Test android tool for building makefile.

Test android tool for building makefile.

Depends on https://code.google.com/p/skia/source/detail?r=13344

BUG=skia:1975

[borenet, 2014-02-06 22:44:03]

This looks fine, but the CL needs to include the updated factory configuration
expectation files.  You can generate them by running:
$ python tools/tests/factory_configuration/factory_configuration_test.py
--rebaseline

For the record, I think the Housekeeping bot is the wrong place for these.  I'd
rather have a "self-tests" bot which runs on the main master, and reserve the
housekeepers for purely housekeeping tasks, rather than performing any tests. 
But that shouldn't block this.

[scroggo, 2014-02-06 23:00:20]

On 2014/02/06 22:44:03, borenet wrote:
> This looks fine, but the CL needs to include the updated factory configuration
> expectation files.  You can generate them by running:
> $ python tools/tests/factory_configuration/factory_configuration_test.py
> --rebaseline

Done.

> 
> For the record, I think the Housekeeping bot is the wrong place for these. 
I'd
> rather have a "self-tests" bot which runs on the main master, and reserve the
> housekeepers for purely housekeeping tasks, rather than performing any tests. 
> But that shouldn't block this.

Added a TODO. Who should I assign as the owner? (Should I create a bug?)

[borenet, 2014-02-06 23:04:21]

On 2014/02/06 23:00:20, scroggo wrote:
> On 2014/02/06 22:44:03, borenet wrote:
> > This looks fine, but the CL needs to include the updated factory
configuration
> > expectation files.  You can generate them by running:
> > $ python tools/tests/factory_configuration/factory_configuration_test.py
> > --rebaseline
> 
> Done.
> 
> > 
> > For the record, I think the Housekeeping bot is the wrong place for these. 
> I'd
> > rather have a "self-tests" bot which runs on the main master, and reserve
the
> > housekeepers for purely housekeeping tasks, rather than performing any
tests. 
> > But that shouldn't block this.
> 
> Added a TODO. Who should I assign as the owner? (Should I create a bug?)

Thanks.  If you'd like to file a bug, just assign it to me.  LGTM

[scroggo, 2014-02-06 23:12:37]

On 2014/02/06 23:04:21, borenet wrote:
> On 2014/02/06 23:00:20, scroggo wrote:
> > On 2014/02/06 22:44:03, borenet wrote:
> > > This looks fine, but the CL needs to include the updated factory
> configuration
> > > expectation files.  You can generate them by running:
> > > $ python tools/tests/factory_configuration/factory_configuration_test.py
> > > --rebaseline
> > 
> > Done.
> > 
> > > 
> > > For the record, I think the Housekeeping bot is the wrong place for these.

> > I'd
> > > rather have a "self-tests" bot which runs on the main master, and reserve
> the
> > > housekeepers for purely housekeeping tasks, rather than performing any
> tests. 
> > > But that shouldn't block this.
> > 
> > Added a TODO. Who should I assign as the owner? (Should I create a bug?)
> 
> Thanks.  If you'd like to file a bug, just assign it to me.  LGTM

Done: https://code.google.com/p/skia/issues/detail?id=2144

[epoger, 2014-02-07 14:49:40]

LGTM with or without one small change

Thanks for adding this to the bots, Leon!

https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
File master/skia_master_scripts/housekeeping_percommit_factory.py (right):

https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
master/skia_master_scripts/housekeeping_percommit_factory.py:34: # TODO
(borenet): Move these to a self-tests bot.
Please add a link to the bug, http://skbug.com/2144

Also, I think our TODOs usually don't have a space before the parentheses, so
this might be easier for Eric to search for:

TODO(borenet): blah

[scroggo, 2014-02-10 14:30:17]

On 2014/02/07 14:49:40, epoger wrote:
> LGTM with or without one small change
> 
> Thanks for adding this to the bots, Leon!
> 
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> File master/skia_master_scripts/housekeeping_percommit_factory.py (right):
> 
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> master/skia_master_scripts/housekeeping_percommit_factory.py:34: # TODO
> (borenet): Move these to a self-tests bot.
> Please add a link to the bug, http://skbug.com/2144

Done.

> 
> Also, I think our TODOs usually don't have a space before the parentheses, so
> this might be easier for Eric to search for:
> 
> TODO(borenet): blah

Done.

[borenet, 2014-02-10 14:33:14]

On 2014/02/10 14:30:17, scroggo wrote:
> On 2014/02/07 14:49:40, epoger wrote:
> > LGTM with or without one small change
> > 
> > Thanks for adding this to the bots, Leon!
> > 
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > File master/skia_master_scripts/housekeeping_percommit_factory.py (right):
> > 
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > master/skia_master_scripts/housekeeping_percommit_factory.py:34: # TODO
> > (borenet): Move these to a self-tests bot.
> > Please add a link to the bug, http://skbug.com/2144
> 
> Done.
> 
> > 
> > Also, I think our TODOs usually don't have a space before the parentheses,
so
> > this might be easier for Eric to search for:
> > 
> > TODO(borenet): blah
> 
> Done.

LGTM.  Elliot: for the record, I'm most likely to just grep for my username.

[scroggo, 2014-02-10 14:38:19]

On 2014/02/10 14:33:14, borenet wrote:
> On 2014/02/10 14:30:17, scroggo wrote:
> > On 2014/02/07 14:49:40, epoger wrote:
> > > LGTM with or without one small change
> > > 
> > > Thanks for adding this to the bots, Leon!
> > > 
> > >
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > > File master/skia_master_scripts/housekeeping_percommit_factory.py (right):
> > > 
> > >
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > > master/skia_master_scripts/housekeeping_percommit_factory.py:34: # TODO
> > > (borenet): Move these to a self-tests bot.
> > > Please add a link to the bug, http://skbug.com/2144
> > 
> > Done.
> > 
> > > 
> > > Also, I think our TODOs usually don't have a space before the parentheses,
> so
> > > this might be easier for Eric to search for:
> > > 
> > > TODO(borenet): blah
> > 
> > Done.
> 
> LGTM.  Elliot: for the record, I'm most likely to just grep for my username.

I'm getting a strange message when I try to submit (using git cl dcommit on a
git-svn checkout):

ERROR from SVN:
Filesystem has no item: Could not resolve path
/buildbot/master/skia_master_scripts/housekeeping_percommit_factory.py
ERROR: Not all changes have been committed into SVN.
Please see the above messages for details.

I'm not sure what the "above message" is, but here is a presubmit warning:

** Presubmit Warnings **
One or more buildbot tests failed.

***************
Searching for tests.
Found 9 tests.
Traceback (most recent call last):
  File
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/csv_comparer_test.py",
line 14, in <module>
    import csv_comparer
  File
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/csv_comparer.py",
line 14, in <module>
    from django.template import loader
ImportError: No module named django.template

Traceback (most recent call last):
  File
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/telemetry_master_scripts/json_summary_combiner_test.py",
line 14, in <module>
    import json_summary_combiner
  File
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/telemetry_master_scripts/json_summary_combiner.py",
line 15, in <module>
    from django.template import loader
ImportError: No module named django.template

Failed 2 of 1.
***************

Presubmit checks took 18.4s to calculate.


(Is this something I need to fix before submitting?)

In case it's related, when I tried to do a sync (git svn fetch && git svn
rebase), I get the following mysterious message:

W: Filesystem has no item: '/svn/!svn/bc/13379/buildbot' path not found at
/usr/share/perl5/Git/SVN.pm line 153

W: Ignoring error from SVN, path probably does not exist: (160013): Filesystem
has no item: '/svn/!svn/bc/13379/buildbot' path not found
W: Do not be alarmed at the above message git-svn is just searching aggressively
for old history.
This may take a while on large repositories
Current branch master is up to date.

Has buildbot/ moved?

[borenet, 2014-02-10 14:41:57]

On 2014/02/10 14:38:19, scroggo wrote:
> On 2014/02/10 14:33:14, borenet wrote:
> > On 2014/02/10 14:30:17, scroggo wrote:
> > > On 2014/02/07 14:49:40, epoger wrote:
> > > > LGTM with or without one small change
> > > > 
> > > > Thanks for adding this to the bots, Leon!
> > > > 
> > > >
> > >
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > > > File master/skia_master_scripts/housekeeping_percommit_factory.py
(right):
> > > > 
> > > >
> > >
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > > > master/skia_master_scripts/housekeeping_percommit_factory.py:34: # TODO
> > > > (borenet): Move these to a self-tests bot.
> > > > Please add a link to the bug, http://skbug.com/2144
> > > 
> > > Done.
> > > 
> > > > 
> > > > Also, I think our TODOs usually don't have a space before the
parentheses,
> > so
> > > > this might be easier for Eric to search for:
> > > > 
> > > > TODO(borenet): blah
> > > 
> > > Done.
> > 
> > LGTM.  Elliot: for the record, I'm most likely to just grep for my username.
> 
> I'm getting a strange message when I try to submit (using git cl dcommit on a
> git-svn checkout):
> 
> ERROR from SVN:
> Filesystem has no item: Could not resolve path
> /buildbot/master/skia_master_scripts/housekeeping_percommit_factory.py
> ERROR: Not all changes have been committed into SVN.
> Please see the above messages for details.
> 
> I'm not sure what the "above message" is, but here is a presubmit warning:
> 
> ** Presubmit Warnings **
> One or more buildbot tests failed.
> 
> ***************
> Searching for tests.
> Found 9 tests.
> Traceback (most recent call last):
>   File
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/csv_comparer_test.py",
> line 14, in <module>
>     import csv_comparer
>   File
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/csv_comparer.py",
> line 14, in <module>
>     from django.template import loader
> ImportError: No module named django.template
> 
> Traceback (most recent call last):
>   File
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/telemetry_master_scripts/json_summary_combiner_test.py",
> line 14, in <module>
>     import json_summary_combiner
>   File
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/telemetry_master_scripts/json_summary_combiner.py",
> line 15, in <module>
>     from django.template import loader
> ImportError: No module named django.template
> 
> Failed 2 of 1.
> ***************
> 
> Presubmit checks took 18.4s to calculate.
> 
> 
> (Is this something I need to fix before submitting?)
> 
> In case it's related, when I tried to do a sync (git svn fetch && git svn
> rebase), I get the following mysterious message:
> 
> W: Filesystem has no item: '/svn/!svn/bc/13379/buildbot' path not found at
> /usr/share/perl5/Git/SVN.pm line 153
> 
> W: Ignoring error from SVN, path probably does not exist: (160013): Filesystem
> has no item: '/svn/!svn/bc/13379/buildbot' path not found
> W: Do not be alarmed at the above message git-svn is just searching
aggressively
> for old history.
> This may take a while on large repositories
> Current branch master is up to date.
> 
> Has buildbot/ moved?

We're using pure git now.  The CQ works for buildbot though.

[scroggo, 2014-02-10 14:59:38]

On 2014/02/10 14:41:57, borenet wrote:
> On 2014/02/10 14:38:19, scroggo wrote:
> > On 2014/02/10 14:33:14, borenet wrote:
> > > On 2014/02/10 14:30:17, scroggo wrote:
> > > > On 2014/02/07 14:49:40, epoger wrote:
> > > > > LGTM with or without one small change
> > > > > 
> > > > > Thanks for adding this to the bots, Leon!
> > > > > 
> > > > >
> > > >
> > >
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > > > > File master/skia_master_scripts/housekeeping_percommit_factory.py
> (right):
> > > > > 
> > > > >
> > > >
> > >
> >
>
https://codereview.chromium.org/147953006/diff/120001/master/skia_master_scri...
> > > > > master/skia_master_scripts/housekeeping_percommit_factory.py:34: #
TODO
> > > > > (borenet): Move these to a self-tests bot.
> > > > > Please add a link to the bug, http://skbug.com/2144
> > > > 
> > > > Done.
> > > > 
> > > > > 
> > > > > Also, I think our TODOs usually don't have a space before the
> parentheses,
> > > so
> > > > > this might be easier for Eric to search for:
> > > > > 
> > > > > TODO(borenet): blah
> > > > 
> > > > Done.
> > > 
> > > LGTM.  Elliot: for the record, I'm most likely to just grep for my
username.
> > 
> > I'm getting a strange message when I try to submit (using git cl dcommit on
a
> > git-svn checkout):
> > 
> > ERROR from SVN:
> > Filesystem has no item: Could not resolve path
> > /buildbot/master/skia_master_scripts/housekeeping_percommit_factory.py
> > ERROR: Not all changes have been committed into SVN.
> > Please see the above messages for details.
> > 
> > I'm not sure what the "above message" is, but here is a presubmit warning:
> > 
> > ** Presubmit Warnings **
> > One or more buildbot tests failed.
> > 
> > ***************
> > Searching for tests.
> > Found 9 tests.
> > Traceback (most recent call last):
> >   File
> >
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/csv_comparer_test.py",
> > line 14, in <module>
> >     import csv_comparer
> >   File
> >
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/csv_comparer.py",
> > line 14, in <module>
> >     from django.template import loader
> > ImportError: No module named django.template
> > 
> > Traceback (most recent call last):
> >   File
> >
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/telemetry_master_scripts/json_summary_combiner_test.py",
> > line 14, in <module>
> >     import json_summary_combiner
> >   File
> >
>
"/usr/local/ssd/skia/buildbot/compute_engine_scripts/telemetry/telemetry_master_scripts/json_summary_combiner.py",
> > line 15, in <module>
> >     from django.template import loader
> > ImportError: No module named django.template
> > 
> > Failed 2 of 1.
> > ***************
> > 
> > Presubmit checks took 18.4s to calculate.
> > 
> > 
> > (Is this something I need to fix before submitting?)
> > 
> > In case it's related, when I tried to do a sync (git svn fetch && git svn
> > rebase), I get the following mysterious message:
> > 
> > W: Filesystem has no item: '/svn/!svn/bc/13379/buildbot' path not found at
> > /usr/share/perl5/Git/SVN.pm line 153
> > 
> > W: Ignoring error from SVN, path probably does not exist: (160013):
Filesystem
> > has no item: '/svn/!svn/bc/13379/buildbot' path not found
> > W: Do not be alarmed at the above message git-svn is just searching
> aggressively
> > for old history.
> > This may take a while on large repositories
> > Current branch master is up to date.
> > 
> > Has buildbot/ moved?
> 
> We're using pure git now.  The CQ works for buildbot though.

Woohoo!

Prepare for several messages telling you the status.

[scroggo, 2014-02-10 14:59:43]

The CQ bit was checked by scroggo@google.com

[scroggo, 2014-02-10 19:53:17]

On 2014/02/10 14:59:43, scroggo wrote:
> The CQ bit was checked by mailto:scroggo@google.com

Any idea why this hasn't done anything? I would have expected it to attempt to
commit by now (or at least have done a try?).

[rmistry, 2014-02-10 19:56:54]

On 2014/02/10 19:53:17, scroggo wrote:
> On 2014/02/10 14:59:43, scroggo wrote:
> > The CQ bit was checked by mailto:scroggo@google.com
> 
> Any idea why this hasn't done anything? I would have expected it to attempt to
> commit by now (or at least have done a try?).

Ah yes, it is because the buildbot CQ only accepts Git base urls and not SVN.

[scroggo, 2014-02-10 19:58:04]

On 2014/02/10 19:56:54, rmistry wrote:
> On 2014/02/10 19:53:17, scroggo wrote:
> > On 2014/02/10 14:59:43, scroggo wrote:
> > > The CQ bit was checked by mailto:scroggo@google.com
> > 
> > Any idea why this hasn't done anything? I would have expected it to attempt
to
> > commit by now (or at least have done a try?).
> 
> Ah yes, it is because the buildbot CQ only accepts Git base urls and not SVN.

Haha, I was afraid that might be the case. I'll check out using Git and try
again.

[borenet, 2014-02-10 19:58:10]

On 2014/02/10 19:56:54, rmistry wrote:
> On 2014/02/10 19:53:17, scroggo wrote:
> > On 2014/02/10 14:59:43, scroggo wrote:
> > > The CQ bit was checked by mailto:scroggo@google.com
> > 
> > Any idea why this hasn't done anything? I would have expected it to attempt
to
> > commit by now (or at least have done a try?).
> 
> Ah yes, it is because the buildbot CQ only accepts Git base urls and not SVN.

Ok - try doing a pure-git checkout and then patching this CL in.

[scroggo, 2014-02-10 21:49:05]

The CQ bit was checked by scroggo@google.com

[scroggo, 2014-02-11 00:49:45]

On 2014/02/10 21:49:05, scroggo wrote:
> The CQ bit was checked by mailto:scroggo@google.com

I used a Git checkout, and it still seems to be doing nothing. (If it makes any
difference, I used git cl patch in my Git checkout, and then uploaded it as
patch set 5.)

[epoger, 2014-02-11 03:13:06]

On 2014/02/11 00:49:45, scroggo wrote:
> On 2014/02/10 21:49:05, scroggo wrote:
> > The CQ bit was checked by mailto:scroggo@google.com
> 
> I used a Git checkout, and it still seems to be doing nothing. (If it makes
any
> difference, I used git cl patch in my Git checkout, and then uploaded it as
> patch set 5.)

I recommend uploading it as a totally new CL, with TBR=borenet in the
description, and then checking the commit box.  Just to get a fresh start...

[borenet, 2014-02-11 13:02:22]

On 2014/02/11 03:13:06, epoger wrote:
> On 2014/02/11 00:49:45, scroggo wrote:
> > On 2014/02/10 21:49:05, scroggo wrote:
> > > The CQ bit was checked by mailto:scroggo@google.com
> > 
> > I used a Git checkout, and it still seems to be doing nothing. (If it makes
> any
> > difference, I used git cl patch in my Git checkout, and then uploaded it as
> > patch set 5.)
> 
> I recommend uploading it as a totally new CL, with TBR=borenet in the
> description, and then checking the commit box.  Just to get a fresh start...

Yeah, I don't think the baseurl changes in Rietveld after the initial upload.

[scroggo, 2014-02-11 15:14:01]

On 2014/02/11 13:02:22, borenet wrote:
> On 2014/02/11 03:13:06, epoger wrote:
> > On 2014/02/11 00:49:45, scroggo wrote:
> > > On 2014/02/10 21:49:05, scroggo wrote:
> > > > The CQ bit was checked by mailto:scroggo@google.com
> > > 
> > > I used a Git checkout, and it still seems to be doing nothing. (If it
makes
> > any
> > > difference, I used git cl patch in my Git checkout, and then uploaded it
as
> > > patch set 5.)
> > 
> > I recommend uploading it as a totally new CL, with TBR=borenet in the
> > description, and then checking the commit box.  Just to get a fresh start...
> 
> Yeah, I don't think the baseurl changes in Rietveld after the initial upload.

Lesson learned. Uploaded from git and submitted from
https://codereview.chromium.org/159793002