platform/upstream/v8.git
10 years agoARM: Do not set FPSCR when converting to clamped uint8
rmcilroy@chromium.org [Tue, 15 Apr 2014 09:58:09 +0000 (09:58 +0000)]
ARM: Do not set FPSCR when converting to clamped uint8

Setting the FPSCR flags is expensive on some CPUs. Get rid of repeated
setting of the FPSCR by relying on the correct default flags being set
when doing uint8 clamping. Also use vcvt_u32_f64 instead of vcvt_s32_f64,
which enables removing the check against zero (vcvt_u32_f64 will clamp to
zero).

To be on the safe side, add asserts to check that the VFP rounding mode
flags are set to default as expected.

This increases performance of a hot loop repeatedly setting
Uint8ClampedArray values on some CPUs by as much as a factor of 12.

BUG=v8:3253
LOG=N
R=jacob.bramley@arm.com, rmcilroy@chromium.org, ulan@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=20676

Review URL: https://codereview.chromium.org/230473005

Patch from Olli Etuaho <oetuaho@nvidia.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20755 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoBump kMaxFastProperties (non-keyed assignments) to 128.
verwaest@chromium.org [Tue, 15 Apr 2014 08:59:51 +0000 (08:59 +0000)]
Bump kMaxFastProperties (non-keyed assignments) to 128.

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/238683004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20754 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoremove some isolate::current uses from api.cc
dcarney@chromium.org [Tue, 15 Apr 2014 08:30:42 +0000 (08:30 +0000)]
remove some isolate::current uses from api.cc

R=svenpanne@chromium.org

BUG=

Review URL: https://codereview.chromium.org/238843002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20753 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRefactor ParseFunctionLiteral.
marja@chromium.org [Tue, 15 Apr 2014 08:29:24 +0000 (08:29 +0000)]
Refactor ParseFunctionLiteral.

It was a pretty monstrous 500 line function.

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/237243003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20752 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland r20692 "Check stack limit in ArgumentAdaptorTrampoline."
ulan@chromium.org [Tue, 15 Apr 2014 08:26:26 +0000 (08:26 +0000)]
Reland r20692 "Check stack limit in ArgumentAdaptorTrampoline."

BUG=353058
LOG=N
TEST=mjsunit/regress/regress-353058
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/236633006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20751 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoFix bogus Object.isSealed check in some Array builtins.
mstarzinger@chromium.org [Tue, 15 Apr 2014 08:25:42 +0000 (08:25 +0000)]
Fix bogus Object.isSealed check in some Array builtins.

R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/237253002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20750 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify TypedArray setters.
yangguo@chromium.org [Tue, 15 Apr 2014 08:15:33 +0000 (08:15 +0000)]
Handlify TypedArray setters.

R=dslomov@chromium.org

Review URL: https://codereview.chromium.org/235923013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20749 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemove assertion from callers of TryCall.
yangguo@chromium.org [Tue, 15 Apr 2014 07:56:00 +0000 (07:56 +0000)]
Remove assertion from callers of TryCall.

R=ishell@chromium.org
BUG=363280
LOG=N

Review URL: https://codereview.chromium.org/238753003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20748 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemoved GetDefaultIsolate{Debugger,ForLocking,StackGuard}.
svenpanne@chromium.org [Tue, 15 Apr 2014 07:47:33 +0000 (07:47 +0000)]
Removed GetDefaultIsolate{Debugger,ForLocking,StackGuard}.

Some first steps towards removing the default Isolate. Fixed argument
order on the way, incl. temporary helpers.

BUG=359977
LOG=y
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/238803002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20747 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "Track field types.".
bmeurer@chromium.org [Tue, 15 Apr 2014 07:36:47 +0000 (07:36 +0000)]
Reland "Track field types.".

This is an initial step towards tracking the exact types instead of just
the representations of fields. It adds support to track up to one map of
heap object field values, eliminating various map checks on values
loaded from such fields, at the cost of making stores to such fields
slightly more expensive.

Issues with transitioning stores and fast object literals in Crankshaft
fixed.

TEST=mjsunit/field-type-tracking
R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/238773002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20746 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "Handlify GetDeclaredAccessorProperty."
yangguo@chromium.org [Tue, 15 Apr 2014 06:54:56 +0000 (06:54 +0000)]
Reland "Handlify GetDeclaredAccessorProperty."

R=jarin@chromium.org

Review URL: https://codereview.chromium.org/236613006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20745 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemove unnecessary %UnwrapGlobalProxy calls from object-observe.js
adamk@chromium.org [Mon, 14 Apr 2014 20:52:20 +0000 (20:52 +0000)]
Remove unnecessary %UnwrapGlobalProxy calls from object-observe.js

The intent of these calls was to properly key the WeakMap get/set calls
on the underlying global object, not the proxy, since that is the object
actually being observed. But unwrapping at this layer is unnecessary
since GetIdentityHash will already do the unwrapping (via its call to
GetHiddenProperty).

Also remove the runtime function itself, as these were the only callers,
and remove the now-redundant IS_SPEC_OBJECT() checks from object-observe.js's
MapWrapper type.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/234143002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20740 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "Handlify DebugLookupResultValue."
yangguo@chromium.org [Mon, 14 Apr 2014 19:38:29 +0000 (19:38 +0000)]
Reland "Handlify DebugLookupResultValue."

R=jarin@chromium.org

Review URL: https://codereview.chromium.org/237013003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20739 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "Handlify Runtime::InitializeIntrinsicFunctionNames."
yangguo@chromium.org [Mon, 14 Apr 2014 19:37:51 +0000 (19:37 +0000)]
Reland "Handlify Runtime::InitializeIntrinsicFunctionNames."

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/236823003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20738 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoExport [Dis]allowJavascriptExecutionScope
yangguo@chromium.org [Mon, 14 Apr 2014 19:34:32 +0000 (19:34 +0000)]
Export [Dis]allowJavascriptExecutionScope

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/236833009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20737 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoInitial steps for moving code onto Map
verwaest@chromium.org [Mon, 14 Apr 2014 16:21:38 +0000 (16:21 +0000)]
Initial steps for moving code onto Map

BUG=
R=ishell@chromium.org

Review URL: https://codereview.chromium.org/237143004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20736 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify Heap::AllocateJSArrayStorage and friends.
mstarzinger@chromium.org [Mon, 14 Apr 2014 16:05:19 +0000 (16:05 +0000)]
Handlify Heap::AllocateJSArrayStorage and friends.

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/236983002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20735 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "HashTable::Shrink() handlified and derived template parameter added to HashTa...
ishell@chromium.org [Mon, 14 Apr 2014 15:56:57 +0000 (15:56 +0000)]
Reland "HashTable::Shrink() handlified and derived template parameter added to HashTable hierarchy."

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/227423004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20734 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRe-reland "More tests for Union & Intersect"
rossberg@chromium.org [Mon, 14 Apr 2014 15:35:36 +0000 (15:35 +0000)]
Re-reland "More tests for Union & Intersect"

R=jarin@chromium.org
BUG=

Review URL: https://codereview.chromium.org/237143002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20733 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Handlify DebugLookupResultValue."
yangguo@chromium.org [Mon, 14 Apr 2014 14:34:37 +0000 (14:34 +0000)]
Revert "Handlify DebugLookupResultValue."

This reverts r20729.

TBR=jarin@chromium.org

Review URL: https://codereview.chromium.org/237093003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20732 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Handlify GetDeclaredAccessorProperty."
yangguo@chromium.org [Mon, 14 Apr 2014 14:31:38 +0000 (14:31 +0000)]
Revert "Handlify GetDeclaredAccessorProperty."

This reverts r20721.

TBR=jarin@chromium.org
BUG=

Review URL: https://codereview.chromium.org/234623004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20731 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoAllow allocation and GC in access check callbacks.
yangguo@chromium.org [Mon, 14 Apr 2014 14:03:20 +0000 (14:03 +0000)]
Allow allocation and GC in access check callbacks.

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/234913003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20730 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify DebugLookupResultValue.
yangguo@chromium.org [Mon, 14 Apr 2014 14:00:35 +0000 (14:00 +0000)]
Handlify DebugLookupResultValue.

R=ulan@chromium.org

Review URL: https://codereview.chromium.org/236683003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20729 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoDon't run tests that rely on compaction when compaction is turned off.
hpayer@chromium.org [Mon, 14 Apr 2014 13:52:41 +0000 (13:52 +0000)]
Don't run tests that rely on compaction when compaction is turned off.

BUG=
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/236203010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20728 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoClean up default-platform gyp defines
jochen@chromium.org [Mon, 14 Apr 2014 13:38:16 +0000 (13:38 +0000)]
Clean up default-platform gyp defines

The files that are excluded do not exist at that location, so the rule
is not effective.

We'll first want to move the platform stuff into a separate library, so
I'm just removing those lines for now

BUG=v8:3015
R=dcarney@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/237053002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20727 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Reland "HashTable::Shrink() handlified and derived template parameter added...
ishell@chromium.org [Mon, 14 Apr 2014 13:25:16 +0000 (13:25 +0000)]
Revert "Reland "HashTable::Shrink() handlified and derived template parameter added to HashTable hierarchy.""

This reverts r20724.

TBR=jarin@chromium.org

Review URL: https://codereview.chromium.org/237043002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20726 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoFix initial FPSCR state in ARM simulator
rmcilroy@chromium.org [Mon, 14 Apr 2014 12:50:24 +0000 (12:50 +0000)]
Fix initial FPSCR state in ARM simulator

Rounding mode should be round-to-nearest by default, and NaN mode bit
should be 0, like in a real ARM system.

BUG=3253
LOG=N
R=rmcilroy@chromium.org

Review URL: https://codereview.chromium.org/235893005

Patch from Olli Etuaho <oetuaho@nvidia.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20725 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "HashTable::Shrink() handlified and derived template parameter added to HashTa...
ishell@chromium.org [Mon, 14 Apr 2014 12:50:04 +0000 (12:50 +0000)]
Reland "HashTable::Shrink() handlified and derived template parameter added to HashTable hierarchy."

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/235643002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20724 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoFix result of LCodeGen::DoWrapReceiver for strict functions and builtins.
ulan@chromium.org [Mon, 14 Apr 2014 11:58:18 +0000 (11:58 +0000)]
Fix result of LCodeGen::DoWrapReceiver for strict functions and builtins.

BUG=362128
LOG=Y
TEST=mjsunit/regress/regress-362128
R=jacob.bramley@arm.com

Review URL: https://codereview.chromium.org/226363007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20723 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify AddToOptimizedCodeMap.
yangguo@chromium.org [Mon, 14 Apr 2014 11:46:35 +0000 (11:46 +0000)]
Handlify AddToOptimizedCodeMap.

R=ulan@chromium.org

Review URL: https://codereview.chromium.org/236963002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20722 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify GetDeclaredAccessorProperty.
yangguo@chromium.org [Mon, 14 Apr 2014 11:46:02 +0000 (11:46 +0000)]
Handlify GetDeclaredAccessorProperty.

R=dcarney@chromium.org

Review URL: https://codereview.chromium.org/236943002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20721 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlification in ic.cc
mvstanton@chromium.org [Mon, 14 Apr 2014 11:44:13 +0000 (11:44 +0000)]
Handlification in ic.cc

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/235453010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20720 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Handlify Runtime::InitializeIntrinsicFunctionNames."
yangguo@chromium.org [Mon, 14 Apr 2014 11:43:40 +0000 (11:43 +0000)]
Revert "Handlify Runtime::InitializeIntrinsicFunctionNames."

This reverts r20718.

TBR=jarin@chromium.org

Review URL: https://codereview.chromium.org/236343005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20719 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify Runtime::InitializeIntrinsicFunctionNames.
yangguo@chromium.org [Mon, 14 Apr 2014 11:32:12 +0000 (11:32 +0000)]
Handlify Runtime::InitializeIntrinsicFunctionNames.

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/236133005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20718 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoTry to switch Array builtins into strict mode.
mstarzinger@chromium.org [Mon, 14 Apr 2014 11:24:40 +0000 (11:24 +0000)]
Try to switch Array builtins into strict mode.

R=rossberg@chromium.org
TEST=mjsunit,test262,webkit

Review URL: https://codereview.chromium.org/233083003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20717 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoClean up the public interface of Map.
verwaest@chromium.org [Mon, 14 Apr 2014 11:19:37 +0000 (11:19 +0000)]
Clean up the public interface of Map.

BUG=
R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/234573005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20716 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoARM64: Avoid iterating through all unresolved branch infos when many are pending.
alexandre.rames@arm.com [Mon, 14 Apr 2014 11:08:14 +0000 (11:08 +0000)]
ARM64: Avoid iterating through all unresolved branch infos when many are pending.

Instead, inspect the label chain and delete pending information for every branch
in the chain.

R=ulan@chromium.org

Review URL: https://codereview.chromium.org/227043010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20715 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoARM64: Fix EnsureSpace to not emit pools.
alexandre.rames@arm.com [Mon, 14 Apr 2014 11:04:46 +0000 (11:04 +0000)]
ARM64: Fix EnsureSpace to not emit pools.

The EnsureSpace scope must not trigger any code emission, so it should only grow
the buffer, and not check for pools.

R=ulan@chromium.org

Review URL: https://codereview.chromium.org/230223005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20714 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoMake V8 releases script add useful links.
machenbach@chromium.org [Mon, 14 Apr 2014 10:32:00 +0000 (10:32 +0000)]
Make V8 releases script add useful links.

BUG=
R=jarin@chromium.org

Review URL: https://codereview.chromium.org/227303005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20713 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Reland "More tests for Union & Intersect""
rossberg@chromium.org [Mon, 14 Apr 2014 09:46:56 +0000 (09:46 +0000)]
Revert "Reland "More tests for Union & Intersect""

Need to reproduce wrong result only occurring with ASAN.

TBR=jarin@chromium.org
BUG=

Review URL: https://codereview.chromium.org/236873002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20712 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify Object::ToObject.
yangguo@chromium.org [Mon, 14 Apr 2014 09:19:09 +0000 (09:19 +0000)]
Handlify Object::ToObject.

R=bmeurer@chromium.org, mvstanton@chromium.org

Review URL: https://codereview.chromium.org/235943007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20711 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoSome tests and simplified TransitionArray copying
mvstanton@chromium.org [Mon, 14 Apr 2014 08:58:00 +0000 (08:58 +0000)]
Some tests and simplified TransitionArray copying

Tests for verifying that we deal correctly with shrinking transition
arrays while allocating a copy of one.

Also, we can rely on a transition array only shrinking and not
disappearing during gc while copying one.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/232883003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20710 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoParser cleanup: PreParser doesn't need to produce symbol data any more.
marja@chromium.org [Mon, 14 Apr 2014 08:49:23 +0000 (08:49 +0000)]
Parser cleanup: PreParser doesn't need to produce symbol data any more.

State of the art:
- Chromium doesn't do a separate preparsing phase any more.
- We start parsing with Parser, and when it sees a lazy function, it falls back
to PreParser for that function.
- The symbol data should contain symbols which are *outside* lazy functions.
- So Parser should always produce symbol data, and PreParser should never.
- Because it's this simple now, we don't need to keep track of "should
produce symbol data" (i.e., whether we're inside a lazy func or not).

R=ulan@chromium.org
BUG=

Review URL: https://codereview.chromium.org/222123003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20707 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "More tests for Union & Intersect"
rossberg@chromium.org [Mon, 14 Apr 2014 08:46:39 +0000 (08:46 +0000)]
Reland "More tests for Union & Intersect"

Fixes size approximation in Intersect. Also lowers the number of fuzzed types in test, to address time-outs.

R=bmeurer@chromium.org
BUG=

Review URL: https://codereview.chromium.org/226523004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20706 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoAllow GetScriptNameOrSourceURL to be called with exception pending.
yangguo@chromium.org [Mon, 14 Apr 2014 08:27:00 +0000 (08:27 +0000)]
Allow GetScriptNameOrSourceURL to be called with exception pending.

R=jarin@chromium.org, ishell@chromium.org

Review URL: https://codereview.chromium.org/235943006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20705 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Track field types."
jarin@chromium.org [Mon, 14 Apr 2014 08:24:15 +0000 (08:24 +0000)]
Revert "Track field types."

Revert r20701.

TBR=bmeurer@chromium.org
BUG=

Review URL: https://codereview.chromium.org/236843002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20704 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoFail the compilation if the cached data is invalid.
marja@chromium.org [Mon, 14 Apr 2014 07:35:46 +0000 (07:35 +0000)]
Fail the compilation if the cached data is invalid.

R=svenpanne@chromium.org
BUG=

Review URL: https://codereview.chromium.org/234953002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20703 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemove ConfigureDefaults version that does not take the amount of virtual mem
jochen@chromium.org [Mon, 14 Apr 2014 06:55:42 +0000 (06:55 +0000)]
Remove ConfigureDefaults version that does not take the amount of virtual mem

R=dcarney@chromium.org
LOG=n
BUG=none

Review URL: https://codereview.chromium.org/230553003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20702 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoTrack field types.
bmeurer@chromium.org [Mon, 14 Apr 2014 06:29:15 +0000 (06:29 +0000)]
Track field types.

This is an initial step towards tracking the exact types instead of just the representations of fields. It adds support to track up to one map of heap object field values, eliminating various map checks on values loaded from such fields, at the cost of making stores to such fields slightly more expensive.

TEST=mjsunit/field-type-tracking
R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/167303005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20701 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "HashTable::Shrink() handlified and derived template parameter added to HashTa...
ishell@chromium.org [Fri, 11 Apr 2014 17:55:52 +0000 (17:55 +0000)]
Revert "HashTable::Shrink() handlified and derived template parameter added to HashTable hierarchy."

This reverts commit r20697 for breaking ASAN build.

TBR=yangguo@chromium.org

Review URL: https://codereview.chromium.org/235353003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20698 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHashTable::Shrink() handlified and derived template parameter added to HashTable...
ishell@chromium.org [Fri, 11 Apr 2014 17:44:58 +0000 (17:44 +0000)]
HashTable::Shrink() handlified and derived template parameter added to HashTable hierarchy.

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/234663004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20697 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoSkip mjsunit/regress/regress-353058 for ASAN and ARM until r20692 is relanded.
ulan@chromium.org [Fri, 11 Apr 2014 17:03:35 +0000 (17:03 +0000)]
Skip mjsunit/regress/regress-353058 for ASAN and ARM until r20692 is relanded.

TBR=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/232463005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20696 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert r20692 "Check stack limit in ArgumentAdaptorTrampoline."
ulan@chromium.org [Fri, 11 Apr 2014 15:59:12 +0000 (15:59 +0000)]
Revert r20692 "Check stack limit in ArgumentAdaptorTrampoline."

Reason: mjsunit/regress/regress-353058 failure on GC-stress.

TBR=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/234813004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20695 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoGcstress bug fix: Transition arrays may get smaller during gc.
mvstanton@chromium.org [Fri, 11 Apr 2014 14:25:00 +0000 (14:25 +0000)]
Gcstress bug fix: Transition arrays may get smaller during gc.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/234873004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20694 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify Map::RawCopy
verwaest@chromium.org [Fri, 11 Apr 2014 13:48:52 +0000 (13:48 +0000)]
Handlify Map::RawCopy

BUG=
R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/235183002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20693 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoCheck stack limit in ArgumentAdaptorTrampoline.
ulan@chromium.org [Fri, 11 Apr 2014 13:39:19 +0000 (13:39 +0000)]
Check stack limit in ArgumentAdaptorTrampoline.
BUG=353058
LOG=N
TEST=mjsunit/regress/regress-353058
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/215853005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20692 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoDo not call user defined getter of Error.stackTraceLimit.
ulan@chromium.org [Fri, 11 Apr 2014 13:16:36 +0000 (13:16 +0000)]
Do not call user defined getter of Error.stackTraceLimit.

Handlify GetNormalizedProperty.

BUG=360733
LOG=N
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/233243005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20691 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoInline immutable property loads
verwaest@chromium.org [Fri, 11 Apr 2014 13:07:10 +0000 (13:07 +0000)]
Inline immutable property loads

When a non-configurable, non-writable field is
read from a constant holder, the load is
eliminated and replaced with the direct value
of the field

BUG=
R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/232853002

Patch from Petka Antonov <p.antonov@partner.samsung.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20690 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReland "Handlify GetProperty."
yangguo@chromium.org [Fri, 11 Apr 2014 12:47:34 +0000 (12:47 +0000)]
Reland "Handlify GetProperty."

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/235083002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20689 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoMIPS: Implement handlified String::Equals and Name::Equals.
palfia@homejinni.com [Fri, 11 Apr 2014 12:40:26 +0000 (12:40 +0000)]
MIPS: Implement handlified String::Equals and Name::Equals.

Port r20669 (29a6ec29)

Note:
This commit fixes a typo.

BUG=
R=danno@chromium.org

Review URL: https://codereview.chromium.org/234993003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20688 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "More tests for Union & Intersect"
dcarney@chromium.org [Fri, 11 Apr 2014 12:33:20 +0000 (12:33 +0000)]
Revert "More tests for Union & Intersect"

This reverts r20684.

TBR=rossberg@chromium.org

BUG=

Review URL: https://codereview.chromium.org/235133002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20687 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify Map::CopyDropDescriptors().
mvstanton@chromium.org [Fri, 11 Apr 2014 12:13:53 +0000 (12:13 +0000)]
Handlify Map::CopyDropDescriptors().

* And contain knowledge better in TransitionArray and DescriptorArray (for example WhitenessWitness is now private to DescriptorArray).
* And remove some factory methods
* And handlify some other things.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/234783002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20686 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Handlify GetProperty."
yangguo@chromium.org [Fri, 11 Apr 2014 11:56:54 +0000 (11:56 +0000)]
Revert "Handlify GetProperty."

This reverts r20682.

TBR=dcarney@chromium.org

Review URL: https://codereview.chromium.org/234893003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20685 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoMore tests for Union & Intersect
rossberg@chromium.org [Fri, 11 Apr 2014 11:53:35 +0000 (11:53 +0000)]
More tests for Union & Intersect

Some fixes of corner cases on the way

R=bmeurer@chromium.org
BUG=

Review URL: https://codereview.chromium.org/230923005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20684 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemove the PreCompile API and ScriptData.
marja@chromium.org [Fri, 11 Apr 2014 11:44:49 +0000 (11:44 +0000)]
Remove the PreCompile API and ScriptData.

The new compilation API (ScriptCompiler::Compile) can produce the same data, so
the separate precompilation phase is not needed. ScriptData is replaced by
ScriptCompiler::CachedData.

R=svenpanne@chromium.org
BUG=

Review URL: https://codereview.chromium.org/225753004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20683 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify GetProperty.
yangguo@chromium.org [Fri, 11 Apr 2014 11:26:22 +0000 (11:26 +0000)]
Handlify GetProperty.

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/233233004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20682 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemove MUST_USE_RESULT for Execution::TryCall.
yangguo@chromium.org [Fri, 11 Apr 2014 11:09:49 +0000 (11:09 +0000)]
Remove MUST_USE_RESULT for Execution::TryCall.

TBR=ishell@chromium.org

Review URL: https://codereview.chromium.org/235003002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20681 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoReturn MaybeHandle from Invoke.
yangguo@chromium.org [Fri, 11 Apr 2014 10:41:09 +0000 (10:41 +0000)]
Return MaybeHandle from Invoke.

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/231883007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20680 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoMake maps in monomorphic IC stubs weak.
ulan@chromium.org [Fri, 11 Apr 2014 10:36:09 +0000 (10:36 +0000)]
Make maps in monomorphic IC stubs weak.

Maps in monomorphic Load, KeyedLoad, Store, KeyedStore, and CompareNil IC
stubs are treated as weak references by the marking visitor.

During generation of an IC stub with a weak map, the stub is appended to the
dependent code array of the map. When the map dies, all stubs in its dependent
code array are invalidated by setting embedded maps to undefined.

BUG=v8:2073
LOG=Y
TEST=cctest/test-heap/WeakMapInMonomorphic*IC
R=mstarzinger@chromium.org, verwaest@chromium.org

Review URL: https://codereview.chromium.org/188783003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20679 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "ARM: Do not set FPSCR when converting to clamped uint8"
rmcilroy@chromium.org [Fri, 11 Apr 2014 10:04:25 +0000 (10:04 +0000)]
Revert "ARM: Do not set FPSCR when converting to clamped uint8"

This reverts commit r20676.

TBR=danno@chromium.org

Review URL: https://codereview.chromium.org/233013005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20678 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoInline TypeImpl::NowContains().
bmeurer@chromium.org [Fri, 11 Apr 2014 09:24:35 +0000 (09:24 +0000)]
Inline TypeImpl::NowContains().

R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/234743003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20677 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoARM: Do not set FPSCR when converting to clamped uint8
rmcilroy@chromium.org [Fri, 11 Apr 2014 09:22:14 +0000 (09:22 +0000)]
ARM: Do not set FPSCR when converting to clamped uint8

Setting the FPSCR flags is expensive on some CPUs. Get rid of repeated
setting of the FPSCR by relying on the correct default flags being set
when doing uint8 clamping. Also use vcvt_u32_f64 instead of vcvt_s32_f64,
which enables removing the check against zero (vcvt_u32_f64 will clamp to
zero).

To be on the safe side, add asserts to check that the VFP rounding mode
flags are set to default as expected.

This increases performance of a hot loop repeatedly setting
Uint8ClampedArray values on some CPUs by as much as a factor of 12.

BUG=v8:3253
LOG=N
R=jacob.bramley@arm.com, rmcilroy@chromium.org, ulan@chromium.org

Review URL: https://codereview.chromium.org/230473005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20676 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoUnify mechanism to find trailing AllocationMementos
danno@chromium.org [Fri, 11 Apr 2014 09:20:56 +0000 (09:20 +0000)]
Unify mechanism to find trailing AllocationMementos

R=hpayer@chromium.org

Review URL: https://codereview.chromium.org/234703003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20675 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoMake sure that ranges are not accessed after range analysis. Remove HValue::PrintRangeTo.
svenpanne@chromium.org [Fri, 11 Apr 2014 09:17:18 +0000 (09:17 +0000)]
Make sure that ranges are not accessed after range analysis. Remove HValue::PrintRangeTo.

The ranges are simply wrong after range analysis, and we should only rely on computed flags.

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/234583005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20674 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoAdd rmcilroy@chromium.org to src/OWNERS.
ulan@chromium.org [Fri, 11 Apr 2014 08:55:18 +0000 (08:55 +0000)]
Add rmcilroy@chromium.org to src/OWNERS.

BUG=
R=danno@chromium.org

Review URL: https://codereview.chromium.org/234793002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20673 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemove remnant from deferred error formatting on GC.
yangguo@chromium.org [Fri, 11 Apr 2014 08:31:10 +0000 (08:31 +0000)]
Remove remnant from deferred error formatting on GC.

Deferred error formatting on GC was a failed experiment. It has been
removed, except for this part.  See r13371.

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/233163004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20672 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoFix chromium ranges output in v8 releases script.
machenbach@chromium.org [Fri, 11 Apr 2014 07:40:48 +0000 (07:40 +0000)]
Fix chromium ranges output in v8 releases script.

Add space after commas for auto line break in spreadsheets.
Add default empty strings to avoid undefined values in spreadsheets.

BUG=
R=jarin@chromium.org

Review URL: https://codereview.chromium.org/233343002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20671 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoImplement handlified String::Equals and Name::Equals.
yangguo@chromium.org [Fri, 11 Apr 2014 07:27:25 +0000 (07:27 +0000)]
Implement handlified String::Equals and Name::Equals.

R=ulan@chromium.org

Review URL: https://codereview.chromium.org/225823003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20669 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoDrop unused name parameter of SetPropertyToField().
bmeurer@chromium.org [Fri, 11 Apr 2014 07:15:17 +0000 (07:15 +0000)]
Drop unused name parameter of SetPropertyToField().

R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/233673007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20667 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoAvoid type assertion on object comparison in Hydrogen - the comparison is unreachable...
jarin@chromium.org [Fri, 11 Apr 2014 06:45:24 +0000 (06:45 +0000)]
Avoid type assertion on object comparison in Hydrogen - the comparison is unreachable because of previous checks.

BUG=
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/232053004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20666 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoUse the right kind of guard for debug code.
svenpanne@chromium.org [Fri, 11 Apr 2014 06:40:36 +0000 (06:40 +0000)]
Use the right kind of guard for debug code.

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/232883002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20665 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agox64: Make sure that the upper half of a 64bit register contains 0 for int32 values.
svenpanne@chromium.org [Fri, 11 Apr 2014 06:32:06 +0000 (06:32 +0000)]
x64: Make sure that the upper half of a 64bit register contains 0 for int32 values.

BUG=360611
LOG=y
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/225393005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20664 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoThere is no definition for HArgumentsObject, so LDummyUse confuses the register alloc...
jarin@chromium.org [Fri, 11 Apr 2014 06:29:51 +0000 (06:29 +0000)]
There is no definition for HArgumentsObject, so LDummyUse confuses the register allocator. I have recently made similar fix for HCapturedObject (see https://codereview.chromium.org/222283002/).

BUG=
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/226613007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20663 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRename PushInt64AsTwoSmis to PushRegisterAsTwoSmis and support x32 as well.
haitao.feng@intel.com [Fri, 11 Apr 2014 02:02:56 +0000 (02:02 +0000)]
Rename PushInt64AsTwoSmis to PushRegisterAsTwoSmis and support x32 as well.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/232533002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20662 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoFix android_arm64 target for experimental ndk
rmcilroy@chromium.org [Thu, 10 Apr 2014 21:28:40 +0000 (21:28 +0000)]
Fix android_arm64 target for experimental ndk

BUG=354405
LOG=N
R=ulan@chromium.org

Review URL: https://codereview.chromium.org/233163003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20659 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoARM64: Preserve x8 and x9 when necessary.
Jacob.Bramley@arm.com [Thu, 10 Apr 2014 15:47:45 +0000 (15:47 +0000)]
ARM64: Preserve x8 and x9 when necessary.

Fix a couple of places were x8 and x9 are excluded from lists of saved
registers. These are caller-saved registers, so C code can corrupt them.

x8 and x9 were originally reserved for debug code in the ARM64 port, so
we didn't bother preserving them, but they are now normal allocatable
registers.

BUG=v8:3263
LOG=N
R=rmcilroy@chromium.org, ulan@chromium.org

Review URL: https://codereview.chromium.org/233373002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20658 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoAdd support for --raw and --omit to js2c.
vogelheim@chromium.org [Thu, 10 Apr 2014 14:53:32 +0000 (14:53 +0000)]
Add support for --raw and --omit to js2c.
  --raw writes the raw source data to a separate file
  --omit allows omitting the source data from the generated files.

The intention is (future) support for having the embedder optionally
store the source data 'blob' and handling it to V8::Initialize, with
the goal of reducing the binary size of V8.

The patch also contains numerous unrelated changes/refactorings in the hope of increasing maintainability. Let me know whether you agree. In particular:
- Remove some unused code.
- Do not overwrite Python built-ins (e.g. the type() function)
- Do not use a string as exception object (no longer supported in python 2.7)
- Add command line argument handling + help text.
- Split logic into:
  - PrepareSources - which reads + preprocesses the source files
  - BuildMetadata - which takes the prepared sources and generates the data the code template needs.

BUG=355539
LOG=N
R=jochen@chromium.org

Review URL: https://codereview.chromium.org/225723002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20657 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRemove V8_ALLOW_ACCESS_TO_RAW_HANDLE_CONSTRUCTOR.
vogelheim@chromium.org [Thu, 10 Apr 2014 13:51:30 +0000 (13:51 +0000)]
Remove V8_ALLOW_ACCESS_TO_RAW_HANDLE_CONSTRUCTOR.

The usage of this define has been obsoleted by removal of UnsafePersistent
from Chromium.

Depends on https://codereview.chromium.org/230613005

R=dcarney@chromium.org
BUG=276323
LOG=N

Review URL: https://codereview.chromium.org/230443004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20656 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoPartially fix semantics of Array.push()
danno@chromium.org [Thu, 10 Apr 2014 13:17:48 +0000 (13:17 +0000)]
Partially fix semantics of Array.push()

Semantics of elements accessors are now preserved in all optimized code paths
through Array.push(). Previously it was possible to have inconsistent behavior
between optimized and unoptimized code, and there were cases where element
accessors were completely ingored.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/232873002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20655 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoBugfix: A TransitionArray can disappear during copy.
mvstanton@chromium.org [Thu, 10 Apr 2014 13:06:52 +0000 (13:06 +0000)]
Bugfix: A TransitionArray can disappear during copy.

During handlification of TransitionArray code, an error was introduced
in TransitionArray::CopyInsert because after creating a copy of a
TransitionArray, it may be that the array disappears during GC
because it is modified during the marking of the owning map.

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/228483005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20654 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert r20652 "Handlify and convert string.length to new API-style accessor."
ulan@chromium.org [Thu, 10 Apr 2014 12:40:55 +0000 (12:40 +0000)]
Revert r20652 "Handlify and convert string.length to new API-style accessor."

Reason: test failure with --noenable-sse3.

R=dcarney@chromium.org
TBR=dcarney@chromium.org

Review URL: https://codereview.chromium.org/232933003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20653 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoHandlify and convert string.length to new API-style accessor.
ulan@chromium.org [Thu, 10 Apr 2014 12:00:36 +0000 (12:00 +0000)]
Handlify and convert string.length to new API-style accessor.

BUG=
R=dcarney@chromium.org, yangguo@chromium.org

Review URL: https://codereview.chromium.org/230693003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20652 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoWork towards unifying descriptor array handling.
verwaest@chromium.org [Thu, 10 Apr 2014 11:59:10 +0000 (11:59 +0000)]
Work towards unifying descriptor array handling.

BUG=
R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/231283006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20651 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoTreat uninitialized as internal type.
bmeurer@chromium.org [Thu, 10 Apr 2014 11:51:03 +0000 (11:51 +0000)]
Treat uninitialized as internal type.

TEST=cctest/test-types
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/232913002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20650 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoYet more type system tests
rossberg@chromium.org [Thu, 10 Apr 2014 11:49:18 +0000 (11:49 +0000)]
Yet more type system tests

R=bmeurer@chromium.org
BUG=

Review URL: https://codereview.chromium.org/232843002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20649 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoRevert "Populate receiver types when there is no type feedback"
dcarney@chromium.org [Thu, 10 Apr 2014 11:01:09 +0000 (11:01 +0000)]
Revert "Populate receiver types when there is no type feedback"

This reverts r20646.

TBR=verwaest@chromium.org

BUG=

Review URL: https://codereview.chromium.org/232903002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20648 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoAdd random_seed parameter to run-deopt-fuzzer.
bmeurer@chromium.org [Thu, 10 Apr 2014 10:43:13 +0000 (10:43 +0000)]
Add random_seed parameter to run-deopt-fuzzer.

R=machenbach@chromium.org

Review URL: https://codereview.chromium.org/226843003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20647 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoPopulate receiver types when there is no type feedback
verwaest@chromium.org [Thu, 10 Apr 2014 09:49:53 +0000 (09:49 +0000)]
Populate receiver types when there is no type feedback

When there is no type feedback yet, ComputeReceiverTypes
should still populate the SmallMapList when the receiver
is a HConstant.

BUG=
R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/230363003

Patch from Petka Antonov <p.antonov@partner.samsung.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20646 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoDo not use ranges after range analysis.
svenpanne@chromium.org [Thu, 10 Apr 2014 09:40:17 +0000 (09:40 +0000)]
Do not use ranges after range analysis.

Due to the SSA vs. SSI difference, we are only allowed to use the
flags computed during range analysis, not the ranges themselves. For
the case at hand, there is no such flag, so the condition is simply
remvoed.

BUG=361608
LOG=y
R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/232553004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20645 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

10 years agoObject::GetElements() and friends maybehandlification.
ishell@chromium.org [Thu, 10 Apr 2014 09:20:11 +0000 (09:20 +0000)]
Object::GetElements() and friends maybehandlification.

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/231103002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@20644 ce2b1a6d-e550-0410-aec6-3dcde31c8c00