Roll openmax_dl 79e64bc:6d58d90.

Roll openmax_dl 79e64bc:6d58d90.

Adds support for non-Android ARM. Requires a GN change corresponding
to the gyp change here:
https://webrtc-codereview.appspot.com/29639004

BUG=415393
Committed: https://crrev.com/0a9d636948b331b86c226df4771616215175fbc5
Cr-Commit-Position: refs/heads/master@{#298635}

[ajm, 2014-10-03 22:23:51]

ajm@chromium.org changed reviewers:
+ rtoy@chromium.org

[ajm, 2014-10-03 22:23:52]

Rolling in Chromium first, as the openmax dep in webrtc is now pinned to the
Chromium rev.

[Raymond Toy, 2014-10-03 22:28:35]

lgtm

[ajm, 2014-10-03 23:08:01]

The CQ bit was checked by ajm@chromium.org

[commit-bot: I haz the power, 2014-10-03 23:09:10]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/626183003/1

[commit-bot: I haz the power, 2014-10-04 00:05:55]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-10-04 00:05:56]

Try jobs failed on following builders:
  android_chromium_gn_compile_rel on tryserver.chromium.linux ()

[ajm, 2014-10-04 01:25:46]

ajm@chromium.org changed reviewers:
+ kjellander@chromium.org

[ajm, 2014-10-04 01:27:16]

The Android GN bot caught a failure on this. My try jobs are just giving
exceptions for some reason (following up with chrome-infra) but I could repro
locally and fixed it with the GN change in PS#2.

Raymond please have another look, and Henrik please review the GN change.

[kjellander_chromium, 2014-10-04 11:28:27]

On 2014/10/04 01:27:16, ajm wrote:
> The Android GN bot caught a failure on this. My try jobs are just giving
> exceptions for some reason (following up with chrome-infra) but I could repro
> locally and fixed it with the GN change in PS#2.
> 
> Raymond please have another look, and Henrik please review the GN change.

build/secondary/third_party/openmax_dl/dl/BUILD.gn: lgtm

[kjellander_chromium, 2014-10-04 11:28:45]

...and here are my minor comments/nits.

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
File build/secondary/third_party/openmax_dl/dl/BUILD.gn (right):

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:5: if (cpu_arch == "arm") {
You can skip this condition as it's already in the arm.gni

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:21: if (!arm_use_neon &&
is_android) {
replace with:
} else if (is_android) {

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:248: sources += [
"sp/src/arm/detect.c", ]
nit: skip comma after source entry.

[ajm, 2014-10-06 05:53:28]

Actually, this required a blink change as well. Please have a look here:
https://codereview.chromium.org/635443002

That will have to be rolled in to Chromium before this lands.

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
File build/secondary/third_party/openmax_dl/dl/BUILD.gn (right):

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:5: if (cpu_arch == "arm") {
On 2014/10/04 11:28:45, kjellander wrote:
> You can skip this condition as it's already in the arm.gni

Done.

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:21: if (!arm_use_neon &&
is_android) {
On 2014/10/04 11:28:45, kjellander wrote:
> replace with:
> } else if (is_android) {

Done.

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:248: sources += [
"sp/src/arm/detect.c", ]
On 2014/10/04 11:28:45, kjellander wrote:
> nit: skip comma after source entry.

Done.

[Raymond Toy, 2014-10-06 16:21:29]

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
File build/secondary/third_party/openmax_dl/dl/BUILD.gn (left):

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:56: ":openmax_dl_armv7"
This dependency seems to have been removed. It's still required on Android I
think to get the openmax_dl library to include the armv7 versions of the FFTs.

[ajm, 2014-10-06 16:35:59]

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
File build/secondary/third_party/openmax_dl/dl/BUILD.gn (left):

https://codereview.chromium.org/626183003/diff/20001/build/secondary/third_pa...
build/secondary/third_party/openmax_dl/dl/BUILD.gn:56: ":openmax_dl_armv7"
On 2014/10/06 16:21:29, Raymond Toy wrote:
> This dependency seems to have been removed. It's still required on Android I
> think to get the openmax_dl library to include the armv7 versions of the FFTs.

It's been moved above to line 66. Sorry for the bad diff.

[Raymond Toy, 2014-10-06 16:53:00]

lgtm.

[ajm, 2014-10-07 17:25:42]

My blink change was rolled; trying to land this one again.

[ajm, 2014-10-07 17:25:46]

The CQ bit was checked by ajm@chromium.org

[commit-bot: I haz the power, 2014-10-07 17:26:49]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/626183003/40001

[commit-bot: I haz the power, 2014-10-07 17:59:41]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-10-07 17:59:42]

Try jobs failed on following builders:
  android_arm64_dbg_recipe on tryserver.chromium.linux
(http://build.chromium.org/p/tryserver.chromium.linux/builders/android_arm64_d...)

[ajm, 2014-10-07 22:06:58]

Urgh, ARM64 is still failing. Raymond, I have a fix here:
https://webrtc-codereview.appspot.com/26739004/

Should be the last one. Sorry for all the iterations.

[Raymond Toy, 2014-10-07 22:16:41]

On 2014/10/07 at 22:06:58, ajm wrote:
> Urgh, ARM64 is still failing. Raymond, I have a fix here:
> https://webrtc-codereview.appspot.com/26739004/
> 
> Should be the last one. Sorry for all the iterations.

No problems. I should have tried a full arm64 build instead of just running the
tests.

[ajm, 2014-10-07 22:25:09]

The CQ bit was checked by ajm@chromium.org

[commit-bot: I haz the power, 2014-10-07 22:26:16]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/626183003/80001

[ajm, 2014-10-08 00:23:25]

The CQ bit was unchecked by ajm@chromium.org

[ajm, 2014-10-08 00:23:28]

The CQ bit was checked by ajm@chromium.org

[commit-bot: I haz the power, 2014-10-08 00:23:48]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/626183003/80001

[commit-bot: I haz the power, 2014-10-08 01:39:41]

Committed patchset #5 (id:80001) as ec013dd21a9b682cf2354b17274efcafe98bedfe

[commit-bot: I haz the power, 2014-10-08 01:42:28]

Patchset 5 (id:??) landed as
https://crrev.com/0a9d636948b331b86c226df4771616215175fbc5
Cr-Commit-Position: refs/heads/master@{#298635}