This avoid the worst case of patching ICs repeatedly with the
same stub code object.

BUG=v8:1489
Review URL: http://codereview.chromium.org/7230007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8364 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

are short of virtual memory.  This should make V8 work in 64 bit
on OpenBSD in its default configuration.  It is a simplified
version of the reverted 8133 which also lowered the non-code
heap size, causing test failures in Chromium on MacOS.
Review URL: http://codereview.chromium.org/7234004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8363 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Now only uses the common superclass of the two scanner classes.
Updated comment on KeywordMatcher.

Review URL: http://codereview.chromium.org/7211013

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8362 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The declaration of the ToBoolean class moved to the platform-independent part
and its implementations are now structurally very similar. This is just an
intermediate cleanup step to add type recording at the call site.

Note that the MIPS implementation has not really been touched, so it should
continue to work, too.
Review URL: http://codereview.chromium.org/7218012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8359 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This is a follow-on to r8340, http://codereview.chromium.org/7216004/

Søren had commited changes and closed issue, as Karl was commenting about
problems in the code.

Thanks Karl, for your detailed code review. I have corrected the issues
you raised.

BUG=
TEST=

Review URL: http://codereview.chromium.org//7232004
Patch from Paul Lind <plind44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8358 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=none
TEST=none

Review URL: http://codereview.chromium.org//7217011
Patch from Martyn Capewell <m.m.capewell@googlemail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8357 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=82437
TEST=Compile browser_tests for Arm against http://codereview.chromium.org/7087014/

Review URL: http://codereview.chromium.org/7212014

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8356 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7200024

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8355 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The destination is a GP register, not an XMM register.  ia32 did not
have this problem.

BUG=
TEST=Check --print_code output for code that tests for -0.0.

Review URL: http://codereview.chromium.org/7191015

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8352 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

A multi-line comment containing a newline is considered a line-terminator for
other purposes, but a "-->" following such a comment is considered as being
on the same line as the text preceeding the multi-line comment.
This behavior matches JSC matching Firefox.

TEST=cctest/test-parsing/ScanHTMLEndComments

Review URL: http://codereview.chromium.org/7218009

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8351 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

We had two instructions HCompareJsObjectEq and HCompareSymbolEq that behave
exactly the same. I removed one and renamed the remaining instruction into
HCompareObjectEq.
Review URL: http://codereview.chromium.org/7206040

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8349 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TEST=added to test/mjsunit/arguments.js

Review URL: http://codereview.chromium.org/7217005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8348 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=v8:1491
TEST=mjsunit/regress/regress-1491.js

Review URL: http://codereview.chromium.org/7206038

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8343 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This is a preparation for removing aggregated heap snapshots.
W/o this API, counting object instances in a snapshot is very hard.

R=sgjesse@chromium.org
BUG=1481
TEST=cctest/test-heap-profiler/NodesIteration

Review URL: http://codereview.chromium.org/7204040

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8342 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7216007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8341 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Ported r8315 (d70d395).

BUG=
TEST=

Review URL: http://codereview.chromium.org//7216004
Patch from Paul Lind <plind44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8340 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

There were some changes in ARM r8172 that we needed to port to MIPS,
but missed.

The bug was exposed with the Arm & Mips fix in r8321, "Fix wrong
parenthesis in stub ARM and MIPS generation code."

BUG=
TEST=

Review URL: http://codereview.chromium.org//7211006
Patch from Paul Lind <plind44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8339 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Following r8289 (bf0c0cf).

BUG=
TEST=

Review URL: http://codereview.chromium.org//7211007
Patch from Paul Lind <plind44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8338 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Prepare push to trunk.  Now working on V8 version 3.4.6.  Revision 8324, which also claims this, has been superseded by this one. 
Review URL: http://codereview.chromium.org/7210016

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8336 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7205030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8335 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

[I've accidentally lost my gcl_info directory and I see no way to commit the old CL, so I simply deleted the old CL opened a new one. :-P ]
Review URL: http://codereview.chromium.org/7210012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8332 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Was broken by the recent JumpIfSmi() cleanup.

TEST=es5conform

Review URL: http://codereview.chromium.org/7206023

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8331 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

with AST IDs. Previously 3 different places had to match in how they handle a
given case, now we are down to 2, with an even simpler logic.

The downside is that due to this simpler logic the allocated dictionary could be
larger than before, but test have shown that this happens *very* rarely, because
its capacity is rounded to the next power of 2, anyway. Furthermore, the oracle
doesn't live long enough that we should really care.

The whole oracle is probably still a bit too tricky in its details, but this is
at least a step into the right direction.
Review URL: http://codereview.chromium.org/7204003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8330 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

In simple unconditional loops - like the following - the hydrogen stack check elimination did not detect the call as the loop header block itself was not considered.

function f(o) {
  while(true) {
    o.a();
  }
}

R=fschneider@chromium.org

BUG=none
TEST=none

Review URL: http://codereview.chromium.org//7210010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8329 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

unless the caller has indicated that pretenuring is desirable.
This is intended to fix http://jsperf.com/scope-lookups/3
Review URL: http://codereview.chromium.org/7210011

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8328 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This is a regression relative to the original behavior, when "native" was a keyword,
since keywords cannot contain esacpes.

Added tests for escapes and for not allowing line-terminators betwen "native" and "function".

Review URL: http://codereview.chromium.org/7206022

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8327 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

We now only recognize "native function" when it occurs in extension scripts
(parsing with a non-NULL extension), and only if there is no line-terminator
between "native" and "function" (so that it would otherwise be a Syntax Error).
Preparsing never recognizes native functions, which is acceptable since we
never preparse extension scripts (because we don't allow lazy functions
anyway).

BUG=v8:1097

Review URL: http://codereview.chromium.org/7206020

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8326 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7170012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8325 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7212007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8324 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

added a few test cases.
Review URL: http://codereview.chromium.org/7212006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8323 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7191007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8322 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

I think the problem did not manifest as a bug, since the only
consequence was to generate extra code to return heap number
results in case of the INT32-stub.
Review URL: http://codereview.chromium.org/7192006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8321 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=kasperl@chromium.org,kmillikin@chromium.org
BUG=1417
TEST=test-log/Issue23768

Review URL: http://codereview.chromium.org/7184027

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8320 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7189030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8319 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

report, but was not general enough to catch all cases.  This is a new
approach.  Includes regression test!
Review URL: http://codereview.chromium.org/7193007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8318 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Now follows the specification. Follows WebKit change in revision 89100.

BUG=86431
TEST=regress-892742

Review URL: http://codereview.chromium.org/7184034

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8317 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7172026

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8316 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7187007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8315 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7189007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8312 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=v8:1473

Review URL: http://codereview.chromium.org/7147014

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8311 ce2b1a6d-e550-0410-aec6-3dcde31c8c00