arv [Fri, 13 Feb 2015 18:33:42 +0000 (10:33 -0800)]
TF: Add support for [[HomeObject]]
BUG=None
LOG=N
R=mstarzinger@chromium.org, dslomov@chromium.org
Review URL: https://codereview.chromium.org/
926013002
Cr-Commit-Position: refs/heads/master@{#26641}
bratell [Fri, 13 Feb 2015 10:56:41 +0000 (02:56 -0800)]
Make the statistics generation data driven.
StatisticsExtension::GetCounters() has grown to repeat a lot of code
so to make it more maintainable and generate less machine code,
make it data driven. This makes a 64 bit Linux build 50 KB smaller.
Total change: -51677 bytes
==========================
1 shrunk, for a net change of -51677 bytes (64222 bytes before, 12545 bytes after) across 1 sources
279691 unchanged, totalling
51423668 bytes
Per-source Analysis:
-------------------------------------------------------------------------------------------------------------------
-51677 - Source: /home/bratell/src/chromium/src/v8/src/extensions/statistics-extension.cc - (gained 0, lost 51677)
-------------------------------------------------------------------------------------------------------------------
Shrunk symbols:
-51677: v8::internal::StatisticsExtension::GetCounters(v8::FunctionCallbackInfo<v8::Value> const&) type=t, (was 64222 bytes, now 12545 bytes)
BUG=
Review URL: https://codereview.chromium.org/
912413003
Cr-Commit-Position: refs/heads/master@{#26640}
loislo [Fri, 13 Feb 2015 10:45:26 +0000 (02:45 -0800)]
Move identical code from platform specific assemblers to assembler.cc
BUG=none
LOG=n
Review URL: https://codereview.chromium.org/
922153002
Cr-Commit-Position: refs/heads/master@{#26639}
Ross McIlroy [Fri, 13 Feb 2015 10:35:37 +0000 (10:35 +0000)]
Protect access to the external_snapshot_blob global with a lock.
The external_snapshot_blob is a global and might be accessed from multiple
threads. Protect it with a lock.
BUG=457656
LOG=N
R=vogelheim@chromium.org
Review URL: https://codereview.chromium.org/
918213002
Cr-Commit-Position: refs/heads/master@{#26638}
Sven Panne [Fri, 13 Feb 2015 10:35:18 +0000 (11:35 +0100)]
Refactored the final parts of the TurboFan pipeline a bit.
This removed some inconsistent/duplicated code and made the testing
side entrance much simpler: Now it simply constructs a pipeline and
hands off to the "tail" of the normal pipeline.
R=danno@chromium.org
Review URL: https://codereview.chromium.org/
924823002
Cr-Commit-Position: refs/heads/master@{#26637}
loislo [Fri, 13 Feb 2015 10:29:11 +0000 (02:29 -0800)]
Fix compilation for the case dcheck_always_on=1.
It is a default option for some try bots.
otherwise compilation fails on
DCHECK(layout_descriptor->IsConsistentWithMap(*map));
BUG=none
LOG=n
Review URL: https://codereview.chromium.org/
920993003
Cr-Commit-Position: refs/heads/master@{#26636}
Akos Palfi [Fri, 13 Feb 2015 01:25:58 +0000 (02:25 +0100)]
MIPS: new classes: no longer experimental.
Port
2707d4c9f4b6e6e600e708b3b08b09d27d594ce7
Original commit message:
This CL fixes tests that no longer valid and also fixes two issues:
1. 'super()' in non derived constructors.
2. Failure to step into derived constructors.
BUG=
R=paul.lind@imgtec.com
Review URL: https://codereview.chromium.org/
920173002
Cr-Commit-Position: refs/heads/master@{#26635}
Akos Palfi [Fri, 13 Feb 2015 01:21:08 +0000 (02:21 +0100)]
MIPS: new classes: implement correct check for uninitialized this in 'super()'
Port
fdcf3e59ba4af1aa3c5b31a18e615169c97958ab
BUG=
R=paul.lind@imgtec.com
Review URL: https://codereview.chromium.org/
920543004
Cr-Commit-Position: refs/heads/master@{#26634}
Erik Arvidsson [Thu, 12 Feb 2015 23:04:40 +0000 (18:04 -0500)]
Make super() a syntax error in base class constructor
BUG=v8:3330
LOG=N
R=dslomov@chromium.org
Review URL: https://codereview.chromium.org/
917933007
Cr-Commit-Position: refs/heads/master@{#26633}
Erik Arvidsson [Thu, 12 Feb 2015 20:53:03 +0000 (15:53 -0500)]
Fix location of super syntax errors
BUG=v8:3885
LOG=N
R=adamk@chromium.org, dslomov@chromium.org
Review URL: https://codereview.chromium.org/
916293004
Cr-Commit-Position: refs/heads/master@{#26632}
Dmitry Lomov [Thu, 12 Feb 2015 20:06:52 +0000 (21:06 +0100)]
Remove --experimental-classes flag and related dead code.
R=arv@chromium.org
BUG=v8:3834
LOG=N
Review URL: https://codereview.chromium.org/
919643008
Cr-Commit-Position: refs/heads/master@{#26631}
loislo [Thu, 12 Feb 2015 19:51:26 +0000 (11:51 -0800)]
CPUProfiler: Push deopt reason further to ProfileNode.
1) create beefy RelocInfo table when cpu profiler is active, so if a function
was optimized when profiler was active RelocInfo would get separate DeoptInfo
for the each deopt case.
2) push DeoptInfo from CodeEntry to ProfileNode.
When deopt happens we put the info collected on #1 into CodeEntry and record stack sample.
On the sampling thread we grab the deopt data and append it to the corresponding ProfileNode deopts list.
Sample profile dump.
[Top down]:
0 (root) 0 #1
1 29 #2
1 test 29 #3
2 opt_function 29 #4
2 opt_function 29 #5
deopted at 118 with reason 'not a heap number'
deopted at 137 with reason 'division by zero'
BUG=452067
LOG=n
Committed: https://crrev.com/
ce8701b247d3c6604f24f17a90c02d17b4417f54
Cr-Commit-Position: refs/heads/master@{#26615}
Review URL: https://codereview.chromium.org/
919953002
Cr-Commit-Position: refs/heads/master@{#26630}
machenbach [Thu, 12 Feb 2015 18:34:21 +0000 (10:34 -0800)]
Create v8 roll branches in refs/heads.
This is needed as normal chromium checkouts don't fetch
branch-heads.
These branches will also be replicated by GitHub and are
fetched when cloning from the repository.
The expected object size will be the same or smaller than
before as we currently push the same information onto the
candidates branch (which already is in refs/heads). The only
difference is the increasing number of refs in refs/heads.
BUG=chromium:451975
TBR=tandrii@chromium.org
NOTRY=true
LOG=n
TEST=./script_test.py
TEST=tools/release/create_release.py -r machenbach@chromium.org -a machenbach@chromium.org --dry-run
Review URL: https://codereview.chromium.org/
914973002
Cr-Commit-Position: refs/heads/master@{#26629}
Dmitry Lomov [Thu, 12 Feb 2015 18:06:24 +0000 (19:06 +0100)]
new classes: no longer experimental.
This CL fixes tests that no longer valid and also fixes two issues:
1. 'super()' in non derived constructors.
2. Failure to step into derived constructors.
R=arv@chromium.org, yurys@chromium.org
BUG=v8:3834
LOG=Y
Review URL: https://codereview.chromium.org/
923443003
Cr-Commit-Position: refs/heads/master@{#26628}
ishell@chromium.org [Thu, 12 Feb 2015 17:26:57 +0000 (18:26 +0100)]
Fix for the compilation issue in shared mode introduced by r26624
TBR=verwaest@chromium.org
Review URL: https://codereview.chromium.org/
919043002
Cr-Commit-Position: refs/heads/master@{#26627}
mstarzinger [Thu, 12 Feb 2015 17:08:47 +0000 (09:08 -0800)]
Add missing FrameState to JSTypedLoweringTester::Binop.
R=titzer@chromium.org
TEST=cctest/test-js-typed-lowering
Review URL: https://codereview.chromium.org/
925583002
Cr-Commit-Position: refs/heads/master@{#26626}
yangguo [Thu, 12 Feb 2015 16:47:50 +0000 (08:47 -0800)]
Mark pages created during bootstrapping as never-evacuate.
This is to ensure that immutable immortal objects created during
bootstrapping are not relocated.
R=hpayer@chromium.org
Review URL: https://codereview.chromium.org/
905773004
Cr-Commit-Position: refs/heads/master@{#26625}
ishell [Thu, 12 Feb 2015 16:40:26 +0000 (08:40 -0800)]
Enable layout descriptor consistency checks in heap verification mode.
Review URL: https://codereview.chromium.org/
917743004
Cr-Commit-Position: refs/heads/master@{#26624}
rossberg [Thu, 12 Feb 2015 16:29:42 +0000 (08:29 -0800)]
Properly thread language mode to compilation cache
Fixes the TSAN issue.
R=mstarzinger@chromium.org
NOTRY=true
BUG=
Review URL: https://codereview.chromium.org/
923573002
Cr-Commit-Position: refs/heads/master@{#26623}
ulan [Thu, 12 Feb 2015 16:00:25 +0000 (08:00 -0800)]
Fix GCMole after
b79b985988fd
NOTREECHECKS=true
BUG=
TBR=verwaest@chromium.org
Review URL: https://codereview.chromium.org/
917193002
Cr-Commit-Position: refs/heads/master@{#26622}
jarin [Thu, 12 Feb 2015 15:21:17 +0000 (07:21 -0800)]
[turbofan] Make the representation type component independent of the semantic component.
R=rossberg@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
904863002
Cr-Commit-Position: refs/heads/master@{#26621}
marja [Thu, 12 Feb 2015 15:12:32 +0000 (07:12 -0800)]
Parser / PreParser: trivial unifications
(To minimize the diff of actually interesting unifications.)
R=rossberg@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
915383002
Cr-Commit-Position: refs/heads/master@{#26620}
rossberg [Thu, 12 Feb 2015 15:09:55 +0000 (07:09 -0800)]
[strong] no sloppy equality
R=marja@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
917703003
Cr-Commit-Position: refs/heads/master@{#26619}
chunyang.dai [Thu, 12 Feb 2015 14:40:01 +0000 (06:40 -0800)]
X87: silence C++ compiler warnings.
To solence C++ compiler warnings because is_mutable_ is only for CHECK.
BUG=
Review URL: https://codereview.chromium.org/
922673002
Cr-Commit-Position: refs/heads/master@{#26618}
jochen [Thu, 12 Feb 2015 14:38:54 +0000 (06:38 -0800)]
Make it possible to define arguments for CompileFunctionInContext
Also make sure that the function body really produces only one function literal.
LOG=y
BUG=none
R=yangguo@chromium.org
Review URL: https://codereview.chromium.org/
925433002
Cr-Commit-Position: refs/heads/master@{#26617}
loislo [Thu, 12 Feb 2015 14:36:21 +0000 (06:36 -0800)]
Revert of CPUProfiler: Push deopt reason further to ProfileNode. (patchset #1 id:1 of https://codereview.chromium.org/
919953002/)
Reason for revert:
static initializers broke the build
Original issue's description:
> CPUProfiler: Push deopt reason further to ProfileNode.
>
> 1) create beefy RelocInfo table when cpu profiler is active, so if a function
> was optimized when profiler was active RelocInfo would get separate DeoptInfo
> for the each deopt case.
>
> 2) push DeoptInfo from CodeEntry to ProfileNode.
> When deopt happens we put the info collected on #1 into CodeEntry and record stack sample.
> On the sampling thread we grab the deopt data and append it to the corresponding ProfileNode deopts list.
>
> Sample profile dump.
> [Top down]:
> 0 (root) 0 #1
> 1 29 #2
> 5 test 29 #3
> 3 opt_function 29 #4
> deopted at 52 with reason 'not a heap number'
> deopted at 71 with reason 'division by zero'
>
> BUG=452067
> LOG=n
>
> Committed: https://crrev.com/
ce8701b247d3c6604f24f17a90c02d17b4417f54
> Cr-Commit-Position: refs/heads/master@{#26615}
TBR=jarin@chromium.org,svenpanne@chromium.org,yurys@chromium.org,alph@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=452067
Review URL: https://codereview.chromium.org/
915173005
Cr-Commit-Position: refs/heads/master@{#26616}
loislo [Thu, 12 Feb 2015 13:24:59 +0000 (05:24 -0800)]
CPUProfiler: Push deopt reason further to ProfileNode.
1) create beefy RelocInfo table when cpu profiler is active, so if a function
was optimized when profiler was active RelocInfo would get separate DeoptInfo
for the each deopt case.
2) push DeoptInfo from CodeEntry to ProfileNode.
When deopt happens we put the info collected on #1 into CodeEntry and record stack sample.
On the sampling thread we grab the deopt data and append it to the corresponding ProfileNode deopts list.
Sample profile dump.
[Top down]:
0 (root) 0 #1
1 29 #2
5 test 29 #3
3 opt_function 29 #4
deopted at 52 with reason 'not a heap number'
deopted at 71 with reason 'division by zero'
BUG=452067
LOG=n
Review URL: https://codereview.chromium.org/
919953002
Cr-Commit-Position: refs/heads/master@{#26615}
ulan [Thu, 12 Feb 2015 13:21:06 +0000 (05:21 -0800)]
Use weak cells in dependent code.
BUG=
Review URL: https://codereview.chromium.org/
871253005
Cr-Commit-Position: refs/heads/master@{#26614}
mstarzinger [Thu, 12 Feb 2015 13:14:33 +0000 (05:14 -0800)]
Fix missing visitation of FrameState in simplified lowering.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/
918143002
Cr-Commit-Position: refs/heads/master@{#26613}
marja [Thu, 12 Feb 2015 13:02:30 +0000 (05:02 -0800)]
Parsing: Make Parser not know about Isolate during background parsing.
Parser must be able to operate independent of Isolate and the V8 heap during
parsing. After the heap-independent phase, there is a heap dependent phase,
during which we internalize strings, handle errors, etc.
This makes Isolate (also via CompilationInfo) unaccessible during parsing, and
thus decreases the probability of accidental code changes which would add
heap-dependent operations into the heap-independent phase.
Since Isolate is also accessible via CompilationInfo, now CompilationInfo is
only passed to the entry points of parsing, and not stored in Parser.
R=rossberg@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
908173003
Cr-Commit-Position: refs/heads/master@{#26612}
bmeurer [Thu, 12 Feb 2015 13:01:21 +0000 (05:01 -0800)]
[turbofan] Merge node-aux-data-inl.h into node-aux-data.h.
Google style guide forbids -inl.h headers.
R=svenpanne@chromium.org
Review URL: https://codereview.chromium.org/
922633002
Cr-Commit-Position: refs/heads/master@{#26611}
svenpanne [Thu, 12 Feb 2015 12:57:07 +0000 (04:57 -0800)]
Removed one bogus CompilationInfo constructor.
Use a fake code stub instead, basically following the null object pattern.
Review URL: https://codereview.chromium.org/
918973002
Cr-Commit-Position: refs/heads/master@{#26610}
bmeurer [Thu, 12 Feb 2015 12:46:58 +0000 (04:46 -0800)]
Fix invalid use of int in Zone.
Review URL: https://codereview.chromium.org/
924453002
Cr-Commit-Position: refs/heads/master@{#26609}
danno [Thu, 12 Feb 2015 12:15:14 +0000 (04:15 -0800)]
Remove redundant source position information in RelocInfo
Previously, emitting two more more unique source positions at the same pc would
generate two or more RelocInfo entries. Now, only the last emitted source
position for any pc is added to the RelocInfo.
Review URL: https://codereview.chromium.org/
908443002
Cr-Commit-Position: refs/heads/master@{#26608}
hpayer [Thu, 12 Feb 2015 11:58:24 +0000 (03:58 -0800)]
Fix JSArrayBufferView visitor template to fix ARM build.
NOTREECHECKS=true
TBR=ulan@chromium.org
Review URL: https://codereview.chromium.org/
918113002
Cr-Commit-Position: refs/heads/master@{#26607}
Ben L. Titzer [Thu, 12 Feb 2015 11:56:57 +0000 (12:56 +0100)]
Inline the ParseInfo structure as parameters to the Parser constructor.
R=mstarzinger@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
913183005
Cr-Commit-Position: refs/heads/master@{#26606}
hpayer [Thu, 12 Feb 2015 10:00:26 +0000 (02:00 -0800)]
Just visit young array buffers during scavenge. Additionally keep the views in new space in a separate global list and move them to the corresponding array buffers when they get promoted.
This reduces young generation garbage collections when many array buffers are allocated.
BUG=
Review URL: https://codereview.chromium.org/
904633003
Cr-Commit-Position: refs/heads/master@{#26605}
mstarzinger [Thu, 12 Feb 2015 09:16:18 +0000 (01:16 -0800)]
Fix nasty off-by-one in the TurboFan JSON tracer.
R=danno@chromium.org
TEST=cctest/test-run-intrinsics/CallFunction
Review URL: https://codereview.chromium.org/
916873002
Cr-Commit-Position: refs/heads/master@{#26604}
cdai2 [Thu, 12 Feb 2015 08:10:56 +0000 (16:10 +0800)]
X87: new classes: implement correct check for uninitialized this in 'super()'
port
fdcf3e59ba4af1aa3c5b31a18e615169c97958ab (r26599)
original commit message:
new classes: implement correct check for uninitialized this in 'super()'
BUG=
R=weiliang.lin@intel.com
Review URL: https://codereview.chromium.org/
913383002
Cr-Commit-Position: refs/heads/master@{#26603}
bmeurer [Thu, 12 Feb 2015 07:21:06 +0000 (23:21 -0800)]
[ycm] The YouCompleteMe configuration is not VIM specific.
Rename tools/vim/v8.ycm_extra_conf.py to .ycm_extra_conf.py, so every
YCM plugin (no matter if it's VIM or Emacs) should pick it up
automatically.
Also move the common tools/vim/ninja_output.py to tools/ninja/ as it's
also not related to VIM.
Review URL: https://codereview.chromium.org/
904143003
Cr-Commit-Position: refs/heads/master@{#26602}
cdai2 [Thu, 12 Feb 2015 03:51:22 +0000 (11:51 +0800)]
X87: new classes: implement default constructors
port
9b158fa79a510bd1fee2f663163d9116879cb33a (r26594)
original commit message:
new classes: implement default constructors
BUG=
R=weiliang.lin@intel.com
Review URL: https://codereview.chromium.org/
914403004
Cr-Commit-Position: refs/heads/master@{#26601}
dcarney [Wed, 11 Feb 2015 21:29:17 +0000 (13:29 -0800)]
fix typo in mjsunit expectations
TBR=machenbach@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
915113004
Cr-Commit-Position: refs/heads/master@{#26600}
dslomov [Wed, 11 Feb 2015 20:47:22 +0000 (12:47 -0800)]
new classes: implement correct check for uninitialized this in 'super()'
R=arv@chromium.org
BUG=v8:3834
LOG=N
Review URL: https://codereview.chromium.org/
918603004
Cr-Commit-Position: refs/heads/master@{#26599}
akos.palfi [Wed, 11 Feb 2015 20:41:53 +0000 (12:41 -0800)]
MIPS: new classes: implement default constructors.
Port
9b158fa79a510bd1fee2f663163d9116879cb33a
BUG=
Review URL: https://codereview.chromium.org/
914303003
Cr-Commit-Position: refs/heads/master@{#26598}
akos.palfi [Wed, 11 Feb 2015 19:54:13 +0000 (11:54 -0800)]
MIPS: new classes: implement new.target passing to superclass constructor.
Port
bf49be39f333bf1ce14202663595d455aa4b24cb
BUG=
Review URL: https://codereview.chromium.org/
906813003
Cr-Commit-Position: refs/heads/master@{#26597}
chunyang.dai [Wed, 11 Feb 2015 17:41:53 +0000 (09:41 -0800)]
X87: new classes: implement new.target passing to superclass constructor
port
bf49be39f333bf1ce14202663595d455aa4b24cb (r26572)
original commit message:
new classes: implement new.target passing to superclass constructor
BUG=
Review URL: https://codereview.chromium.org/
912233003
Cr-Commit-Position: refs/heads/master@{#26596}
chunyang.dai [Wed, 11 Feb 2015 17:31:59 +0000 (09:31 -0800)]
Copy the corresponding opt_count_and_bailout_reason info when invoking Runtime_SetCode.
This error is exposed by this CL: https://codereview.chromium.org/
817293005.
This CL set all Buildin JS function to optimization_disabled.
And set the bailout reason to "kBuiltinFunctionCannotBeOptimized". But in Runtime_SetCode() function.
It copied the the compiler_hints which include the "bool optimization_disabled" value,
But the corresponding Bailout reason is not copied.
This leads to the check error in "CompilationInfo::AbortOptimization(...)" function.
This issue is exposed on turbofan unsupported X87 platform. Crankshaft is invoked to compile
the typed array function Uint8Array.
BUG=
Review URL: https://codereview.chromium.org/
917643002
Cr-Commit-Position: refs/heads/master@{#26595}
dslomov [Wed, 11 Feb 2015 17:22:50 +0000 (09:22 -0800)]
new classes: implement default constructors.
R=arv@chromium.org,rossberg@chromium.org
BUG=v8:3834
LOG=N
Review URL: https://codereview.chromium.org/
917753002
Cr-Commit-Position: refs/heads/master@{#26594}
loislo [Wed, 11 Feb 2015 17:11:23 +0000 (09:11 -0800)]
Fix for arm64 after v8:r26448
The offensive cl is https://codereview.chromium.org/
874323003/
Test:
./out/arm64.debug/d8 --test --random-seed=-
235865360 --turbo-deoptimization --turbo-filter=* --always-opt --debug-code --verify-heap --gc-interval=500 --stress-compaction test/mjsunit/mjsunit.js test/mjsunit/regress/regress-builtinbust-7.js --trace-deopt
BUG=452067
TBR=dcarney, svenpanne
LOG=n
Review URL: https://codereview.chromium.org/
917823002
Cr-Commit-Position: refs/heads/master@{#26593}
rodolph.perfetta [Wed, 11 Feb 2015 16:52:20 +0000 (08:52 -0800)]
Fix issue when compiling with gcc 4.6.
BUG=
Review URL: https://codereview.chromium.org/
905383003
Cr-Commit-Position: refs/heads/master@{#26592}
mstarzinger [Wed, 11 Feb 2015 16:21:01 +0000 (08:21 -0800)]
Remove redundant check from ControlEquivalence traversal.
R=jarin@chromium.org
TEST=unittests/ControlEquivalenceTest
Review URL: https://codereview.chromium.org/
916823002
Cr-Commit-Position: refs/heads/master@{#26591}
cdai2 [Wed, 11 Feb 2015 15:50:17 +0000 (23:50 +0800)]
X87: Parsing: Make Scope not know about Isolate.
port
5d68529be240a1fc9fcc67c8950ead4baf7604e4 (r26546)
original commit message:
Parsing: Make Scope not know about Isolate.
Scope, like Parser, must be able to operate independent of Isolate and the V8
heap (for background parsing). After the heap-independent phase, there is a heap
dependent phase, during which we do operations such as scope anaylysis.
This CL makes the phases explicit by not telling Scope about the Isolate too
early (during the heap-independent phase, Scope should know nothing about
Isolate). This decreases the probability of accidental code changes which would
add heap-dependent operations into the heap-independent phase.
BUG=
R=weiliang.lin@intel.com
Review URL: https://codereview.chromium.org/
913163003
Cr-Commit-Position: refs/heads/master@{#26590}
titzer [Wed, 11 Feb 2015 15:49:02 +0000 (07:49 -0800)]
Reduce the number of #includes of compiler.h.
R=mstarzinger@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
917583004
Cr-Commit-Position: refs/heads/master@{#26589}
cdai2 [Wed, 11 Feb 2015 15:36:48 +0000 (23:36 +0800)]
X87: Propagate Deopt reason to cpu-profiler
port
86cae1633caa31f2225b6932850fb13d3a21cf59 (r26545)
original commit message:
1) Deoptimizer::Reason was replaced with Deoptimizer::DeoptInfo
because it also has raw position. Also the old name clashes with DeoptReason enum.
2) c_entry_fp assignment call was added to EntryGenerator::Generate
So we can calculate sp and have a chance to record the stack for the deopting function.
btw it makes the test stable.
3) new kind of CodeEvents was added to cpu-profiler
4) GetDeoptInfo method was extracted from PrintDeoptLocation.
So it could be reused in cpu profiler.
BUG=
R=weiliang.lin@intel.com
Review URL: https://codereview.chromium.org/
912403003
Cr-Commit-Position: refs/heads/master@{#26588}
adamk [Wed, 11 Feb 2015 15:36:27 +0000 (07:36 -0800)]
Remove stray comment from ast.cc
This comment used to have a (commented-out) ASSERT attached to it.
But that ASSERT was removed years ago, so the comment is now a
non-sequitar.
Review URL: https://codereview.chromium.org/
918453002
Cr-Commit-Position: refs/heads/master@{#26587}
arv [Wed, 11 Feb 2015 15:13:05 +0000 (07:13 -0800)]
Fix issue with super and computed property names
We did not set up the [[HomeObject]] for properties created for
computed property names.
BUG=v8:3879
LOG=N
R=dslomov@chromium.org
Review URL: https://codereview.chromium.org/
914773002
Cr-Commit-Position: refs/heads/master@{#26586}
arv [Wed, 11 Feb 2015 15:11:54 +0000 (07:11 -0800)]
Remove dead code related to new super
Support for `new super` has been removed so this code in the full
code gen is now unreachable.
BUG=None
LOG=N
R=dslomov@chromium.org
Review URL: https://codereview.chromium.org/
911223002
Cr-Commit-Position: refs/heads/master@{#26585}
mstarzinger [Wed, 11 Feb 2015 15:02:32 +0000 (07:02 -0800)]
Mark some common operator with Property::kNoThrow.
R=bmeurer@chromium.org
TEST=unittests/CommonOperatorTest
Review URL: https://codereview.chromium.org/
912393002
Cr-Commit-Position: refs/heads/master@{#26584}
titzer [Wed, 11 Feb 2015 14:51:59 +0000 (06:51 -0800)]
Move SetFunctionInfo() from compiler.cc to objects.cc. Rationale: not related to generating code.
R=verwaest@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
917433007
Cr-Commit-Position: refs/heads/master@{#26583}
dcarney [Wed, 11 Feb 2015 14:35:52 +0000 (06:35 -0800)]
skip failing test on arm64
LOG=N
BUG=v8:3884
Review URL: https://codereview.chromium.org/
916803002
Cr-Commit-Position: refs/heads/master@{#26582}
Yang Guo [Wed, 11 Feb 2015 14:32:15 +0000 (15:32 +0100)]
Fix shared library build.
TBR=vogelheim@chromium.org
Review URL: https://codereview.chromium.org/
913173002
Cr-Commit-Position: refs/heads/master@{#26581}
svenpanne [Wed, 11 Feb 2015 14:12:03 +0000 (06:12 -0800)]
Removed most of the bogus CompilationInfo constructor calls.
A CompilationInfo constructed from just an Isolate* and a Zone* is in
weird an inconsistent state (calling e.g. flags() on it will crash),
so we need to avoid them. This CL removes almost all of them, the
remaining 2 call sites in (for testing only) will be handled in a
separate CL. Things which have been changed:
* Linkage is basically a decorator for CallDescriptor now.
* ChangeLowering doesn't need Linkage at all.
* JSGenericLowering doesn't need a full CompilationInfo*, just a
single flag.
* JSContextSpecializer doesn't need the full CompilationInfo, just a
Context.
* Removed unused CompilationInfo from SimplifiedLoweringTester.
This nicely decouples things already a bit more, but there's still
work to do...
Review URL: https://codereview.chromium.org/
899803003
Cr-Commit-Position: refs/heads/master@{#26580}
yangguo [Wed, 11 Feb 2015 14:10:02 +0000 (06:10 -0800)]
Catch graceful failures when deserializing in d8.
R=vogelheim@chromium.org
Review URL: https://codereview.chromium.org/
917653002
Cr-Commit-Position: refs/heads/master@{#26579}
yangguo [Wed, 11 Feb 2015 14:03:10 +0000 (06:03 -0800)]
Whitelist serialized objects wrt MSAN.
R=jochen@chromium.org
BUG=chromium:457459
LOG=N
Review URL: https://codereview.chromium.org/
919613002
Cr-Commit-Position: refs/heads/master@{#26578}
hpayer [Wed, 11 Feb 2015 13:39:29 +0000 (05:39 -0800)]
Use just one to-space page for the promotion queue.
BUG=454725
LOG=n
Review URL: https://codereview.chromium.org/
919473008
Cr-Commit-Position: refs/heads/master@{#26577}
titzer [Wed, 11 Feb 2015 13:26:35 +0000 (05:26 -0800)]
[turbofan] Use heavy-handed graph duplication to do loop peeling for OSR.
BUG=
Review URL: https://codereview.chromium.org/
898353002
Cr-Commit-Position: refs/heads/master@{#26576}
yangguo [Wed, 11 Feb 2015 12:06:11 +0000 (04:06 -0800)]
Fix NewStringRangeError failure due to failing malloc.
R=jkummerow@chromium.org
Review URL: https://codereview.chromium.org/
912333002
Cr-Commit-Position: refs/heads/master@{#26575}
yangguo [Wed, 11 Feb 2015 10:21:18 +0000 (02:21 -0800)]
Throw on range error when creating a string via API.
R=jkummerow@chromium.org
BUG=v8:3853
LOG=Y
Review URL: https://codereview.chromium.org/
867373003
Cr-Commit-Position: refs/heads/master@{#26574}
ulan [Wed, 11 Feb 2015 09:59:41 +0000 (01:59 -0800)]
Reland r26549 "Fix IsWeakObjectInOptimizedCode predicate to also include property cells."
BUG=v8:3882
LOG=NO
Review URL: https://codereview.chromium.org/
913093002
Cr-Commit-Position: refs/heads/master@{#26573}
dslomov [Wed, 11 Feb 2015 09:47:32 +0000 (01:47 -0800)]
new classes: implement new.target passing to superclass constructor.
R=arv@chromium.org,rossberg@chromium.org
BUG=v8:3834
LOG=N
Committed: https://crrev.com/
8aed43e82c6d2742fe5988603cb8841324cc942b
Cr-Commit-Position: refs/heads/master@{#26560}
Review URL: https://codereview.chromium.org/
908883002
Cr-Commit-Position: refs/heads/master@{#26572}
dcarney [Wed, 11 Feb 2015 09:28:06 +0000 (01:28 -0800)]
remove undetectable strings
BUG=
Review URL: https://codereview.chromium.org/
916753002
Cr-Commit-Position: refs/heads/master@{#26571}
mstarzinger [Wed, 11 Feb 2015 09:25:37 +0000 (01:25 -0800)]
Remove obsolete Malloced::FatalProcessOutOfMemory.
R=hpayer@chromium.org
Review URL: https://codereview.chromium.org/
897213003
Cr-Commit-Position: refs/heads/master@{#26570}
dcarney [Wed, 11 Feb 2015 09:15:19 +0000 (01:15 -0800)]
add transitions for global properties in ics
R=verwaest@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
911713003
Cr-Commit-Position: refs/heads/master@{#26569}
dusan.milosavljevic [Wed, 11 Feb 2015 09:05:16 +0000 (01:05 -0800)]
Make SNaN verification universal for all arches.
The mips HW prior to revision 5 has opposite encoding for NaNs.
TEST=mjsunit/regress/*, regress-1167, regress-undefined-nan2
BUG=
Review URL: https://codereview.chromium.org/
912833003
Cr-Commit-Position: refs/heads/master@{#26568}
svenpanne [Wed, 11 Feb 2015 08:52:32 +0000 (00:52 -0800)]
Relaxed parameter types a bit.
Somehow these changes didn't make it into https://codereview.chromium.org/
915583002 ... :-/
Review URL: https://codereview.chromium.org/
919473005
Cr-Commit-Position: refs/heads/master@{#26567}
Benedikt Meurer [Wed, 11 Feb 2015 07:42:53 +0000 (08:42 +0100)]
[vim] Pass correct -std flags to YCM libclang.
TBR=verwaest@chromium.org
Review URL: https://codereview.chromium.org/
915933003
Cr-Commit-Position: refs/heads/master@{#26566}
loislo [Wed, 11 Feb 2015 07:26:59 +0000 (23:26 -0800)]
Remove obsolete method ProfileTree::AddPathFromStart and the corresponding test.
BUG=
LOG=n
Review URL: https://codereview.chromium.org/
907353003
Cr-Commit-Position: refs/heads/master@{#26565}
bmeurer [Wed, 11 Feb 2015 07:11:56 +0000 (23:11 -0800)]
[vim] Add YouCompleteMe configuration.
R=verwaest@chromium.org
Review URL: https://codereview.chromium.org/
915983002
Cr-Commit-Position: refs/heads/master@{#26564}
bmeurer [Wed, 11 Feb 2015 06:59:00 +0000 (22:59 -0800)]
[vim] Add ninja-powered CrBuild/CrCompileFile functions.
These are based on the ninja-build.vim file found in Chromium, adapted
to V8, and enhanced with goma support (automatically increase job limit
for ninja if goma is detected).
Review URL: https://codereview.chromium.org/
915963002
Cr-Commit-Position: refs/heads/master@{#26563}
v8-autoroll [Wed, 11 Feb 2015 04:34:47 +0000 (20:34 -0800)]
Update V8 DEPS.
Rolling v8/tools/clang to
e64bdee4434b184d13a8f2cce79b65cd18e02ee2
TBR=machenbach@chromium.org
Review URL: https://codereview.chromium.org/
919473003
Cr-Commit-Position: refs/heads/master@{#26562}
dslomov [Wed, 11 Feb 2015 01:36:17 +0000 (17:36 -0800)]
Revert of new classes: implement new.target passing to superclass constructor. (patchset #9 id:150001 of https://codereview.chromium.org/
908883002/)
Reason for revert:
Breaks Linux64 release
Original issue's description:
> new classes: implement new.target passing to superclass constructor.
>
> R=arv@chromium.org,rossberg@chromium.org
> BUG=v8:3834
> LOG=N
>
> Committed: https://crrev.com/
8aed43e82c6d2742fe5988603cb8841324cc942b
> Cr-Commit-Position: refs/heads/master@{#26560}
TBR=arv@chromium.org,rossberg@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:3834
Review URL: https://codereview.chromium.org/
911363002
Cr-Commit-Position: refs/heads/master@{#26561}
dslomov [Wed, 11 Feb 2015 01:23:03 +0000 (17:23 -0800)]
new classes: implement new.target passing to superclass constructor.
R=arv@chromium.org,rossberg@chromium.org
BUG=v8:3834
LOG=N
Review URL: https://codereview.chromium.org/
908883002
Cr-Commit-Position: refs/heads/master@{#26560}
arv [Tue, 10 Feb 2015 22:13:29 +0000 (14:13 -0800)]
Remove Function.prototype.toMethod
Function.prototype.toMethod was removed from ES6.
This removes the function and updates the tests to either
use %ToMethod or a dedicated syntax (using concise method
or a class).
BUG=v8:3330
LOG=N
R=dslomov@chromium.org, adamk
Review URL: https://codereview.chromium.org/
914713002
Cr-Commit-Position: refs/heads/master@{#26559}
Erik Arvidsson [Tue, 10 Feb 2015 21:15:42 +0000 (16:15 -0500)]
Fix usage of super in js perf test
BUG=none
LOG=N
R=adamk@chromium.org, adamk, dslomov@chromium.org
Review URL: https://codereview.chromium.org/
916573002
Cr-Commit-Position: refs/heads/master@{#26558}
arv [Tue, 10 Feb 2015 19:51:18 +0000 (11:51 -0800)]
super is only allowed in methods, accessors and constructor
super() is only allowed in a class constructor.
super.p is allowed in methods, accessors and constructors.
The parser now checks the FunctionState to see what kind of function
we are currently inside.
BUG=v8:3330
LOG=N
R=dslomov@chromium.org, marja@chromium.org
Review URL: https://codereview.chromium.org/
915563003
Cr-Commit-Position: refs/heads/master@{#26557}
rossberg [Tue, 10 Feb 2015 19:12:51 +0000 (11:12 -0800)]
[strong] Introduce --use-strong flag
R=marja@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
907403002
Cr-Commit-Position: refs/heads/master@{#26556}
adamk [Tue, 10 Feb 2015 19:11:44 +0000 (11:11 -0800)]
Enable compiling mjsunit tests as ES6 modules
Adding the line "// MODULE" to an mjsunit file will now cause
run-tests.py to prefix the test case with "--module" in the
d8 commandline.
d8 has itself been updated to treat files preceded with "--module" as
modules (that is, it compiles them with ScriptCompiler::CompileModule,
and turns on --harmony-modules).
Review URL: https://codereview.chromium.org/
902263002
Cr-Commit-Position: refs/heads/master@{#26555}
balazs.kilvady [Tue, 10 Feb 2015 19:10:37 +0000 (11:10 -0800)]
MIPS64: Propagate DeoptInfo to cpu-profiler
Port
86cae1633caa31f2225b6932850fb13d3a21cf59
Original commit message:
1) Deoptimizer::Reason was replaced with Deoptimizer::DeoptInfo
because it also has raw position. Also the old name clashes with DeoptReason enum.
2) c_entry_fp assignment call was added to EntryGenerator::Generate
So we can calculate sp and have a chance to record the stack for the deopting function.
btw it makes the test stable.
3) new kind of CodeEvents was added to cpu-profiler
4) GetDeoptInfo method was extracted from PrintDeoptLocation.
So it could be reused in cpu profiler.
BUG=452067
LOG=n
Review URL: https://codereview.chromium.org/
907383003
Cr-Commit-Position: refs/heads/master@{#26554}
balazs.kilvady [Tue, 10 Feb 2015 19:09:25 +0000 (11:09 -0800)]
MIPS: Parsing: Make Scope not know about Isolate.
Port
5d68529be240a1fc9fcc67c8950ead4baf7604e4
Original commit message:
Scope, like Parser, must be able to operate independent of Isolate and the V8
heap (for background parsing). After the heap-independent phase, there is a heap
dependent phase, during which we do operations such as scope anaylysis.
This CL makes the phases explicit by not telling Scope about the Isolate too
early (during the heap-independent phase, Scope should know nothing about
Isolate). This decreases the probability of accidental code changes which would
add heap-dependent operations into the heap-independent phase.
BUG=
Review URL: https://codereview.chromium.org/
911073004
Cr-Commit-Position: refs/heads/master@{#26553}
Michael Achenbach [Tue, 10 Feb 2015 17:00:41 +0000 (18:00 +0100)]
Revert "Whitespace CL to test poller."
This reverts commit
f27de256c12ab0d5f152ab3ad66950caa5158d03.
Cr-Commit-Position: refs/heads/master@{#26552}
machenbach [Tue, 10 Feb 2015 16:58:39 +0000 (08:58 -0800)]
Revert of Fix IsWeakObjectInOptimizedCode predicate to also include property cells. (patchset #1 id:1 of https://codereview.chromium.org/
909123002/)
Reason for revert:
Seems to fail several slow asserts (not on the try server):
http://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20debug/builds/1841
Original issue's description:
> Fix IsWeakObjectInOptimizedCode predicate to also include property cells.
>
> BUG=v8:3882
> LOG=NO
>
> Committed: https://crrev.com/
f8c404849dd01b538e52d95412225c06a0101808
> Cr-Commit-Position: refs/heads/master@{#26549}
TBR=dcarney@chromium.org,ulan@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:3882
Review URL: https://codereview.chromium.org/
909973004
Cr-Commit-Position: refs/heads/master@{#26551}
ishell [Tue, 10 Feb 2015 16:42:16 +0000 (08:42 -0800)]
Revert of Temporarily disable double fields unboxing. (patchset #1 id:1 of https://codereview.chromium.org/
893843004/)
Reason for revert:
Experiment is over, turning it on back.
Original issue's description:
> Temporarily disable double fields unboxing.
>
> Committed: https://crrev.com/
51c59ede18831d1c1a1d70fa9fa456f946ac486b
> Cr-Commit-Position: refs/heads/master@{#26492}
TBR=jochen@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/
912673003
Cr-Commit-Position: refs/heads/master@{#26550}
ulan [Tue, 10 Feb 2015 16:09:12 +0000 (08:09 -0800)]
Fix IsWeakObjectInOptimizedCode predicate to also include property cells.
BUG=v8:3882
LOG=NO
Review URL: https://codereview.chromium.org/
909123002
Cr-Commit-Position: refs/heads/master@{#26549}
Michael Achenbach [Tue, 10 Feb 2015 16:03:39 +0000 (17:03 +0100)]
Whitespace CL to test poller.
Cr-Commit-Position: refs/heads/master@{#26548}
vogelheim [Tue, 10 Feb 2015 15:38:01 +0000 (07:38 -0800)]
Fix cctest + unittest to work with an external snapshot.
To do so, extract startup_data_util from d8 and use it those executables.
BUG=
Review URL: https://codereview.chromium.org/
913703002
Cr-Commit-Position: refs/heads/master@{#26547}
marja [Tue, 10 Feb 2015 14:39:00 +0000 (06:39 -0800)]
Parsing: Make Scope not know about Isolate.
Scope, like Parser, must be able to operate independent of Isolate and the V8
heap (for background parsing). After the heap-independent phase, there is a heap
dependent phase, during which we do operations such as scope anaylysis.
This CL makes the phases explicit by not telling Scope about the Isolate too
early (during the heap-independent phase, Scope should know nothing about
Isolate). This decreases the probability of accidental code changes which would
add heap-dependent operations into the heap-independent phase.
R=rossberg@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
909093003
Cr-Commit-Position: refs/heads/master@{#26546}
loislo [Tue, 10 Feb 2015 14:32:42 +0000 (06:32 -0800)]
Propagate DeoptInfo to cpu-profiler
1) Deoptimizer::Reason was replaced with Deoptimizer::DeoptInfo
because it also has raw position. Also the old name clashes with DeoptReason enum.
2) c_entry_fp assignment call was added to EntryGenerator::Generate
So we can calculate sp and have a chance to record the stack for the deopting function.
btw it makes the test stable.
3) new kind of CodeEvents was added to cpu-profiler
4) GetDeoptInfo method was extracted from PrintDeoptLocation.
So it could be reused in cpu profiler.
BUG=452067
LOG=n
Review URL: https://codereview.chromium.org/
910773002
Cr-Commit-Position: refs/heads/master@{#26545}
marja [Tue, 10 Feb 2015 13:27:08 +0000 (05:27 -0800)]
Get rid of PreParserScope.
It's unnecessary; PreParser can just use normal Scopes for the things it needs
to track. Note: the only functionalities of PreParserScope were keeping track of
the scope stack, and for each scope, the scope type and language mode. Those are
now done by Scope. PreParser doesn't yet put variables into scopes (that will be
done in a follow up).
R=rossberg@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
912563004
Cr-Commit-Position: refs/heads/master@{#26544}
v8-autoroll [Tue, 10 Feb 2015 12:52:46 +0000 (04:52 -0800)]
Update V8 DEPS.
Rolling v8/tools/clang to
4b5142322face2df2374a246dba98ae82be317b0
TBR=machenbach@chromium.org
Review URL: https://codereview.chromium.org/
872893010
Cr-Commit-Position: refs/heads/master@{#26543}
svenpanne [Tue, 10 Feb 2015 11:45:59 +0000 (03:45 -0800)]
Massage the CodeStub class hierarchy a bit.
Several tiny changes for the upcoming TurboFan-genrated handlers/stubs:
* Relaxed the type of code_stub().
* Made GetCodeKind() public, it effectly was like this, anyway.
* Const-corrected GetStubType().
Review URL: https://codereview.chromium.org/
915583002
Cr-Commit-Position: refs/heads/master@{#26542}