Add a new chromium.mojo master

Add a new chromium.mojo master.

This adds a new master to master1. It will run as one of the
main chromium waterfalls, and will build the Mojo services that
live in the chromium repo and run the tests for them.

These services can only be built using GN, and hence we cannot
simply build and test them on the normal builders.

The master config is generated from builders.py by the new buildbot-tool
script.


R=jamesr@chromium.org, aa@chromium.org, luqui@chromium.org
BUG=439281
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=293411

[iannucci, 2014-12-08 10:11:34]

iannucci@chromium.org changed reviewers:
+ iannucci@chromium.org

[iannucci, 2014-12-08 10:11:35]

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
File masters/master.chromium.mojo/buildbot (right):

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
masters/master.chromium.mojo/buildbot:1: #!/usr/bin/python
is this file actually used by anything?

[Dirk Pranke, 2014-12-08 18:41:47]

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
File masters/master.chromium.mojo/buildbot (right):

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
masters/master.chromium.mojo/buildbot:1: #!/usr/bin/python
On 2014/12/08 10:11:35, iannucci wrote:
> is this file actually used by anything?

I have no idea :).

[jamesr, 2014-12-08 20:28:07]

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
File masters/master.chromium.mojo/Makefile (right):

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
masters/master.chromium.mojo/Makefile:4: # DO NOT EDIT BY HAND!
where is the input to this generator?

[Dirk Pranke, 2014-12-08 20:46:56]

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
File masters/master.chromium.mojo/Makefile (right):

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
masters/master.chromium.mojo/Makefile:4: # DO NOT EDIT BY HAND!
On 2014/12/08 20:28:07, jamesr wrote:
> where is the input to this generator?

builders.json is the input. 

I can update the comment to reflect that.

[jamesr, 2014-12-08 20:58:47]

Yeah - that'd be useful.

The config lgtm as far as I can tell!

[agable, 2014-12-09 20:58:21]

agable@chromium.org changed reviewers:
+ agable@chromium.org

[agable, 2014-12-09 20:58:22]

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
File masters/master.chromium.mojo/builders.json (right):

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
masters/master.chromium.mojo/builders.json:24: "slaves": ["vm9999-m1"]
Is this a real hostname?

https://codereview.chromium.org/782043003/diff/1/tests/masters_test.py
File tests/masters_test.py (right):

https://codereview.chromium.org/782043003/diff/1/tests/masters_test.py#newcod...
tests/masters_test.py:216: 'master.chromium.mojo': 'ChromiumMojo',
I assume this can't be done by buildbot-tool yet.

[Dirk Pranke, 2014-12-09 21:03:43]

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
File masters/master.chromium.mojo/builders.json (right):

https://codereview.chromium.org/782043003/diff/1/masters/master.chromium.mojo...
masters/master.chromium.mojo/builders.json:24: "slaves": ["vm9999-m1"]
On 2014/12/09 20:58:22, agable wrote:
> Is this a real hostname?

No. I believe we're still waiting on real hostnames.

https://codereview.chromium.org/782043003/diff/1/tests/masters_test.py
File tests/masters_test.py (right):

https://codereview.chromium.org/782043003/diff/1/tests/masters_test.py#newcod...
tests/masters_test.py:216: 'master.chromium.mojo': 'ChromiumMojo',
On 2014/12/09 20:58:22, agable wrote:
> I assume this can't be done by buildbot-tool yet.

Correct.

[Dirk Pranke, 2014-12-10 01:19:50]

Patchset #3 (id:40001) has been deleted

[luqui, 2014-12-10 07:53:23]

lgtm % json -> py ast, and assuming you tested it locally, of course.

[Dirk Pranke, 2014-12-10 22:19:52]

Okay, this adds the actual recipe logic we'll want to run on these bots, in a
new 'chromium_mojo' recipe.

James or Aaron, can you take a look at the recipe just to confirm that it builds
and runs what you want?

Robbie, can you take a look at the changes in chromium_mojo.py, chromium_gn.py,
and the chromium module to make sure my refactoring is sane?

[jamesr, 2014-12-10 22:32:25]

chromium_mojo.py lgtm - seems fine

https://codereview.chromium.org/782043003/diff/80001/scripts/slave/recipes/ch...
File scripts/slave/recipes/chromium_mojo.py (right):

https://codereview.chromium.org/782043003/diff/80001/scripts/slave/recipes/ch...
scripts/slave/recipes/chromium_mojo.py:44: # TODO(dpranke): goma doesn't work on
windows GN builds yet.
interesting - we were using GN + goma on the windows mojo bots (when those were
active) and it seemed to work fine

[Dirk Pranke, 2014-12-10 22:35:31]

On 2014/12/10 22:32:25, jamesr wrote:
> interesting - we were using GN + goma on the windows mojo bots (when those
were
> active) and it seemed to work fine

I think it's a Win8 specific thing, and you're probably running win7. I'm
leaving it off everywhere for now,
as we're not building enough for it to be a huge win.

[Dirk Pranke, 2014-12-10 22:36:44]

I'm actually going to split the recipe-side work into a separate CL so we can
land it a bit earlier and minimize the amount of stuff changing at once.

See https://codereview.chromium.org/791733005/ .

[Dirk Pranke, 2014-12-10 22:43:16]

dpranke@chromium.org changed reviewers:
- aa@chromium.org, agable@chromium.org, luqui@chromium.org

[Dirk Pranke, 2014-12-17 00:08:34]

The CQ bit was checked by dpranke@chromium.org

[commit-bot: I haz the power, 2014-12-17 00:08:58]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/782043003/160001

[commit-bot: I haz the power, 2014-12-17 00:11:45]

Committed patchset #8 (id:160001) as
http://src.chromium.org/viewvc/chrome?view=rev&revision=293411