Promote ArrayBuffer, DataView and typed arrays to non-experimental.

Promote ArrayBuffer, DataView and typed arrays to non-experimental.

The primary reason for this change is to bake these guys into the
snapshot.

Flag definitions (--harmony-typed-arrays, --harmony-array-buffer) are
still there so that Blink does not complain, but they are noop and
default to true.

R=mstarzinger@chromium.org
BUG=270527
Committed: https://code.google.com/p/v8/source/detail?r=16137

Committed: https://code.google.com/p/v8/source/detail?r=16228

Committed: https://code.google.com/p/v8/source/detail?r=16248

[Dmitry Lomov (no reviews), 2013-08-09 13:03:34]

PTAL. 
This should help http://crbug.com/270572.

[Michael Starzinger, 2013-08-09 15:42:06]

LGTM (with nits)

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc
File src/bootstrapper.cc (right):

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc#newcode1070
src/bootstrapper.cc:1070: global, "ArrayBuffer", JS_ARRAY_BUFFER_TYPE,
This change should finally make ArrayBuffer.prototype non-writable, could we add
a separate test file to mjsunit that checks that using something like the
following:

Object.getOwnPropertyDescriptor(ArrayBuffer, "prototype")

Lets choose a name for the test suite unrelated to typed arrays (e.g.
mjsunit/function-prototype-primordial.js) as this applies to all experimental
features. Or add it to the mjsunit/function-prototype.js file.

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc#newcode1113
src/bootstrapper.cc:1113: }
nit: Add an empty newline here for readability.

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc#newcode1380
src/bootstrapper.cc:1380: 
nit: Only one empty newline here.

[Michael Starzinger, 2013-08-09 15:49:06]

I take back one of my comments ...

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc
File src/bootstrapper.cc (right):

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc#newcode1070
src/bootstrapper.cc:1070: global, "ArrayBuffer", JS_ARRAY_BUFFER_TYPE,
On 2013/08/09 15:42:06, Michael Starzinger wrote:
> This change should finally make ArrayBuffer.prototype non-writable, could we
add
> a separate test file to mjsunit that checks that using something like the
> following:
> 
> Object.getOwnPropertyDescriptor(ArrayBuffer, "prototype")
> 
> Lets choose a name for the test suite unrelated to typed arrays (e.g.
> mjsunit/function-prototype-primordial.js) as this applies to all experimental
> features. Or add it to the mjsunit/function-prototype.js file.

I just realized that this is already covered by mjsunit/builtins.js for all
functions that are inside the snapshot (i.e. not behind a flag). Please ignore
this comment. Sorry for the noise.

[Dmitry Lomov (no reviews), 2013-08-09 15:53:21]

Nits addressed, landing.

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc
File src/bootstrapper.cc (right):

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc#newcode1113
src/bootstrapper.cc:1113: }
On 2013/08/09 15:42:06, Michael Starzinger wrote:
> nit: Add an empty newline here for readability.

Done.

https://codereview.chromium.org/22390008/diff/1/src/bootstrapper.cc#newcode1380
src/bootstrapper.cc:1380: 
On 2013/08/09 15:42:06, Michael Starzinger wrote:
> nit: Only one empty newline here.

Done.

[Dmitry Lomov (no reviews), 2013-08-09 15:57:41]

Committed patchset #1 manually as r16137 (presubmit successful).

[Dmitry Lomov (no reviews), 2013-08-20 00:50:22]

Committed patchset #1 manually as r16228.

[Dmitry Lomov (no reviews), 2013-08-20 13:56:10]

Committed patchset #2 manually as r16248.

[Toon Verwaest, 2013-09-04 13:20:42]

This CL breaks test/webkit/regexp-in-and-foreach-handling.js (from stress-GC 2):

out/ia32.debug/d8 --stress-opt --always-opt --test --nobreak-on-abort
--nodead-code-elimination --nofold-constants --enable-slow-asserts --debug-code
--verify-heap test/webkit/resources/standalone-pre.js
test/webkit/regexp-in-and-foreach-handling.js
test/webkit/resources/standalone-post.js --gc-interval=500 --stress-compaction
--concurrent-recompilation-queue-length=64 --concurrent-recompilation-delay=500
--concurrent-recompilation

[Toon Verwaest, 2013-09-04 13:52:52]

Nevermind, it seems to just be so sensitive to timing that this CL probably just
changes that timing enough to make it break. The problem seems to be that the
regexp matches has a hole in the array where there should be an undefined. Yang
is looking into it.