floitschV8@gmail.com [Wed, 3 Nov 2010 12:05:01 +0000 (12:05 +0000)]
Fix assumptions in DoubleToFixed.
By luck two errors cancelled each other out.
Review URL: http://codereview.chromium.org/4135014
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5760
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
sandholm@chromium.org [Wed, 3 Nov 2010 10:30:15 +0000 (10:30 +0000)]
Minor regexp test cleanup.
Review URL: http://codereview.chromium.org/4371002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5759
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Wed, 3 Nov 2010 08:11:13 +0000 (08:11 +0000)]
Prepare push to trunk. Now working on version 2.5.5.
Review URL: http://codereview.chromium.org/4298003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5756
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Tue, 2 Nov 2010 13:37:59 +0000 (13:37 +0000)]
Remove regexp caching.
Review URL: http://codereview.chromium.org/4308001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5755
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
sgjesse@chromium.org [Tue, 2 Nov 2010 12:08:58 +0000 (12:08 +0000)]
Remove all remaining code related to exposing C-frames to the JavaScript debugger
BUG=906
Review URL: http://codereview.chromium.org/4104009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5754
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vegorov@chromium.org [Tue, 2 Nov 2010 11:56:10 +0000 (11:56 +0000)]
Fix converting NULL to Failure bugs introduced in r5698.
Review URL: http://codereview.chromium.org/4293002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5753
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Tue, 2 Nov 2010 11:45:47 +0000 (11:45 +0000)]
Remove old preparser option and behavior from the parser.
Review URL: http://codereview.chromium.org/4244003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5752
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Tue, 2 Nov 2010 08:26:33 +0000 (08:26 +0000)]
Landing for Rodolph Perfetta.
Improve V8 VFPv3 runtime detection, to address issue 914.
This patch will check for the exact word vfpv3 as well as 0xc08 (CortexA8
part number) as some earlier kernel didn't report vfpv3 for A8.
Codereview URL: http://codereview.chromium.org/4103013/show
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5751
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Tue, 2 Nov 2010 07:21:37 +0000 (07:21 +0000)]
Stand-alone parser template.
Uses existing Scanner and ParserLog.
Generates same preparse-data as existing preparser.
Review URL: http://codereview.chromium.org/4112012
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5750
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
fschneider@chromium.org [Mon, 1 Nov 2010 12:17:58 +0000 (12:17 +0000)]
Prepare push to trunk. Now working on version 2.5.4.
Review URL: http://codereview.chromium.org/4158015
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5747
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
serya@chromium.org [Mon, 1 Nov 2010 10:51:44 +0000 (10:51 +0000)]
Refactoring of v8:Arguments similary we did with v8::AccessorInfo (codereview.chromium.org/242050). GC-controlled values moved to a separate array.
Review URL: http://codereview.chromium.org/4117010
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5746
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Mon, 1 Nov 2010 09:16:39 +0000 (09:16 +0000)]
Remove invalid asserts from last commit.
TBR=erik.corry@gmail.com
Review URL: http://codereview.chromium.org/4094010
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5745
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Mon, 1 Nov 2010 08:55:40 +0000 (08:55 +0000)]
Landing for Rodolph Perfetta.
Reduces the number of movw/movt instructions generated in ProbeTable. It
improves code size for CompileCallMegamorphic by about 10%.
BUG=none
TEST=none
Review URL: http://codereview.chromium.org/4220004/show
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5744
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vegorov@chromium.org [Fri, 29 Oct 2010 08:13:19 +0000 (08:13 +0000)]
Clear JS function result caches in all global contexts.
Original patch by Mark Lam <mark.lam@palm.com> from Hewlett-Packard Development Company, LP. (http://codereview.chromium.org/4187007)
Fix memory corruption in JSFunctionResultCache::Clear caused by out of bounds writes which was revealed by the patch.
Review URL: http://codereview.chromium.org/4200009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5738
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Fri, 29 Oct 2010 07:10:41 +0000 (07:10 +0000)]
Fix some memory leaks in the serialization tests.
Review URL: http://codereview.chromium.org/4095009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5736
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Fri, 29 Oct 2010 07:10:14 +0000 (07:10 +0000)]
Stop running some liveedit tests (see issue 915).
Plug tiny memory leaks in test-lock.cc.
Review URL: http://codereview.chromium.org/4161005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5735
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Fri, 29 Oct 2010 06:15:34 +0000 (06:15 +0000)]
Landing for Justin Schuh.
This switches out the existing constant splitting with masking that works like this:
1. Generate a random 32-bit value at compilation time.
2. XOR the 32-bit constant with the random value.
3. Emit the resulting immediate value along with the XOR operation to generate the original value.
BUG=http://code.google.com/p/v8/issues/detail?id=908
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5734
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vitalyr@chromium.org [Thu, 28 Oct 2010 16:48:10 +0000 (16:48 +0000)]
ll_prof: minor fixes to the issues found while running with chromium.
Review URL: http://codereview.chromium.org/4097011
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5727
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
sgjesse@chromium.org [Thu, 28 Oct 2010 11:50:50 +0000 (11:50 +0000)]
Remove the patching of the debug break stub
This is no longer required to break at the function entry. Actually has not been required since the inlined stack check was introduced long time ago.
Review URL: http://codereview.chromium.org/4106005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5726
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Thu, 28 Oct 2010 08:43:47 +0000 (08:43 +0000)]
Add gc() calls to some tests to please valgrind.
Review URL: http://codereview.chromium.org/4116005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5724
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Thu, 28 Oct 2010 07:35:07 +0000 (07:35 +0000)]
Landing for Rodolph Perfetta.
Fix the ARM simulator, the ARM disassembler and extend the stop feature.
The stop feature in the simulator now support enabling, disabling and
counting.
BUG=None
TEST=None
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5723
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Thu, 28 Oct 2010 07:33:51 +0000 (07:33 +0000)]
Fix the last USE_SIMULATOR mips issue introduced with my last arm
simulator change.
Review URL: http://codereview.chromium.org/4151005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5722
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Wed, 27 Oct 2010 17:03:01 +0000 (17:03 +0000)]
Fix reporting of impossible nested calls of DOM functions.
It turns out that having PC pointing _exactly_ at callback entry
point is quite probable.
BUG=crbug/60753
Review URL: http://codereview.chromium.org/4169005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5721
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vitalyr@chromium.org [Wed, 27 Oct 2010 12:45:46 +0000 (12:45 +0000)]
Profiler on linux: use SA_RESTART signal flag to have most of the syscalls restarted for us.
This change keeps the interactive shell alive.
Review URL: http://codereview.chromium.org/4197002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5720
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Wed, 27 Oct 2010 12:33:48 +0000 (12:33 +0000)]
Make Parser class have no friends and fewer things to depend on it.
Review URL: http://codereview.chromium.org/4146006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5719
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
karlklose@chromium.org [Wed, 27 Oct 2010 11:37:59 +0000 (11:37 +0000)]
Fix a bug that prevents constants from overwriting function values in object literals.
BUG=http://code.google.com/p/v8/issues/detail?id=907
Review URL: http://codereview.chromium.org/4004006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5718
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Wed, 27 Oct 2010 09:19:43 +0000 (09:19 +0000)]
Separate JSON parsing from the JavaScript parser.
Switch JSON parsing to creating the value directly instead of createing
code to create the value.
Review URL: http://codereview.chromium.org/4135004
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5715
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Wed, 27 Oct 2010 08:09:34 +0000 (08:09 +0000)]
Prepare push to trunk. Now working on version 2.5.3.
Review URL: http://codereview.chromium.org/4128008
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5714
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vitalyr@chromium.org [Tue, 26 Oct 2010 18:14:45 +0000 (18:14 +0000)]
Faster ascii string case conversion.
Review URL: http://codereview.chromium.org/4189001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5713
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Tue, 26 Oct 2010 14:51:31 +0000 (14:51 +0000)]
Improve sampler resolution on Linux.
Instead of relying on itimer signals from kernel, send them
ourselves from a separate thread. This disables an ability
to profile multiple VM threads on Linux, but it anyway doesn't
work on other platforms, so we need a common solution for
it (issue 913 created to track this).
Review URL: http://codereview.chromium.org/4000007
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5711
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Tue, 26 Oct 2010 14:13:21 +0000 (14:13 +0000)]
Don't create empty vectors using SubVector.
Fixes build-breakage.
TBR: whesse@chromium.org
Review URL: http://codereview.chromium.org/4168002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5710
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Tue, 26 Oct 2010 13:30:31 +0000 (13:30 +0000)]
Use Vector::SubVector instead of using the constructor.
Review URL: http://codereview.chromium.org/4148003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5708
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Tue, 26 Oct 2010 08:12:17 +0000 (08:12 +0000)]
Allow forcing the use of a simulator from the build script
independently of the host architecture.
Fix build issue for the mips simulator. The mips simulator does not
currently work, but at least this change does not make it worse.
Review URL: http://codereview.chromium.org/4090003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5701
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Mon, 25 Oct 2010 16:40:41 +0000 (16:40 +0000)]
Clean up definitions in simulator.
Clean up the definition of USE_SIMULATOR and its uses.
Review URL: http://codereview.chromium.org/4044003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5700
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Mon, 25 Oct 2010 15:29:06 +0000 (15:29 +0000)]
Last minute typos from the last change.
TBR=vegorov
Review URL: http://codereview.chromium.org/4037006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5699
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Mon, 25 Oct 2010 15:22:03 +0000 (15:22 +0000)]
Make Failure inherit from MaybeObject instead of Object.
Review URL: http://codereview.chromium.org/3970005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5698
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Mon, 25 Oct 2010 15:04:08 +0000 (15:04 +0000)]
Make windows-tick-processor respect D8_PATH.
Review URL: http://codereview.chromium.org/4078002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5697
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Mon, 25 Oct 2010 09:53:39 +0000 (09:53 +0000)]
A little piece of code for aligning heap snapshots vs. object inspection.
Use original names for function infos.
Review URL: http://codereview.chromium.org/4040008
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5696
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Mon, 25 Oct 2010 08:26:12 +0000 (08:26 +0000)]
Use size_t for return value from fwrite.
TBR=sgjesse@chromium.org
Review URL: http://codereview.chromium.org/4033005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5695
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Mon, 25 Oct 2010 08:22:23 +0000 (08:22 +0000)]
Use return value from fwrite in log.cc to please compilers.
BUG=453
Review URL: http://codereview.chromium.org/4002005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5694
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Sun, 24 Oct 2010 16:09:44 +0000 (16:09 +0000)]
Fix FreeBSD port.
TBR=kasperl@chromium.org
BUG=912
Review URL: http://codereview.chromium.org/4084001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5693
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
serya@chromium.org [Thu, 21 Oct 2010 14:53:19 +0000 (14:53 +0000)]
Fixing build failure - added a missed file.
Review URL: http://codereview.chromium.org/3965004
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5690
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
serya@chromium.org [Thu, 21 Oct 2010 14:21:00 +0000 (14:21 +0000)]
Optimizing HandleScope. Also fixed HandleScope destruction when API getter throws an exception.
Review URL: http://codereview.chromium.org/3792003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5689
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
fschneider@chromium.org [Thu, 21 Oct 2010 14:19:07 +0000 (14:19 +0000)]
Clean up the log-stack-tracer test.
Create the functions using the normal API and remove some
unnecessary helpers.
Review URL: http://codereview.chromium.org/4024003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5688
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Thu, 21 Oct 2010 13:15:12 +0000 (13:15 +0000)]
Implement --noinline-new flag fully on x64 and ia32 platforms.
Review URL: http://codereview.chromium.org/3961005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5687
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Thu, 21 Oct 2010 11:54:32 +0000 (11:54 +0000)]
Strtod fast-case that uses DiyFps and cached powers of ten.
This is a fixed version of r5677.
Review URL: http://codereview.chromium.org/3898007
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5686
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Thu, 21 Oct 2010 06:16:52 +0000 (06:16 +0000)]
Revert r5455 from bleeding_edge: attempt to collect more garbage
before panicking with out of memory.
TBR=antonm@chromium.org
Review URL: http://codereview.chromium.org/4034002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5681
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Wed, 20 Oct 2010 15:30:59 +0000 (15:30 +0000)]
Fix build-breakage.
Revert "Strtod fast-case that uses DiyFps and cached powers of ten."
This reverts commit
493da023514021a63e1d3ba3f70348a275ac4042.
TBR: whesse@chromium.org
Review URL: http://codereview.chromium.org/3870003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5678
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Wed, 20 Oct 2010 15:13:34 +0000 (15:13 +0000)]
Strtod fast-case that uses DiyFps and cached powers of ten.
Review URL: http://codereview.chromium.org/3760013
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5677
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Wed, 20 Oct 2010 13:19:03 +0000 (13:19 +0000)]
Fix GC error in ES5 read-only properties implementation.
Review URL: http://codereview.chromium.org/3920005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5676
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Wed, 20 Oct 2010 12:01:17 +0000 (12:01 +0000)]
Fix the --noinline-new flag on ARM so that it forces us into C++ code
on every allocation. Fix three places where the generated code couldn't
cope with an unlucky GC.
Review URL: http://codereview.chromium.org/3872003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5674
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Wed, 20 Oct 2010 11:54:20 +0000 (11:54 +0000)]
Prepare push to trunk. Now working on version 2.5.2. Fixed past tense in ChangeLog.
Review URL: http://codereview.chromium.org/3913007
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5673
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Wed, 20 Oct 2010 11:36:42 +0000 (11:36 +0000)]
Prepare push to trunk. Now working on version 2.5.2.
Review URL: http://codereview.chromium.org/3898006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5672
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Wed, 20 Oct 2010 08:52:46 +0000 (08:52 +0000)]
Update x64 RecordWrite comment. When the offset is zero the index
should be an untagged 32-bit integer on x64. There are debug mode
asserts in the code that catch tagged integers so we get test
failures if we get it wrong.
Review URL: http://codereview.chromium.org/3928002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5668
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Wed, 20 Oct 2010 08:32:24 +0000 (08:32 +0000)]
CPU Profiler: postpone moved functions registration until GC completes.
An attempt to retrieve security context for a function may fail if the
destination heap space is in an incomplete state. To fix this, we only
record unknown functions discovered at GC object moves, and then
register them after GC completes.
BUG=crbug/59627
Review URL: http://codereview.chromium.org/3763012
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5667
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Wed, 20 Oct 2010 06:47:49 +0000 (06:47 +0000)]
Remove vm-state.cc from build files. Forgot to do that when
removing the file from the repository.
Review URL: http://codereview.chromium.org/3950001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5666
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Wed, 20 Oct 2010 05:54:23 +0000 (05:54 +0000)]
Make state stack thread local. When using Lockers the state stacks of
multiple threads got mixed up so that the current state could be an
already deallocated state from another thread.
Review URL: http://codereview.chromium.org/3828016
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5665
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vitalyr@chromium.org [Tue, 19 Oct 2010 16:56:44 +0000 (16:56 +0000)]
Try to fix win64 build.
TBR=mikhail.naganov@gmail.com
Review URL: http://codereview.chromium.org/3797016
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5664
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vitalyr@chromium.org [Tue, 19 Oct 2010 16:45:11 +0000 (16:45 +0000)]
Support profiling based on linux kernel performance events.
Since 2.6.31 perf_events interface has been available in the
kernel. There's a nice tool called "perf" (linux-2.6/tools/perf) that
uses this interface and provides capabilities similar to oprofile. The
simplest form of its usage is just dumping the raw log (trace) of
events generated by the kernel. In this patch I'm adding a script
(tools/ll_prof.py) to build profiles based on perf trace and our code
log. All the heavy-lifting is done by perf. Compared to oprofile agent
this approach does not require recompilation and supports code moving
garbage collections.
Expected usage is documented in the ll_prof's help. Basically one
should run V8 under perf passing --ll-prof flag and then the produced
logs can be analyzed by tools/ll_prof.py.
The new --ll-prof flag enables logging of generated code object
locations and names (like --log-code), and also of their bodies, which
can be later disassembled and annotated by the script.
Review URL: http://codereview.chromium.org/3831002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5663
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Tue, 19 Oct 2010 14:00:01 +0000 (14:00 +0000)]
Limit end-anchored regexps to testing end of string where possible.
Review URL: http://codereview.chromium.org/3844006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5661
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
kmillikin@chromium.org [Tue, 19 Oct 2010 11:14:03 +0000 (11:14 +0000)]
Change the StackGuard runtime function to take no arguments.
For some reason the StackGuard runtime function took a dummy argument
that it always ignored. Change it to take no arguments.
Review URL: http://codereview.chromium.org/3838003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5660
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Tue, 19 Oct 2010 10:37:54 +0000 (10:37 +0000)]
Revert revision 5657.
TBR: erik.corry
Review URL: http://codereview.chromium.org/3812012
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5658
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Tue, 19 Oct 2010 09:42:40 +0000 (09:42 +0000)]
Use finite-length end-anchored regexps to reduce part of regexp that is searched.
Review URL: http://codereview.chromium.org/3850005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5657
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Tue, 19 Oct 2010 09:16:57 +0000 (09:16 +0000)]
Return false from PagedSpace::Contains if the page from an address is not valid.
This can happen on Mac where C++ code can be in the 4-8K range.
Review URL: http://codereview.chromium.org/3781020
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5656
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
sgjesse@chromium.org [Tue, 19 Oct 2010 08:23:47 +0000 (08:23 +0000)]
Make randomized allocations along 64k granularity boundaries to avoid comitting unused memory.
BUG=56036
TEST=None.
Patch by Justin Schuh <jschuh@chromium.org>
Review URL: http://codereview.chromium.org/3849004
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5655
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vegorov@chromium.org [Tue, 19 Oct 2010 08:14:41 +0000 (08:14 +0000)]
Clear normalized map caches in all global contexts not just in those reachable from ThreadLocalTops.
Issue reported by Mark Lam <mark.lam@palm.com> from Hewlett-Packard Development Company, LP.
Review URL: http://codereview.chromium.org/3828011
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5654
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Mon, 18 Oct 2010 17:34:27 +0000 (17:34 +0000)]
Update sync profiling changes for Solaris platform.
Patch by Maciej Blizinski <blizinski@google.com>
TBR=kasperl@chromium.org
BUG=901
Review URL: http://codereview.chromium.org/3794008
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5653
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vegorov@chromium.org [Mon, 18 Oct 2010 17:01:30 +0000 (17:01 +0000)]
Fix TestInternalWeakLists test
Releasing context handle is not enough to release context, it might be indirectly reachable through compilation cache.
TBR=sgjesse@chromium.org
Review URL: http://codereview.chromium.org/3750017
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5652
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Mon, 18 Oct 2010 16:37:19 +0000 (16:37 +0000)]
Disable double strtod in ARM-simulator.
The ARM simulator is compiled for 32bits. On Linux the floating-point
operations are hence invalid for the simulator too.
Fixes build-breakage.
Review URL: http://codereview.chromium.org/3813009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5651
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Mon, 18 Oct 2010 15:19:39 +0000 (15:19 +0000)]
Fix double-rounding in strtod.
Don't use floating-point operations on Linux,x86 to compute strtod. Since the
floating-point stack on Linux is set to 80bit double rounding may occure.
When falling back to gay_strtod append several '0's so that Gay doesn't take
the same shortcut either.
BUG=
TEST=
Review URL: http://codereview.chromium.org/3851003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5650
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vegorov@chromium.org [Mon, 18 Oct 2010 14:59:03 +0000 (14:59 +0000)]
Link all global contexts into a weak list.
Review URL: http://codereview.chromium.org/3764011
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5649
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vitalyr@chromium.org [Mon, 18 Oct 2010 14:39:59 +0000 (14:39 +0000)]
SConstruct: make help work even if guesses failed.
Review URL: http://codereview.chromium.org/3811010
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5648
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Mon, 18 Oct 2010 14:04:28 +0000 (14:04 +0000)]
Fix bug in comparison of two smis that differ by MIN_SMI on full compiler on all platforms.
Review URL: http://codereview.chromium.org/3767016
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5643
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
antonm@chromium.org [Mon, 18 Oct 2010 12:58:56 +0000 (12:58 +0000)]
Get rid of requested size in RetryAfterGC.
This was only used by Heap::CollectGarbage to check if after GC
available space is bigger than was requested, but nobody checked
the value returned by Heap::CollectGarbage, so requested size
was efficiently unused.
However, it may trigger spurious out of memory exceptions if requested
size is big enough.
BUG=http://code.google.com/p/chromium/issues/detail?id=54580
Review URL: http://codereview.chromium.org/3836001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5642
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
kasperl@chromium.org [Mon, 18 Oct 2010 12:37:07 +0000 (12:37 +0000)]
Try to simplify the semantics of the profiling code by making
sure to suspend the thread (if necessary) on mac/win32 before
reading the VM state.
Avoid dealing with signals delivered to non-VM threads on
linux no matter if we're profiling or not.
Review URL: http://codereview.chromium.org/3845006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5641
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Mon, 18 Oct 2010 12:16:38 +0000 (12:16 +0000)]
Fix missing check for GC failure in number dictionary code.
Review URL: http://codereview.chromium.org/3781014
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5640
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
vegorov@chromium.org [Mon, 18 Oct 2010 11:06:51 +0000 (11:06 +0000)]
Prepare push to trunk. Now working on version 2.5.1.
Review URL: http://codereview.chromium.org/3748013
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5637
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Mon, 18 Oct 2010 10:23:45 +0000 (10:23 +0000)]
Fix compilation error on ARM with gcc 4.4. Remove NULL check of pointer to member, which was dead code (never failed).
Review URL: http://codereview.chromium.org/3793011
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5636
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Mon, 18 Oct 2010 09:15:38 +0000 (09:15 +0000)]
New Heap profiler: add dumping HeapNumbers and InternalFields to snapshot.
HeapNumbers do consume memory, so it's worth dumping them. However, we
don't dump their values, as they are not as self-descriptive as values
of strings, and they will increase snapshot size. Storing heap numbers
values can be added if we will feel a sufficient demand for that.
InternalFields are used, e.g. for storing references to DOM nodes
event handlers.
Review URL: http://codereview.chromium.org/3769007
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5635
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Fri, 15 Oct 2010 14:50:49 +0000 (14:50 +0000)]
Do not return the constant singleton float -0.0 from UnaryOpStub, because it may be overwritten, on X64 platform.
Review URL: http://codereview.chromium.org/3748007
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5631
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ricow@chromium.org [Fri, 15 Oct 2010 13:03:59 +0000 (13:03 +0000)]
Remove trailing spaces from js files.
This is a cosmetic change only.
Review URL: http://codereview.chromium.org/3797009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5630
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Fri, 15 Oct 2010 11:45:05 +0000 (11:45 +0000)]
Add stop-at flag implementation to full compiler on all platforms.
Review URL: http://codereview.chromium.org/3791009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5629
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Fri, 15 Oct 2010 10:45:18 +0000 (10:45 +0000)]
Fix new-space-too-small test crash on 64 bit VM.
Review URL: http://codereview.chromium.org/3778009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5628
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
erik.corry@gmail.com [Fri, 15 Oct 2010 07:54:20 +0000 (07:54 +0000)]
Fix creation of an exception to avoid rare GC corner case.
Review URL: http://codereview.chromium.org/3782009
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5627
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Thu, 14 Oct 2010 18:54:53 +0000 (18:54 +0000)]
Fix the version number in the ChangeLog.
TBR=kmillikin@chromium.org
Review URL: http://codereview.chromium.org/3757004
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5625
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Thu, 14 Oct 2010 15:54:51 +0000 (15:54 +0000)]
Work around Windows bug. Use different constants.
Windows' strtod doesn't correctly read 3e-324 a the lowest denormal, but returns 0.0 instead.
Using 4e-324 is still the same value and works.
BUG=
TEST=
Review URL: http://codereview.chromium.org/3744008
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5624
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Thu, 14 Oct 2010 13:47:48 +0000 (13:47 +0000)]
Weed out extreme exponents in strtod.
If a decimal exponent is less than -309 return 0.0.
If a decimal exponent is greater than +324 return +infinity.
BUG=
TEST=
Review URL: http://codereview.chromium.org/3519017
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5623
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Thu, 14 Oct 2010 12:54:00 +0000 (12:54 +0000)]
Restructure RegExp exec cache code.
Review URL: http://codereview.chromium.org/3778004
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5622
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
sgjesse@chromium.org [Thu, 14 Oct 2010 11:46:55 +0000 (11:46 +0000)]
Remove line added by accident.
TBR=ager@chromium.org
Review URL: http://codereview.chromium.org/3761002
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5621
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
sgjesse@chromium.org [Thu, 14 Oct 2010 11:39:48 +0000 (11:39 +0000)]
Added USE_SIMULATOR macro that explicitly indicates that we wish to use the simulator as the execution engine. For example, this allows us to run with the ARM simulator on ARM.
Patch by Mark Lam <mark.lam@palm.com> from Hewlett-Packard Development Company, LP
Review URL: http://codereview.chromium.org/3825001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5620
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
lrn@chromium.org [Thu, 14 Oct 2010 08:51:20 +0000 (08:51 +0000)]
Fix bug in cache handling of lastIndex on global regexps.
Review URL: http://codereview.chromium.org/3745005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5615
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
mikhail.naganov@gmail.com [Wed, 13 Oct 2010 14:57:00 +0000 (14:57 +0000)]
Better align heap snapshots contents with debugger info.
I created a heap snapshot in Chromium, and then started comparing
it side-by-side with representations of objects provided by
the debugger, fixing discrepancies.
Review URL: http://codereview.chromium.org/3590029
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5612
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Wed, 13 Oct 2010 08:12:58 +0000 (08:12 +0000)]
Prepare push to trunk. Now working on version 2.5.0.
Review URL: http://codereview.chromium.org/3776001
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5609
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
kmillikin@chromium.org [Fri, 8 Oct 2010 13:07:55 +0000 (13:07 +0000)]
Fix a bug in our handling of conditional expressions in test contexts.
In the FullCodeGenerator, we compile the true subexpression of a
conditional (ternary) expression in the inherited context of the
entire expression. This is correct for effect and value contexts, but
not for test contexts where the context includes a possible
fall-through label.
Review URL: http://codereview.chromium.org/3621013
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5607
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Thu, 7 Oct 2010 13:02:36 +0000 (13:02 +0000)]
Simplify powers-of-ten cache.
The existing cache was build to support changes to fast-dtoa. This complicated the cache. Since the cached numbers are now in a .cc file they can be shared with other parts of v8.
Review URL: http://codereview.chromium.org/3608011
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5606
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Thu, 7 Oct 2010 12:53:59 +0000 (12:53 +0000)]
Implement fast case for strtod.
Reapply r5603 with additional fix: use OS::StrNCpy instead of posix strncpy.
BUG=
TEST=
Review URL: http://codereview.chromium.org/3557010
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5605
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
ager@chromium.org [Thu, 7 Oct 2010 12:24:02 +0000 (12:24 +0000)]
Revert strtod fast-case change (r5603) because of Windows compilation
errors.
TBR=floitschV8@gmail.com
Review URL: http://codereview.chromium.org/3582017
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5604
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Thu, 7 Oct 2010 12:01:25 +0000 (12:01 +0000)]
Implement fast case for strtod.
If there are few non-zero digits and the 10^exponent fits into a double then
we can compute the result using 1 (or 2) double operations.
BUG=
TEST=
Review URL: http://codereview.chromium.org/3584015
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5603
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
whesse@chromium.org [Thu, 7 Oct 2010 09:40:26 +0000 (09:40 +0000)]
Fix new strtod on Windows platform, using OS::SNPrintF instead of snprintf. Fixes error in r5600.
Review URL: http://codereview.chromium.org/3601018
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5602
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00
floitschV8@gmail.com [Thu, 7 Oct 2010 09:12:21 +0000 (09:12 +0000)]
Fix build-breakage. Add new line to end of strtod.h. TBR: whesse@chromium.org
BUG=
TEST=
Review URL: http://codereview.chromium.org/3584018
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@5601
ce2b1a6d-e550-0410-aec6-
3dcde31c8c00