AuroraMiddleware/v8
1
0
Fork 0
Code Issues 2 Pull Requests Releases Wiki Activity
16,676 Commits 1 Branch 0 Tags 839 MiB
848a9af6b4
Commit Graph

16676 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
vogelheim@chromium.org
06086a90e0 Fix the "PersistentValueMap" memory leak reported here: 
http://build.chromium.org/p/client.v8/builders/V8%20Linux64%20ASAN

The bug:
The code assumed that a weak Persistent whose weak callback is being
called would still be weak. That isn't true since the persistent is
un-weakened by the garbage collector before calling the weak callback. [1]

Specifically, PersistentValueMap would funnel all 'remove' actions
through its Release method, which uses PersistentBase::ClearWeak to
obtain the callback data. [2] For 'removes' caused by the weak callback,
ClearWeak always returns a NULL-pointer since by that time the weak
persistent was already un-weakend. The result was a memory leak in
the test, since the code to delete the weak callback data would
delete NULL.

The fix:
I explicity call Traits::DisposeCallbackData from the weak callback
with the data obtained from the v8::WeakCallbackData. To avoid invalid
calls to DisposeCallbackData, I also check whether this instance is
(still) weak before calling it. (That check could easily be elided
if it's expensive, for the price of having two 'remove' code paths.)

Severety:
Probably low. At least in Chromium, noone uses the API in a way to
trigger this; only the test does.

[1] https://code.google.com/p/chromium/codesearch#chromium/src/v8/src/global-handles.cc&q=global-handles.cc&sq=package:chromium&type=cs&l=231
[2] https://code.google.com/p/chromium/codesearch#chromium/src/v8/include/v8-util.h&sq=package:chromium&l=332-345

R=dcarney@chromium.org, dcarney

BUG=

Review URL: https://codereview.chromium.org/297193004

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21514 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 09:31:06 +00:00
hpayer@chromium.org
0fde744d00 Escape % in SNPrintF call in pretenuring test. 
BUG=
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/298183002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21513 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 08:41:12 +00:00
jochen@chromium.org
84e078e561 Reland 21502 - "Move OS::MemCopy and OS::MemMove out of platform to utils" 
Verified that arm builds locally.

BUG=none
TBR=jkummerow@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/306473004

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21512 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 07:57:22 +00:00
hpayer@chromium.org
b3c067ff03 Remove global pretenuring mode. 
BUG=
R=bmeurer@chromium.org, mvstanton@chromium.org

Review URL: https://codereview.chromium.org/304493002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21511 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 07:48:36 +00:00
yangguo@chromium.org
c03b98484a Fix d8's interactive shell. 
R=danno@chromium.org

Review URL: https://codereview.chromium.org/305553002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21510 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 07:19:32 +00:00
bmeurer@chromium.org
7e8678bbaa Avoid HeapObject check in HStoreNamedField. 
This way an HStoreNamedField instruction can never deoptimize
itself, which is another important step towards a working
store elimination.

R=jarin@chromium.org

Review URL: https://codereview.chromium.org/299373005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21509 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 07:17:08 +00:00
yangguo@chromium.org
94b4aef7d6 Fix arm64 gc stress issue. 
R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/306483002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21506 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 06:35:45 +00:00
bmeurer@chromium.org
6ae4a89e79 Relax register constraints for LMathSqrt. 
R=jarin@chromium.org

Review URL: https://codereview.chromium.org/299423005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21505 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-27 04:19:18 +00:00
jochen@chromium.org
eabd5a19b9 Revert 21502 - "Move OS::MemCopy and OS::MemMove out of platform to utils" 
TBR=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/297303004

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21504 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 19:56:27 +00:00
jochen@chromium.org
6640ed7235 Buildfix for arm 
TBR=jkummerow@chromium.org
LOG=n
BUG=none

Review URL: https://codereview.chromium.org/299353006

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21503 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 19:43:01 +00:00
jochen@chromium.org
a5a21a0da4 Move OS::MemCopy and OS::MemMove out of platform to utils 
Since both are jitted on some platforms and depend on codegen, they
don't belong to the platform abstraction. At the same time, I can't put
them to codegen.h, as this would introduce cyclic dependencies.

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

Review URL: https://codereview.chromium.org/302563004

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21502 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 19:33:15 +00:00
jochen@chromium.org
0f73456d3f Move NumberOfProcessorsOnline from CPU to OS 
It's really more an OS-level information, and this way the default
platform doesn't depend on CPU-level details

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

Review URL: https://codereview.chromium.org/300713002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21501 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 15:18:45 +00:00
plind44@gmail.com
4fd78904be MIPS: Skip write barriers in the fast case when setting up local context. 
Port r21481 (5973b48)

Original commit message:
The FastNewContextStub always allocates in new space, so we don't
need to update the write barrier when copying the parameters to
the newly allocated context.

BUG=
R=plind44@gmail.com

Review URL: https://codereview.chromium.org/302633002

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21500 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 14:40:55 +00:00
mvstanton@chromium.org
d755611e93 Reland "Customized support for feedback on calls to Array." and follow-up fixes. 
Comparing one CallIC::State to another was not done correctly, leading to a failure to patch a CallIC when transitioning from monomorphic Array to megamorphic.

BUG=chromium:377198,chromium:377290
LOG=Y
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/305493003

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21499 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 13:59:24 +00:00
machenbach@chromium.org
2dc61205fd Send idle notification on ASAN builder for proper tear down. 
BUG=
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/300003005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21498 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 13:34:23 +00:00
rossberg@chromium.org
9516c181a1 More nuisance 
TBR=mstarzinger@chromium.org
BUG=

Review URL: https://codereview.chromium.org/297323004

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21497 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 13:33:06 +00:00
yangguo@chromium.org
f24e37d4e4 D8 should send idle notification even after the last run. 
This, in combination with changes to the bot, should weed out false negatives.

R=marja@chromium.org

Review URL: https://codereview.chromium.org/293393003

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21496 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 13:31:55 +00:00
rossberg@chromium.org
2fbc98a65a Shut up VS 
TBR=mstarzinger@chromium.org
BUG=

Review URL: https://codereview.chromium.org/298073005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21495 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 13:24:59 +00:00
rossberg@chromium.org
dbf7d8528c Various extensions to types 
R=bmeurer@chromium.org
BUG=

Review URL: https://codereview.chromium.org/296213005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21494 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 13:10:52 +00:00
hpayer@chromium.org
6ed0102b1d Remove high promotion mode. 
BUG=
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/296413004

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21493 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 12:58:55 +00:00
mstarzinger@chromium.org
82b3b2a367 Make 'name' property on functions configurable. 
R=rossberg@chromium.org
BUG=v8:3333
LOG=N

Review URL: https://codereview.chromium.org/296413003

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21492 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 11:42:56 +00:00
yangguo@chromium.org
865c54725f Actually fix the memory leak in debugger wrt promises. 
In the fuzz test the debugger is not actually loaded...

R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/304453003

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21491 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 11:40:34 +00:00
jochen@chromium.org
0d1dc27eda Reland 21482 - "Merge v8globals.h and globals.h" 
> BUG=none
> R=mstarzinger@chromium.org
> LOG=n
>
> Review URL: https://codereview.chromium.org/293363006

BUG=none
TBR=mstarzinger@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/295373003

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21490 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 11:28:08 +00:00
yangguo@chromium.org
e08b33eaac Fix memory leak caused by fuzzing. 
%DebugPromiseHandlePrologue expects to be followed up by %..Epilogue,
otherwise we get a leak.

R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/301553002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21489 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 11:13:18 +00:00
marja@chromium.org
2b7810c09c Fix mem leaks in tests & ScriptCompiler::CompileUnbound. 
- Leak in test-parsing/DontRegressPreParserDataSizes
- Leak in test-api/EventLogging
- Leak in ScriptCompiler::CompileUnbound which won't happen during normal
operation, but exposed by test-apöi/CompiledWithInvalidCachedData.

R=yangguo@chromium.org
BUG=

Review URL: https://codereview.chromium.org/302603002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21488 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 11:04:32 +00:00
jochen@chromium.org
f0527ae685 Always fully qualify the namespace to use in v8.h 
Also, remove some unused methods.

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

Review URL: https://codereview.chromium.org/297223006

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21487 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 09:36:13 +00:00
jkummerow@chromium.org
60e665627d Revert "Customized support for feedback on calls to Array." and follow-up fixes. 
This reverts r21429, r21434, r21435, r21440, r21445.

BUG=chromium:377198
LOG=y
R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/300693002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21484 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 09:04:00 +00:00
verwaest@chromium.org
53bbe2aec9 Revert "Merge v8globals.h and globals.h" 
Because of tree redness.

TBR=jochen@chromium.org

Review URL: https://codereview.chromium.org/306443002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21483 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 08:39:04 +00:00
jochen@chromium.org
cd818d697d Merge v8globals.h and globals.h 
BUG=none
R=mstarzinger@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/293363006

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21482 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 08:15:13 +00:00
bmeurer@chromium.org
683bc694ad Skip write barriers in the fast case when setting up local context. 
The FastNewContextStub always allocates in new space, so we don't
need to update the write barrier when copying the parameters to
the newly allocated context.

R=hpayer@chromium.org

Review URL: https://codereview.chromium.org/297203002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21481 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 08:13:09 +00:00
ulan@chromium.org
3fcda0e576 Make let variables fresh in each iteration of a for-loop. 
BUG=v8:2198
LOG=N
TEST=mjsunit/harmony/block-for
R=rossberg@chromium.org

Review URL: https://codereview.chromium.org/292743009

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21480 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 08:07:02 +00:00
yangguo@chromium.org
a5f0489e5d Fix memory leak in the debugger. 
R=jarin@chromium.org

Review URL: https://codereview.chromium.org/297373002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21479 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 08:05:04 +00:00
jochen@chromium.org
b4394e5b13 Some progress on translating toolchain.gypi to gn 
BUG=none
R=brettw@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/299173002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21478 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 07:29:39 +00:00
yangguo@chromium.org
32f433c12e Fix leak in debug mirror cache. 
When fetching loaded scripts, mirror objects are created and cached.
If the cache is not cleared, it holds script objects alive.

This also fixes a minor issue with script unloading.

R=ulan@chromium.org
BUG=376534
LOG=N

Review URL: https://codereview.chromium.org/296953005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21477 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 07:05:56 +00:00
bmeurer@chromium.org
cdcee29ef7 LoadUint32() doesn't need a scratch register. 
R=jarin@chromium.org

Review URL: https://codereview.chromium.org/293363005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21476 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-26 06:41:21 +00:00
plind44@gmail.com
b2fc45955c MIPS: Fix r21470 “Reland r21442 Inobject slack tracking is done on a per-closure basis instead of per-shared info basis." 
BUG=
R=plind44@gmail.com

Review URL: https://codereview.chromium.org/297093002

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21471 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 19:32:48 +00:00
plind44@gmail.com
50e9d22a44 MIPS: Reland r21442 "Inobject slack tracking is done on a per-closure basis instead of per-shared info basis." 
Port r21457 (8db39a8)

Original commit message:
This fixes inobject slack tracking for prototype inheritance pattern that uses closures.

BUG=
R=plind44@gmail.com

Review URL: https://codereview.chromium.org/294973013

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21470 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 18:05:40 +00:00
danno@chromium.org
9c485e182b Introduce x87 port 
Support x87-only platform (ia32 without SSE)

R=danno@chromium.org

Review URL: https://codereview.chromium.org/293743005

Patch from Weiliang Lin <weiliang.lin@intel.com>.

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21469 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 16:37:27 +00:00
alexandre.rames@arm.com
8d334ed1dc Allow HPushArgument to handle more than one argument. 
R=ulan@chromium.org

Review URL: https://codereview.chromium.org/296113008

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21468 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 14:06:42 +00:00
mstarzinger@chromium.org
cf448aa15f Fix representation inference for mutable double boxes. 
R=jarin@chromium.org
BUG=v8:3307
TEST=mjsunit/regress/regress-3307
LOG=N

Review URL: https://codereview.chromium.org/298723014

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21467 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 14:02:08 +00:00
dslomov@chromium.org
45ab7d5266 Check for cached transition to ExternalArray elements kind. 
R=ishell@chromium.org, verwaest@chromium.org
BUG=v8:3337
LOG=Y

Review URL: https://codereview.chromium.org/291193011

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21466 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 14:01:17 +00:00
alexandre.rames@arm.com
18412a83b0 Provide a helper to generate multiple Lithium instructions for one Hydrogen instruction. 
R=jkummerow@chromium.org, ulan@chromium.org

Review URL: https://codereview.chromium.org/296993002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21465 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 13:15:07 +00:00
ishell@chromium.org
7c55f645d5 Cleanup after inobject slack tracking improvement. 
1) %SetExpectedNumberOfProperties() function removed.
2) Obsolete SharedFunctionInfo::BeforeVisitingPointers() removed.

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/289283018

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21464 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 12:55:57 +00:00
bmeurer@chromium.org
238a2489e2 Skip dead blocks/instructions in store elimination. 
Also improve tracing.

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/297933002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21463 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 11:48:40 +00:00
svenpanne@chromium.org
988642494f Tiny steps towards a non-local 'Merge removable simulates' phase. 
Distinguish copying/merging of the state. Better tracing.

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/299883003

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21462 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 11:25:23 +00:00
mstarzinger@chromium.org
a2082b256d Revert "Make v8::TryCatch able to consume natively thrown exceptions (again)." 
TBR=machenbach@chromium.org

Review URL: https://codereview.chromium.org/296133013

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21461 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 11:12:25 +00:00
yurys@chromium.org
7741fe0466 Support ES6 weak collections in heap profiler 
BUG=chromium:376196
LOG=Y
R=alph@chromium.org, yangguo@chromium.org

Review URL: https://codereview.chromium.org/294163005

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21460 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 11:05:22 +00:00
rodolph.perfetta@arm.com
123608ab9b ARM64 simulator fix for EXTR 
Fixes extract when imms = 0 because a left shift of 64 is not valid.

R=rodolph.perfetta@arm.com, svenpanne@chromium.org

Review URL: https://codereview.chromium.org/286193004

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21459 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 09:41:41 +00:00
bmeurer@chromium.org
a56bbe9203 Avoid dynamic initial map check when inlining call-new. 
This improves check elimination and removes a load plus
a map check for every inlined call-new.

R=hpayer@chromium.org

Review URL: https://codereview.chromium.org/293223002

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21458 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 09:30:47 +00:00
ishell@chromium.org
4b9c03050a Reland r21442 "Inobject slack tracking is done on a per-closure basis instead of per-shared info basis." 
This fixes inobject slack tracking for prototype inheritance pattern that uses closures.

R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/292183008

git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@21457 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
 2014-05-23 08:52:05 +00:00