bmeurer [Fri, 18 Sep 2015 13:20:25 +0000 (06:20 -0700)]
[es6] Use the correct ToPrimitive in the Date Constructor.
This way we can finally remove the ES5 ToPrimitive builtin from
runtime.js, and the Date Constructor now properly supports
@@toPrimitive for the single argument case as well.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_layout_dbg,v8_linux_nosnap_dbg
R=rossberg@chromium.org
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/
1346893003
Cr-Commit-Position: refs/heads/master@{#30832}
mlippautz [Fri, 18 Sep 2015 12:41:19 +0000 (05:41 -0700)]
[base] Fix check that makes sure we commit in the virtual memory range.
R=jochen@chromium.org
BUG=chromium:533342
LOG=N
Review URL: https://codereview.chromium.org/
1349723005
Cr-Commit-Position: refs/heads/master@{#30831}
vogelheim [Fri, 18 Sep 2015 12:07:16 +0000 (05:07 -0700)]
Fix incorrect buffer length.
R=jochen@chromium.org
BUG=chromium:533243
LOG=N
Review URL: https://codereview.chromium.org/
1356863002
Cr-Commit-Position: refs/heads/master@{#30830}
chunyang.dai [Fri, 18 Sep 2015 12:00:48 +0000 (05:00 -0700)]
X87: Remove --pretenure-call-new
port
b5588f48fd0b4e3ee43be1fe6c19d7ddd8b8b5f1 (r30767).
original commit message:
There isn't a plan to turn it on soon, so we'll take it out in favor of cleaner code.
BUG=
Review URL: https://codereview.chromium.org/
1346043005
Cr-Commit-Position: refs/heads/master@{#30829}
chunyang.dai [Fri, 18 Sep 2015 11:59:31 +0000 (04:59 -0700)]
X87: [runtime] Initial step towards switching Execution::Call to callable.
port
d5bbd45f044ae6796c0d0f7bd8732069d74418de (r30808).
oringial commit message:
Currently Execution::Call (and friends) still duplicate a lot of the
Call sequence logic that should be encapsulated in the Call and
CallFunction builtins. So the plan now is to switch Execution::Call
to accept any Callable and just pass that through to the Call builtin.
BUG=
Review URL: https://codereview.chromium.org/
1350183005
Cr-Commit-Position: refs/heads/master@{#30828}
chunyang.dai [Fri, 18 Sep 2015 11:47:23 +0000 (04:47 -0700)]
X87: Vector ICs: Hook up vectors in platform builtins to their SharedFunctionInfos.
port
905e008c52ba06120f4a523aab00a53bc50830f1 (r30758)
BUG=
Review URL: https://codereview.chromium.org/
1352173002
Cr-Commit-Position: refs/heads/master@{#30827}
chunyang.dai [Fri, 18 Sep 2015 11:46:12 +0000 (04:46 -0700)]
X87: [builtins] Unify the String constructor.
port
a3d6f6cce317dbe1d31079eb81e15e49f3fb687a (r30759).
original commit message:
Implement the String constructor completely as native builtin,
avoiding the need to do gymnastics in JavaScript builtin to
properly detect the no argument case (which is different from
the undefined argument case) and also allowing to just
tailcall through to ToString or SymbolDescriptiveString for
the common case. Also the JavaScript builtin was misleading
since the case for construct call was unused, but could be
triggered in a wrong way once we support tail calls from
constructor functions.
This refactoring allows us to properly implement subclassing
for String builtins, once we have the correct initial_map on
derived classes (it's merely a matter of using NewTarget
instead of the target register now).
This introduces a new %SymbolDescriptiveString runtime
entry, which is also used by Symbol.toString() now.
BUG=
Review URL: https://codereview.chromium.org/
1349403002
Cr-Commit-Position: refs/heads/master@{#30826}
cbruni [Fri, 18 Sep 2015 11:15:34 +0000 (04:15 -0700)]
elements.cc cleanup
Reuse code for copying arguments and pushing/shifting elements.
BUG=
Review URL: https://codereview.chromium.org/
1346013005
Cr-Commit-Position: refs/heads/master@{#30825}
balazs.kilvady [Fri, 18 Sep 2015 11:08:17 +0000 (04:08 -0700)]
MIPS64: Optimize simulator.
Port
09f41681ef83fe9e9d79748e2a60f16b70d5934c
Original commit message:
The patch decreases the calls of huge switch instructions making the
DecodeType*() functions to work in one phase and optimizing
Instruction::InstructionType(). Speed gain in release full check is
about 33% (6:13 s -> 4:09 s) and in optdebug full test is about 50%
(12:29 -> 6:17)
BUG=
Review URL: https://codereview.chromium.org/
1356693002
Cr-Commit-Position: refs/heads/master@{#30824}
cbruni [Fri, 18 Sep 2015 11:03:30 +0000 (04:03 -0700)]
Fix --hydrogen-stats crashing on null_ptr for shared_info
BUG=
Review URL: https://codereview.chromium.org/
1350293002
Cr-Commit-Position: refs/heads/master@{#30823}
mlippautz [Fri, 18 Sep 2015 10:49:11 +0000 (03:49 -0700)]
[heap] Cleanup: Align naming of parallel sweeping with parallel compaction.
Pure refactoring.
R=hpayer@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
1354613002
Cr-Commit-Position: refs/heads/master@{#30822}
bmeurer [Fri, 18 Sep 2015 10:18:41 +0000 (03:18 -0700)]
[turbofan] Use StringCompareStub for string comparisons.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/
1353103002
Cr-Commit-Position: refs/heads/master@{#30821}
brettw [Fri, 18 Sep 2015 09:32:30 +0000 (02:32 -0700)]
Use public_deps for v8_base in GN.
The previous code took advantage of the fact that a group's deps are implicitly public, but I'm trying to fix that. This also cleans up some duplicated code between component and non-component builds.
Review URL: https://codereview.chromium.org/
1356723002
Cr-Commit-Position: refs/heads/master@{#30820}
jkummerow [Fri, 18 Sep 2015 09:15:39 +0000 (02:15 -0700)]
[hydrogen] Add crash-hunting instrumentation to Hydrogen too
This extends instrumentation added in r30683 and r30768 to cover
the possibility that the root cause we're after is in optimized code.
This CL is intended to be reverted in a couple of days, but should
cause no harm while it's in the tree (we would crash anyway).
BUG=chromium:527994
LOG=n
Review URL: https://codereview.chromium.org/
1348823003
Cr-Commit-Position: refs/heads/master@{#30819}
bmeurer [Fri, 18 Sep 2015 08:30:22 +0000 (01:30 -0700)]
[stubs] Refactor StringCompareStub and use it for HStringCompareAndBranch.
The StringCompareStub used to take its parameters on the (JavaScript)
stack, which made it impossible to use in TurboFan. Actually
StringCompareStub was currently completely unused. This changes the
calling convention to something TurboFan compatible and introduces a
CallInterfaceDescriptor for StringCompareStub. It also changes
HStringCompareAndBranch to use the StringCompareStub instead of using
the full blown CompareICStub for a stupid string comparison.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/
1347913003
Cr-Commit-Position: refs/heads/master@{#30818}
machenbach [Fri, 18 Sep 2015 08:06:14 +0000 (01:06 -0700)]
[test] Allow passing extra flags to perf tryjobs.
NOTRY=true
Review URL: https://codereview.chromium.org/
1342263003
Cr-Commit-Position: refs/heads/master@{#30817}
bmeurer [Fri, 18 Sep 2015 06:35:36 +0000 (23:35 -0700)]
[runtime] Replace COMPARE/COMPARE_STRONG with proper Object::Compare.
This removes the weird COMPARE and COMPARE_STRONG JavaScript builtins
and replaces them with a proper C++ implementation in Object::Compare
and appropriate wrappers Object::LessThan, Object::GreaterThan, and
friends that are intended to be used by a true/false returning CompareIC
in the future, as well as the interpreter. As a short-term solution we
provide %Compare and %Compare_Strong entry points for the current
CompareIC that return the appropriate integer values expected by
fullcodegen currently.
Now the Abstract Relational Comparison is also using the correct
ToPrimitive implementation, which properly supports @@toPrimitive.
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/
1350113002
Cr-Commit-Position: refs/heads/master@{#30816}
v8-autoroll [Fri, 18 Sep 2015 03:26:40 +0000 (20:26 -0700)]
Update V8 DEPS.
Rolling v8/tools/clang to
0e7dbd100f91096de47919f394ac0b32dd7c21a2
TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org
Review URL: https://codereview.chromium.org/
1355633003
Cr-Commit-Position: refs/heads/master@{#30815}
aperez [Thu, 17 Sep 2015 23:59:47 +0000 (16:59 -0700)]
Use a kMaxSafeInteger instead of Number.MAX_SAFE_INTEGER
Defines and uses a kMaxSafeInteger macro (which expands to the constant
2^53-1) instead of accessing Number.MAX_SAFE_INTEGER. This saves loading
the attribute from the Number object, which is slightly faster. This also
makes it clearer from reading the code that tha value being compared is
constant.
BUG=
LOG=N
Review URL: https://codereview.chromium.org/
1353953002
Cr-Commit-Position: refs/heads/master@{#30814}
mbrandy [Thu, 17 Sep 2015 20:39:45 +0000 (13:39 -0700)]
PPC: [runtime] Initial step towards switching Execution::Call to callable.
Port
d5bbd45f044ae6796c0d0f7bd8732069d74418de
Original commit message:
Currently Execution::Call (and friends) still duplicate a lot of the
Call sequence logic that should be encapsulated in the Call and
CallFunction builtins. So the plan now is to switch Execution::Call
to accept any Callable and just pass that through to the Call builtin.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, dstence@us.ibm.com
BUG=v8:4413
LOG=n
Review URL: https://codereview.chromium.org/
1347213003
Cr-Commit-Position: refs/heads/master@{#30813}
gdeepti [Thu, 17 Sep 2015 17:54:48 +0000 (10:54 -0700)]
[simdjs] Update spec version to 0.8.4
Merge ShiftRightArithmetic and ShiftRightLogical Functions.
BUG=v8:4124
LOG=Y
Review URL: https://codereview.chromium.org/
1351663002
Cr-Commit-Position: refs/heads/master@{#30812}
rmcilroy [Thu, 17 Sep 2015 17:24:13 +0000 (10:24 -0700)]
[arm64]: Fix bug introduced accidentally in r30710
Uncomment the if (!serializer_enabled()) check which I accidentially left
commented out after debugging during the CLs development.
BUG=chromium:532969
R=bmeurer@chromium.org
LOG=n
Review URL: https://codereview.chromium.org/
1351943002
Cr-Commit-Position: refs/heads/master@{#30811}
mbrandy [Thu, 17 Sep 2015 17:16:43 +0000 (10:16 -0700)]
PPC: [runtime] Replace the EQUALS builtin with proper Object::Equals.
Port
54bab695f5de5bf5948c5b50b217628a00d60f91
Original commit message:
Move the implementation of the Abstract Equality Comparison to the
runtime and thereby remove the EQUALS dispatcher builtin. Also remove
the various runtime entry points that were only used to support the
EQUALS builtin.
Now the Abstract Equality Comparison is also using the correct
ToPrimitive implementation, which properly supports @@toPrimitive.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, dstence@us.ibm.com
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/
1357493002
Cr-Commit-Position: refs/heads/master@{#30810}
mbrandy [Thu, 17 Sep 2015 17:14:06 +0000 (10:14 -0700)]
PPC: [builtins] Unify the String constructor.
Port
a3d6f6cce317dbe1d31079eb81e15e49f3fb687a
Original commit message:
Implement the String constructor completely as native builtin,
avoiding the need to do gymnastics in JavaScript builtin to
properly detect the no argument case (which is different from
the undefined argument case) and also allowing to just
tailcall through to ToString or SymbolDescriptiveString for
the common case. Also the JavaScript builtin was misleading
since the case for construct call was unused, but could be
triggered in a wrong way once we support tail calls from
constructor functions.
This refactoring allows us to properly implement subclassing
for String builtins, once we have the correct initial_map on
derived classes (it's merely a matter of using NewTarget
instead of the target register now).
This introduces a new %SymbolDescriptiveString runtime
entry, which is also used by Symbol.toString() now.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, dstence@us.ibm.com
BUG=
Review URL: https://codereview.chromium.org/
1354663002
Cr-Commit-Position: refs/heads/master@{#30809}
bmeurer [Thu, 17 Sep 2015 17:11:38 +0000 (10:11 -0700)]
[runtime] Initial step towards switching Execution::Call to callable.
Currently Execution::Call (and friends) still duplicate a lot of the
Call sequence logic that should be encapsulated in the Call and
CallFunction builtins. So the plan now is to switch Execution::Call
to accept any Callable and just pass that through to the Call builtin.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_nosnap_dbg
R=jarin@chromium.org
BUG=v8:4413
LOG=n
Committed: https://crrev.com/
359645f48156e15f235e9a9ede7910e0bcd9ae45
Cr-Commit-Position: refs/heads/master@{#30791}
Review URL: https://codereview.chromium.org/
1353723002
Cr-Commit-Position: refs/heads/master@{#30808}
mbrandy [Thu, 17 Sep 2015 17:03:57 +0000 (10:03 -0700)]
PPC: Remove --pretenure-call-new
Port
b5588f48fd0b4e3ee43be1fe6c19d7ddd8b8b5f1
Original commit message:
There isn't a plan to turn it on soon, so we'll take it out in favor of cleaner code.
R=mvstanton@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, dstence@us.ibm.com
BUG=
Review URL: https://codereview.chromium.org/
1347253003
Cr-Commit-Position: refs/heads/master@{#30807}
mbrandy [Thu, 17 Sep 2015 17:02:45 +0000 (10:02 -0700)]
PPC: Vector ICs: Hook up vectors in platform builtins to their SharedFunctionInfos.
Port
905e008c52ba06120f4a523aab00a53bc50830f1
R=mvstanton@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com, dstence@us.ibm.com
BUG=v8:4423
LOG=N
Review URL: https://codereview.chromium.org/
1350923003
Cr-Commit-Position: refs/heads/master@{#30806}
tandrii [Thu, 17 Sep 2015 14:59:24 +0000 (07:59 -0700)]
Whitespace change.
R=machenbach@chromium.org,jochen@chromium.org
NOTRY=True
BUG=
Review URL: https://codereview.chromium.org/
1357453002
Cr-Commit-Position: refs/heads/master@{#30805}
ben [Thu, 17 Sep 2015 14:47:37 +0000 (07:47 -0700)]
Pretenure builtin typed arrays.
Typed arrays from the snapshot start out in the young space but they
all seem to end up in the old space sooner or later anyway. Let's
expedite that by allocating them in the old space right away.
Review URL: https://codereview.chromium.org/
1347263003
Cr-Commit-Position: refs/heads/master@{#30804}
mlippautz [Thu, 17 Sep 2015 14:43:15 +0000 (07:43 -0700)]
[heap] Fix waiting for parallel tasks
R=hpayer@chromium.org
BUG=chromium:524425
LOG=N
Review URL: https://codereview.chromium.org/
1356663002
Cr-Commit-Position: refs/heads/master@{#30803}
ishell [Thu, 17 Sep 2015 14:05:30 +0000 (07:05 -0700)]
Fix for deopt fuzzer which was broken by https://codereview.chromium.org/
1352803002
Review URL: https://codereview.chromium.org/
1347073004
Cr-Commit-Position: refs/heads/master@{#30802}
machenbach [Thu, 17 Sep 2015 13:41:31 +0000 (06:41 -0700)]
[test] Switch perf try wrapper to buildbucket.
NOTRY=true
Review URL: https://codereview.chromium.org/
1351093002
Cr-Commit-Position: refs/heads/master@{#30801}
hpayer [Thu, 17 Sep 2015 13:38:14 +0000 (06:38 -0700)]
[heap] Scalable slots buffer for parallel compaction.
BUG=chromium:524425
LOG=N
Review URL: https://codereview.chromium.org/
1341973003
Cr-Commit-Position: refs/heads/master@{#30800}
ishell [Thu, 17 Sep 2015 13:02:01 +0000 (06:02 -0700)]
Disable tests that are known to be non-deterministic in --verify-predictable mode.
Review URL: https://codereview.chromium.org/
1352803002
Cr-Commit-Position: refs/heads/master@{#30799}
jkummerow [Thu, 17 Sep 2015 13:00:57 +0000 (06:00 -0700)]
Reland "[test] Fix cctest path separators on Windows"
Now run-tests.py understands "suite/foo/bar" with forward slashes for
command-line test selection on all test suites on all platforms.
Previously, file-based suites like mjsunit also accepted "mjsunit/foo\bar";
that behavior is sacrificed here in favor of unification. For the cctest
suite, OTOH, it wasn't possible on Windows to select specific tests at all.
Original review: https://codereview.chromium.org/
1348653003/
This reverts commit
5f44a9105980e7ca3a444c9c4293ee9442a8139f.
NOTRY=true
Review URL: https://codereview.chromium.org/
1356613002
Cr-Commit-Position: refs/heads/master@{#30798}
cbruni [Thu, 17 Sep 2015 12:52:37 +0000 (05:52 -0700)]
Improve JSReceiver::GetKeys Speed
The core bottleneck lies in N-square cost of array union. Depending on the size
of the arrays involved it makes sense to rely on a hash-set/table for the lookup.
LOG=N
BUG=v8:2904
Review URL: https://codereview.chromium.org/
1316213008
Cr-Commit-Position: refs/heads/master@{#30797}
mlippautz [Thu, 17 Sep 2015 12:23:46 +0000 (05:23 -0700)]
Reland "[heap] Introduce parallel compaction algorithm."
This reverts commit
7a0a0b8b85e4cdf06795ffea01855b345776b932.
- The number of parallel tasks is still 1, i.e., we only compact on the main
thread.
- Remove emergency memory (PagedSpace, and CodeRange)
- Introduce partial compaction of pages.
- Logic for multiple tasks is in place.
BUG=chromium:524425
LOG=N
Review URL: https://codereview.chromium.org/
1356533002
Cr-Commit-Position: refs/heads/master@{#30796}
jkummerow [Thu, 17 Sep 2015 12:00:13 +0000 (05:00 -0700)]
Revert of [test] Fix cctest path separators on Windows (patchset #2 id:20001 of https://codereview.chromium.org/
1348653003/ )
Reason for revert:
mozilla tests are failing on Windows
Original issue's description:
> [test] Fix cctest path separators on Windows
>
> Now run-tests.py understands "suite/foo/bar" with forward slashes for
> command-line test selection on all test suites on all platforms.
>
> Previously, file-based suites like mjsunit also accepted "mjsunit/foo\bar";
> that behavior is sacrificed here in favor of unification. For the cctest
> suite, OTOH, it wasn't possible on Windows to select specific tests at all.
>
> Committed: https://crrev.com/
b36cfdb39ae648b49a1396c4f669df9b1f57996c
> Cr-Commit-Position: refs/heads/master@{#30794}
TBR=machenbach@google.com,machenbach@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/
1349163002
Cr-Commit-Position: refs/heads/master@{#30795}
jkummerow [Thu, 17 Sep 2015 11:23:56 +0000 (04:23 -0700)]
[test] Fix cctest path separators on Windows
Now run-tests.py understands "suite/foo/bar" with forward slashes for
command-line test selection on all test suites on all platforms.
Previously, file-based suites like mjsunit also accepted "mjsunit/foo\bar";
that behavior is sacrificed here in favor of unification. For the cctest
suite, OTOH, it wasn't possible on Windows to select specific tests at all.
Review URL: https://codereview.chromium.org/
1348653003
Cr-Commit-Position: refs/heads/master@{#30794}
machenbach [Thu, 17 Sep 2015 10:11:37 +0000 (03:11 -0700)]
Revert of [runtime] Initial step towards switching Execution::Call to callable. (patchset #1 id:1 of https://codereview.chromium.org/
1353723002/ )
Reason for revert:
[Sheriff] Causes a dcheck failure in layout tests (and some test changes in release):
https://storage.googleapis.com/chromium-layout-test-archives/V8-Blink_Linux_64__dbg_/1442/layout-test-results/virtual/android/fullscreen/api/element-request-fullscreen-top-stderr.txt
from
http://build.chromium.org/p/client.v8.fyi/builders/V8-Blink%20Linux%2064%20%28dbg%29/builds/1442
Original issue's description:
> [runtime] Initial step towards switching Execution::Call to callable.
>
> Currently Execution::Call (and friends) still duplicate a lot of the
> Call sequence logic that should be encapsulated in the Call and
> CallFunction builtins. So the plan now is to switch Execution::Call
> to accept any Callable and just pass that through to the Call builtin.
>
> CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_nosnap_dbg
> R=jarin@chromium.org
> BUG=v8:4413
> LOG=n
>
> Committed: https://crrev.com/
359645f48156e15f235e9a9ede7910e0bcd9ae45
> Cr-Commit-Position: refs/heads/master@{#30791}
TBR=jarin@chromium.org,bmeurer@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:4413
Review URL: https://codereview.chromium.org/
1346763005
Cr-Commit-Position: refs/heads/master@{#30793}
conradw [Thu, 17 Sep 2015 09:38:22 +0000 (02:38 -0700)]
Fix temp_zone scoping when parsing inner function literals
BUG=v8:4392
LOG=Y
Review URL: https://codereview.chromium.org/
1354523003
Cr-Commit-Position: refs/heads/master@{#30792}
bmeurer [Thu, 17 Sep 2015 09:05:28 +0000 (02:05 -0700)]
[runtime] Initial step towards switching Execution::Call to callable.
Currently Execution::Call (and friends) still duplicate a lot of the
Call sequence logic that should be encapsulated in the Call and
CallFunction builtins. So the plan now is to switch Execution::Call
to accept any Callable and just pass that through to the Call builtin.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_nosnap_dbg
R=jarin@chromium.org
BUG=v8:4413
LOG=n
Review URL: https://codereview.chromium.org/
1353723002
Cr-Commit-Position: refs/heads/master@{#30791}
neis [Thu, 17 Sep 2015 08:51:58 +0000 (01:51 -0700)]
Intersection of certain constants with bitsets was wrongly non-empty.
R=jarin
BUG=
Review URL: https://codereview.chromium.org/
1343933002
Cr-Commit-Position: refs/heads/master@{#30790}
hpayer [Thu, 17 Sep 2015 08:44:06 +0000 (01:44 -0700)]
[heap] Inline record slot methods.
BUG=chromium:532784
LOG=n
Review URL: https://codereview.chromium.org/
1347363002
Cr-Commit-Position: refs/heads/master@{#30789}
mlippautz [Thu, 17 Sep 2015 07:58:18 +0000 (00:58 -0700)]
Revert of [heap] Introduce parallel compaction algorithm. (patchset #9 id:160001 of https://codereview.chromium.org/
1343333002/ )
Reason for revert:
Check failed: https://chromegw.corp.google.com/i/client.v8/builders/V8%20Win64/builds/5535/steps/Check%20%28flakes%29/logs/IndependentWeakHandle
Original issue's description:
> [heap] Introduce parallel compaction algorithm.
>
> - The number of parallel tasks is still 1, i.e., we only compact on the main
> thread.
> - Remove emergency memory (PagedSpace, and CodeRange)
> - Introduce partial compaction of pages.
> - Logic for multiple tasks is in place.
>
> BUG=chromium:524425
> LOG=N
>
> Committed: https://crrev.com/
61ea4f55616d3f7bc2ce049a678f16f7475e03e0
> Cr-Commit-Position: refs/heads/master@{#30787}
TBR=hpayer@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:524425
Review URL: https://codereview.chromium.org/
1347873003
Cr-Commit-Position: refs/heads/master@{#30788}
mlippautz [Thu, 17 Sep 2015 07:35:59 +0000 (00:35 -0700)]
[heap] Introduce parallel compaction algorithm.
- The number of parallel tasks is still 1, i.e., we only compact on the main
thread.
- Remove emergency memory (PagedSpace, and CodeRange)
- Introduce partial compaction of pages.
- Logic for multiple tasks is in place.
BUG=chromium:524425
LOG=N
Review URL: https://codereview.chromium.org/
1343333002
Cr-Commit-Position: refs/heads/master@{#30787}
bmeurer [Thu, 17 Sep 2015 06:31:12 +0000 (23:31 -0700)]
Revert "[profiler] Make no frame region detection code more robust", "Fix ASAN after r30777" and "Fix MSAN warning after r30777 (try 2)"
This reverts commits
12c7bc9a226859c3200609495689592a675a21af,
cb0b3592258173c4d20e1500cbd5731e15b9e8b1, and
a6e00c6a9f5a8abf2747293d7452dd0cf572c99e, because they introduced weird
flaky crashes in random places now at least in the arm simulator, where
it see that quite often now on different change sets, i.e. see
http://build.chromium.org/p/tryserver.v8/builders/v8_linux_arm_rel/builds/8138/steps/Check%20%28flakes%29/logs/LoadICFastApi_DirectC..
and
https://chromegw.corp.google.com/i/client.v8/builders/V8%20Linux%20-%20mipsel%20-%20sim/builds/2566/steps/Check%20%28flakes%29/logs/LoadICFastApi_DirectC..
for example.
TBR=alph@chromium.org
BUG=chromium:529931
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
LOG=N
Review URL: https://codereview.chromium.org/
1354573002
Cr-Commit-Position: refs/heads/master@{#30786}
alph [Thu, 17 Sep 2015 05:41:01 +0000 (22:41 -0700)]
Fix MSAN warning after r30777 (try 2)
TBR=bmeurer,yurys
NOTRY=true
Review URL: https://codereview.chromium.org/
1348493003
Cr-Commit-Position: refs/heads/master@{#30785}
mtrofin [Thu, 17 Sep 2015 03:54:57 +0000 (20:54 -0700)]
[tubofan] Greedy: groupper -> grouper.
Small spelling fix.
Review URL: https://codereview.chromium.org/
1352673002
Cr-Commit-Position: refs/heads/master@{#30784}
mtrofin [Thu, 17 Sep 2015 03:53:43 +0000 (20:53 -0700)]
[turbofan] Greedy: faster compile time.
Avoiding unnecessarily traversing conflicts when doing weight
comparisons. This reduced compile time regressions from a few
multiples to under 10% - at least for zlib.
Review URL: https://codereview.chromium.org/
1346263004
Cr-Commit-Position: refs/heads/master@{#30783}
v8-autoroll [Thu, 17 Sep 2015 03:26:17 +0000 (20:26 -0700)]
Update V8 DEPS.
Rolling v8/tools/clang to
eea56c7ed84778edadbcd43f06793b0311a56b28
TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org
Review URL: https://codereview.chromium.org/
1350993003
Cr-Commit-Position: refs/heads/master@{#30782}
chunyang.dai [Thu, 17 Sep 2015 01:25:23 +0000 (18:25 -0700)]
X87: Reland VectorICs: ia32 store ics need a virtual register.
port
1e00bb57a2969e3e428a1d552116752a95c06022 (r30737).
original commit message:
(reason for revert/reland: patch incorrectly left --vector-stores flag
on, helpfully revealing some gcstress issues to look at, but they
don't need to block this CL).
Some pretty hacky code was used to carry out the tail-call
handler dispatch on ia32 vector stores due to a lack
of free registers. It really tanks performance. A better
approach is to use a virtual register on the isolate.
BUG=
Review URL: https://codereview.chromium.org/
1344383002
Cr-Commit-Position: refs/heads/master@{#30781}
chunyang.dai [Thu, 17 Sep 2015 01:21:46 +0000 (18:21 -0700)]
X87: [runtime] Replace the EQUALS builtin with proper Object::Equals.
port
54bab695f5de5bf5948c5b50b217628a00d60f91 (r30747).
original commit message:
Move the implementation of the Abstract Equality Comparison to the
runtime and thereby remove the EQUALS dispatcher builtin. Also remove
the various runtime entry points that were only used to support the
EQUALS builtin.
Now the Abstract Equality Comparison is also using the correct
ToPrimitive implementation, which properly supports @@toPrimitive.
BUG=
Review URL: https://codereview.chromium.org/
1349623002
Cr-Commit-Position: refs/heads/master@{#30780}
alph [Thu, 17 Sep 2015 00:56:33 +0000 (17:56 -0700)]
Fix ASAN after r30777
TBR=bmeurer,yurys
NOTRY=true
Review URL: https://codereview.chromium.org/
1349953002
Cr-Commit-Position: refs/heads/master@{#30779}
ofrobots [Thu, 17 Sep 2015 00:55:24 +0000 (17:55 -0700)]
improve allocation accounting for incremental mark
Add an assertion that allocated_bytes >= 0 in IncrementalMark::Step and then
make it pass. We were not being diligent in maintaining top_on_previous_step_
and as a result inaccurate, and even negative values of allocated_bytes were
being reported to Step.
BUG=
R=hpayer@chromium.org
Review URL: https://codereview.chromium.org/
1274453002
Cr-Commit-Position: refs/heads/master@{#30778}
alph [Thu, 17 Sep 2015 00:12:08 +0000 (17:12 -0700)]
[profiler] Make no frame region detection code more robust
Upon collection of the stack trace if the current PC falls into
the frame building code, the top frame might be in a non-consistent
state. That leads to some of the frames could be missing from the
stack trace.
The patch makes it check instructions under current PC and if they
look like the frame setup/destroy code, it skips the entire sample.
Support for x86/x64
BUG=chromium:529931
LOG=N
Review URL: https://codereview.chromium.org/
1341413002
Cr-Commit-Position: refs/heads/master@{#30777}
mtrofin [Wed, 16 Sep 2015 21:43:19 +0000 (14:43 -0700)]
[turbofan] Greedy: small fix in groupping algo.
This is a performance bug, not a functional bug: we were
losing grouping opportunities.
BUG=
Review URL: https://codereview.chromium.org/
1342243003
Cr-Commit-Position: refs/heads/master@{#30776}
caitpotter88 [Wed, 16 Sep 2015 21:27:19 +0000 (14:27 -0700)]
[cleanup] refactor ParsePropertyDefinition for clarity
Some cleanup of ParsePropertyDefinition --- Replaces certain hacks with
more structured, clean code, and adds additional comments to aid in
comprehension of this tricky area of the ambiguous recursive descent
parser.
BUG=v8:3583
LOG=N
R=adamk, aperez, wingo, rossberg
Review URL: https://codereview.chromium.org/
1348773004
Cr-Commit-Position: refs/heads/master@{#30775}
adamk [Wed, 16 Sep 2015 21:19:21 +0000 (14:19 -0700)]
Disallow Object.observe calls on access-checked objects
We already disallowed observing the global proxy; now we also
disallow any observation of access-checked objects (regardless
of whether the access check would succeed or fail, since there's
not a good way to tell the embedder what kind of access is being
requested).
Also disallow Object.getNotifier for the same reasons.
BUG=chromium:531891
LOG=y
Review URL: https://codereview.chromium.org/
1346813002
Cr-Commit-Position: refs/heads/master@{#30774}
domenic [Wed, 16 Sep 2015 21:00:45 +0000 (14:00 -0700)]
Implement V8 extras utils object
This adds a utils object meant specifically for V8 extras, presenting a limited
API surface for doing things that would otherwise require %-functions.
BUG=v8:4276
LOG=Y
R=jochen@chromium.org,yangguo@chromium.org
Review URL: https://codereview.chromium.org/
1343113003
Cr-Commit-Position: refs/heads/master@{#30773}
aperez [Wed, 16 Sep 2015 18:01:38 +0000 (11:01 -0700)]
ES6: Array.prototype.slice and friends should use ToLength instead of ToUint32
Defines a new --harmony-tolength flag, and a ToLengthFlagged() runtime function,
that is used where ES6 requires ToLength(), but a pre-ES6 conversion existed
before. When the flag is disabled, the function uses TO_UINT32(), which is
the pre-ES6 behaviour. When the flag enabled, the ES6-compliant ToLength()
conversion is used.
Based on a patch initially from Diego Pino <dpino@igalia.com>
BUG=v8:3087
LOG=Y
Review URL: https://codereview.chromium.org/
1309243003
Cr-Commit-Position: refs/heads/master@{#30772}
fedor [Wed, 16 Sep 2015 17:27:40 +0000 (10:27 -0700)]
[objects] do not visit ArrayBuffer's backing store
ArrayBuffer's backing store is a pointer to external heap, and can't be
treated as a heap object. Doing so will result in crashes, when the
backing store is unaligned.
See: https://github.com/nodejs/node/issues/2791
BUG=chromium:530531
R=mlippautz@chromium.org
LOG=N
Review URL: https://codereview.chromium.org/
1327403002
Cr-Commit-Position: refs/heads/master@{#30771}
karl [Wed, 16 Sep 2015 16:21:33 +0000 (09:21 -0700)]
[es6] Optimize TypedArray.subarray()
````
var array = new Uint8Array(65000);
var startDate = Date.now();
var counter = 0;
while (counter++ <
50000000) {
array.subarray(start, end);
}
var endDate = Date.now();
print(endDate - startDate);
````
4200 ms -> 3500 ms (16.67%)
BUG=
Review URL: https://codereview.chromium.org/
1331993004
Cr-Commit-Position: refs/heads/master@{#30770}
caitpotter88 [Wed, 16 Sep 2015 16:01:47 +0000 (09:01 -0700)]
[es6] support `get` and `set` in shorthand properties
Add support for `get` and `set` as shorthand properties. Also
supports them for CoverInitializedName in BindingPatterns and (once implemented)
AssignmentPatterns.
BUG=v8:4412, v8:3584
LOG=N
R=adamk, aperez, wingo, rossberg
Review URL: https://codereview.chromium.org/
1328083002
Cr-Commit-Position: refs/heads/master@{#30769}
mvstanton [Wed, 16 Sep 2015 15:38:37 +0000 (08:38 -0700)]
Extra code to diagnose a crash bug.
This will catch an invalid receiver before being passed to a load ic miss
handler in the runtime.
BUG=
R=jkummerow@chromium.org
Review URL: https://codereview.chromium.org/
1351493002
Cr-Commit-Position: refs/heads/master@{#30768}
mvstanton [Wed, 16 Sep 2015 15:12:24 +0000 (08:12 -0700)]
Remove --pretenure-call-new
There isn't a plan to turn it on soon, so we'll take it out in favor of cleaner code.
BUG=
Review URL: https://codereview.chromium.org/
1202173002
Cr-Commit-Position: refs/heads/master@{#30767}
mstarzinger [Wed, 16 Sep 2015 13:04:25 +0000 (06:04 -0700)]
[turbofan] Make arguments object materialization inlinable.
This makes sure that the arguments object materialization in the method
prologue is composable with respect to inlining. The generic runtime
functions materializing those objects now respect the deoptimization
information when reconstructing the original arguments.
R=mvstanton@chromium.org
Review URL: https://codereview.chromium.org/
1340313003
Cr-Commit-Position: refs/heads/master@{#30766}
ivica.bogosavljevic [Wed, 16 Sep 2015 12:15:15 +0000 (05:15 -0700)]
MIPS: Fixing floating point register clobbering
Fixing floating point register clobbering for MIPSr6 (32 and 64)
due to using of f31 floating point register as double compare register,
without saving the value of the register before using it.
TEST=cctest/test-debug/*
BUG=
Review URL: https://codereview.chromium.org/
1346623002
Cr-Commit-Position: refs/heads/master@{#30765}
jarin [Wed, 16 Sep 2015 11:55:27 +0000 (04:55 -0700)]
[turbofan] Get rid of type lower bounds.
Review URL: https://codereview.chromium.org/
1348073002
Cr-Commit-Position: refs/heads/master@{#30764}
vogelheim [Wed, 16 Sep 2015 11:37:04 +0000 (04:37 -0700)]
Avoid excessive data copying for ExternalStreamingStream::SetBookmark.
BUG=v8:4422
R=jochen@chromium.org
LOG=Y
Review URL: https://codereview.chromium.org/
1346613002
Cr-Commit-Position: refs/heads/master@{#30763}
bmeurer [Wed, 16 Sep 2015 11:35:15 +0000 (04:35 -0700)]
[builtins] Also simplify the Symbol constructor.
No need to rely on the %_IsConstructCall magic here, we can just
implement the Symbol constructor in C++ altogether (it was just a
stupid wrapper around %CreateSymbol anyway).
R=mstarzinger@chromium.org
Review URL: https://codereview.chromium.org/
1349643002
Cr-Commit-Position: refs/heads/master@{#30762}
mstarzinger [Wed, 16 Sep 2015 11:32:54 +0000 (04:32 -0700)]
[turbofan] Add inlining guards to Runtime_NewArguments.
This adds debug code that makes sure that the runtime functions that
materialize arguments objects, {Runtime_New[Sloppy|Strict]Arguments},
are not being called from within an inlined scope. They would produce
wrong results and we should avoid producing code that does this.
R=bmeurer@chromium.org
Review URL: https://codereview.chromium.org/
1343763002
Cr-Commit-Position: refs/heads/master@{#30761}
mlippautz [Wed, 16 Sep 2015 11:18:07 +0000 (04:18 -0700)]
Reland of "[heap] Concurrency support for heap book-keeping info"
Adds concurrency support for:
- MemoryChunk: Fragmentation counters
- MemoryChunk: High-water mark
- MemoryAllocator: Lowest and highest ever allocated addresses, size, and
capacity
R=hpayer@chromium.org
BUG=chromium:524425
LOG=N
This reverts commit
0db34dbe8111f8670c82bb4c42110400a9050d08.
BUG=
Review URL: https://codereview.chromium.org/
1346973002
Cr-Commit-Position: refs/heads/master@{#30760}
bmeurer [Wed, 16 Sep 2015 10:44:36 +0000 (03:44 -0700)]
[builtins] Unify the String constructor.
Implement the String constructor completely as native builtin,
avoiding the need to do gymnastics in JavaScript builtin to
properly detect the no argument case (which is different from
the undefined argument case) and also allowing to just
tailcall through to ToString or SymbolDescriptiveString for
the common case. Also the JavaScript builtin was misleading
since the case for construct call was unused, but could be
triggered in a wrong way once we support tail calls from
constructor functions.
This refactoring allows us to properly implement subclassing
for String builtins, once we have the correct initial_map on
derived classes (it's merely a matter of using NewTarget
instead of the target register now).
This introduces a new %SymbolDescriptiveString runtime
entry, which is also used by Symbol.toString() now.
R=mstarzinger@chromium.org
Review URL: https://codereview.chromium.org/
1344893002
Cr-Commit-Position: refs/heads/master@{#30759}
mvstanton [Wed, 16 Sep 2015 10:08:50 +0000 (03:08 -0700)]
Vector ICs: Hook up vectors in platform builtins to their SharedFunctionInfos.
BUG=v8:4423
LOG=N
Review URL: https://codereview.chromium.org/
1342013003
Cr-Commit-Position: refs/heads/master@{#30758}
jochen [Wed, 16 Sep 2015 07:31:40 +0000 (00:31 -0700)]
Add myself to heap owners
R=hpayer@chromium.org
BUG=
NOTRY=true
Review URL: https://codereview.chromium.org/
1350633003
Cr-Commit-Position: refs/heads/master@{#30757}
paul.lind [Tue, 15 Sep 2015 21:28:13 +0000 (14:28 -0700)]
MIPS64: Fix unittests (to not use invalid load representation).
Same as https://codereview.chromium.org/
1340303002/
BUG=
Review URL: https://codereview.chromium.org/
1339763005
Cr-Commit-Position: refs/heads/master@{#30756}
binji [Tue, 15 Sep 2015 21:24:50 +0000 (14:24 -0700)]
[Atomics] Remove support for atomic accesses on floating-point values.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/
1318713007
Cr-Commit-Position: refs/heads/master@{#30755}
adamk [Tue, 15 Sep 2015 16:43:29 +0000 (09:43 -0700)]
Fix spread operator in ArrayLiterals when nested in other literals
Mark ArrayLiterals utilizing the spread operator as non-simple.
This causes them to return false for IsCompileTimeValue, and thus
causes spread to work as expected in nested literals.
BUG=v8:4417
LOG=y
Review URL: https://codereview.chromium.org/
1336123002
Cr-Commit-Position: refs/heads/master@{#30754}
jarin [Tue, 15 Sep 2015 15:50:32 +0000 (08:50 -0700)]
[mips] Fix mips unittests (to not use invalid load representation).
Review URL: https://codereview.chromium.org/
1340303002
Cr-Commit-Position: refs/heads/master@{#30753}
mlippautz [Tue, 15 Sep 2015 15:11:36 +0000 (08:11 -0700)]
Revert of [heap] Concurrency support for heap book-keeping info (patchset #4 id:60001 of https://codereview.chromium.org/
1340923004/ )
Reason for revert:
crashing: http://build.chromium.org/p/client.v8/builders/V8%20Win32%20-%20debug%20-%203/builds/4716
Original issue's description:
> [heap] Concurrency support for heap book-keeping info.
>
> Adds concurrency support for:
> - MemoryChunk: Fragmentation counters
> - MemoryChunk: High-water mark
> - MemoryAllocator: Lowest and highest ever allocated addresses, size, and
> capacity
>
> R=hpayer@chromium.org
> BUG=chromium:524425
> LOG=N
>
> Committed: https://crrev.com/
63190721cda4966e01d71e92a730ce48ea789fbc
> Cr-Commit-Position: refs/heads/master@{#30749}
TBR=hpayer@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=chromium:524425
Review URL: https://codereview.chromium.org/
1340323002
Cr-Commit-Position: refs/heads/master@{#30752}
mlippautz [Tue, 15 Sep 2015 14:58:44 +0000 (07:58 -0700)]
Add barriers to atomic utils.
Loads get an acquire, store/cas a release. Increment gets a full barrier.
R=hpayer@chromium.org
BUG=
Review URL: https://codereview.chromium.org/
1343883004
Cr-Commit-Position: refs/heads/master@{#30751}
bmeurer [Tue, 15 Sep 2015 14:32:25 +0000 (07:32 -0700)]
[es6] Move builtin constructors for primitives to strict mode.
The ES6 specification says that "Built-in functions that are ECMAScript
function objects must be strict mode functions", which in particular
means that you can never test for them using the "caller" field of a
sloppy mode function.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_layout_dbg,v8_linux_nosnap_dbg
R=mstarzinger@chromium.org
BUG=v8:105
LOG=n
Review URL: https://codereview.chromium.org/
1347663002
Cr-Commit-Position: refs/heads/master@{#30750}
mlippautz [Tue, 15 Sep 2015 13:50:26 +0000 (06:50 -0700)]
[heap] Concurrency support for heap book-keeping info.
Adds concurrency support for:
- MemoryChunk: Fragmentation counters
- MemoryChunk: High-water mark
- MemoryAllocator: Lowest and highest ever allocated addresses, size, and
capacity
R=hpayer@chromium.org
BUG=chromium:524425
LOG=N
Review URL: https://codereview.chromium.org/
1340923004
Cr-Commit-Position: refs/heads/master@{#30749}
mlippautz [Tue, 15 Sep 2015 13:45:18 +0000 (06:45 -0700)]
[heap] Extend mutex guards for CodeRange.
Previously the mutex only guarded free_list_. The extension covers also
allocation_list_.
BUG=chromium:524425
LOG=N
Review URL: https://codereview.chromium.org/
1341293002
Cr-Commit-Position: refs/heads/master@{#30748}
bmeurer [Tue, 15 Sep 2015 13:14:36 +0000 (06:14 -0700)]
[runtime] Replace the EQUALS builtin with proper Object::Equals.
Move the implementation of the Abstract Equality Comparison to the
runtime and thereby remove the EQUALS dispatcher builtin. Also remove
the various runtime entry points that were only used to support the
EQUALS builtin.
Now the Abstract Equality Comparison is also using the correct
ToPrimitive implementation, which properly supports @@toPrimitive.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_layout_dbg,v8_linux_nosnap_dbg
R=mstarzinger@chromium.org
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/
1337993005
Cr-Commit-Position: refs/heads/master@{#30747}
hpayer [Tue, 15 Sep 2015 12:35:05 +0000 (05:35 -0700)]
[heap] Move slots buffer into a separate file.
BUG=
Review URL: https://codereview.chromium.org/
1343043002
Cr-Commit-Position: refs/heads/master@{#30746}
chunyang.dai [Tue, 15 Sep 2015 12:24:48 +0000 (05:24 -0700)]
X87: [Interpreter] Add support for JS calls.
port
e7fb233946b990ecbbbd76cc6529f62bd5da64e3 (r30710).
original commit message:
Adds support for JS calls to the interpreter. In order to support
calls from the interpreter, the PushArgsAndCall builtin is added
which pushes a sequence of arguments onto the stack and calls
builtin::Call.
Adds the Call bytecode.
BUG=
Review URL: https://codereview.chromium.org/
1334153004
Cr-Commit-Position: refs/heads/master@{#30745}
chunyang.dai [Tue, 15 Sep 2015 12:16:35 +0000 (05:16 -0700)]
X87: [builtins] Simplify String constructor code.
port
eadfd666318c1dd959d7fac7c27538ef478a8b72 (r30706).
original commit message:
The String constructor was somewhat complex with a lot of micro
optimizations that are not relevant or even misguided. It would be
really hard to port that code to ES6, which requires String to be
subclassable. So as a first step we reduced the necessary complexity
to the bare minimum (also removing the last user of the fairly complex
MacroAssembler::LookupNumberStringCache method).
This also removes the counters for the String constructor, which
were not properly exposed anymore (and not kept in sync with inlined
versions of the String constructor anyway).
BUG=
Review URL: https://codereview.chromium.org/
1336133003
Cr-Commit-Position: refs/heads/master@{#30744}
chunyang.dai [Tue, 15 Sep 2015 11:51:38 +0000 (04:51 -0700)]
X87: Make FlushICache part of Assembler(Base) and take Isolate as parameter.
port
9fc4fc141fa1d1aa1a3ccb901518cdffa0d6d268 (r30695).
BUG=
Review URL: https://codereview.chromium.org/
1339293002
Cr-Commit-Position: refs/heads/master@{#30743}
mlippautz [Tue, 15 Sep 2015 11:44:48 +0000 (04:44 -0700)]
[heap] Let caller figure out target space for evacuation when compacting.
R=hpayer@chromium.org
BUG=chromium:524425
LOG=N
Review URL: https://codereview.chromium.org/
1340253004
Cr-Commit-Position: refs/heads/master@{#30742}
chunyang.dai [Tue, 15 Sep 2015 11:31:13 +0000 (04:31 -0700)]
X87: [builtins] Remove the weird STACK_OVERFLOW builtin.
port
39604dda567b9ba9f4dcfe42f95471f30ec7cb6b (r30693).
original commit message:
Just use a %ThrowStackOverflow runtime function instead, which
does the trick, especially since the Isolate already has a
preallocated StackOverflow error for that.
BUG=
Review URL: https://codereview.chromium.org/
1344793002
Cr-Commit-Position: refs/heads/master@{#30741}
chunyang.dai [Tue, 15 Sep 2015 11:26:51 +0000 (04:26 -0700)]
X87: [stubs] Simplify the non-function case of CallConstructStub.
port
622fa0ea21b3517b1feaed447c87e580bafac865 (r30691).
original commit message:
Currently we do this dance between the CallConstructStub, the
CALL_* builtins and the %GetConstructorDelegate, %GetProxyTrap,
and %Apply runtime functions for every [[Construct]] operation on
non-function callables. This is complexity is unnecessary, and can
be simplified to work without any JS builtin. This will also make it
a lot easier to implement ES6 compliant [[Construct]] for proxies.
Also sanitize the invariant for CallConstructStub, which up until now
always restored the context itself, but that force us to always create
another copy of all arguments in case of proxies and other callables,
so we can relax that constraint by making the caller restore the context
(this only affects fullcodegen, since the optimizing compilers already
properly restore the context anyway).
BUG=
Review URL: https://codereview.chromium.org/
1341233002
Cr-Commit-Position: refs/heads/master@{#30740}
chunyang.dai [Tue, 15 Sep 2015 11:20:36 +0000 (04:20 -0700)]
X87: Vector ICs: The Oracle needs to report feedback for the object literals and the count operation.
port
752b0308df72461bfeb644bf6dd8dd331fcdb722 (r30686).
original commit message:
The refactoring is because it's awkward and error-prone to deterimine which IC slot an
ObjectLiteralProperty uses for feedback. The fix is for each one to know it's own slot. In the
numbering pass, we allocate slots for the ObjectLiteral, then hand out those slots into the
properties.
It adds one word to the ObjectLiteralProperty expression - I'm investigating if thats a
problem.
This changes makes compiling the object literal cleaner across the three compilers. Also, the
slot allocation logic in ObjectLiteral::ComputeFeedbackRequirements() was refactoring to mimic
the style in full-codegen. This is useful since it must remain in sync with
FullCodegen::VisitObjectLiteral().
BUG=
Review URL: https://codereview.chromium.org/
1341223002
Cr-Commit-Position: refs/heads/master@{#30739}
mstarzinger [Tue, 15 Sep 2015 11:19:02 +0000 (04:19 -0700)]
[turbofan] Fix JSInliner to handle non-returning bodies.
The assumption that every function body produces a value does not hold
for functions that e.g. unconditionally throw or endlessly loop. This
fixes the inlining logic to handle such cases.
R=bmeurer@chromium.org
TEST=mjsunit/regress/regress-crbug-530598
BUG=chromium:530598
LOG=n
Review URL: https://codereview.chromium.org/
1333193005
Cr-Commit-Position: refs/heads/master@{#30738}
mvstanton [Tue, 15 Sep 2015 10:14:11 +0000 (03:14 -0700)]
Reland VectorICs: ia32 store ics need a virtual register.
(reason for revert/reland: patch incorrectly left --vector-stores flag
on, helpfully revealing some gcstress issues to look at, but they
don't need to block this CL).
Some pretty hacky code was used to carry out the tail-call
handler dispatch on ia32 vector stores due to a lack
of free registers. It really tanks performance. A better
approach is to use a virtual register on the isolate.
BUG=
TBR=jkummerow@chromium.org, vogelheim@chromium.org
Review URL: https://codereview.chromium.org/
1346573002
Cr-Commit-Position: refs/heads/master@{#30737}
cbruni [Tue, 15 Sep 2015 09:08:28 +0000 (02:08 -0700)]
builtins.cc return PackedElementsKind where applicable
Returning a result array with holey elements kind in where we actually have a packed kind causes performance regressions.
LOG=N
BUG=chromium:531357
Review URL: https://codereview.chromium.org/
1340033002
Cr-Commit-Position: refs/heads/master@{#30736}
mstarzinger [Tue, 15 Sep 2015 09:07:21 +0000 (02:07 -0700)]
[turbofan] Model arguments object materialization in graph.
This models the materialization of arguments objects in the prologue
within the IR graph. It will in turn allow us to optimize access to
these objects and also correctly handle them with inlining.
R=bmeurer@chromium.org,mvstanton@chromium.org
TEST=cctest/test-run-jsobjects/Arguments*
Review URL: https://codereview.chromium.org/
1344553003
Cr-Commit-Position: refs/heads/master@{#30735}
machenbach [Tue, 15 Sep 2015 08:45:34 +0000 (01:45 -0700)]
Port cfi blacklist from chromium.
BUG=chromium:515782
LOG=n
Review URL: https://codereview.chromium.org/
1333133003
Cr-Commit-Position: refs/heads/master@{#30734}
brettw [Tue, 15 Sep 2015 08:43:25 +0000 (01:43 -0700)]
Remove transitional GN code.
The conditionals were added to facilitate a two-sided build GN config update
where the default optimization config was renamed. This is complete so the
transitional code can be removed.
Review URL: https://codereview.chromium.org/
1338163003
Cr-Commit-Position: refs/heads/master@{#30733}