MIPS64: Add big-endian support for mips64.

MIPS64: Add big-endian support for mips64.

TEST=
BUG=

Committed: https://crrev.com/8bd431de92d4032c67f51be750987967691a77c1
Cr-Commit-Position: refs/heads/master@{#31011}

[dusan.milosavljevic, 2015-09-10 18:31:30]

dusan.m.milosavljevic@gmail.com changed reviewers:
+ bmeurer@chromium.org, danno@chromium.org, jkummerow@chromium.org,
paul.lind@imgtec.com, titzer@chromium.org

[dusan.milosavljevic, 2015-09-10 18:31:30]

PTAL.

[Benedikt Meurer, 2015-09-11 05:06:19]

bmeurer@chromium.org changed required reviewers:
+ jkummerow@chromium.org

[Benedikt Meurer, 2015-09-11 05:06:20]

LGTM

[Jakob Kummerow, 2015-09-11 09:23:24]

LGTM (reviewed cross-platform files, rubberstamping mips64-specific code)

[paul.l..., 2015-09-15 00:06:41]

Looking good. First round of comments on build issues. PTAL.

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi
File build/toolchain.gypi (right):

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi#newcod...
build/toolchain.gypi:344: or v8_target_arch=="mips64el") and
v8_target_arch==target_arch', {
Please add 'or v8_target_arch=="mips64"' clause here. We are not getting the
r-path and ldso-path set properly.

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi#newcod...
build/toolchain.gypi:758: 'V8_TARGET_ARCH_MIPS64_LE',
Per MIPS conventions, this should probably be called V8_TARGET_ARCH_MIPS64_EL.

Also, this defintion does not appear to be used, so maybe unnecessary?

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi#newcod...
build/toolchain.gypi:763: 'V8_TARGET_ARCH_MIPS64_BE',
Should be  ..._EB, as above. While you use this in build_config.h, I don't
understand the reason why. Maybe you can elaborate on the comment in that file.

https://codereview.chromium.org/1334793004/diff/1/src/base/build_config.h
File src/base/build_config.h (right):

https://codereview.chromium.org/1334793004/diff/1/src/base/build_config.h#new...
src/base/build_config.h:167: #if defined(__MIPSEB__) ||
defined(V8_TARGET_ARCH_MIPS64_BE)
What's the reason for the || V8_TARGET_ARCH_MIPS64_BE ? Do some toolchains not
support the __MIPSEB__ definition? (I'm comparing this to mips32 big-endian just
above)

https://codereview.chromium.org/1334793004/diff/1/src/base/build_config.h#new...
src/base/build_config.h:168: define V8_TARGET_BIG_ENDIAN 1
Typo: define needs leading #

[dusan.milosavljevic, 2015-09-15 12:58:09]

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi
File build/toolchain.gypi (right):

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi#newcod...
build/toolchain.gypi:344: or v8_target_arch=="mips64el") and
v8_target_arch==target_arch', {
On 2015/09/15 00:06:41, paul.l... wrote:
> Please add 'or v8_target_arch=="mips64"' clause here. We are not getting the
> r-path and ldso-path set properly.

Done.

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi#newcod...
build/toolchain.gypi:758: 'V8_TARGET_ARCH_MIPS64_LE',
On 2015/09/15 00:06:41, paul.l... wrote:
> Per MIPS conventions, this should probably be called V8_TARGET_ARCH_MIPS64_EL.
> 
> Also, this defintion does not appear to be used, so maybe unnecessary?

Not based on MIPS convention, I used these names to be consistent with other
ports (ppc).

https://codereview.chromium.org/1334793004/diff/1/build/toolchain.gypi#newcod...
build/toolchain.gypi:763: 'V8_TARGET_ARCH_MIPS64_BE',
On 2015/09/15 00:06:41, paul.l... wrote:
> Should be  ..._EB, as above. While you use this in build_config.h, I don't
> understand the reason why. Maybe you can elaborate on the comment in that
file.

This is used to detect host byte-order in addition to detection of target
toolchain flags. This enables correct build on be hosts.

https://codereview.chromium.org/1334793004/diff/1/src/base/build_config.h
File src/base/build_config.h (right):

https://codereview.chromium.org/1334793004/diff/1/src/base/build_config.h#new...
src/base/build_config.h:168: define V8_TARGET_BIG_ENDIAN 1
On 2015/09/15 00:06:41, paul.l... wrote:
> Typo: define needs leading #

Done.

[paul.l..., 2015-09-29 16:53:54]

This one is ready now. I don't think any of the arch-indep files changed after
your guy's previous review, except for rebasing.

Thanks for the new feature, and the fixes, Dusan! LGTM.

[dusan.milosavljevic, 2015-09-29 16:59:58]

The CQ bit was checked by dusan.m.milosavljevic@gmail.com

[dusan.milosavljevic, 2015-09-29 16:59:59]

The patchset sent to the CQ was uploaded after l-g-t-m from
bmeurer@chromium.org, jkummerow@chromium.org
Link to the patchset: https://codereview.chromium.org/1334793004/#ps200001
(title: "Rebase ToT.")

[commit-bot: I haz the power, 2015-09-29 17:00:09]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1334793004/200001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1334793004/200001

[commit-bot: I haz the power, 2015-09-29 17:23:49]

Committed patchset #11 (id:200001)

[commit-bot: I haz the power, 2015-09-29 17:24:06]

Patchset 11 (id:??) landed as
https://crrev.com/8bd431de92d4032c67f51be750987967691a77c1
Cr-Commit-Position: refs/heads/master@{#31011}