Handle misaligned loads and stores in load elimination. Do not track misaligned loads and be conservative about invalidating misaligned stores. Add more tests for number conversion to string (NumberToStringStub exhibits misaligned loads)

BUG=v8:2934
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/28383003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17294 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/32243003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17291 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This fixes a performance regression that was caused by converting the
BinaryOpStub to a Hydrogen code stub. It also fixes a leftover TODO wrt.
the handling of Number*String or String*Number versions of the stub.

R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/27674002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17290 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Revert "Add a soft-deopt in keyed element access when current IC is pre-monomorphic and no type feedback was collected."

This reverts commit r17288 for breaking the mjsunit/unbox-double-arrays
test on almost every platform.

TBR=hpayer@chromium.org

Review URL: https://codereview.chromium.org/32373002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17289 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Add a soft-deopt in keyed element access when current IC is pre-monomorphic and no type feedback was collected.

BUG=
R=danno@chromium.org

Review URL: https://codereview.chromium.org/32263002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17288 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=v8:2938
R=dslomov@chromium.org

Review URL: https://codereview.chromium.org/28793002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17286 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This reverts commit r17281.

TBR=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/32113002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17284 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/31923003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17281 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=dslomov@chromium.org
BUG=v8:2938

Review URL: https://codereview.chromium.org/28723002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17279 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/31913002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@17276 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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