Add recipe config for android clang x86 tester.

Add recipe config for android clang x86 tester.

BUG=534815
Committed: https://chromium.googlesource.com/chromium/tools/build/+/f43b7753c335488f3e09a67462f8bb98641acdf5

[bpastene, 2016-10-15 00:18:42]

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

[Nico, 2016-10-17 14:44:50]

Thanks so much for this!

In addition to the comments below: I was wondering that if we should have a
"regular" FYI bot that builds and runs chrome/android/intel with the default
build settings (i.e. not clang/tot) -- that way, we can see what the state of
tests is with the default toolchain. That bot would use gcc for now, and then
clang once we flip the default. It could use the same slave.

https://codereview.chromium.org/2414253003/diff/40001/masters/master.chromium...
File masters/master.chromium.fyi/master.cfg (right):

https://codereview.chromium.org/2414253003/diff/40001/masters/master.chromium...
masters/master.chromium.fyi/master.cfg:97: 'ClangToTAndroid x86 Tester',
Sicne it's a builder_tester, it probably should just be called ClangToTAndroid
x86 (cf ClangToTAndroid (dbg)).

Also, these devices are capable of x64 and since we want to switch x64 first, we
should probably start with x64, not x86.

https://codereview.chromium.org/2414253003/diff/40001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py (right):

https://codereview.chromium.org/2414253003/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py:1456: 'TARGET_BITS':
32,
64

[bpastene, 2016-10-18 23:22:29]

On 2016/10/17 14:44:50, Nico (mostly afk until Oct 23) wrote:
> Thanks so much for this!
> 
> In addition to the comments below: I was wondering that if we should have a
> "regular" FYI bot that builds and runs chrome/android/intel with the default
> build settings (i.e. not clang/tot) -- that way, we can see what the state of
> tests is with the default toolchain. That bot would use gcc for now, and then
> clang once we flip the default. It could use the same slave.

I'd actually find that super useful as well (for my own selfish reasons; I have
a bot on fyi that's already building x86 chrome-on-android w/ gcc & default
flags and testing it on gce instances of android. It'd be really interesting if
I could get the same binaries on these VMs to also run on real devices and
compare the results.)

Once we've got this bot out there, and got it stable, I'll add this 2nd bot. (I
think I'll just add a 2nd slave, we got plenty of the tablets.)

[bpastene, 2016-10-18 23:22:49]

https://codereview.chromium.org/2414253003/diff/40001/masters/master.chromium...
File masters/master.chromium.fyi/master.cfg (right):

https://codereview.chromium.org/2414253003/diff/40001/masters/master.chromium...
masters/master.chromium.fyi/master.cfg:97: 'ClangToTAndroid x86 Tester',
On 2016/10/17 14:44:50, Nico (mostly afk until Oct 23) wrote:
> Sicne it's a builder_tester, it probably should just be called ClangToTAndroid
> x86 (cf ClangToTAndroid (dbg)).

Done.

> Also, these devices are capable of x64 and since we want to switch x64 first,
we
> should probably start with x64, not x86.

You sure about that? From the device itself:

./adb -s 18E64E8A shell getprop | grep abi
[ro.product.cpu.abi]: [x86]
[ro.product.cpu.abilist32]: [x86,armeabi-v7a,armeabi]
[ro.product.cpu.abilist64]: []
[ro.product.cpu.abilist]: [x86,armeabi-v7a,armeabi]

[Nico, 2016-10-18 23:40:13]

Huh, that's surprising to me; the thread sounded like these were x64
tablets. I know we had at least some x64 phones about a year ago. If these
aren't those, do we have x64 devices somewhere? The bug this refers to is
about x64 coverage iirc.

On Oct 18, 2016 6:22 PM, <bpastene@chromium.org> wrote:

>
> https://codereview.chromium.org/2414253003/diff/40001/
> masters/master.chromium.fyi/master.cfg
> File masters/master.chromium.fyi/master.cfg (right):
>
> https://codereview.chromium.org/2414253003/diff/40001/
> masters/master.chromium.fyi/master.cfg#newcode97
> masters/master.chromium.fyi/master.cfg:97: 'ClangToTAndroid x86 Tester',
> On 2016/10/17 14:44:50, Nico (mostly afk until Oct 23) wrote:
> > Sicne it's a builder_tester, it probably should just be called
> ClangToTAndroid
> > x86 (cf ClangToTAndroid (dbg)).
>
> Done.
>
> > Also, these devices are capable of x64 and since we want to switch x64
> first, we
> > should probably start with x64, not x86.
>
> You sure about that? From the device itself:
>
> ./adb -s 18E64E8A shell getprop | grep abi
> [ro.product.cpu.abi]: [x86]
> [ro.product.cpu.abilist32]: [x86,armeabi-v7a,armeabi]
> [ro.product.cpu.abilist64]: []
> [ro.product.cpu.abilist]: [x86,armeabi-v7a,armeabi]
>
> https://codereview.chromium.org/2414253003/
>

-- 
You received this message because you are subscribed to the Google Groups
"Chromium-reviews" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[bpastene, 2016-10-18 23:44:09]

On 2016/10/18 23:40:13, Nico (mostly afk until Oct 23) wrote:
> Huh, that's surprising to me; the thread sounded like these were x64
> tablets. I know we had at least some x64 phones about a year ago. If these
> aren't those, do we have x64 devices somewhere? The bug this refers to is
> about x64 coverage iirc.

Hmmm, yeah you're right. Let me dig through that shipment we got. I thought they
were all the same device, but I could have gotten that wrong; maybe we got more
than one type.

[Nico, 2016-10-18 23:44:48]

On 2016/10/18 23:40:13, Nico (mostly afk until Oct 23) wrote:
> Huh, that's surprising to me; the thread sounded like these were x64
> tablets. I know we had at least some x64 phones about a year ago. If these
> aren't those, do we have x64 devices somewhere? The bug this refers to is
> about x64 coverage iirc.
> 
> On Oct 18, 2016 6:22 PM, <mailto:bpastene@chromium.org> wrote:
> 
> >
> > https://codereview.chromium.org/2414253003/diff/40001/
> > masters/master.chromium.fyi/master.cfg
> > File masters/master.chromium.fyi/master.cfg (right):
> >
> > https://codereview.chromium.org/2414253003/diff/40001/
> > masters/master.chromium.fyi/master.cfg#newcode97
> > masters/master.chromium.fyi/master.cfg:97: 'ClangToTAndroid x86 Tester',
> > On 2016/10/17 14:44:50, Nico (mostly afk until Oct 23) wrote:
> > > Sicne it's a builder_tester, it probably should just be called
> > ClangToTAndroid
> > > x86 (cf ClangToTAndroid (dbg)).
> >
> > Done.
> >
> > > Also, these devices are capable of x64 and since we want to switch x64
> > first, we
> > > should probably start with x64, not x86.
> >
> > You sure about that? From the device itself:
> >
> > ./adb -s 18E64E8A shell getprop | grep abi
> > [ro.product.cpu.abi]: [x86]
> > [ro.product.cpu.abilist32]: [x86,armeabi-v7a,armeabi]
> > [ro.product.cpu.abilist64]: []
> > [ro.product.cpu.abilist]: [x86,armeabi-v7a,armeabi]
> >
> > https://codereview.chromium.org/2414253003/
> >
> 
> -- 
> You received this message because you are subscribed to the Google Groups
> "Chromium-reviews" group.
> To unsubscribe from this group and stop receiving emails from it, send an
email
> to mailto:chromium-reviews+unsubscribe@chromium.org.

About the cl: the bot seems to not run the "build clang locally" code as far as
I can tell from the json file (compare to existing boy; it should set some magic
env var. Maybe I missed it though, on phone)

[Nico, 2016-10-18 23:44:59]

https://codereview.chromium.org/2414253003/diff/60001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py (right):

https://codereview.chromium.org/2414253003/diff/60001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py:1451:
'chromium_config': 'android',
clang_tot_android not needed?

[Nico, 2016-10-24 14:08:59]

On Tue, Oct 18, 2016 at 7:44 PM, <bpastene@chromium.org> wrote:

> On 2016/10/18 23:40:13, Nico (mostly afk until Oct 23) wrote:
> > Huh, that's surprising to me; the thread sounded like these were x64
> > tablets. I know we had at least some x64 phones about a year ago. If
> these
> > aren't those, do we have x64 devices somewhere? The bug this refers to is
> > about x64 coverage iirc.
>
> Hmmm, yeah you're right. Let me dig through that shipment we got. I
> thought they
> were all the same device, but I could have gotten that wrong; maybe we got
> more
> than one type.
>

^ did you have a chance to look at this?


>
>
> https://codereview.chromium.org/2414253003/
>

-- 
You received this message because you are subscribed to the Google Groups
"Chromium-reviews" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[bpastene, 2016-10-24 16:30:46]

On 2016/10/24 14:08:59, Nico (mostly afk until Oct 23) wrote:
> ^ did you have a chance to look at this?

https://bugs.chromium.org/p/chromium/issues/detail?id=658327

[Nico, 2016-10-25 18:39:02]

On 2016/10/24 16:30:46, bpastene wrote:
> On 2016/10/24 14:08:59, Nico (mostly afk until Oct 23) wrote:
> > ^ did you have a chance to look at this?
> 
> https://bugs.chromium.org/p/chromium/issues/detail?id=658327

Looks like that's done now :-) The only thing left is to land this CL (after
changing it to x64), right?

[bpastene, 2016-10-25 22:24:42]

https://codereview.chromium.org/2414253003/diff/40001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py (right):

https://codereview.chromium.org/2414253003/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py:1456: 'TARGET_BITS':
32,
On 2016/10/17 14:44:50, Nico wrote:
> 64

Done.

https://codereview.chromium.org/2414253003/diff/60001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py (right):

https://codereview.chromium.org/2414253003/diff/60001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/chromium_fyi.py:1451:
'chromium_config': 'android',
On 2016/10/18 23:44:59, Nico wrote:
> clang_tot_android not needed?

Added it.

I'm 90% sure a lot of this becomes obsolete when the bot is configured via mb.
The only thing that it really affects, I think, is env vars on gclient hooks,
and those seem to look right in the expectation?

Man, we have too many ways to configure a recipe...

[Nico, 2016-10-26 02:49:55]

lgtm, thanks!

https://codereview.chromium.org/2414253003/diff/120001/scripts/slave/recipes/...
File
scripts/slave/recipes/chromium.expected/full_chromium_fyi_ClangToTAndroid_x64.json
(right):

https://codereview.chromium.org/2414253003/diff/120001/scripts/slave/recipes/...
scripts/slave/recipes/chromium.expected/full_chromium_fyi_ClangToTAndroid_x64.json:136:
"LLVM_FORCE_HEAD_REVISION": "YES",
Ah you're right. This env var is the important bit that the recipe needs to set.

[Nico, 2016-10-27 17:10:07]

can this land?

[bpastene, 2016-10-27 17:21:32]

The CQ bit was checked by bpastene@chromium.org

[commit-bot: I haz the power, 2016-10-27 17:21:42]

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 17:58:17]

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

[commit-bot: I haz the power, 2016-10-27 17:58:19]

Try jobs failed on following builders:
  Build Presubmit on luci.infra.try (JOB_FAILED,
https://luci-milo.appspot.com/swarming/task/321ffb2499a1bb10)

[bpastene, 2016-10-27 19:17:54]

The CQ bit was checked by bpastene@chromium.org

[bpastene, 2016-10-27 19:17:54]

The patchset sent to the CQ was uploaded after l-g-t-m from thakis@chromium.org
Link to the patchset: https://codereview.chromium.org/2414253003/#ps140001
(title: "fix namesss")

[commit-bot: I haz the power, 2016-10-27 19:18:00]

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 19:27:12]

Description was changed from

==========
Add recipe config for android clang x86 tester.

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

to

==========
Add recipe config for android clang x86 tester.

BUG=534815

Committed:
https://chromium.googlesource.com/chromium/tools/build/+/f43b7753c335488f3e09...
==========

[commit-bot: I haz the power, 2016-10-27 19:27:14]

Committed patchset #8 (id:140001) as
https://chromium.googlesource.com/chromium/tools/build/+/f43b7753c335488f3e09...