MIPS: Inline Math.fround in optimized code.

MIPS: Inline Math.fround in optimized code.

Port r22665 (7e3d70d)

BUG=v8:3469
LOG=N
R=paul.lind@imgtec.com

Committed: https://code.google.com/p/v8/source/detail?r=22682

[paul.l..., 2014-07-29 16:54:26]

small change needed....

https://codereview.chromium.org/425053002/diff/1/src/mips64/lithium-codegen-m...
File src/mips64/lithium-codegen-mips64.cc (right):

https://codereview.chromium.org/425053002/diff/1/src/mips64/lithium-codegen-m...
src/mips64/lithium-codegen-mips64.cc:3861: __ cvt_d_s(result, result.low());
Please remove use of result.low() in both instructions, and just use 'result'.
In fp64 mode, we pass full 64-bit FP regs that can be used as target of either
single or double-precision ops. We no longer use pairs, and should then no
longer use .low() or .high(). (They will be removed in a future mips64 cleanup
CL)

[balazs.kilvady, 2014-07-29 18:36:11]

https://codereview.chromium.org/425053002/diff/1/src/mips64/lithium-codegen-m...
File src/mips64/lithium-codegen-mips64.cc (right):

https://codereview.chromium.org/425053002/diff/1/src/mips64/lithium-codegen-m...
src/mips64/lithium-codegen-mips64.cc:3861: __ cvt_d_s(result, result.low());
On 2014/07/29 16:54:26, paul.l... wrote:
> Please remove use of result.low() in both instructions, and just use 'result'.
> In fp64 mode, we pass full 64-bit FP regs that can be used as target of either
> single or double-precision ops. We no longer use pairs, and should then no
> longer use .low() or .high(). (They will be removed in a future mips64 cleanup
> CL)

Done.

[paul.l..., 2014-07-29 18:41:20]

lgtm

[balazs.kilvady, 2014-07-29 19:26:11]

Committed patchset #2 manually as r22682 (presubmit successful).