiple times.
- Normally we create VAR parameters with a name, or (for destructuring
parameters), TEMPORARY parmeters with an empty name. *Except* for
destructuring rest parameters; then we create VAR a parameter with an empty
name. This CL makes the empty-named parameter TEMPORARY instead of VAR.
- This makes it clear that Parser::DeclareFormalParameters declares exactly
those params which Parser::BuildParamerterInitializationBlock doesn't declare.
- This unification doesn't change any functionality, but it makes sense to do
since I'll need to make PreParser emulate what Parser does; this way I don't
need to emulate the weird behavior.
BUG=v8:5501
Change-Id: Ifa6c116bc5908f4e03a36e74f47558888d1582bd
Reviewed-on: https://chromium-review.googlesource.com/443106
Reviewed-by: Daniel Vogelheim <vogelheim@chromium.org>
Commit-Queue: Marja Hölttä <marja@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43220}
This handles all arithmetic binary operations except addition during the
early type-hint lowering (i.e. during graph construction). We still use
static type information to potentially further reduce the speculative
operations down to pure operations during the typed lowering phase.
R=bmeurer@chromium.org
Change-Id: I8b93fd7c46ec8e5b81234a49624d503520c3d082
Reviewed-on: https://chromium-review.googlesource.com/443105
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43218}
This CL changes the datastructure to store live variables from a
std::bitset<256> to a std::vector<bool> to support an arbitrary number
of locals. Unfortunately, std::vector<bool> does not define |= and &=
operators, so I added them on the Environment class.
R=vegorov@chromium.org, mstarzinger@chromium.org, machenbach@chromium.org
BUG=v8:5970
Review-Url: https://codereview.chromium.org/2694103005
Cr-Commit-Position: refs/heads/master@{#43216}
The API class is v8::ArrayBuffer; JSArrayBuffer is the internal counterpart,
but its name should not appear in a public API comment.
Review-Url: https://codereview.chromium.org/2692853007
Cr-Commit-Position: refs/heads/master@{#43214}
- Removes shared InnerArrayCopyWithin JS builtin from src/js/array.js
- Implements %TypedArray%.prototype.copyWithin as a C++ builtin, which
relies on std::memmove rather than accessing individual eleements.
- Fixes the case where copyWithin is invoked on a TypedArray with a
detached buffer.
- Add tests to ensure that +/-Infinity (for all 3 parameters) is handled
correctly by the
algorithm
The C++ version gets through the benchmark more than 25000 times as
quickly as the JS implementation.
BUG=v8:5925, v8:5929, v8:4648
R=cbruni@chromium.org, adamk@chromium.org, littledan@chromium.org
Review-Url: https://codereview.chromium.org/2697593002
Cr-Commit-Position: refs/heads/master@{#43213}
Down from 16 * kPointerSize.
Modern compilers have good inlining support for memcpy(), so our
custom C++ loop is only beneficial for very short loops (if at all).
BUG=v8:5395
Review-Url: https://codereview.chromium.org/2438583002
Cr-Commit-Position: refs/heads/master@{#43212}
Instead of calling the LoadGlobalIC stub, bytecode handlers now inline
logic for LoadGlobalIC. The LoadGlobalICData case takes a fast path
which omits name loading and frame construction.
BUG=v8:5917
Review-Url: https://codereview.chromium.org/2684973002
Cr-Commit-Position: refs/heads/master@{#43210}
... and don't clear ICs during GC. The IC clearing used to prevent memory
leaks but it's not necessary anymore because all the handlers that need
to embed objects already use weak cells.
This CL unblocks inlining of IC dispatchers into bytecode handlers.
BUG=v8:5917
Change-Id: I229b9ba8dba44f431dfbe8ac5370d855e3e84dd6
Reviewed-on: https://chromium-review.googlesource.com/442127
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43209}
This allows part of typed lowering that is solely based on type-hints to
run as part of the graph construction. The lowering in question does not
inspect types and hence doesn't require the typer to have run before. We
insert the speculative simplied-level operations in favor of the generic
JavaScript-level variants.
R=bmeurer@chromium.org
Change-Id: I5f0549fc1e4ff607622ee9059e6232a32f77db2e
Reviewed-on: https://chromium-review.googlesource.com/442584
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43208}
ParserBase::is_any_identifier currently does not recognise
Token::ESCAPED_STRICT_RESERVED_WORD as an identifier. This seems different
from what ParserBase::ParseIdentifierName does, and also prevents
"l\u0065t", unlike "let", from becoming a label.
This CL extends is_any_identifier to also accept ESCAPED_STRICT_RESERVED_WORD.
BUG=v8:5692
Review-Url: https://codereview.chromium.org/2695973003
Cr-Commit-Position: refs/heads/master@{#43204}
The method ExpressionUnexpectedToken is not referenced anywhere apart from its
definition. This CL removes it.
The association with the bug below is only through discovering the dead code
when working on a fix for that bug.
BUG=v8:5692
Review-Url: https://codereview.chromium.org/2688413009
Cr-Commit-Position: refs/heads/master@{#43203}
Previously we captured the PC before calling EnsureSpace in
RecordProtectedInstruction. Sometimes EnsureSpace would resize and move
the buffer, which would invalidate the previously captured PC and trip an
assert when writing RelocInfo. With this change, we do not capture the PC
until after we've ensured there's enough space, which ensures the PC will
be valid.
BUG=
Review-Url: https://codereview.chromium.org/2690523003
Cr-Commit-Position: refs/heads/master@{#43202}
Use them to encode int32 constants properly.
This reduces the generated wasm size in the unity benchmark from
21.6 MB to 16.8 MB (-22.2%).
This hopefully also translates to increased performance especially on
mobile because of lower memory usage.
R=bradnelson@chromium.org, titzer@chromium.org
Review-Url: https://codereview.chromium.org/2692943002
Cr-Commit-Position: refs/heads/master@{#43200}
because that's what it actually does. In the next step, I will
introduce EmitVarInt again with the correct semantics, and use it in
EmitI32Const and other places where we don't do proper varint encoding
currently.
R=bradnelson@chromium.org, titzer@chromium.org
Review-Url: https://codereview.chromium.org/2694633003
Cr-Commit-Position: refs/heads/master@{#43198}
CreateNumberFormat of src/js/i18n.js implements
http://www.ecma-international.org/ecma-402/1.0/#sec-11.1.1.1, but has a typo
in step 33a. The spec says that the default value for minimumSignificantDigits
should be 1, while the script set it to 0.
This CL fixes that typo and adds a test for that.
BUG=v8:5554
Review-Url: https://codereview.chromium.org/2694673003
Cr-Commit-Position: refs/heads/master@{#43197}
It's needed by code-stubs.h and it was defined in objects-inl.h.
That cannot work without violating the include rules.
BUG=v8:5402
R=mstarzinger@chromium.org
Change-Id: Icb84b97de5622df8cf76e9fc4d117982901c99d9
Reviewed-on: https://chromium-review.googlesource.com/441845
Commit-Queue: Marja Hölttä <marja@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43195}
After https://github.com/tc39/ecma262/pull/642, Date.UTC no longer requires
the month argument to be specified. The spec provides 0 as its default value.
This CL updates the builtins-date.cc code to reflect that and drops the test
suppression for test262/built-ins/Date/UTC/return-value.
BUG=v8:5534
Review-Url: https://codereview.chromium.org/2689173003
Cr-Commit-Position: refs/heads/master@{#43193}
Add the function literal to eager_literals before recursing into the literal.
This ensures that the order of eager_literals reflects the order in which
the functions will be called, which in-turn ensures it is compiled first
by the compiler-dispatcher.
BUG=v8:5203
Change-Id: I0e334a7d1ee93cf67936adbadd2883e40acb1cc5
Reviewed-on: https://chromium-review.googlesource.com/441825
Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Reviewed-by: Marja Hölttä <marja@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43192}
BUG=
Change-Id: I4a5db9bc045a63e710d0115523ab23b98e7c7ae6
Reviewed-on: https://chromium-review.googlesource.com/442504
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43189}
... and revert "[runtime] Mark maps created as a result of field type or
representation generalization as migration targets." because it regresses
performance of too many map checks in optimized code.
BUG=v8:5444, chromium:680995
Change-Id: I84038d75425df3709a81e5fdcc6c0db32939060d
Reviewed-on: https://chromium-review.googlesource.com/442125
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43188}
The string will be internalized anyhow when the first object is instantiated,
and since templates are eternal, the string will be eternal anyhow even if
the template is never instantiated.
Review-Url: https://codereview.chromium.org/2697683002
Cr-Commit-Position: refs/heads/master@{#43187}
My hypothesis is that the OOM crash is caused by a loop of forgetting and reallocating a VirtualObject inside of a VirtualState.
R=mstarzinger@chromium.org
BUG=chromium:691487
Review-Url: https://codereview.chromium.org/2694723002
Cr-Commit-Position: refs/heads/master@{#43184}
CQ will only allow CL authors (owners in Gerrit) and V8 committers to trigger
CQ in Gerrit codereview. Thus, CQ voting restriction to just committers
can be lifted, letting every authenticated user attempt to trigger CQ.
R=machenbach@chromium.org,agable@chromium.org
BUG=641422,685318
NOTRY=True
Change-Id: Ied310a65277d6fefa44a9945cc780cb8fe827e03
Reviewed-on: https://chromium-review.googlesource.com/442124
Commit-Queue: Andrii Shyshkalov <tandrii@chromium.org>
Reviewed-by: Aaron Gable <agable@chromium.org>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#43183}
If the Reflect.construct receives an argument expected to be a constructor,
and the argument is not a constructor, V8 currently declares that
Reflect.construct is not a function. It should instead say that the offending
argument is not a constructor.
This is the case for all ports of builtins
(Builtins::Generate_ReflectConstruct). All of them make an
attempt to at least pass the right argument to the TypeError parametrised
message, calling out the offending Reflect.construct argument. However,
Runtime::kThrowCalledNonCallable extracts the callsite from those arguments,
discarding the precise information.
This CL adds Runtime::kNotConstructor, which reports the arguments passed
to it, and the CL also modifies the ports of builtins to make use of
Runtime::kNotConstructor
BUG=v8:5671
Review-Url: https://codereview.chromium.org/2688393003
Cr-Commit-Position: refs/heads/master@{#43182}
