MIPS: pre-crankshaft updates to code-stubs and stub-cache (3/3)

MIPS: pre-crankshaft updates to code-stubs and stub-cache (3/3)

Highlights:
- code-stubs-mips.cc
-- use EmitFPUTruncate in place of inline code in several places.
-- use BranchF macro rather than lower-level FP cmp and branch for readability.
-- Port of Sven's r8859 (Implement type recording for ToBoolean) and r8886
(Simplify and optimize ToBoolean handling.)
-- Fix bug in TranscendentalCacheStub::Generate where some regs were not
saved across CFunction call.
-- use updated xxxCFunction macros.
-- update InstanceOfStub to support crankshaft
DoDeferredLInstanceOfKnownGlobal
-- Provide code-patching and I-cache flushing support for generated
code, used for InstanceOfStub under crankshaft (not submitted here).
This requires adding new ExternalReference to src/assember.cc,h

- stub-cache-mips.cc
-- port Danno's r8901 (Create a common base class for Fixed-, FixedDouble-
and ExternalArrays) to mips crankshaft branch.

BUG=
TEST=

Committed: http://code.google.com/p/v8/source/detail?r=9308

[danno, 2011-09-14 09:14:35]

adding erik

[Erik Corry, 2011-09-14 13:48:56]

LGTM

http://codereview.chromium.org/7890001/diff/1/src/mips/code-stubs-mips.cc
File src/mips/code-stubs-mips.cc (right):

http://codereview.chromium.org/7890001/diff/1/src/mips/code-stubs-mips.cc#new...
src/mips/code-stubs-mips.cc:1244: __ PrepareCallCFunction(0, 2, t4);  // Two
doubles count as 4 arguments.
Comment out of date?

http://codereview.chromium.org/7890001/diff/1/src/mips/code-stubs-mips.cc#new...
src/mips/code-stubs-mips.cc:6495: // Don't base result on status bits when a NaN
is involved.
Comment seems out of date.

[Paul Lind, 2011-09-15 05:49:25]

Updated per review comments. Thanks.

http://codereview.chromium.org/7890001/diff/1/src/mips/code-stubs-mips.cc
File src/mips/code-stubs-mips.cc (right):

http://codereview.chromium.org/7890001/diff/1/src/mips/code-stubs-mips.cc#new...
src/mips/code-stubs-mips.cc:1244: __ PrepareCallCFunction(0, 2, t4);  // Two
doubles count as 4 arguments.
On 2011/09/14 13:48:57, Erik Corry wrote:
> Comment out of date?

Yep, Done.

http://codereview.chromium.org/7890001/diff/1/src/mips/code-stubs-mips.cc#new...
src/mips/code-stubs-mips.cc:6495: // Don't base result on status bits when a NaN
is involved.
On 2011/09/14 13:48:57, Erik Corry wrote:
> Comment seems out of date.
Yes, this comment was bad, as were some others nearby, and then I optimized the
code a bit as well. Better now. Thanks.

[danno, 2011-09-16 09:23:38]

LGTM, I'll land this now.