Enable transitional non-SFI browser tests on ARM

Enable transitional non-SFI browser tests on ARM

- Produce nacl_helper_nonsfi binary in out/{Debug,Release}
- Enable all transitional non-SFI browser tests on ARM

BUG=433201, 372049
CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32,linux_arm_compile,linux_arm

Committed: https://crrev.com/f4aadf3f424b10f8f9e0111229ccf114e598b84e
Cr-Commit-Position: refs/heads/master@{#311490}

[mazda, 2014-12-12 18:45:29]

mazda@chromium.org changed reviewers:
+ hidehiko@chromium.org, mseaborn@chromium.org

[mazda, 2014-12-12 18:45:30]

PTAL

NaClBrowserTestPnaclTransitionalNonSfi.* tests passed with two NaCl side changes
applied.
https://codereview.chromium.org/771593002/
https://codereview.chromium.org/722423003/

[mazda, 2015-01-13 03:41:16]

Hi, happy new year!

Now Chromium ToT uses the NaCl revision that includes changes needed for this
CL, and the tests pass without local changes.
Could you review the patch?

Thanks,

[hidehiko, 2015-01-14 05:56:45]

LGTM.

Could you run linux_rel_precise32 bot just in case?
Also, could you add 

CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32

to the commit message, so that CQ will run the bot, too?

On 2015/01/13 03:41:16, mazda wrote:
> Hi, happy new year!
> 
> Now Chromium ToT uses the NaCl revision that includes changes needed for this
> CL, and the tests pass without local changes.
> Could you review the patch?
> 
> Thanks,

[Mark Seaborn, 2015-01-14 06:28:18]

LGTM

https://codereview.chromium.org/794573002/diff/80001/chrome/test/ppapi/ppapi_...
File chrome/test/ppapi/ppapi_browsertest.cc (right):

https://codereview.chromium.org/794573002/diff/80001/chrome/test/ppapi/ppapi_...
chrome/test/ppapi/ppapi_browsertest.cc:94: MAYBE_PNACL_NONSFI(test_name)) { \
Will it end up being necessary to change all of these back to
MAYBE_PNACL_TRANSITIONAL_NONSFI when we enable x86-32-nonsfi testing on the
x86-64 Linux bots?  (What's the status of testing on those bots, BTW?)

If so, it might be desirable to leave MAYBE_PNACL_TRANSITIONAL_NONSFI in place,
maybe with a TODO about x86-64 bots.

[mazda, 2015-01-14 08:59:28]

Patchset #6 (id:100001) has been deleted

[mazda, 2015-01-14 09:55:02]

Thank you for the review.

> Could you run linux_rel_precise32 bot just in case?
> Also, could you add 
> 
> CQ_EXTRA_TRYBOTS=tryserver.chromium.linux:linux_rel_precise32
> 
> to the commit message, so that CQ will run the bot, too?
Done

> Will it end up being necessary to change all of these back to
> MAYBE_PNACL_TRANSITIONAL_NONSFI when we enable x86-32-nonsfi testing on the
> x86-64 Linux bots?  (What's the status of testing on those bots, BTW?)
> 
> If so, it might be desirable to leave MAYBE_PNACL_TRANSITIONAL_NONSFI in
place,
> maybe with a TODO about x86-64 bots.
Right. I restored MAYBE_PNACL_TRANSITIONAL_NONSFI and added TODO about x86-64.

As to the status of testing on x86-64 bots, we can build nacl_helper_nonsfi with
target_arch=x64 build configuration, but a few tweaks of build rules are
necessary to run tests with it on x86-64 bots.

[mazda, 2015-01-14 10:14:37]

mazda@chromium.org changed reviewers:
+ csharp@chromium.org

[mazda, 2015-01-14 10:14:37]

csharp@
Could you do OWNERS review for chrome/chrome.isolate?

Thanks,

[csharp, 2015-01-14 14:24:44]

lgtm

[mazda, 2015-01-14 15:08:15]

The CQ bit was checked by mazda@chromium.org

[commit-bot: I haz the power, 2015-01-14 15:09:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/794573002/120001

[commit-bot: I haz the power, 2015-01-14 16:39:51]

Committed patchset #6 (id:120001)

[commit-bot: I haz the power, 2015-01-14 16:40:54]

Patchset 6 (id:??) landed as
https://crrev.com/f4aadf3f424b10f8f9e0111229ccf114e598b84e
Cr-Commit-Position: refs/heads/master@{#311490}

[hidehiko, 2015-01-15 02:25:54]

FYI: as for x64 PPAPI browser_tests, it was blocked by NaCl roll. Unfortunately,
it didn't happen before my Year-End vacation last year, so the task was
suspended. Now it should be ready. I'll look into it.