platform/upstream/v8.git
14 years agoAll hidden properties of an object are stored in a value of a regular property with...
yurys@chromium.org [Tue, 10 Nov 2009 16:13:21 +0000 (16:13 +0000)]
All hidden properties of an object are stored in a value of a regular property with empty name. This property may confuse user if returned among regular properties. It should not be exposed directly by ObjectMirror. Should we want an access to these properties from debugger we need to implement an explicit method for that.

Current patch filters the hidden_symbol from property names returned to ObjectMirror.

See http://crbug.com/26491

Review URL: http://codereview.chromium.org/390001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3265 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoTest mjsunit/string-indexof times out on ARM simulator
sgjesse@chromium.org [Tue, 10 Nov 2009 15:50:23 +0000 (15:50 +0000)]
Test mjsunit/string-indexof times out on ARM simulator

Removal of string slices made this test time out on the ARM simulator. Temporaly mark this as pass or timeout.

BUG=http://code.google.com/p/v8/issues/detail?id=500
TBR=christian.plesner.hansen@gmail.com
Review URL: http://codereview.chromium.org/390002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3264 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRevert r3032 that uses push instead of 'sub esp, size'. This change
ager@chromium.org [Tue, 10 Nov 2009 14:58:16 +0000 (14:58 +0000)]
Revert r3032 that uses push instead of 'sub esp, size'.  This change
leads to stack corruption in 32-bit version of V8.

See http://code.google.com/p/chromium/issues/detail?id=27227 for a
reproducible case.

Since this is only an issue on 32-bit V8 I think this has got
something to do with the UnsafeSmi handling that we do on ia32.  I'm
reverting for now so we can push a fix, but we should track down the
issue and create a regression test for this.

Review URL: http://codereview.chromium.org/383005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3263 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoModify test using string slices
sgjesse@chromium.org [Tue, 10 Nov 2009 14:41:53 +0000 (14:41 +0000)]
Modify test using string slices

With slices string string type removed generating a large amount of sub-strings takes more time. Change a test to avoid timeout in debug mode.
Review URL: http://codereview.chromium.org/385006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3260 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRemove sliced string string type
sgjesse@chromium.org [Tue, 10 Nov 2009 13:23:05 +0000 (13:23 +0000)]
Remove sliced string string type

As a first step to reduce the complexity of the string hierachy the sliced string type is removed. Whenever a sub-string is created it is allocated as a fresh flat string.
Review URL: http://codereview.chromium.org/385004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3259 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoMarked cyrillic as expected to fail.
christian.plesner.hansen@gmail.com [Tue, 10 Nov 2009 12:19:36 +0000 (12:19 +0000)]
Marked cyrillic as expected to fail.

Review URL: http://codereview.chromium.org/388002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3258 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoReapply 3246 and 3247 after fixing valgrind warning.
lrn@chromium.org [Tue, 10 Nov 2009 10:23:23 +0000 (10:23 +0000)]
Reapply 3246 and 3247 after fixing valgrind warning.

Review URL: http://codereview.chromium.org/386001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3257 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoReport both size and count from the heap profile processor
sgjesse@chromium.org [Tue, 10 Nov 2009 10:18:41 +0000 (10:18 +0000)]
Report both size and count from the heap profile processor

Added two options for controlling this --size and --count. Default is --size to match the original behaviour.
Review URL: http://codereview.chromium.org/385001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3256 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFixed bug in while-loops that caused an assertion to fail.
fschneider@chromium.org [Tue, 10 Nov 2009 09:57:13 +0000 (09:57 +0000)]
Fixed bug in while-loops that caused an assertion to fail.

We forgot resetting true-/false-label to NULL after evaluating the
condition expression in dowhile- and while-loops.
This change fixes this.

This causes an assertion to fail in VisitIfStatement whenever there is an
if-statement after a while-loop before. e.g. like in:

  var i=0, j=0;
  while(j<5) { j++; }
  if (i ==0 ) { j++; }

Review URL: http://codereview.chromium.org/371070

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3255 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRevert 3245 and 3246 because they cause valgrind failures.
erik.corry@gmail.com [Mon, 9 Nov 2009 17:05:51 +0000 (17:05 +0000)]
Revert 3245 and 3246 because they cause valgrind failures.
TBR=lrn
Review URL: http://codereview.chromium.org/372059

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3254 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoTemporarily de-activate while-loops in the top-level compiler because
fschneider@chromium.org [Mon, 9 Nov 2009 15:31:15 +0000 (15:31 +0000)]
Temporarily de-activate while-loops in the top-level compiler because
it makes some debug tests fail.

TBR=christian.plesner.hansen@gmail.com
Review URL: http://codereview.chromium.org/371069

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3253 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix small memory leak in new serialization code.
erik.corry@gmail.com [Mon, 9 Nov 2009 13:30:50 +0000 (13:30 +0000)]
Fix small memory leak in new serialization code.
Review URL: http://codereview.chromium.org/371068

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3252 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRevert change r3240-3241 because it causes ARM tests to fail.
fschneider@chromium.org [Mon, 9 Nov 2009 13:30:35 +0000 (13:30 +0000)]
Revert change r3240-3241 because it causes ARM tests to fail.

(non-lazily compiled function literals)

Review URL: http://codereview.chromium.org/376021

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3251 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFast-compiler: Added trivial implementations of while and do/while.
lrn@chromium.org [Mon, 9 Nov 2009 13:22:38 +0000 (13:22 +0000)]
Fast-compiler: Added trivial implementations of while and do/while.

Review URL: http://codereview.chromium.org/372055

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3250 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agohttp://codereview.chromium.org/371067 thanks to Jan de Mooij
erik.corry@gmail.com [Mon, 9 Nov 2009 13:21:14 +0000 (13:21 +0000)]
codereview.chromium.org/371067 thanks to Jan de Mooij

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3249 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoInline common case of one capture when using replace with a regexp and
ager@chromium.org [Mon, 9 Nov 2009 13:17:50 +0000 (13:17 +0000)]
Inline common case of one capture when using replace with a regexp and
a function.

Review URL: http://codereview.chromium.org/371065

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3248 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix lint issue.
lrn@chromium.org [Mon, 9 Nov 2009 12:18:25 +0000 (12:18 +0000)]
Fix lint issue.

Review URL: http://codereview.chromium.org/379005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3247 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRemove unnecessary buffer doubling and content copying.
lrn@chromium.org [Mon, 9 Nov 2009 11:52:18 +0000 (11:52 +0000)]
Remove unnecessary buffer doubling and content copying.

Review URL: http://codereview.chromium.org/377006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3246 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix SVN merge problem in the top-level compiler.
fschneider@chromium.org [Mon, 9 Nov 2009 10:30:33 +0000 (10:30 +0000)]
Fix SVN merge problem in the top-level compiler.

(Remove bailout on non-global declarations again)

TBR=whesse@chromium.org
Review URL: http://codereview.chromium.org/372054

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3245 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFixed breakage of test running when sputniktests not checked out.
christian.plesner.hansen@gmail.com [Mon, 9 Nov 2009 10:23:50 +0000 (10:23 +0000)]
Fixed breakage of test running when sputniktests not checked out.

Review URL: http://codereview.chromium.org/381002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3244 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years ago* Fix regexp benchmark regression where we were doing work to
erik.corry@gmail.com [Mon, 9 Nov 2009 10:01:23 +0000 (10:01 +0000)]
* Fix regexp benchmark regression where we were doing work to
make standard regexps like \s and . case independent.
* Make use of the fact that the subject string is ASCII only
when making character classes case independent.
* Avoid spending time making large ideogram or punctuation
ranges case independent when there is no case mapping anyway.
Review URL: http://codereview.chromium.org/378024

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3243 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoEnable writes and reads of context slots in fast compiler.
whesse@chromium.org [Mon, 9 Nov 2009 09:56:57 +0000 (09:56 +0000)]
Enable writes and reads of context slots in fast compiler.
Review URL: http://codereview.chromium.org/360054

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3242 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd function literals that always require non-lazy compilation to the top-level compiler.
fschneider@chromium.org [Mon, 9 Nov 2009 09:43:24 +0000 (09:43 +0000)]
Add function literals that always require non-lazy compilation to the top-level compiler.

Review URL: http://codereview.chromium.org/371064

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3241 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdded test runner for sputnik tests.
christian.plesner.hansen@gmail.com [Mon, 9 Nov 2009 09:31:57 +0000 (09:31 +0000)]
Added test runner for sputnik tests.

Review URL: http://codereview.chromium.org/372053

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3240 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix missing () on function call causing warning on some versions
erik.corry@gmail.com [Fri, 6 Nov 2009 14:06:35 +0000 (14:06 +0000)]
Fix missing () on function call causing warning on some versions
of gcc.
Review URL: http://codereview.chromium.org/371020

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3239 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoKeep natives source code in external strings instead of putting
erik.corry@gmail.com [Fri, 6 Nov 2009 13:48:33 +0000 (13:48 +0000)]
Keep natives source code in external strings instead of putting
it in regular flat strings that are part of the snapshot.
After this change we don't need libraries-empty.cc any more.  In
this change libraries-empty.cc is just a the same as libraries.cc
and the scons build builds it but does not use it.  We can move
in stages to a situation where it is not generated at all for all
the build systems that we have.
Review URL: http://codereview.chromium.org/360050

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3238 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFixed problem where popping a handle scope after calling an accessor
christian.plesner.hansen@gmail.com [Fri, 6 Nov 2009 11:35:47 +0000 (11:35 +0000)]
Fixed problem where popping a handle scope after calling an accessor
would clobber the register holding the result.

Review URL: http://codereview.chromium.org/377004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3237 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix bug 486, Cyrillic character ranges in case independent regexps.
erik.corry@gmail.com [Fri, 6 Nov 2009 11:15:20 +0000 (11:15 +0000)]
Fix bug 486, Cyrillic character ranges in case independent regexps.
http://code.google.com/p/v8/issues/detail?id=486
Review URL: http://codereview.chromium.org/361033

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3236 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoEnable mksnapshot on ARM when cross-compiling using gyp
sgjesse@chromium.org [Fri, 6 Nov 2009 07:50:31 +0000 (07:50 +0000)]
Enable mksnapshot on ARM when cross-compiling using gyp

Review URL: http://codereview.chromium.org/351019

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3235 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd support for all declarations in the top-level compiler:
fschneider@chromium.org [Thu, 5 Nov 2009 17:33:50 +0000 (17:33 +0000)]
Add support for all declarations in the top-level compiler:

Until now we only handled global declarations. This change
adds declarations of local variables, consts and functions.

Review URL: http://codereview.chromium.org/368005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3234 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRemove an unused enumeration value.
kmillikin@chromium.org [Thu, 5 Nov 2009 17:25:50 +0000 (17:25 +0000)]
Remove an unused enumeration value.

Review URL: http://codereview.chromium.org/366030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3233 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix a bad commit due to an svn merge error.
kmillikin@chromium.org [Thu, 5 Nov 2009 16:21:27 +0000 (16:21 +0000)]
Fix a bad commit due to an svn merge error.

Review URL: http://codereview.chromium.org/368007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3232 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoChange the special handling of typeof comparisons to strings for
kmillikin@chromium.org [Thu, 5 Nov 2009 16:08:48 +0000 (16:08 +0000)]
Change the special handling of typeof comparisons to strings for
regular expressions to be consistent with typeof in other contexts.
Typeof regular expressions should now be 'function' in all contexts.

In the JS natives, IS_FUNCTION, IS_OBJECT, and IS_REGEXP return the
same answers as before.

Review URL: http://codereview.chromium.org/360053

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3231 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoMade iteration of global handles more efficient on scavenges.
christian.plesner.hansen@gmail.com [Thu, 5 Nov 2009 15:12:36 +0000 (15:12 +0000)]
Made iteration of global handles more efficient on scavenges.

Review URL: http://codereview.chromium.org/355041

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3230 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix issue 493: Infinite loop when debug break is set when entering function.apply.
sgjesse@chromium.org [Thu, 5 Nov 2009 13:59:40 +0000 (13:59 +0000)]
Fix issue 493: Infinite loop when debug break is set when entering function.apply.

In the generated code for function.apply there was a loop checking the stack limit for interruption. This loop would call into the runtime system to handle interuption and keep running until there was no interruption. However if the interuption was debug break the runtime system would never clear the interruption as debug break is prevented in builtins are prevented and the assumption here was that returning with the debug break flag set would move execution forward.

Renamed initial_jslimit and initial_climit to real_jslimit and real_climit. Renamed a few external references related to the stack limit as well.

Exposed the real stack limit to generated code to make the stack check when entering function.apply use the real stack limit and not the stack limit which is changed to signal interruption.

Added the real stack limit to the roots array.

BUG=http://code.google.com/p/v8/issues/detail?id=493
TEST=cctest/test-debug/DebugBreakFunctionApply
Review URL: http://codereview.chromium.org/345048

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3229 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRework the way we handle the fact that the ARM simulator uses a
ager@chromium.org [Thu, 5 Nov 2009 13:27:21 +0000 (13:27 +0000)]
Rework the way we handle the fact that the ARM simulator uses a
separate JS stack.

In exception handling, we need to be able to compare addresses into
the JavaScript portion of the stack with the address of a C++ handler
on the stack.  Since the stacks are separate on the simulator, we need
a JavaScript stack address corresponding to a C++ try catch handler in
order to perform valid address comparisons.

On the simulator, we now link the C++ try catch handlers indirectly
through the JS stack and use the JS stack indirection address for
comparisons.

      JS                    C++

                           handler
 [C++ address]   <------    next_
                \
                 \
                  \---->   handler
 [C++ address]   <------    next_

On actual hardware the C++ try catch handlers continue to be directly
linked.

BUG=http://code.google.com/p/v8/issues/detail?id=271
Review URL: http://codereview.chromium.org/360004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3228 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoMake "typeof this" return object in fast compiler.
whesse@chromium.org [Thu, 5 Nov 2009 12:45:53 +0000 (12:45 +0000)]
Make "typeof this" return object in fast compiler.
Review URL: http://codereview.chromium.org/366028

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3227 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix case where we treat an unaliased call to eval as an aliased call
ager@chromium.org [Thu, 5 Nov 2009 11:19:37 +0000 (11:19 +0000)]
Fix case where we treat an unaliased call to eval as an aliased call
to eval.

BUG=http://code.google.com/p/v8/issues/detail?id=496
Review URL: http://codereview.chromium.org/366027

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3225 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoIn the toplevel code generator, support local context allocation
kmillikin@chromium.org [Thu, 5 Nov 2009 10:38:26 +0000 (10:38 +0000)]
In the toplevel code generator, support local context allocation
provided that none of the parameters need to be copied into the
context.

Review URL: http://codereview.chromium.org/369003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3224 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRemoved (now) unused Token-lookup code.
lrn@chromium.org [Thu, 5 Nov 2009 10:24:20 +0000 (10:24 +0000)]
Removed (now) unused Token-lookup code.

Review URL: http://codereview.chromium.org/361026

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3223 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoImplement typeof in fast compiler.
whesse@chromium.org [Thu, 5 Nov 2009 10:15:25 +0000 (10:15 +0000)]
Implement typeof in fast compiler.
Review URL: http://codereview.chromium.org/354027

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3222 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoChanged keyword token recognition to be done inline in the identifier scanner.
lrn@chromium.org [Thu, 5 Nov 2009 10:11:38 +0000 (10:11 +0000)]
Changed keyword token recognition to be done inline in the identifier scanner.

Review URL: http://codereview.chromium.org/360048

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3221 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd support for debugger statement to top-level compiler.
fschneider@chromium.org [Wed, 4 Nov 2009 19:04:29 +0000 (19:04 +0000)]
Add support for debugger statement to top-level compiler.

Review URL: http://codereview.chromium.org/366004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3220 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix a bug where if a non-lazily-compiled function is a candidate for
kmillikin@chromium.org [Wed, 4 Nov 2009 18:39:44 +0000 (18:39 +0000)]
Fix a bug where if a non-lazily-compiled function is a candidate for
toplevel compilation and the toplevel compiler cannot handle the
function, we would erroneously signal a stack overflow instead of
falling back on the optimizing compiler.

Review URL: http://codereview.chromium.org/366005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3219 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoObey the flag --lazy in the toplevel code generator.
kmillikin@chromium.org [Wed, 4 Nov 2009 17:59:24 +0000 (17:59 +0000)]
Obey the flag --lazy in the toplevel code generator.

There were two separate implementations of the function
'BuildBoilerplate' that is used to compile function declarations and
function literals.  The implementations did not do exactly the same
thing.  In particular, one ignored the flag --lazy.

Combine the two implementations.

Review URL: http://codereview.chromium.org/360011

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3218 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoSort tests to increase predictability of mozilla test runs
erik.corry@gmail.com [Wed, 4 Nov 2009 16:55:24 +0000 (16:55 +0000)]
Sort tests to increase predictability of mozilla test runs
in the face of suspected python bugs.  Also remove some
unused code.
Review URL: http://codereview.chromium.org/365001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3217 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix ARM debug build
sgjesse@chromium.org [Wed, 4 Nov 2009 15:24:11 +0000 (15:24 +0000)]
Fix ARM debug build

TBR=fschneider@chromium.org

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3216 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix issue 491: constantpool dump violates ARM debugger assertion for return point
sgjesse@chromium.org [Wed, 4 Nov 2009 14:45:50 +0000 (14:45 +0000)]
Fix issue 491: constantpool dump violates ARM debugger assertion for return point

The generation of the return sequence is now protected from having the constant pool emitted inside of it in both compilers.

BUG=http://code.google.com/p/v8/issues/detail?id=491
TEST=test/mjsunit/regress/regress-491.js
Review URL: http://codereview.chromium.org/362003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3215 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix compilation error in debug build for IA-32 and ARM.
fschneider@chromium.org [Wed, 4 Nov 2009 14:33:37 +0000 (14:33 +0000)]
Fix compilation error in debug build for IA-32 and ARM.

TBR=sgjesse@chromium.org
Review URL: http://codereview.chromium.org/361005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3214 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoEmitting the common return sequence in the top-level compiler in one function for...
fschneider@chromium.org [Wed, 4 Nov 2009 14:11:39 +0000 (14:11 +0000)]
Emitting the common return sequence in the top-level compiler in one function for each platform.

I factored out the code for emitting the return sequence since we had this code duplicated in the top-level compiler.

Review URL: http://codereview.chromium.org/354024

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3213 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRemove the typeof state threaded through the code generator. It was
kmillikin@chromium.org [Wed, 4 Nov 2009 13:56:41 +0000 (13:56 +0000)]
Remove the typeof state threaded through the code generator.  It was
used to signal that an expression was the immediate subexpression of
typeof, or (?) in the arm of a conditional expression itself in the
typeof state.

It was inconsistently consulted.  It was not used for property loads,
but only for slot loads.  This means that we matched the Webkit JSC
(not Spidermonkey) behavior for:

typeof(true ? x : y)  // throws ReferenceError

and we matched the SpiderMonkey behavior (not JSC) for:

with ({}) { typeof(true ? x : y) }  // ==> "undefined"

Now we are expected to match the JSC behavior in all cases.

Review URL: http://codereview.chromium.org/362004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3212 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix xssue 492: ARM debug crash: mozilla/ecma/FunctionObjects/15.3.1.1-3
sgjesse@chromium.org [Wed, 4 Nov 2009 10:04:22 +0000 (10:04 +0000)]
Fix xssue 492: ARM debug crash: mozilla/ecma/FunctionObjects/15.3.1.1-3

When the number of parameters times 4 (kPointerSize) to a function cannot be encoded in 12 bits the return sequence gets one more instruction. Changed the assertion to check for this case.

BUG=http://code.google.com/p/v8/issues/detail?id=492
TEST=test/mjsunit/regress/regress-492.js
Review URL: http://codereview.chromium.org/354028

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3211 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoHate. Hate. Hate.
christian.plesner.hansen@gmail.com [Wed, 4 Nov 2009 09:19:30 +0000 (09:19 +0000)]
Hate.  Hate.  Hate.

Review URL: http://codereview.chromium.org/360005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3210 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoReverting 3174. Aka reapplying 3150, 3151 and 3159. Aka api accessor
christian.plesner.hansen@gmail.com [Wed, 4 Nov 2009 08:51:48 +0000 (08:51 +0000)]
Reverting 3174.  Aka reapplying 3150, 3151 and 3159.  Aka api accessor
ics.

Review URL: http://codereview.chromium.org/341082

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3209 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years ago* Do a GC in mksnapshot to get rid of some extraneous junk.
erik.corry@gmail.com [Tue, 3 Nov 2009 21:00:43 +0000 (21:00 +0000)]
* Do a GC in mksnapshot to get rid of some extraneous junk.
* Make snapshot more compact by coding the tag and the space
in one byte.  Contract some common sequences to one byte.
* Use back references only within one page.  Index from the
start of the space otherwise.
* Serialize Smis as raw data rather than int-encoding them.
This takes a little more space but is faster.
Review URL: http://codereview.chromium.org/341079

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3208 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoInsert call to CodeForFunctionPosition in the ARM code generator.
fschneider@chromium.org [Tue, 3 Nov 2009 16:08:35 +0000 (16:08 +0000)]
Insert call to CodeForFunctionPosition in the ARM code generator.

(== SetFunctionPosition in the top-level compiler)

IA32 and X64 already do this at the beginning of CodeGenerator::gencode

Review URL: http://codereview.chromium.org/354026

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3207 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoBegin using the top-level code generator for code that is inside
kmillikin@chromium.org [Tue, 3 Nov 2009 14:48:59 +0000 (14:48 +0000)]
Begin using the top-level code generator for code that is inside
directly-applied function literals that are themselves compiled with
the top-level code generator.

The choice is guarded by a test that the function is anonymous (thus
not expected to be recursive) and not in a loop.

A compilation hint is set in the shared function info and used to make
the choice.

Review URL: http://codereview.chromium.org/341081

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3206 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoMarked mjsunit/apply as expected to fail on mac debug.
christian.plesner.hansen@gmail.com [Tue, 3 Nov 2009 12:53:30 +0000 (12:53 +0000)]
Marked mjsunit/apply as expected to fail on mac debug.

Review URL: http://codereview.chromium.org/351021

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3205 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix a latent bug in the top-level compilation of function calls.
kmillikin@chromium.org [Tue, 3 Nov 2009 12:47:19 +0000 (12:47 +0000)]
Fix a latent bug in the top-level compilation of function calls.

Calls to a non-global variable would go through the "call to a global"
path, rather than the "call to an arbitrary (other) expression" path.

Review URL: http://codereview.chromium.org/355009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3204 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoMade rethrow test not depend on stack overflow behavior.
christian.plesner.hansen@gmail.com [Tue, 3 Nov 2009 09:52:06 +0000 (09:52 +0000)]
Made rethrow test not depend on stack overflow behavior.

Review URL: http://codereview.chromium.org/345046

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3202 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdded TryCatch::ReThrow method.
christian.plesner.hansen@gmail.com [Tue, 3 Nov 2009 08:53:34 +0000 (08:53 +0000)]
Added TryCatch::ReThrow method.

Review URL: http://codereview.chromium.org/342078

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3201 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoCommiting changelist issue 348039 outside the codereview tool. Add comparisons to...
whesse@chromium.org [Mon, 2 Nov 2009 15:01:06 +0000 (15:01 +0000)]
Commiting changelist issue 348039 outside the codereview tool.  Add comparisons to fast compiler

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3200 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd support for for loops to the toplevel code generator.
kmillikin@chromium.org [Mon, 2 Nov 2009 14:12:48 +0000 (14:12 +0000)]
Add support for for loops to the toplevel code generator.

Review URL: http://codereview.chromium.org/340059

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3199 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd support for initialization block assignments in the toplevel code
kmillikin@chromium.org [Mon, 2 Nov 2009 13:30:24 +0000 (13:30 +0000)]
Add support for initialization block assignments in the toplevel code
generator, mimicing the behavior of the optimizing compiler.

Initialization blocks can only contain (thus begin and end) with a
property assignment in toplevel code.

Review URL: http://codereview.chromium.org/348038

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3198 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoDon't use string slices when processing RexExp replace (re-apply r3153)
sgjesse@chromium.org [Mon, 2 Nov 2009 12:21:43 +0000 (12:21 +0000)]
Don't use string slices when processing RexExp replace (re-apply r3153)

Re-apply r3153 with a fix for issue 490. Except for the change in line 1756 and the added test this change is identical to http://codereview.chromium.org/342015.

BUG=490
TEST=test/mjsunit/regress/regress-490.js
Review URL: http://codereview.chromium.org/341064

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3197 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoSupport for function calls on an arbitrary expression that returns
fschneider@chromium.org [Mon, 2 Nov 2009 12:04:35 +0000 (12:04 +0000)]
Support for function calls on an arbitrary expression that returns
a function in the top-level compiler.

e.g.

function f() { return (function() { return true; }) }
f()()

Review URL: http://codereview.chromium.org/346029

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3196 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRefactor the somewhat complicated code generation for assignments into
kmillikin@chromium.org [Mon, 2 Nov 2009 11:01:47 +0000 (11:01 +0000)]
Refactor the somewhat complicated code generation for assignments into
a platform-independent structure and a few platform-specific helpers
to do the heavy lifting.

Review URL: http://codereview.chromium.org/342073

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3195 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoSupport for post-fix count operations (x++, x--) where x is a global
fschneider@chromium.org [Mon, 2 Nov 2009 10:22:22 +0000 (10:22 +0000)]
Support for post-fix count operations (x++, x--) where x is a global
variable for the top-level compiler.

Review URL: http://codereview.chromium.org/342058

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3194 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd unary not operator to fast compiler.
whesse@chromium.org [Mon, 2 Nov 2009 10:17:12 +0000 (10:17 +0000)]
Add unary not operator to fast compiler.
Review URL: http://codereview.chromium.org/343057

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3193 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd conditional expressions (ternary choice operator) to fast compiler.
whesse@chromium.org [Mon, 2 Nov 2009 08:44:19 +0000 (08:44 +0000)]
Add conditional expressions (ternary choice operator) to fast compiler.
Review URL: http://codereview.chromium.org/340058

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3192 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoReport function source position via JSON protocol
peter.rybin@gmail.com [Fri, 30 Oct 2009 16:38:37 +0000 (16:38 +0000)]
Report function source position via JSON protocol

Review URL: http://codereview.chromium.org/341024

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3189 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoImplement new support for if statements in top-level code.
kmillikin@chromium.org [Fri, 30 Oct 2009 15:34:09 +0000 (15:34 +0000)]
Implement new support for if statements in top-level code.

Review URL: http://codereview.chromium.org/346022

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3188 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoInitial implementation of top-level compilation of expressions in test
kmillikin@chromium.org [Fri, 30 Oct 2009 14:06:48 +0000 (14:06 +0000)]
Initial implementation of top-level compilation of expressions in test
context.  Test contexts are used for the left subexpressions of
short-circuited boolean operators.  The right subexpressions inherit
their expression context from the binary op expression.

Compilation of short-circuited operations in effect and test context
is straightforward:

effect(e0 || e1) =
 test(e0, L0, L1)
 L1: effect(e1)
 L0:

test(e0 || e1, L0, L1) =
 test(e0, L0, L2)
 L2: test(e1, L0, L1)

Because the value of the first subexpression may be needed as the
value of the whole expression in a value context, we introduce a
hybrid value/test contest (the value is needed if true, but not if
false).

value(e0 || e1) =
 value/test(e0, L0, L1)
 L1: value(e1)
 L0:

The compilation of value/test and test/value (introduced by boolean
AND) is:

value/test(e0 || e1, L0, L1) =
 value/test(e0, L0, L2)
 L2: value/test(e1, L0, L1)

test/value(e0 || e1, L0, L1) =
 test(e0, L0, L2)
 L2: test/value(e1, L0, L1)

Boolean AND is the dual.  The AST nodes themselves (not their parents)
are responsible for producing the proper result (effect, value, or
control flow) depending on their context.

Review URL: http://codereview.chromium.org/339082

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3187 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd void operator to fast compiler.
whesse@chromium.org [Fri, 30 Oct 2009 11:32:42 +0000 (11:32 +0000)]
Add void operator to fast compiler.
Review URL: http://codereview.chromium.org/342055

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3186 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix new snapshot compilation on ARM.
erik.corry@gmail.com [Fri, 30 Oct 2009 11:28:49 +0000 (11:28 +0000)]
Fix new snapshot compilation on ARM.
Review URL: http://codereview.chromium.org/349001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3185 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoIntroduce a switch for the new snapshot code and switch
erik.corry@gmail.com [Fri, 30 Oct 2009 10:23:12 +0000 (10:23 +0000)]
Introduce a switch for the new snapshot code and switch
it on by default.  Includes bug fixes for new snapshots.
Review URL: http://codereview.chromium.org/342054

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3184 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoSupport for calls on named and keyed properties in the fast compiler of the form:
fschneider@chromium.org [Fri, 30 Oct 2009 10:22:31 +0000 (10:22 +0000)]
Support for calls on named and keyed properties in the fast compiler of the form:

o.x() and o[expr]()

other changes:
- Fix missing relocation info for StoreIC on global object.
- Generate only one common return sequence instead of always appending
  "return <undefined>" at the end of each function: The first JS
  return-statement will generate the common return sequence. All
  other return-statements will generate a unconditional branch to the common
  return sequence.

Review URL: http://codereview.chromium.org/340037

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3183 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix thinko in assertion.
kmillikin@chromium.org [Fri, 30 Oct 2009 09:57:08 +0000 (09:57 +0000)]
Fix thinko in assertion.

TBR=whesse@chromium.org

Review URL: http://codereview.chromium.org/343055

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3182 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoMove the Location class into the AST Expression class as a member.
kmillikin@chromium.org [Fri, 30 Oct 2009 08:36:46 +0000 (08:36 +0000)]
Move the Location class into the AST Expression class as a member.
Since it is (currently) only an enum, change it to an enum (for now).

Review URL: http://codereview.chromium.org/342035

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3181 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoSupport for property assignment in the fast compiler.
fschneider@chromium.org [Fri, 30 Oct 2009 07:56:38 +0000 (07:56 +0000)]
Support for property assignment in the fast compiler.

The code for .result = (b.y = 99) where b is a global variable is:

  push [esi+0x17]
  mov ecx,0xf5c229ad          ;; object: 0xf5c229ad <String[1]: b>
  call LoadIC_Initialize
  nop
  mov [esp],eax
  mov eax,0xc6
  mov ecx,0xf5c25c41          ;; object: 0xf5c25c41 <String[1]: y>
  call StoreIC_Initialize
  nop
  mov [esp],eax
  pop [ebp+0xf4]

There is still some room for improvement in the generated code.

Other changes:
 - Replaced switch-statement in FastCodeGenerator::VisitProperty with DropAndMove(...)
 - Do not emit nop after IC calls on ARM.

Review URL: http://codereview.chromium.org/347001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3180 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd a test which verifies that weak reference callbacks cannot be invoked while scave...
antonm@chromium.org [Thu, 29 Oct 2009 23:06:59 +0000 (23:06 +0000)]
Add a test which verifies that weak reference callbacks cannot be invoked while scavenging.

BUG=25819

Review URL: http://codereview.chromium.org/334043

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3179 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoPrepare to push to trunk. We are now working on 1.3.19
sgjesse@chromium.org [Thu, 29 Oct 2009 14:33:51 +0000 (14:33 +0000)]
Prepare to push to trunk. We are now working on 1.3.19
Review URL: http://codereview.chromium.org/348010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3176 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRename the Location type tags to be consistent with our current naming
kmillikin@chromium.org [Thu, 29 Oct 2009 13:58:04 +0000 (13:58 +0000)]
Rename the Location type tags to be consistent with our current naming
scheme for enumerations (eg, EFFECT => kEffect).

Remove the ability to move from one Location to another, which should
never be necessary.

Review URL: http://codereview.chromium.org/340034

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3175 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoReverting 3159, 3151 and 3150
sgjesse@chromium.org [Thu, 29 Oct 2009 13:17:11 +0000 (13:17 +0000)]
Reverting 3159, 3151 and 3150

TBR=christian.plesner.hansen@gmail.com
Review URL: http://codereview.chromium.org/343035

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3174 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix x64 build.
kmillikin@chromium.org [Thu, 29 Oct 2009 12:19:20 +0000 (12:19 +0000)]
Fix x64 build.

TBR=whesse@chromium.org

Review URL: http://codereview.chromium.org/346012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3173 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd binary operations to fast compiler.
whesse@chromium.org [Thu, 29 Oct 2009 11:55:03 +0000 (11:55 +0000)]
Add binary operations to fast compiler.
Review URL: http://codereview.chromium.org/342019

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3172 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoRename the kinds of locations to be consistent with the (codegen)
kmillikin@chromium.org [Thu, 29 Oct 2009 10:35:29 +0000 (10:35 +0000)]
Rename the kinds of locations to be consistent with the (codegen)
context of the expressions they label.  Introduce an "unintialized"
location to catch failure to assign any location at all.

Changed the object literal initialization on ARM to use a Store IC in
the same cases where it did on the other platforms.  This was required
because the location of the literal property name is given an
"unitialized" location.

Review URL: http://codereview.chromium.org/339045

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3171 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoReverting 3153
sgjesse@chromium.org [Thu, 29 Oct 2009 07:31:23 +0000 (07:31 +0000)]
Reverting 3153

TBR=erik.corry@gmail.com
Review URL: http://codereview.chromium.org/341031

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3170 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoPrepare push to trunk. Now working on version 1.3.18.
ager@chromium.org [Wed, 28 Oct 2009 14:30:55 +0000 (14:30 +0000)]
Prepare push to trunk.  Now working on version 1.3.18.
Review URL: http://codereview.chromium.org/338063

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3166 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoARM still not quite right with new serialization code.
erik.corry@gmail.com [Wed, 28 Oct 2009 14:28:20 +0000 (14:28 +0000)]
ARM still not quite right with new serialization code.
Review URL: http://codereview.chromium.org/339043

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3165 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoIssue 485: Fix leak of builtins object through call and apply functions.
lrn@chromium.org [Wed, 28 Oct 2009 13:51:30 +0000 (13:51 +0000)]
Issue 485: Fix leak of builtins object through call and apply functions.

Review URL: http://codereview.chromium.org/345007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3164 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoIn the toplevel compiler, shift the responsibility of assigning a
kmillikin@chromium.org [Wed, 28 Oct 2009 13:25:40 +0000 (13:25 +0000)]
In the toplevel compiler, shift the responsibility of assigning a
location to an Expression AST node from the node's parent to the node
itself.

This allows an inherited code generation context from a parent node to
be passed arbitrarily far down the tree (eg, the subexpression of a
unary not is in the same context as the unary expression itself, the
then and else subexpressions of the ternary operator are in the same
context as the whole expression, and so forth).

We do not yet take advantage of this in the backend (eg, the right
subexpression of short-circuited OR is still compiled by using the
parent's destination location, rather than the subexpression's
itself).

Review URL: http://codereview.chromium.org/340005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3163 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoDerive string size constants
sgjesse@chromium.org [Wed, 28 Oct 2009 13:10:36 +0000 (13:10 +0000)]
Derive string size constants

* The maximum length of short and medium sized strings is now derived from other constants.
* Remove the redundant String part of their names.
Review URL: http://codereview.chromium.org/347002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3162 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years ago* Fix new snapshot code on ARM.
erik.corry@gmail.com [Wed, 28 Oct 2009 12:37:54 +0000 (12:37 +0000)]
* Fix new snapshot code on ARM.
Review URL: http://codereview.chromium.org/344011

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3161 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoMarking test as expected to flake.
christian.plesner.hansen@gmail.com [Wed, 28 Oct 2009 12:23:35 +0000 (12:23 +0000)]
Marking test as expected to flake.

Review URL: http://codereview.chromium.org/343016

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3160 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoChanged accessor ICs to using direct calls.
christian.plesner.hansen@gmail.com [Wed, 28 Oct 2009 12:14:20 +0000 (12:14 +0000)]
Changed accessor ICs to using direct calls.

Review URL: http://codereview.chromium.org/343015

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3159 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd fast check for NaN in NumberCompare slow case.
ager@chromium.org [Wed, 28 Oct 2009 10:19:56 +0000 (10:19 +0000)]
Add fast check for NaN in NumberCompare slow case.
Review URL: http://codereview.chromium.org/344010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3158 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFix missing recording of source position for property access in fast compiler.
fschneider@chromium.org [Wed, 28 Oct 2009 10:14:32 +0000 (10:14 +0000)]
Fix missing recording of source position for property access in fast compiler.

Review URL: http://codereview.chromium.org/341002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3157 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoFactor out a commonly used code sequence to DropAndMove(Location, Register)
whesse@chromium.org [Wed, 28 Oct 2009 09:56:14 +0000 (09:56 +0000)]
Factor out a commonly used code sequence to DropAndMove(Location, Register)
Review URL: http://codereview.chromium.org/337060

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3156 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

14 years agoAdd snapshot option to test runner.
sgjesse@chromium.org [Wed, 28 Oct 2009 09:38:09 +0000 (09:38 +0000)]
Add snapshot option to test runner.

Passing option --snapshot to test.py has the same effect as passing -S snapshot=on.
Review URL: http://codereview.chromium.org/346004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@3155 ce2b1a6d-e550-0410-aec6-3dcde31c8c00