Instead of relying on eax value after the call load returned handle value directly from the slot that was preallocated for it.

R=vitalyr@chromium.org

Review URL: http://codereview.chromium.org/7307004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8533 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Due to issues relating mostly to chrome extensions we have lately been
running into OOMs that are caused by our executable space running
out. This change introduces flushing of code from regexps if we have
not used the code for 5 mark sweeps.

The approach is different from the normal function code flusing. Here
we make a copy of the code inside the data array, and exchange the
original code with a smi determined by the sweep_generation (a new
heap variable increased everytime we do mark sweep/compact). If we
encounter a smi in EnsureCompiled we simply reinstate the code
object. If, in the marking phase of mark sweep, we find a regexp that
already have a smi in the code field, and this is more than 5
generations old we flush the code from the saved index.
Review URL: http://codereview.chromium.org/7282026

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8532 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

As described in the issue 1532, this may lead to building a 32-bit version over
existing x64-bit. It's better delegate to delegate this task to an user.

R=vitalyr@chromium.org
BUG=1532

Review URL: http://codereview.chromium.org/7309012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8530 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

As part of allowing different contexts for inlined functions, eliminate most explicit reads of the context from the stack frame in ia32 crankshaft codegen.

Eliminates the enum flag RESTORE_CONTEXT and CONTEXT_ADJUSTED, and adds a context HValue and LOperand to many hydrogen and lithium instructions.

Context is still used from the stack from in CallKnownFunction (this seems safe), and in CallRuntimeFromDeferred in lithium-codegen-ia32.cc, which needs to be fixed.

BUG=
TEST=

Review URL: http://codereview.chromium.org/7132002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8529 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Internal HandleScope::CloseAndEscape crashes on NULL handles.

R=kmillikin@chromium.org
BUG=v8:1529
TEST=mjsunit/regress/regress-1529

Review URL: http://codereview.chromium.org/7309004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8527 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7307003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8526 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
TEST=

Review URL: http://codereview.chromium.org/7307002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8525 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Return to previous random number generator, but mix more bits into
output to hide hidden state better.  Keep the multithreading fix that
moves the PNG into isolate.

BUG=
TEST=

Review URL: http://codereview.chromium.org/7250005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8524 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

We were only looking one level up the scope chain to decide which
closure to use in the fresh context.  Instead, we should look to the
first non-catch scope.

R=vegorov@chromium.org
BUG=1528
TEST=regress-1528

Review URL: http://codereview.chromium.org/7309002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8523 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Port r8496 (b7d5ecf0).

Original commit message:
Introduce scopes to keep track of catch blocks at compile time.
The catch variable is bound in the catch scope. For simplicity in this
initial implementation, it is always allocated even if unused and always
allocated to a catch context even if it doesn't escape. The presence of
catch is no longer treated as a with.
In this change, care must be taken to distinguish between the scope where a
var declaration is hoisted to and the scope where the initialization occurs.

BUG=
TEST=

Review URL: http://codereview.chromium.org//7212020
Patch from Paul Lind <plind44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8522 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Do not pass the global object as the receiver to strict-mode and
builtin replace and sort functions.

Port r8488 (6f347b6)

BUG=
TEST=

Review URL: http://codereview.chromium.org//7278024
Patch from Paul Lind <plind44@gmail.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8521 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The constant pool was not taken into account in the test test-compiler/SplitConstantsInFullCompiler which caused random failures.

This also reverts the test code added in r8469 and r8471.

R=ricow@chromium.org

BUG=none
TEST=test-compiler/SplitConstantsInFullCompiler

Review URL: http://codereview.chromium.org//7308001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8520 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Revision 8516 contained a temporary hack that doesn't work on Windows.

TBR: ricow

Review URL: http://codereview.chromium.org/7298008

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8519 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

When optimizing a function defined inside a catch, we did not count
the catch context as part of the context chain.

R=vegorov@chromium.org
BUG=1521
TEST=regress-1521

Review URL: http://codereview.chromium.org/7285032

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8518 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Add tests for duplicate properties of object initialisers to preparser.

TEST=preparser

Review URL: http://codereview.chromium.org/7168016

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8517 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=vegorov@chromium.org

Review URL: http://codereview.chromium.org/7298004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8516 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Made "map_cache" a weak field of global context and added a pass over
all caches late in the marking phase.

R=vegorov@chromium.org
BUG=v8:1516
TEST=cctest/test-api/Regress1516

Review URL: http://codereview.chromium.org/7285031

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8515 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TEST=start with ./d8 --allow-natives-syntax and try %ObjectHasFastElements({})

Review URL: http://codereview.chromium.org/7289010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8514 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This means that ES5 Date Time Strings will default to UTC if timezone is absent.
Handle as many legacy strings as possible the same way as before

BUG=v8:1498
TEST=mjsunit/date

Review URL: http://codereview.chromium.org/7291022

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8513 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

(if the test ran too slow, optimization kicked in earlier than expected by the test)

TEST=mjsunit/assert-opt-and-deopt.js no longer flaky

Review URL: http://codereview.chromium.org/7298001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8511 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=vitalyr@chromium.org
BUG=
TEST=

Review URL: http://codereview.chromium.org/7278017

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8510 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7210056

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8509 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7278018

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8508 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Modules now makes a little sense, as there is only one module.
Tags was a simplistic attempt of supporting nested profiles.

R=sgjesse@chromium.org
BUG=859

Review URL: http://codereview.chromium.org/7276046

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8507 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7285030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8506 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

the latter ones twice.
Review URL: http://codereview.chromium.org/7292030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8505 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7283044

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8504 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Only a single constructor is really needed. Make UnaryOp_Patch's arguments
simpler, avoiding encoding/decoding things in a cryptic way.
Review URL: http://codereview.chromium.org/7248063

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8503 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Fixes the build for the shell on illumos and Solaris. -D__C99FEATURES__ was added to mirror how the build is done on the normal platform. The changes in the platform code are a follow up to a prior review and has the Solaris implementation be more similar to the Linux version as opposed to the FreeBSD.

Contributed by Robert Mustacchi <rm@fingolfin.org>

TEST=Note the test suite uncovered a bug in libm where pow(3M) was not doing the right thing on edge cases. The only test failures are related to this bug.

Review URL: http://codereview.chromium.org/7282034

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8502 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7290006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8501 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7282033

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8500 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=fschneider@chromium.org

Review URL: http://codereview.chromium.org/7291001

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8499 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=
Review URL: http://codereview.chromium.org/7284035

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8498 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=kmillikin@chromium.org
BUG=1514
TEST=mjsunit/regress/regress-1513.js

Review URL: http://codereview.chromium.org/7282029

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8497 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The catch variable is bound in the catch scope.  For simplicity in this
initial implementation, it is always allocated even if unused and always
allocated to a catch context even if it doesn't escape.  The presence of
catch is no longer treated as a with.

In this change, care must be taken to distinguish between the scope where a
var declaration is hoisted to and the scope where the initialization occurs.

R=ager@chromium.org
BUG=
TEST=

Review URL: http://codereview.chromium.org/7280012

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8496 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Instead of generating two instructions and combining them
at lithium translation using the EmitAtUses predicate, we
generate the correct branch instruction right from the start.
Review URL: http://codereview.chromium.org/7237024

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8495 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This helps understanding what is a particular array for by
just looking at its name.

R=vitalyr@chromium.org

Review URL: http://codereview.chromium.org/7248058

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8494 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Note: It has only been verified that compilation is OK, nothing more... ;-)
Review URL: http://codereview.chromium.org/7273097

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8493 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=ricow@chromium.org

Review URL: http://codereview.chromium.org/7284032

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8492 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7283043

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8491 ce2b1a6d-e550-0410-aec6-3dcde31c8c00