A64: Fix write barrier input in KeyedStoreIC::GenerateSloppyArguments.

A64: Fix write barrier input in KeyedStoreIC::GenerateSloppyArguments.

This fixes flaky crashes in gc-stress bot:
> Fatal error in ../src/incremental-marking.cc, line 84
> CHECK(obj->IsHeapObject()) failed


BUG=353551
LOG=N
TEST=test/mjsunit/regress/regress-353551.js
R=m.m.capewell@googlemail.com

Committed: https://code.google.com/p/v8/source/detail?r=20098

[ulan, 2014-03-19 12:05:25]

PTAL

[m.m.capewell, 2014-03-19 14:11:43]

DBC, as Jacob's away.

https://codereview.chromium.org/204453002/diff/40001/src/a64/ic-a64.cc
File src/a64/ic-a64.cc (right):

https://codereview.chromium.org/204453002/diff/40001/src/a64/ic-a64.cc#newcod...
src/a64/ic-a64.cc:395: __ Mov(scratch1, Operand(scratch1, LSL,
kPointerSizeLog2));
Might be clearer as Lsl(scratch1, scratch1, kPointerSizeLog2).

[ulan, 2014-03-19 14:20:27]

https://codereview.chromium.org/204453002/diff/40001/src/a64/ic-a64.cc
File src/a64/ic-a64.cc (right):

https://codereview.chromium.org/204453002/diff/40001/src/a64/ic-a64.cc#newcod...
src/a64/ic-a64.cc:395: __ Mov(scratch1, Operand(scratch1, LSL,
kPointerSizeLog2));
On 2014/03/19 14:11:44, m.m.capewell wrote:
> Might be clearer as Lsl(scratch1, scratch1, kPointerSizeLog2).

Done.

[m.m.capewell, 2014-03-19 15:55:40]

lgtm

[ulan, 2014-03-20 08:33:04]

Committed patchset #2 manually as r20098 (presubmit successful).