1182691 - Run Android 2.3 and Android 4.3 mochitest-chrome on trunk

Status: RESOLVED FIXED

(Release Engineering :: General, defect)

Description

[Geoff Brown [:gbrown]]

We can start running mochitest-chrome on Android 2.3 API9 opt, Android 4.3 API11+ opt, and Android 4.3 API11+ debug on all trunk trees.

Android mochitest-chrome is just like mochitest but --chrome is added to the runtestsremote.py arguments, and we can run all mochitest-chrome in a single chunk.

I think this should display in treeherder as M(c) (like b2g mochitest-chrome). M(oth) is another option (desktop includes mochitest-chrome in "other")...but we have no plans to run additional mochitest flavors on Android.

Comment 1

[Geoff Brown [:gbrown]]

Attachment: (wip) in-tree mozharness config update

Comment 2

[Geoff Brown [:gbrown]]

Attachment: (wip) mozharness config update

Comment 3

[Geoff Brown [:gbrown]]

Attachment: bug1182691-mh

I know jlund is in the middle of moving mozharness in-tree; I will check with him before pushing.

kmoir -- Will you be able to write the buildbot changes to trigger this new job?

Comment 4

[Geoff Brown [:gbrown]]

Attachment: update mozharness in-tree config

Currently --total-chunks is defined in both the mozharness config and the in-tree config. I'm removing it from the in-tree config because that applies to all mochitests, and now mochitest-chrome shares the mochitests config.

Comment 5

[Geoff Brown [:gbrown]]

Attachment: add mochitest-chrome to mozharness configs

Sorry -- attached wrong patch!

Comment 6

[Geoff Brown [:gbrown]]

Verified on try at https://treeherder.mozilla.org/#/jobs?repo=try&revision=13a2ee74657e with this hack to simulate buildbot: http://hg.mozilla.org/users/gbrown_mozilla.com/gbmozharness/rev/7de57f189f1c

Comment 7

[Kim Moir [:kmoir] ET]

Comment on attachment 8632843 [details] [diff] [review]
add mochitest-chrome to mozharness configs

Yes, I can write the buildbot config changes for this

Comment 8

[Geoff Brown [:gbrown]]

https://hg.mozilla.org/build/mozharness/rev/4addb10922bc

Comment 9

[Kim Moir [:kmoir] ET]

Attachment: bug1182691.patch

buildbot configs

Comment 10

[Kim Moir [:kmoir] ET]

Attachment: bug1182691builder.diff

builder diffs

Comment 11

[Kim Moir [:kmoir] ET]

Attachment: bug1182691-2.patch

initially attached wrong patch.  

The loadSkipConfig(BRANCHES,"mobile") had to be moved due to an ordering issue

Comment 12

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/a537c28456f4

Comment 13

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/a537c28456f4

Comment 14

[Kim Moir [:kmoir] ET]

Should have mentioned when I attached the patch that I verified that watch_pending.cfg doesn't need to be modified since it will just fall under mochitest

Comment 15

[Geoff Brown [:gbrown]]

Sorry, forgot to leave-open.

Comment 16

[Jordan Lund (:jlund)]

Comment on attachment 8633074 [details] [diff] [review]
bug1182691-2.patch

Review of attachment 8633074 [details] [diff] [review]:
-----------------------------------------------------------------

that's too bad about loadSkipConfig. Looks good to me anyway :)

::: mozilla-tests/mobile_config.py
@@ +2580,5 @@
>  # schedule instrumentation tests for pandas on ash and cedar
>  # https://bugzilla.mozilla.org/show_bug.cgi?id=1064010
>  remove_suite_from_slave_platform(BRANCHES, PLATFORMS, 'instrumentation', 'panda_android', branches_to_keep=['cedar'])
>  
> +# Bug 1182691 - Run Android 2.3 and Android 4.3 mochitest-chrome on trunk 

white space

Comment 17

[Kim Moir [:kmoir] ET]

Comment on attachment 8633074 [details] [diff] [review]
bug1182691-2.patch

fixed whitespace too

Comment 18

[Rail Aliiev [:rail]]

https://hg.mozilla.org/build/mozharness/rev/94d72bb2ba9b

Comment 19

[[e-mail] (:bhearsum)]

In production: https://hg.mozilla.org/build/buildbot-configs/rev/fa0bbb6e7b44

Comment 20

[[e-mail] (:bhearsum)]

In production: https://hg.mozilla.org/build/buildbot-configs/rev/fa0bbb6e7b44

Comment 21

[Kim Moir [:kmoir] ET]

verified on treeherder tests are green on m-c