It is invalid for OSR deconstruction to leave a graph with a node
representing the OSR normal entry (and no OSR loop entry). Subsequent
lowering phases will not handle {OsrNormalEntry} operators and hence
will lead to serious clogging further down the pipeline.
R=bmeurer@chromium.org
BUG=chromium:641893
Review-Url: https://codereview.chromium.org/2336543002
Cr-Commit-Position: refs/heads/master@{#39340}
With this CL the AstDecoder produces an error if it encounters a
grow_memory instruction in an asmjs module. Additionally asmjs
instructions are not allowed anymore in wasm modules.
BUG=chromium:644674
R=titzer@chromium.org
Review-Url: https://codereview.chromium.org/2324733002
Cr-Commit-Position: refs/heads/master@{#39339}
This patch moves the following parsing method to ParserBase:
- ParseSwitchStatement
It also removes ParseCaseClause and merges it with ParseSwitchStatement,
mainly to avoid the complexity of introducing one more abstract typedef
to be shared between parser implementations, but also because the merged
ParseSwitchStatement is now only 59 lines.
R=adamk@chromium.org, marja@chromium.org
BUG=
LOG=N
Review-Url: https://codereview.chromium.org/2324843005
Cr-Commit-Position: refs/heads/master@{#39337}
Compare the whole word on 64-bit architectures with the kHoleNanInt64 value
since comparing only upper part produces slightly worse code on x64.
Currently TurboFan is not able to fold (array + ((const1 + index*scale) + const))
into single addressing mode: (array + (const + index*scale)).
BUG=chromium:645414
Review-Url: https://codereview.chromium.org/2322283002
Cr-Commit-Position: refs/heads/master@{#39335}
Remove the "Entry"/"Entries" suffix as it doesn't add any information and
ModuleInfoEntry would clash with another name being introduced soon.
R=adamk@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/2300373002
Cr-Commit-Position: refs/heads/master@{#39333}
Reason for revert:
Tests fail when i18n is switched off, trybots do not run this configuration
Original issue's description:
> [builtins] Move StringNormalize to a cpp builtin.
>
> BUG=v8:5364
>
> Committed: https://crrev.com/7f84a6a2e7000bebba49354b4648346ff606ca34
> Cr-Commit-Position: refs/heads/master@{#39331}
TBR=bmeurer@chromium.org,franzih@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:5364
Review-Url: https://codereview.chromium.org/2335553002
Cr-Commit-Position: refs/heads/master@{#39332}
The logic to test whether we already reached --max_inlining_levels
when inlining into some optimized function only checked specifically
for FrameStateType::kJavaScriptFunction, and thereby didn't properly
account for FrameStateType::kInterpretedFunction, which is what we
see when we come in via the bytecode pipeline.
Review-Url: https://codereview.chromium.org/2329923002
Cr-Commit-Position: refs/heads/master@{#39328}
Escape sequences may now be written to a trace file (previously,
any string with a valid escapable character would fail a check).
Also, string properties are now surrounded with quotes.
BUG=v8:4561
Review-Url: https://codereview.chromium.org/2309943005
Cr-Commit-Position: refs/heads/master@{#39319}
This CL introduces a ProtectedLoad instruction with is needed for out
of bounds trap handling. ProtectedLoad behaves like a regular load,
but it takes a context and source position parameter as well. These
are used by an out of line code fragment to generate code to throw a
JS exception for an out of bounds memory reference in Wasm.
These changes a cleaned up subset of
https://codereview.chromium.org/2148743004/
The rest of this feature will follow in future CLs. This includes a
table mapping memory instructions to landing pads as well as the
actual signal handler.
BUG= https://bugs.chromium.org/p/v8/issues/detail?id=5277
Review-Url: https://codereview.chromium.org/2301833004
Cr-Commit-Position: refs/heads/master@{#39318}
The array spread operator is now handled by desugaring in the parser.
Review-Url: https://codereview.chromium.org/2324013002
Cr-Commit-Position: refs/heads/master@{#39317}
GetFunctionNameStr and GetScriptResourceNameStr can be called from a thread
other than isolate VM thread unlike their conterparts GetFunctionName
and GetScriptResourceName.
BUG=406277
Review-Url: https://codereview.chromium.org/2328673003
Cr-Commit-Position: refs/heads/master@{#39313}
Rolling v8/build to 69b6a57635907f95b6302a657d250832e3f91a1f
Rolling v8/third_party/WebKit/Source/platform/inspector_protocol to bfc71ffbbdc7ca43723d655d8d4ae972e9e431da
Rolling v8/third_party/jinja2 to b61a2c009a579593a259c1b300e0ad02bf48fd78
Rolling v8/tools/clang to 7d0756c437d8c3dd68f912f6cc47907b0120fc05
TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org
Review-Url: https://codereview.chromium.org/2319123006
Cr-Commit-Position: refs/heads/master@{#39312}
Reason for revert:
Broke stuff - need to fix the argument
Original issue's description:
> [gn] Switch off gyp in runhooks by default
>
> This will be synced with an announcement on v8-dev. This
> doesn't affect directly calling gyp in our old Makefiles.
>
> BUG=chromium:474921
> NOTRY=true
>
> Committed: https://crrev.com/94f813e8dcd4ed5a67cede8af0f949009beaa501
> Cr-Commit-Position: refs/heads/master@{#39308}
TBR=jochen@chromium.org,jkummerow@chromium.org,vogelheim@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:474921
Review-Url: https://codereview.chromium.org/2324173002
Cr-Commit-Position: refs/heads/master@{#39309}
This will be synced with an announcement on v8-dev. This
doesn't affect directly calling gyp in our old Makefiles.
BUG=chromium:474921
NOTRY=true
Review-Url: https://codereview.chromium.org/2327033002
Cr-Commit-Position: refs/heads/master@{#39308}
... and rename CheckAndGrowElementsCapacity() to TryGrowElementsCapacity().
Review-Url: https://codereview.chromium.org/2324863002
Cr-Commit-Position: refs/heads/master@{#39307}
This fixes a corner-case where the bytecode was using the <new.target>
register directly without going through the local variable. The value
might be clobbered because the deoptimizer doesn't properly restore the
value. The label will causes bytecode pipeline to be flushed and hence
ensure {BytecodeRegisterOptimizer} doesn't reuse <new.target> anymore.
R=rmcilroy@chromium.org
TEST=mjsunit/regress/regress-crbug-645103
BUG=chromium:645103
Review-Url: https://codereview.chromium.org/2325133002
Cr-Commit-Position: refs/heads/master@{#39306}
Reason for revert:
Revert to check if this is causing perf regressions in crbug.com/645411
Original issue's description:
> [heap] Fix a formatting bug in --trace-incremental-marking.
>
> BUG=
>
> Committed: https://crrev.com/212624b7570cd1c1cfad7cf958203b05af961637
> Cr-Commit-Position: refs/heads/master@{#39278}
TBR=mlippautz@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=
Review-Url: https://codereview.chromium.org/2323293002
Cr-Commit-Position: refs/heads/master@{#39305}
For call sites where the target is not a known constant, but potentially
a list of known constants (i.e. a Phi with all HeapConstant inputs), we
still record the call site as a potential candidate for inlining.
In case the heuristic picks that candidate for inlining, we
expand the call site to a dispatched call site and invoke the
actual inlining logic for all the nested call sites.
Like Crankshaft, we currently allow up to 4 targets for polymorphic inlining,
although we might want to refine that later.
This approach is different from what Crankshaft does in
that we don't duplicate the evaluation of the parameters per polymorphic
case. Instead we first perform the load of the target (which usually
dispatches based on the receiver map), then we evaluate all the
parameters, and then we dispatch again based on the known targets. This
might generate better or worse code compared to what Crankshaft does,
and for the cases where we generate worse code (i.e. because we have
only trivial parameters or no parameters at all), we might want to
investigate optimizing away the double dispatch in the
future.
R=mvstanton@chromium.org
BUG=v8:5267,v8:5365
Review-Url: https://codereview.chromium.org/2325943002
Cr-Commit-Position: refs/heads/master@{#39302}
This is a work-around as v8_enable_i18n_support=false does
currently not imply icu_use_data_file_flag=false. The
swarming isolator then tries to find the data file also
in builds without ICU.
Making the implication is non-trivial as icu_use_data_file_flag lives
in ICU and v8_enable_i18n_support lives in V8.
BUG=chromium:474921
NOTRY=true
TBR=petermarshall@chromium.org, vogelheim@chromium.org, jochen@chromium.org
Review-Url: https://codereview.chromium.org/2321563007
Cr-Commit-Position: refs/heads/master@{#39301}
port 9a31162d9d3137d09063d6040865655b2e386384(r39283)
original commit message:
Adds support to collect allocation site feedback for Array function calls
to the call bytecode handler.
BUG=
Review-Url: https://codereview.chromium.org/2319123004
Cr-Commit-Position: refs/heads/master@{#39299}
Without this cast, the integer type isn't promoted before being shifted, and so
for types larger than sizeof(int) there is data loss. This will become an issue
once the host begins using this helper to send 64-bit integers.
BUG=chromium:148757
Review-Url: https://codereview.chromium.org/2326653002
Cr-Commit-Position: refs/heads/master@{#39296}
This patch adds runtime call stats tracing for GC correctly, makes
--runtime-call-stats and tracing mutually exclusive with tracing taking
precedence if both modes are on, and uses only one runtime call stats in
counter.
BUG=v8:5089
Review-Url: https://codereview.chromium.org/2313193002
Cr-Commit-Position: refs/heads/master@{#39295}
Before this change, the spread desugaring would naively call
`%AppendElement($R, the_hole)` and in some cases $R would have
a non-holey elements kind, putting the array into the bad state
of exposing holes to author code.
This patch avoids calling %AppendElement with a hole, instead
simply incrementing $R.length when it sees a hole in the literal
(this is safe because $R is known to be an Array). The existing
logic for elements transitions takes care of giving the array a
holey ElementsKind.
BUG=chromium:644215
Review-Url: https://codereview.chromium.org/2321533003
Cr-Commit-Position: refs/heads/master@{#39294}
This CL fixes %DebugPrint for FAST_HOLEY_DOUBLE_ELEMENTS and now properly
distinguishes TheHole and NaN values.
BUG=
Review-Url: https://codereview.chromium.org/2294913004
Cr-Commit-Position: refs/heads/master@{#39293}
Reason for revert:
There have been no more occurrences of this on dev / beta so we can convert the CHECK back to DCHECK.
Original issue's description:
> [runtime] temporarily transform IsContext check from DCHECK to CHECK
>
> We are enabling this trial on canary to see if we can flush out some missing
> context restores.
>
> BUG=
>
> Committed: https://crrev.com/ec94ad400dc257af396efa3b1899bc3168347d82
> Cr-Commit-Position: refs/heads/master@{#37875}
TBR=jkummerow@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=
Review-Url: https://codereview.chromium.org/2303543003
Cr-Commit-Position: refs/heads/master@{#39292}
