MIPS: Fix 'Add HArrayBufferNotNeutered instruction'.

MIPS: Fix 'Add HArrayBufferNotNeutered instruction'.

BUG=

Committed: https://crrev.com/29bee1604f93341e9f0a564a16a8fc74546a9aa4
Cr-Commit-Position: refs/heads/master@{#28117}

[balazs.kilvady, 2015-04-28 16:54:05]

balazs.kilvady@imgtec.com changed reviewers:
+ akos.palfi@imgtec.com, danno@chromium.org, dusan.milosavljevic@imgtec.com,
jochen@chromium.org, paul.lind@imgtec.com

[paul.l..., 2015-04-28 16:56:18]

LGTM.

[balazs.kilvady, 2015-04-28 16:57:45]

The CQ bit was checked by balazs.kilvady@imgtec.com

[commit-bot: I haz the power, 2015-04-28 16:58:00]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1111003002/1

[commit-bot: I haz the power, 2015-04-28 17:20:54]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-04-28 17:21:10]

Patchset 1 (id:??) landed as
https://crrev.com/29bee1604f93341e9f0a564a16a8fc74546a9aa4
Cr-Commit-Position: refs/heads/master@{#28117}

[MTBrandyberry, 2015-04-28 19:22:56]

On 2015/04/28 16:56:18, paul.l... wrote:
> LGTM.

I think you actually want lw when accessing JSArrayBuffer::kBitFieldOffset. 
Although it's really only important for big-endian mips64 (not sure if this is
supported).

[paul.l..., 2015-04-28 20:16:09]

On 2015/04/28 19:22:56, mtbrandyberry wrote:
> I think you actually want lw when accessing JSArrayBuffer::kBitFieldOffset. 
> Although it's really only important for big-endian mips64 (not sure if this is
> supported).

You're correct Matt, thanks for the comment! We don't support mips64 big-endian,
so I've not been looking at these closely, assuming that accesses should match
their allocation size. What I missed here is that even though the bit-field is
allocated 8-bytes on 64-bit platforms, the C++ accessors are using 32-bit
load/store. So we should do the same.