It caused a regressions in compile time in the optimizing compiler. I'm reverting until
the reason for this is identified and fixed.

TBR=whesse@chromium.org
Review URL: http://codereview.chromium.org/7453024

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8762 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=vitalyr@chromium.org
TEST=cctest/test-dictionary

Review URL: http://codereview.chromium.org/7385006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8761 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7524025

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8760 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7482044

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8759 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

To do this, the Branch instruction needs to carry around a temporary register,
but only when the crankshafted code will make a map access. When the
crankshafted code sees an object of a type it hasn't encountered before, it will
always trigger a deopt. Another option in theses cases would be calling a
ToBooleanStub which can handle all types, but then one has to be careful to
*not* trigger a GC (which is currently a bit tricky to achieve).

Const-corrected ToBoolean::Types. Moved the NeedsMap logic into ToBoolean::Types
itself, where it belongs.

This patch improves a lot of benchmarks, crypto-orig even by 16.7%, but slows
down others. The slowdown has to be investigated, but I'd like to get this patch
out first to fix the flakiness problems we currently have due to the previous
crankshafted ToBoolean.
Review URL: http://codereview.chromium.org/7461107

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8758 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7482042

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8757 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
TEST=mjsunit/scope-calls-eval.js

Review URL: http://codereview.chromium.org/7464030

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8756 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
TEST=

Review URL: http://codereview.chromium.org/7097011

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8755 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=ricow@chromium.org
Review URL: http://codereview.chromium.org/7523025

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8754 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Landing two patches contributed by Andy Wingo:

http://codereview.chromium.org/7514040/ and

http://codereview.chromium.org/7516001/
Review URL: http://codereview.chromium.org/7520022

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8753 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7522015

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8750 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=svenpanne@chromium.org
BUG=none
TEST=none

Review URL: http://codereview.chromium.org/7477047

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8749 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=ricow@chromium.org
BUG=non
TEST=running urls from reliability bots

Review URL: http://codereview.chromium.org/7497010

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8748 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This fixes the issues when GC happens when invoking
ToBooleanStub.
Review URL: http://codereview.chromium.org/7465066

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8747 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=v8:1572
TEST=mjsunit/regress/regress-crbug-l84186.js

Review URL: http://codereview.chromium.org/7480044

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8746 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=fschneider@chromium.org
BUG=none
TEST=none

Review URL: http://codereview.chromium.org/7488016

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8745 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

o Use a more strict limit for old arrays.

o Initial capacity of a slow elements dictionary should be the number
  of used elements and not the old array capacity.

R=danno@chromium.org

Review URL: http://codereview.chromium.org/7464032

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8744 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

TBR=whesse@chromium.org
Review URL: http://codereview.chromium.org/7470016

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8743 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8742 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7470014

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8741 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7198021

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8740 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Landing for yangguo to coordinate with buildbot update.

Original codereview: http://codereview.chromium.org/7282008/
Review URL: http://codereview.chromium.org/7495033

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8739 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

* src/x64/lithium-codegen-x64.cc (DoLoadKeyedFastDoubleElement)
  (DoStoreKeyedFastDoubleElement): Remove set but unused variables.

BUG=
TEST=Fixes a build with gcc-4.6.

Review URL: http://codereview.chromium.org/7491041
Patch from Andy Wingo <wingo@igalia.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8738 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=v8:1539
TEST=cctest test-api/SetPrototypeAttributes

Review URL: http://codereview.chromium.org/7324027

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8737 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Record ToBoolean's type information in Hydrogen's HBranch instruction, so we can use it in LCodeGen::DoBranch later.
Review URL: http://codereview.chromium.org/7491043

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8736 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

If the input range is positive and the shift count is constant
we can replace >>> with >> to compute the output range.

For negative inputs, we can only compute a range if the 
result always fits into a signed int32.

BUG=v8:1510
Review URL: http://codereview.chromium.org/7489043

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8735 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7493055

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8732 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

It causes relibility bot crashes
Review URL: http://codereview.chromium.org/7461061

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8731 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7492042

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8730 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Review URL: http://codereview.chromium.org/7495025

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8729 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=None
TEST=None
Signed-off-by: Thiago Farina <tfarina@chromium.org>

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8728 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Properly handle undefined conversion to zero in Crankshaft.

R=yangguo@chromium.org
BUG=none
TEST=regress-1563.js

Review URL: http://codereview.chromium.org/7461028

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8723 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Loading the map within the stub already implies a check for an undetectable
object, so there is no need to record this separately. Furthermore, this brings
the size of the type info to record down to 8 bits, removing the need to find a
place for the ninth bit in the Code object. ;-)
Review URL: http://codereview.chromium.org/7484022

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8722 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8721 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Since r93412, Chromium no more passes "--logfile=*" to V8.

R=danno@chromium.org
BUG=859

Review URL: http://codereview.chromium.org/7479004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8720 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=ager@chromium.org
BUG=none
TEST=none

Review URL: http://codereview.chromium.org/7491007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8719 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=ager@chromium.org
BUG=none
TEST=cctests, unboxed-double-array.js

Review URL: http://codereview.chromium.org/7473031

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8718 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The test tried to disassemble the last entry in a constant pool, which only
worked by accident until address randomization was introduced.
Review URL: http://codereview.chromium.org/7489005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8717 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

The IC records the set of types it has seen, e.g. {String} or {Boolean,
Undefined}, etc.  Note that in theory this could lead to a large number of
different ToBoolean ICs (512, to be exact, because we distinguish 9 types),
but in practice only a small handful of them are actually generated.

Currently the type recording part is only implemented on ia32, other platforms
continue to work like they did before, though.

Removed some dead code on the way.
Review URL: http://codereview.chromium.org/7473028

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8716 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Used in a separate upcoming patch...
Review URL: http://codereview.chromium.org/7472029

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@8715 ce2b1a6d-e550-0410-aec6-3dcde31c8c00