BUG=
R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/48353007

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18090 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Port r18084 (426bc19)

Original commit message:
Optimize register constraints and code generated for Integer32ToSmi Lithium
instruction.

TEST=none
BUG=
R=gergely@homejinni.com

Review URL: https://codereview.chromium.org/88623004

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18089 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Let the test runner preserve the order of test suites to let suites with long running tests run first.

Mark some tests as slow that can now be skipped via --slow-tests=skip.

BUG=
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/88343002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18086 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This results in a speed-up of around 30% on all architectures. For local developer builds only. The bots will continue to use level 1 for now.

BUG=
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/88563002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18085 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Optimize register constraints and code generated for Integer32ToSmi Lithium
instruction.

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

Review URL: https://codereview.chromium.org/88383002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18084 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=jkummerow@chromium.org
BUG=v8:3025
LOG=Y

Review URL: https://codereview.chromium.org/88583002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18082 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

For substrings of non-flat cons strings, we bail out
to runtime. For single character substrings, we forget
to flatten it. This causes successive bailouts.

R=bmeurer@chromium.org
BUG=323041
LOG=Y

Review URL: https://codereview.chromium.org/88173002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18081 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/88133002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18079 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=titzer@chromium.org

Review URL: https://codereview.chromium.org/86233002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18078 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/88363002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18075 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This way we get all the Crankshaft goodness and avoid always going
through the runtime: Less code + even some small speedup in Kraken.

R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/88053002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18072 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/88073002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18071 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/85163003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18070 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=ishell@chromium.org

Review URL: https://codereview.chromium.org/85473004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18069 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

We cannot yet deprecate the versions that don't take one, because
there's too much test code in v8 using them. To keep this CL small and
managable, I'll first introduce the new APIs.

BUG=none
R=dcarney@chromium.org, svenpanne@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/84833006

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18068 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

It now closely resembles the js2c target, so that we only have
one platform to run it on (either 'host' or 'target'). This way
we do not run into warnings when compiling for e.g. ARM.

R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/85893002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18067 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Port r18059 (b2849c3)

BUG=
R=dslomov@chromium.org, gergely@homejinni.com

Review URL: https://codereview.chromium.org/86053003

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18066 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Port r18054 (59d9d99)

BUG=

Review URL: https://codereview.chromium.org/84943008

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18065 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Port r18046 (99c7352)

Original commit message:
A performance regression in array literal creation was caused by refactoring that eliminated a special fast case for shallow arrays. At the same time the general case got a bit slower. This CL restores most of the performance without coding the special fast case. The virtual dispatching is unnecessary because we know what we want to do at compile time. A flag was added to Runtime::CreateArrayLiteral. The flags delivers information about shallowness but also whether or not allocation mementos should be created. This is useful for crankshafted code.

BUG=v8:3008
LOG=Y
R=gergely@homejinni.com

Review URL: https://codereview.chromium.org/85633004

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18064 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Optimize register constraints and code generated for TypeofIsAndBranch Lithium
instruction.

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

Review URL: https://codereview.chromium.org/78583002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18063 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Transformed variant flags into named variants. Now, all combinations of variants can be specified on the command line. The old command-line flags are kept for backwards compatibility on the bots.

Added two new test groups: slow and pass|fail. Both are implemented similar to the flaky test feature and allow to either skip or run tests marked as slow or as pass|fail.

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/85733003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18062 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=danno@chromium.org

Review URL: https://codereview.chromium.org/59023003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18059 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

We shouldn't have APIs that call Isolate::GetCurrent() internally. This
change removes all remaining occurrences of inlined methods in v8.h

BUG=none
R=svenpanne@chromium.org
LOG=n

Review URL: https://codereview.chromium.org/85693002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18058 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Let change log formatter squash title and bug reference. Repair wrongly indented first line in change log. Add automatic rietveld reload of commit messages to enable late corrections.

BUG=
R=jkummerow@chromium.org

Review URL: https://codereview.chromium.org/83603002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18057 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=mvstanton@chromium.org

Review URL: https://codereview.chromium.org/85793002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18055 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=mvstanton@chromium.org

Review URL: https://chromiumcodereview.appspot.com/82743003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18054 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/85743002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18051 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Usage:
tools/merge-to-branch.sh -R 12345 12346

R=machenbach@chromium.org

Review URL: https://codereview.chromium.org/85413004

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18048 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=svenpanne@chromium.org

Review URL: https://codereview.chromium.org/85483002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18047 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

A performance regression in array literal creation was caused by refactoring that eliminated a special fast case for shallow arrays. At the same time the general case got a bit slower. This CL restores most of the performance without coding the special fast case. The virtual dispatching is unnecessary because we know what we want to do at compile time. A flag was added to Runtime::CreateArrayLiteral. The flags delivers information about shallowness but also whether or not allocation mementos should be created. This is useful for crankshafted code.

BUG=v8:3008
LOG=Y
R=mstarzinger@chromium.org

Review URL: https://codereview.chromium.org/77293003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18046 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

BUG=
R=hpayer@chromium.org

Review URL: https://codereview.chromium.org/48583005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18045 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=verwaest@chromium.org

Review URL: https://codereview.chromium.org/85613003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18044 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=yangguo@chromium.org

Review URL: https://codereview.chromium.org/85623002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18040 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/85343002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18037 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=bmeurer@chromium.org

Review URL: https://codereview.chromium.org/85313002

Patch from Ben Noordhuis <ben@strongloop.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18035 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

In this change, the support comes in two flavours:

--perf_jit_prof - outputs the files in a new perf format that only works with a
patched perf tool (patch obtained from Stephane Eranian). Both 'perf report' and
'perf annotate' are supported (the file format also contains the machine code).

--perf_basic_prof - outputs the files in a format that the existing perf tool
can consume. Only 'perf report' is supported.

In both cases, we have to disable code compaction because the perf tool does not
understand code relocation. (We are told that code relocation should be
supported soon.)

Usage:

perf record -g d8 --perf_jit_prof --no_compact_code_space my.js
perf report

The change itself is straightforward - we simply listen to code events and
write an entry to a log file for every new piece of code.

I am not yet sure whether we should keep both versions or just one (and which
one). My hope is the reviewers can help here.

R=danno@chromium.org
BUG=

Review URL: https://codereview.chromium.org/70013002

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18034 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

R=danno@chromium.org

Review URL: https://codereview.chromium.org/82943005

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18033 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Port r18025 (74fc739)

BUG=

Review URL: https://codereview.chromium.org/83913002

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18032 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

Port r18024 (8956065)

BUG=

Review URL: https://codereview.chromium.org/83483005

Patch from Balazs Kilvady <kilvadyb@homejinni.com>.

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18031 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

This commit fixes a lot of test failures that we saw earlier in the buildbots (http://build.chromium.org/p/client.v8/builders/V8%20Linux%20-%20mips%20-%20sim/builds/3034/steps/Check/logs/stdio).

In some very rare cases the code age stub address can be 0xXXXX0000 and in this case the li maco instruction emits only 1 instruction (instead of the expected 2). Thus the code age sequence will be 6 instructions long instead of 7, which breaks the code aging feature. This change makes sure that li always emits 2 instructions and it also simplifies the code aging sequence.

Also fixes a small mistake in the simulator at the jalr instruction.

BUG=
R=gergely@homejinni.com

Review URL: https://codereview.chromium.org/83583003

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18030 ce2b1a6d-e550-0410-aec6-3dcde31c8c00