Add mb config and testing spec for new android clang x86 tester.

Add mb config and testing spec for new android clang x86 tester.

BUG=534815
Committed: https://crrev.com/b71ee736ca83f78ec3b0cdf324709685b7b89849
Cr-Commit-Position: refs/heads/master@{#428228}

[bpastene, 2016-10-15 00:15:53]

bpastene@chromium.org changed reviewers:
+ dpranke@chromium.org, thakis@chromium.org

[Dirk Pranke, 2016-10-15 17:20:59]

lgtm.

Question, however: why are you running tests locally rather than using swarming?

[Nico, 2016-10-17 14:46:00]

On 2016/10/15 17:20:59, Dirk Pranke wrote:
> lgtm.
> 
> Question, however: why are you running tests locally rather than using
swarming?

Is swarming on android a way to get the tests on the device, or is a "run tests
on a vague cloud-like thing"? If the latter: We only have a handful of intel
devices, all connected to this machine.

(This CL broadly lgtm, but it'll probably need changes after
https://codereview.chromium.org/2414253003/ is through review)

[Dirk Pranke, 2016-10-17 16:45:13]

On 2016/10/17 14:46:00, Nico (mostly afk until Oct 23) wrote:
> On 2016/10/15 17:20:59, Dirk Pranke wrote:
> > Question, however: why are you running tests locally rather than using
> swarming?
> 
> Is swarming on android a way to get the tests on the device, or is a "run
tests
> on a vague cloud-like thing"? If the latter: We only have a handful of intel
> devices, all connected to this machine.

I don't think I understand the distinction you're trying to draw, but swarming
on
android is just like swarming everywhere else: it gives you access to pools of
devices that have particular configurations, so you can share the same devices
across multiple builders. So, your tests would run on real devices, but the
devices
would be attached to a different machine than the one running buildbot code.

[bpastene, 2016-10-17 17:28:41]

On 2016/10/17 16:45:13, Dirk Pranke wrote:
> On 2016/10/17 14:46:00, Nico (mostly afk until Oct 23) wrote:
> > On 2016/10/15 17:20:59, Dirk Pranke wrote:
> > > Question, however: why are you running tests locally rather than using
> > swarming?
> > 
> > Is swarming on android a way to get the tests on the device, or is a "run
> tests
> > on a vague cloud-like thing"? If the latter: We only have a handful of intel
> > devices, all connected to this machine.
> 
> I don't think I understand the distinction you're trying to draw, but swarming
> on
> android is just like swarming everywhere else: it gives you access to pools of
> devices that have particular configurations, so you can share the same devices
> across multiple builders. So, your tests would run on real devices, but the
> devices
> would be attached to a different machine than the one running buildbot code.

Either would theoretically work, but I went with the buildbot setup for a couple
reasons:
a) We don't currently have any x86 devices in the android swarming pool (it's
very limited atm to mainly CQ devices)
b) We could throw this bot into the pool, but I doubt swarming's device
provisioning supports these devices well. We do cpu
throttling/package-management so I'd be surprised if what we've got now would
work on these tablets (We could make it work, but I'm the one that's really
maintaining that code so I'm just trying to save myself a bit of work :P )
c) Many moons ago, I saw a coho bot downstream (it's since been removed) so
buildbot's provisioning should theoretically work here.

[Dirk Pranke, 2016-10-17 17:33:58]

On 2016/10/17 17:28:41, bpastene wrote:
> On 2016/10/17 16:45:13, Dirk Pranke wrote:
> > On 2016/10/17 14:46:00, Nico (mostly afk until Oct 23) wrote:
> > > On 2016/10/15 17:20:59, Dirk Pranke wrote:
> > > > Question, however: why are you running tests locally rather than using
> > > swarming?
> > > 
> > > Is swarming on android a way to get the tests on the device, or is a "run
> > tests
> > > on a vague cloud-like thing"? If the latter: We only have a handful of
intel
> > > devices, all connected to this machine.
> > 
> > I don't think I understand the distinction you're trying to draw, but
swarming
> > on
> > android is just like swarming everywhere else: it gives you access to pools
of
> > devices that have particular configurations, so you can share the same
devices
> > across multiple builders. So, your tests would run on real devices, but the
> > devices
> > would be attached to a different machine than the one running buildbot code.
> 
> Either would theoretically work, but I went with the buildbot setup for a
couple
> reasons:
> a) We don't currently have any x86 devices in the android swarming pool (it's
> very limited atm to mainly CQ devices)
> b) We could throw this bot into the pool, but I doubt swarming's device
> provisioning supports these devices well. We do cpu
> throttling/package-management so I'd be surprised if what we've got now would
> work on these tablets (We could make it work, but I'm the one that's really
> maintaining that code so I'm just trying to save myself a bit of work :P )
> c) Many moons ago, I saw a coho bot downstream (it's since been removed) so
> buildbot's provisioning should theoretically work here.

OK, I was mostly just curious. Thanks!

[bpastene, 2016-10-27 19:32:30]

The CQ bit was checked by bpastene@chromium.org

[bpastene, 2016-10-27 19:32:31]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, thakis@chromium.org
Link to the patchset: https://codereview.chromium.org/2423653002/#ps40001
(title: "fix typo")

[commit-bot: I haz the power, 2016-10-27 19:32:52]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-27 19:41:06]

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

[commit-bot: I haz the power, 2016-10-27 19:41:07]

Try jobs failed on following builders:
  chromium_presubmit on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[Nico, 2016-10-27 20:05:51]

** Presubmit ERRORS **
manage (0.21s) failed
chromium.fyi.json: ClangToTAndroid x64 / media_mojo_unittests is not listed in
gn_isolate_map.pyl.


maybe easiest to remove that test for now i suppose

[bpastene, 2016-10-27 20:54:23]

The CQ bit was checked by bpastene@chromium.org

[bpastene, 2016-10-27 20:54:24]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, thakis@chromium.org
Link to the patchset: https://codereview.chromium.org/2423653002/#ps60001
(title: "remove media_mojo_unittests")

[commit-bot: I haz the power, 2016-10-27 20:55:40]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-27 21:41:23]

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

[commit-bot: I haz the power, 2016-10-27 21:41:24]

Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[bpastene, 2016-10-27 22:04:46]

The CQ bit was checked by bpastene@chromium.org

[commit-bot: I haz the power, 2016-10-27 22:05:41]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-27 23:15:48]

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

[commit-bot: I haz the power, 2016-10-27 23:15:52]

Try jobs failed on following builders:
  linux_android_rel_ng on master.tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/linux_androi...)

[bpastene, 2016-10-27 23:27:20]

The CQ bit was checked by bpastene@chromium.org

[commit-bot: I haz the power, 2016-10-27 23:27:52]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-28 00:12:10]

Committed patchset #4 (id:60001)

[bpastene, 2016-10-28 00:13:57]

On 2016/10/28 00:12:10, commit-bot: I haz the power wrote:
> Committed patchset #4 (id:60001)

\o/

[commit-bot: I haz the power, 2016-10-28 00:16:12]

Description was changed from

==========
Add mb config and testing spec for new android clang x86 tester.

BUG=534815
==========

to

==========
Add mb config and testing spec for new android clang x86 tester.

BUG=534815

Committed: https://crrev.com/b71ee736ca83f78ec3b0cdf324709685b7b89849
Cr-Commit-Position: refs/heads/master@{#428228}
==========

[commit-bot: I haz the power, 2016-10-28 00:16:13]

Patchset 4 (id:??) landed as
https://crrev.com/b71ee736ca83f78ec3b0cdf324709685b7b89849
Cr-Commit-Position: refs/heads/master@{#428228}

[Nico, 2016-10-28 04:21:46]

On 2016/10/28 00:16:13, commit-bot: I haz the power wrote:
> Patchset 4 (id:??) landed as
> https://crrev.com/b71ee736ca83f78ec3b0cdf324709685b7b89849
> Cr-Commit-Position: refs/heads/master@{#428228}

I guess we need an fyi master restart next?

[bpastene, 2016-10-28 17:15:28]

On 2016/10/28 04:21:46, Nico (vacation Fri Oct 28) wrote:
> I guess we need an fyi master restart next?

Someone already got to it:
https://build.chromium.org/p/chromium.fyi/builders/ClangToTAndroid%20x64

Looks like I screwed up the gn args though? Hmmm.....

[Nico, 2016-10-31 14:09:23]

On 2016/10/28 17:15:28, bpastene wrote:
> On 2016/10/28 04:21:46, Nico (vacation Fri Oct 28) wrote:
> > I guess we need an fyi master restart next?
> 
> Someone already got to it:
> https://build.chromium.org/p/chromium.fyi/builders/ClangToTAndroid%20x64
> 
> Looks like I screwed up the gn args though? Hmmm.....

I'll look into this.

[Nico, 2016-11-02 16:30:31]

Ok, the bot builds now. Out of interest, where did you get the list of tests
from? I had to remove several since they didn't exist on Android :-)