Commit Graph

13492 Commits

Author SHA1 Message Date
mstarzinger@chromium.org
701417f955 Make stack_parameter_count a plain register.
R=rossberg@chromium.org, danno@chromium.org

Review URL: https://codereview.chromium.org/28993003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17275 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 14:55:21 +00:00
titzer@chromium.org
4d0f2cdd3b Implement global load elimination based on flow engine.
BUG=
R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/27148004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17273 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 14:28:24 +00:00
yangguo@chromium.org
a428465819 Handlify JSObject::HasReal*Property.
R=ulan@chromium.org
BUG=

Review URL: https://codereview.chromium.org/27518002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17272 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 12:52:07 +00:00
loislo@chromium.org
b50be9ff12 HeapProfiler: integrate FindUntrackedObjects into js code.
In some cases we would like to check untracked objects right from the js code.
Otherwise the objects might be collected rigth before check.

BUG=none
R=hpayer@chromium.org, yurys@chromium.org

Review URL: https://codereview.chromium.org/27717003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17270 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 11:25:43 +00:00
loislo@chromium.org
189d13a3fe HeapProfiler: for the test purposes we would like to trace object movements.
When we have an untracked allocations we need to fix it.
It is a bit tricky process due to lack of logging.

I added FLAG_heap_profiler_trace_objects that enables log for object movements.

Drive by fix: with help of logging I found a bug in LeftTrim method for FixedArray.

BUG=none
R=svenpanne@chromium.org, yurys@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=17258

Committed: https://code.google.com/p/v8/source/detail?r=17263

Review URL: https://codereview.chromium.org/27694004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17269 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 10:59:55 +00:00
jkummerow@chromium.org
4e96d4dee6 Tweak Math.log on ia32/x64
ia32 and x64 more consistent now
1. use non-transcendental cache version of log for x64
2. use negative infinity constant instead of pushing to stack and loading to XMM register
3. remove movdbl, use movsd directly. movdbl seems confusing

BUG=
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/27197013

Patch from Weiliang Lin <weiliang.lin2@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17268 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 10:54:45 +00:00
mstarzinger@chromium.org
a25f75736e Prepare push to trunk. Now working on version 3.22.16.
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/28883002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17265 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 10:32:11 +00:00
mstarzinger@chromium.org
6f77ab124c Revert "HeapProfiler: for the test purposes we would like to trace object movements."
This reverts r17263 because of compile failures accross all builds.

TBR=loislo@chromium.org

Review URL: https://codereview.chromium.org/28773002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17264 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 09:50:06 +00:00
loislo@chromium.org
c748e7bbc9 HeapProfiler: for the test purposes we would like to trace object movements.
When we have an untracked allocations we need to fix it.
It is a bit tricky process due to lack of logging.

I added FLAG_heap_profiler_trace_objects that enables log for object movements.

Drive by fix: with help of logging I found a bug in LeftTrim method for FixedArray.

BUG=none
R=svenpanne@chromium.org, yurys@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=17258

Review URL: https://codereview.chromium.org/27694004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17263 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 09:40:00 +00:00
svenpanne@chromium.org
847f4ce616 Removed dead code from mksnapshot.
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/28683002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17262 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 09:06:58 +00:00
yurys@chromium.org
ed7dea41a9 Always make a copy of a string when adding it to StringsStorage
Otherwise the string passed as const char* may be disposed and we will end up with a dangling pointer.

Also changed StringsStorage::GetCopy so that a copy is not created if the string is already in the cache.

BUG=None
R=alph@chromium.org, svenpanne@chromium.org

Review URL: https://codereview.chromium.org/27627006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17260 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 08:56:14 +00:00
loislo@chromium.org
fd45ddcc15 Revert "HeapProfiler: for the test purposes we would like to trace object movements."
Compilation failed on Win64.

This reverts commit 6909faccca68c04b7ec4537538e6e7a0cc1650b8.

BUG=none
TBR=yurys

Review URL: https://codereview.chromium.org/28573002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17259 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 08:11:36 +00:00
loislo@chromium.org
5ee9a06568 HeapProfiler: for the test purposes we would like to trace object movements.
When we have an untracked allocations we need to fix it.
It is a bit tricky process due to lack of logging.

I added FLAG_heap_profiler_trace_objects that enables log for object movements.

Drive by fix: with help of logging I found a bug in LeftTrim method for FixedArray.

BUG=none
R=svenpanne@chromium.org, yurys@chromium.org

Review URL: https://codereview.chromium.org/27694004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17258 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-18 07:43:50 +00:00
rafaelw@chromium.org
ce2c9b1db1 Prevent changes to hidden properties from being observable via Object.observe
This addresses the leak that mstarzinger points out (https://codereview.chromium.org/26390003/) and includes the test.

Note that this adds a test that observing changes to the empty-string property remains possible.

BUG=
R=mstarzinger@chromium.org, rossberg@chromium.org

Review URL: https://codereview.chromium.org/26592012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17257 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-17 20:49:45 +00:00
svenpanne@chromium.org
882778bf09 Make it possible to run a test only in the standard variant.
Use this for mjsunit/unicode-case-overoptimization, which is not
related to Crankshaft at all and takes ages.

R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/27704002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17255 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-17 13:09:28 +00:00
mstarzinger@chromium.org
0a2b4ecdcc Add regression test for optimized count operation.
This is a regression test for a bug with handling of count operations
that target a JavaScript accessor on the prototype chain in Crankshaft.

R=jkummerow@chromium.org
BUG=chromium:306851
TEST=mjsunit/regress/regress-crbug-306851

Review URL: https://codereview.chromium.org/27702002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17254 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-17 12:48:28 +00:00
marja@chromium.org
f039f334f5 Enable calling the SetReference* & SetObjectGroupId functions with a Persistent<SubclassOfValue>.
This is needed for https://codereview.chromium.org/26792002/

BUG=
R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/27512003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17253 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-17 11:48:03 +00:00
yangguo@chromium.org
8ef4b0ca83 Cosmetic: Add macros for NaN, undefined and Infinity to native js code.
Nobody should need to use $NaN, 0/0, 1/0 and void 0.

R=mvstanton@chromium.org
BUG=

Review URL: https://codereview.chromium.org/27491002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17252 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-17 10:02:45 +00:00
mstarzinger@chromium.org
f8cc87a095 Prepare push to trunk. Now working on version 3.22.15.
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/27676002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17249 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-17 08:03:12 +00:00
palfia@homejinni.com
331850e3eb MIPS: Add support for Token::NE and Token::NE_STRICT to TokenToCondition.
Port r17239 (c203175f)

Original commit message:
This is required to be able to use these tokens with
HCompareNumericAndBranch.

BUG=
R=plind44@gmail.com

Review URL: https://codereview.chromium.org/27516004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17248 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 23:15:30 +00:00
yangguo@chromium.org
e0508615a9 Fix leak in optimizing compiler thread.
R=mvstanton@chromium.org
BUG=

Review URL: https://codereview.chromium.org/27473006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17247 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 16:27:17 +00:00
yangguo@chromium.org
b8dd056fa2 Improve queuing for concurrent OSR.
Specifically, this includes:
- Encapsulating data structure for osr buffer into CyclicBuffer
- Use the new CyclicQueue instead of UnboundedQueue to queue new jobs.
  We can enqueue and dequeue a CyclicQueue on both ends in O(1).
  This allows us to add OSR jobs to the front for lower compile latency.
- Dispose osr buffer by one stale job per GC to avoid leak

R=titzer@chromium.org
BUG=

Review URL: https://codereview.chromium.org/25505002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17244 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 14:47:20 +00:00
olivf@chromium.org
0a6d2dec9f Fix lower bound of shr.
BUG=
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/27503002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17243 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 14:39:27 +00:00
yurys@chromium.org
18482d0da2 Get rid of HEAP_PROFILE macro
All usages of the macro were replaced with direct calls to the heap profiler. The macro does null check for HeapProfiler which is always true.

BUG=None
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/26166004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17242 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 14:33:04 +00:00
mstarzinger@chromium.org
3e0f828b8f Revert "TransitionAndStoreStub bailout needs to transition (and store)."
This reverts commit r17216 breaking fast/js/cross-frame-bad-time.html test.

R=mvstanton@chromium.org
TEST=webkit:fast/js/cross-frame-bad-time.html

Review URL: https://codereview.chromium.org/27516002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17241 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 14:17:31 +00:00
titzer@chromium.org
24a7e06323 Remove dead constants not in block #0.
BUG=
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/27473004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17240 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 13:19:46 +00:00
bmeurer@chromium.org
6054baadd2 Add support for Token::NE and Token::NE_STRICT to TokenToCondition.
This is required to be able to use these tokens with
HCompareNumericAndBranch.

R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/27043010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17239 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 12:20:21 +00:00
yangguo@chromium.org
ada764446a Refactor part of handles.cc
BUG=
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/26955004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17238 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 11:52:03 +00:00
mstarzinger@chromium.org
5eae41e518 Return descriptor ownership after TrimDescriptorArray to allow resharing of descriptor arrays
BUG=None
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/27197002

Patch from Ross McIlroy <rmcilroy@chromium.org>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17237 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 11:16:49 +00:00
mstarzinger@chromium.org
100ab91013 Implement fixpoint iteration for escape analysis.
R=titzer@chromium.org

Review URL: https://codereview.chromium.org/23533003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17235 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 09:16:56 +00:00
yangguo@chromium.org
1eeebd2b6b Handlify GetPropertyWithCallback.
R=mstarzinger@chromium.org
BUG=

Review URL: https://codereview.chromium.org/27335002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17234 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 09:13:31 +00:00
mstarzinger@chromium.org
0bd3e179e5 Prepare push to trunk. Now working on version 3.22.14.
R=jkummerow@chromium.org, bmeurer@chromium.org

Review URL: https://codereview.chromium.org/27288003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17230 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 08:58:53 +00:00
yurys@chromium.org
5b105b248e Fix unexpected stack traces in test-cpu-profiler/JsNativeJsSample
This is a follow-up fix for r17140 which makes sure that cpu profiler is started before profiled js functions are compiled so that the profiler gets information about "no frame" ranges along with add code events.

BUG=v8:2933
R=jkummerow@chromium.org, loislo@chromium.org

Review URL: https://codereview.chromium.org/27121003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17229 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 08:15:06 +00:00
mvstanton@chromium.org
8f9f192f6e AllocationSites for all literals
R=hpayer@chromium.org

Review URL: https://codereview.chromium.org/24250005


Review URL: https://codereview.chromium.org/27366003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17228 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 08:10:36 +00:00
svenpanne@chromium.org
a4f9ef1827 Fix for non-SSE2.
BUG=cctest/test-assembler-ia32/AssemblerIa32Extractps
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/27476004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17227 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 08:08:41 +00:00
haitao.feng@intel.com
d3849b23d4 Refactor deferred_objects_double_values_ a little in the deoptimization
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/25583004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17226 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-16 03:30:06 +00:00
palfia@homejinni.com
f5abb8e52a MIPS: Make TestJSArrayForAllocationMemento less awkward.
Port r17220 (be968d52)

Original commit message:
Generated code ended up having two conditional jump statements in a
row. Also introduce JumpIfJSArrayHasAllocationMemento which handles
most cases more simply.

BUG=
R=gergely@homejinni.com

Review URL: https://codereview.chromium.org/27421002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17225 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 23:12:15 +00:00
rafaelw@chromium.org
2268defb84 Handlify JSObject::SetObserved
BUG=
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/27070002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17224 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 19:28:11 +00:00
olivf@chromium.org
5b9f8f32e7 Properly export uin32_bias double constant.
This makes LoadUint32 snapshot safe.

BUG=
R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/27156004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17223 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 16:12:25 +00:00
mvstanton@chromium.org
f4edc076d8 Revert "AllocationSites for all literals"
This reverts commit r17219 due to WebKit failures.

R=mstarzinger@chromium.org
TBR=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/26539010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17222 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 15:35:23 +00:00
jkummerow@chromium.org
ea4e0a3d28 Simplify array element access in Hydrogen.
No change in functionality intended.

R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/26986004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17221 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 15:06:04 +00:00
mvstanton@chromium.org
f8f4e6c9a1 Make TestJSArrayForAllocationMemento less awkward.
Generated code ended up having two conditional jump statements in a
row. Also introduce JumpIfJSArrayHasAllocationMemento which handles
most cases more simply.

R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/26841009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17220 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 15:04:29 +00:00
mvstanton@chromium.org
362c0cfbca AllocationSites for all literals
BUG=
R=hpayer@chromium.org

Review URL: https://codereview.chromium.org/24250005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17219 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 14:52:58 +00:00
svenpanne@chromium.org
3c9a322eaa Do not rely on actual NaN value for testing extractps.
BUG=cctest/test-assembler-ia32/AssemblerIa32Extractps, cctest/test-assembler-x64/AssemblerX64Extractps
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/26704004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17218 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 14:04:49 +00:00
svenpanne@chromium.org
258cee8f72 Fixed bug in extractps instruction on ia32 and x64
This is a fixed version of https://codereview.chromium.org/27097002/
which was originally written by weiliang.lin@intel.com.

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/27301003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17217 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 12:51:58 +00:00
mvstanton@chromium.org
3769a2d24d TransitionAndStoreStub bailout needs to transition (and store).
Performance regression found in test regress-2185-2.js. The problem was
that the bailout method for TransitionAndStoreStub was not performing
the appropriate transition.

BUG=
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/26911007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17216 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 12:42:03 +00:00
dslomov@chromium.org
5ccd697875 Do not look up ArrayBuffer on global object in typed array constructor.
BUG=v8:2931
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/27238009

Patch from Ben Noordhuis <info@bnoordhuis.nl>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17215 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 11:27:12 +00:00
jkummerow@chromium.org
1d580b4fd0 Delete dead code from HOptimizedGraphBuilder::VisitCompareOperation
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/26419005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17214 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 11:24:37 +00:00
yangguo@chromium.org
cf2aea3ea1 Disable RegExpInterruption test for interpreted irregexp.
Interrupts are not currently supported by interpreted irregexp.

R=mvstanton@chromium.org
BUG=

Review URL: https://codereview.chromium.org/27312002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17209 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 08:59:26 +00:00
mstarzinger@chromium.org
b01a6a4e01 Remove deprecated "i::" prefix from the pre-parser.
R=ulan@chromium.org

Review URL: https://codereview.chromium.org/26296006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17208 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
2013-10-15 08:57:36 +00:00