In the rare case that a tagged template is not initialized before
optimization time, we currently cache this created template in the
feedback vector. If we stop doing this, we simplify the interface
usefully for concurrent compilation and pay little for it.
Bug: v8:7790
Change-Id: Ifc82b0eb931a706767596febd4f4b312e167fd25
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690837
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62590}
This reverts commit 85bc4ef6c2.
Reason for revert: The tag 'e' conflicts with a blink serialization tag: kFileIndexTag.
Original change's description:
> Serialize native errors
>
> Make native errors serializable.
>
> The implementation is mostly straightforward, but there is one
> exception: the stack property. Although the property is not specified,
> the spec for error cloning asks us to preserve the property if
> possible. This implementation serializes the property only when it is
> a string, and otherwise ignores it.
>
> Spec: https://github.com/whatwg/html/pull/4665
> Intent-to-Ship: <TBD>
>
> Bug: chromium:970079
> Change-Id: I7f36b8b4fc5dff22d726d849ccfb9748d0888365
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1649257
> Commit-Queue: Yutaka Hirano <yhirano@chromium.org>
> Reviewed-by: Simon Zünd <szuend@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#62584}
TBR=jbroman@chromium.org,yhirano@chromium.org,adamk@chromium.org,domenic@chromium.org,szuend@chromium.org
Change-Id: Ia0cc902eaa1419cdb0cfec377d8a40fa914612c9
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: chromium:970079
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1692365
Commit-Queue: Simon Zünd <szuend@chromium.org>
Reviewed-by: Simon Zünd <szuend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62589}
EscapeRegExpPattern should return a string representation of a
RegExp instance that in turn can be used to construct a new
RegExp instance with the same internal state as the original one.
Previous versions incorrectly escaped '/' also inside character classes
(e.g. /[/]/ returned "[\/]").
This patch properly escapes '/' when necessary and omits unnecessary
escapes.
Bug: v8:8615, v8:1982, v8:9446
Change-Id: I4ecb993dc69d6976f4637cedf43465cd0c32e427
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1688050
Commit-Queue: Patrick Thier <pthier@google.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Mathias Bynens <mathias@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62587}
A few changes have been made to this feature and disabling it lets us
best see its current performance impact.
Bug: v8:9088
Change-Id: I54d5e09f3fcece215e29d66d5bdb3f19ba07bda0
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690954
Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Commit-Queue: Sigurd Schneider <sigurds@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62586}
Make native errors serializable.
The implementation is mostly straightforward, but there is one
exception: the stack property. Although the property is not specified,
the spec for error cloning asks us to preserve the property if
possible. This implementation serializes the property only when it is
a string, and otherwise ignores it.
Spec: https://github.com/whatwg/html/pull/4665
Intent-to-Ship: <TBD>
Bug: chromium:970079
Change-Id: I7f36b8b4fc5dff22d726d849ccfb9748d0888365
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1649257
Commit-Queue: Yutaka Hirano <yhirano@chromium.org>
Reviewed-by: Simon Zünd <szuend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62584}
This CL changes non-const reference arguments to either a const
reference, or pass-by-value combined with std::move.
Bug: v8:9429
Change-Id: Iabace132f855462612ac31922fbd8b456d8ae20d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690827
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Simon Zünd <szuend@chromium.org>
Auto-Submit: Simon Zünd <szuend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62583}
ResolveExport and Evaluate are the final unimplemented SyntheticModule methods; with this
change the implementation is complete.
Test-api unit tests are also provided.
Bug: v8:9292
Change-Id: Ieb7643cc5b6495dd201a51f04199d2406a703e52
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1681187
Reviewed-by: Adam Klein <adamk@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Commit-Queue: Dan Clark <daniec@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#62582}
Change-Id: Ia506f4741e6ff9f024199d1b1fa7abb7dafe2b25
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1682835
Commit-Queue: Bill Budge <bbudge@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Ben Smith <binji@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62581}
Defers some work in the non-asm.js case
Change-Id: Id20863fb2af83148271408b0242d49e03be29d3b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1682213
Commit-Queue: Bill Budge <bbudge@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62580}
The current integration of crash keys into v8 got the dependencies wrong: it introduced into v8 a dependency on components and base. This change will allow moving the implementation into "gin" (via Platform's abstraction), which is ok to depend on components and base, while providing the default noop implementation for the embedders that don't care to collect crash keys. Gin's side: https://chromium-review.googlesource.com/c/chromium/src/+/1690003.
Bug: v8:9323
Change-Id: I7b6e3e2cdc4b5f14f61ad20d2c362344d53896c6
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1689834
Commit-Queue: Irina Yatsenko <irinayat@microsoft.com>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62579}
This reverts commit 31cd5d83d3.
Reason for revert: It breaks my heart to revert this, but it fails differently on several bots, e.g. https://ci.chromium.org/p/v8/builders/ci/V8%20Linux%20-%20debug/26671.
Original change's description:
> [arraybuffer] Rearchitect backing store ownership
>
> This CL completely rearchitects the ownership of array buffer backing stores,
> consolidating ownership into a {BackingStore} C++ object that is tracked
> throughout V8 using unique_ptr and shared_ptr where appropriate.
>
> Overall, lifetime management is simpler and more explicit. The numerous
> ways that array buffers were initialized have been streamlined to one
> Attach() method on JSArrayBuffer. The array buffer tracker in the
> GC implementation now manages std::shared_ptr<BackingStore> pointers,
> and the construction and destruction of the BackingStore object itself
> handles the underlying page or embedder-allocated memory.
>
> The embedder API remains unchanged for now. We use the
> v8::ArrayBuffer::Contents struct to hide an additional shared_ptr to
> keep the backing store alive properly, even in the case of aliases
> from live heap objects. Thus the embedder has a lower chance of making
> a mistake. Long-term, we should move the embedder to a model where they
> manage backing stores using shared_ptr to an opaque backing store object.
>
> R=mlippautz@chromium.org
> BUG=v8:9380,v8:9221
>
> Change-Id: I48fae5ac85dcf6172a83f252439e77e7c1a16ccd
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1584323
> Commit-Queue: Ben Titzer <titzer@chromium.org>
> Reviewed-by: Ben Titzer <titzer@chromium.org>
> Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
> Reviewed-by: Yang Guo <yangguo@chromium.org>
> Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
> Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#62572}
TBR=ulan@chromium.org,yangguo@chromium.org,mstarzinger@chromium.org,titzer@chromium.org,gdeepti@chromium.org,mlippautz@chromium.org
Change-Id: Ib35788ba8c31192d90cbc72df3dbc41030f109de
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:9380, v8:9221
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1691034
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62578}
Locally tested, octane results are the same or better when using
branchful decompression in generated code.
The biggest one is DeltaBlue in x64 which improves about 25%.
x64's EarleyBoyer is another important one with ~10% improvement.
Bug: v8:7703
Change-Id: Ie1edd6e216b307a1f5263b22472c0a64f76400fc
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690955
Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org>
Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62577}
This reverts commit 907f3a64b7.
Reason for revert: speculative revert for v8:9445
I will reland if the crash is not fixed by the revert.
Original change's description:
> [heap] Replace ConcurrentSweepingState with a MemoryChunk local epoch counter.
>
> Bug: v8:9093
> Change-Id: I7c415fd0ea9e48f7ee189115f164825cb120695b
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1624213
> Commit-Queue: Hannes Payer <hpayer@chromium.org>
> Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#62423}
TBR=ulan@chromium.org,hpayer@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: v8:9093, v8:9445
Change-Id: Ia81a52579dc0a89f57ee41c7d0f8b1ba0f9bba81
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1691025
Commit-Queue: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62575}
This change partially decouples the heap's allocation space numbering
from the snapshot space encoding. This allows encoding read-only
heap allocations when RO_SPACE is removed.
A few pieces of src/snapshot also get cleaner as they no longer need to
check for extraneous AllocationSpace values.
Bug: v8:7464
Change-Id: I984c039b0e50e233209911ac3e655eb39be2551b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690956
Commit-Queue: Maciej Goszczycki <goszczycki@google.com>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Dan Elphick <delphick@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62574}
This moves initialization of heap limits to one place: ConfigureHeap
and also initializes the global limits there.
Change-Id: I7ca76073680a97627a8b1da9a8310af003794a93
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1648259
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62573}
This CL completely rearchitects the ownership of array buffer backing stores,
consolidating ownership into a {BackingStore} C++ object that is tracked
throughout V8 using unique_ptr and shared_ptr where appropriate.
Overall, lifetime management is simpler and more explicit. The numerous
ways that array buffers were initialized have been streamlined to one
Attach() method on JSArrayBuffer. The array buffer tracker in the
GC implementation now manages std::shared_ptr<BackingStore> pointers,
and the construction and destruction of the BackingStore object itself
handles the underlying page or embedder-allocated memory.
The embedder API remains unchanged for now. We use the
v8::ArrayBuffer::Contents struct to hide an additional shared_ptr to
keep the backing store alive properly, even in the case of aliases
from live heap objects. Thus the embedder has a lower chance of making
a mistake. Long-term, we should move the embedder to a model where they
manage backing stores using shared_ptr to an opaque backing store object.
R=mlippautz@chromium.org
BUG=v8:9380,v8:9221
Change-Id: I48fae5ac85dcf6172a83f252439e77e7c1a16ccd
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1584323
Commit-Queue: Ben Titzer <titzer@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62572}
The client API provides a much simpler interface so that we don't have
to deal with producers, consumers etc. directly. This CL removes all the
code that dealt with the more complex API used previously.
The architecture used here requires that the embedder call into
Tracing::Initialize() to set up the tracing backend. The tracing
controller then connects to this backend when calling
DataSource::Register() and Tracing::NewTrace(). This will ultimately
avoid the need for a virtual call (or two) for every trace event that
need to be dispatched over the API - chrome can provide a backend
and V8 will connect to it opaquely with the same code when tracing is
enabled.
Cq-Include-Trybots: luci.v8.try:v8_linux64_perfetto_dbg_ng
Bug: v8:8339
Change-Id: I6b74fbb49ffcc89638caeb59ed3d5cc81238f3e8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1634916
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62568}
This makes the heuristic for computing the number of parallel tasks
in Scavenger consistent with that in Mark-Compactor.
The patch helps mobile devices where even 1 MB new space can take
10ms to scavenge.
Change-Id: I979de5e8485b93808ea079af2756f53d9b720e10
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1685612
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62566}
Performance infra recipes use callstats.py to run web page replays. Split (v8,
infa, perf-infra) repositories make experimenting with callstats.py hard.
This CL creates feature gating for simplifying performance infra experiments.
R=ulan@chromium.org
No-Try: true
Bug: v8:9448
Change-Id: I2c3e139f4b9d6bce1ea4fdda1a44960d74d7d414
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690950
Auto-Submit: Tamer Tas <tmrts@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Tamer Tas <tmrts@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62565}
This fixes the only violation of the 'runtime/references' warning in
src/base by using a forwarding reference.
R=mlippautz@chromium.org
Bug: v8:9429, v8:9396
Change-Id: Ide2d28ab647e1cec78979c3f6d664b53c92a7d6e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690839
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62564}
When I implemented these instructions, I copied the naming scheme of
{GetGlobal}. That's not appropriate for the table.get instruction
though, and I decided I suffered enough from that bad name now.
R=clemensh@chromium.org
Bug: v8:7581, v8:9396
Change-Id: Id1796425458f3d06a2da774374f02c49d665d2c6
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690835
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Commit-Queue: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62563}
This removes two unneeded casts and an unused parameter.
Note that according to the specification, {Allocator}s need to provide
an {allocate} method receiving the size {n}, and *optionally* they can
support {allocate} with a second pointer parameter which they can use
in an unspecified manner. Since we do not use the second argument, we
should just not provide that method.
Drive-by: Remove else-after-return.
R=mlippautz@chromium.org
Bug: v8:9396
Change-Id: I776c5ae2f6652e20c9f9bfd511c41ce6ad1c1329
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690831
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62561}
This is a reland of f24e995809
Speculatively relanding the CL. LLD ignores the added section, so the
output should not be any different. The issue was also not reproducible
locally. Keeping an eye on the bot/roll to quickly revert, in case this
CL really breaks the deterministic build.
Original change's description:
> [snapshot] Add '.note.GNU-stack' marker section for ELF snapshots
>
> This snapshot adds a marker section to the output assembly of
> mksnapshot. Omitting this section can imply an executable stack, which
> can result in linker warnings.
>
> R=jgruber@chromium.org
>
> Change-Id: I6f5dacb24de3a852faf6ce0caf18c32b499ba5a9
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1687669
> Commit-Queue: Simon Zünd <szuend@chromium.org>
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#62504}
TBR=jgruber@chromium.org
Bug: chromium:981138
Change-Id: I2ddbaa7aa33f125ea0a0854cd9a94108c3a3845f
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690825
Reviewed-by: Simon Zünd <szuend@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Simon Zünd <szuend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62560}
This method is not specified as part of the Allocator requirement, as
is deprecated on std::allocator since C++17.
This also removes the only violation of the 'runtime/references'
cpplint check in the zone directory.
R=mlippautz@chromium.org
Bug: v8:9429, v8:9396
Change-Id: I77c28afbf3c0e122c5ac27796b9f46b6c9455e61
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690830
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62559}
This CL aims to address the regressions that we saw in Octane 2.1,
particularly in the DeltaBlue test.
This CL brings a 5% improvement in said test by doing
CompressedSigned -> Word32 conversion (instead of
CompressedSigned -> TaggedSigned -> Word32).
There seems to be room for optimizations doing more specialized conversions
regarding representation changes.
Cq-Include-Trybots: luci.v8.try:v8_linux64_pointer_compression_rel_ng
Cq-Include-Trybots: luci.v8.try:v8_linux64_arm64_pointer_compression_rel_ng
Bug: v8:7703
Change-Id: I24e5b6c06436fdda9fa6a1ac4699dc55c3d67abd
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1684075
Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org>
Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62557}
This adds a test case for using constructed {WebAssembly.Function}
objects in non-zero tables. Due to a recent refactoring that unifies
handling of dispatch tables, this works out of the box. The test
coverage however is still useful, since code paths are slightly
different for non-zero tables.
R=ahaas@chromium.org
TEST=mjsunit/wasm/type-reflection-with-anyref
BUG=v8:7742
Change-Id: I0cf4b0a8039bbef0422b06ee23744a949be8f1b1
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1690821
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62555}
This is a reland of 89d93e3851
Original change's description:
> Reland "Let all early errors be SyntaxErrors."
>
> This is a reland of 99fd5b9b9d which includes a missed update to
> test/test262/test262.status.
>
> Implement the spec change from the following TC39 PR:
> https://github.com/tc39/ecma262/pull/1527
>
> Bug: v8:9326
> Change-Id: Ie3aac60db550e90fb648fc30886a05419fa41afe
> TBR: adamk@chromium.org
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1682989
> Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
> Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#62500}
Bug: v8:9326
Change-Id: Ic30280400dfa5b83a4a397888e563eee479446c5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1688271
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Tamer Tas <tmrts@chromium.org>
Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62553}
Cpplint usually checks for non-const reference arguments. They are
forbidden in the style guide, and v8 does not explicitly make an
exception here.
This CL re-enables that warning, and fixes all current violations by
adding an explicit "NOLINT(runtime/references)" comment. In follow-up
CLs, we should aim to remove as many of them as possible.
TBR=mlippautz@chromium.org
Bug: v8:9429
Change-Id: If7054d0b366138b731972ed5d4e304b5ac8423bb
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1687891
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62551}
Introduce the enum class to expand a boolean parameter in
ErrorUtils::Construct. This is a preliminary change for error
serialization: we want to create an error with the given stack string.
Bug: chromium:970079
Change-Id: Ic55993d39d5d7b92197e2062a2be7cd8e87e552a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1689674
Reviewed-by: Simon Zünd <szuend@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Yutaka Hirano <yhirano@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62550}
This CL mitigates the worst performance regressions for stack trace
symbolization.
The first fix is to hold the StackFrameBase instance in a local
variable. The instance is produced by an iterator but newly created
everytime FrameArrayIterator::Frame is called.
The second fix is to skip symbolization of "MethodName" and
"TypeName" for non-method calls.
R=ulan@chromium.org
Bug: chromium:981541,v8:8742
Change-Id: I58b2e0c87693c3914cf1946ce56341fbd4a797ac
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1688927
Commit-Queue: Simon Zünd <szuend@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62549}
Feedback shall not be updated by the deoptimizer. Although this
mechanism exists, it shall not be used if possible. This CL changes
how V8 learns from BigInt deopts: Previously we updated feedback on
the BinaryOperations in the deoptimizer, now we let the interpreter
widen the feedback type from BigInt to Any after the deopt has occurred.
Bug: v8:9407
Change-Id: I92e5e733085b433fd8ab452674d02404b81b2796
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1687419
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Commit-Queue: Nico Hartmann <nicohartmann@google.com>
Cr-Commit-Position: refs/heads/master@{#62548}
This CL teaches the deoptimizer about JavaScriptBuiltinContinuation
frames that are not preceded by argument adapter frames. This pattern
is used when calling C++ API functions from TurboFan.
This CL fixes a crash when the deoptimizer encounters the pattern
described above. The crash was caused when the deoptimizer tried to
read the arguments of the continuation frame. As no adapter frame
was present, the argument count was read from the SharedFunctionInfo
which had the kDontAdaptArgumentsSentinel value. This translated to
an argument count of ~65000 later down the line, which caused a
FATAL error when the deoptimizer tried to re-construct ~65000
non-existent values.
Bug: chromium:980529
Change-Id: Id2de3bf7607102ab5a16de344c649015e968b185
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1687417
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Commit-Queue: Simon Zünd <szuend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62547}
Got rid of the following circular header dependency chains:
compilation-dependencies <-> js-heap-broker <-> access-info
types <-> js-heap-broker <-> access-info
Extracted former CompilationDependencies::Dependency class into its own header.
Extracted *Ref classes into their own header.
This should enable building on older GCC versions, e.g. 5.4.0.
Bug: v8:9440
Change-Id: Ia345bc227d8f7806d0b8622b706346a7ce6d01ea
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1687415
Commit-Queue: Maya Lekova <mslekova@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Michael Stanton <mvstanton@chromium.org>
Cr-Commit-Position: refs/heads/master@{#62541}
