X64: Add inline cache stub for storing to globals.

X64: Add inline cache stub for storing to globals.  

Committed: http://code.google.com/p/v8/source/detail?r=2543

[William Hesse, 2009-07-27 13:55:55]

A small review for you.

[Kevin Millikin (Chromium), 2009-07-27 14:31:42]

LGTM.

http://codereview.chromium.org/160160/diff/1/3
File src/x64/stub-cache-x64.cc (right):

http://codereview.chromium.org/160160/diff/1/3#newcode499
Line 499: __ DecrementCounter(&Counters::named_store_global_inline, 1);
It seems weird to increment this counter on all paths and then decrement it
here.  Why not increment it on only the fast path?

http://codereview.chromium.org/160160/diff/1/2
File test/mjsunit/mjsunit.status (right):

http://codereview.chromium.org/160160/diff/1/2#newcode101
Line 101: new: PASS, CRASH || FAIL if ($mode == debug)
I'm surprised by the test expectation change (the commit message doesn't mention
it).  Do you really expect this test to pass in release mode now, or do you mean
that the old CRASH || FAIL behavior is still OK?

[William Hesse, 2009-07-27 15:01:48]

http://codereview.chromium.org/160160/diff/1/3
File src/x64/stub-cache-x64.cc (right):

http://codereview.chromium.org/160160/diff/1/3#newcode499
Line 499: __ DecrementCounter(&Counters::named_store_global_inline, 1);
On 2009/07/27 14:31:42, Kevin Millikin wrote:
> It seems weird to increment this counter on all paths and then decrement it
> here.  Why not increment it on only the fast path?

Changed per suggestion.

http://codereview.chromium.org/160160/diff/1/2
File test/mjsunit/mjsunit.status (right):

http://codereview.chromium.org/160160/diff/1/2#newcode101
Line 101: new: PASS, CRASH || FAIL if ($mode == debug)
On 2009/07/27 14:31:42, Kevin Millikin wrote:
> I'm surprised by the test expectation change (the commit message doesn't
mention
> it).  Do you really expect this test to pass in release mode now, or do you
mean
> that the old CRASH || FAIL behavior is still OK?

changed to PASS || CRASH || FAIL