Mark kasko as a hard dependency since it depends on kasko_features which generates a header file.

Mark kasko as a hard dependency since it depends on kasko_features which generates a header file.

BUG=611849
Committed: https://crrev.com/a81d8c56784867ada51a97a9dbac4ffd0200eb74
Cr-Commit-Position: refs/heads/master@{#393891}

[Evan Stade, 2016-05-13 22:44:18]

Description was changed from

==========
Mark kasko_features as a hard dependency since it generates a header
file.

BUG=611849
==========

to

==========
Mark kasko as a hard dependency since it depends on kasko_features which
generates a header file.

BUG=611849
==========

[Evan Stade, 2016-05-13 22:44:18]

estade@chromium.org changed reviewers:
+ dpranke@chromium.org, pmonette@chromium.org

[Evan Stade, 2016-05-13 22:45:01]

(this is a stab in the dark)

[Evan Stade, 2016-05-13 22:48:34]

estade@chromium.org changed reviewers:
+ blundell@chromium.org

[Dirk Pranke, 2016-05-13 23:13:59]

dpranke@chromium.org changed reviewers:
+ scottmg@chromium.org

[Dirk Pranke, 2016-05-13 23:14:00]

lgtm. 

+scottmg ?

[scottmg, 2016-05-13 23:24:27]

I thought it was the dependent that had to be the hard_dependency (i.e.
kasko_features)?

[scottmg, 2016-05-13 23:25:16]

On 2016/05/13 23:24:27, scottmg wrote:
> I thought it was the dependent that had to be the hard_dependency (i.e.
> kasko_features)?

oh, but it's export_dependent_settings'ing kasko_features? ok, seems.. plausible
enough to lgtm then.

[Evan Stade, 2016-05-14 00:47:02]

The CQ bit was checked by estade@chromium.org

[commit-bot: I haz the power, 2016-05-14 00:47:33]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1976113003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1976113003/20001

[commit-bot: I haz the power, 2016-05-14 00:55:07]

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

[commit-bot: I haz the power, 2016-05-14 00:55:08]

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

[Evan Stade, 2016-05-16 18:05:59]

ping pmonette for owners

[Patrick Monette, 2016-05-16 18:09:08]

lgtm

[Evan Stade, 2016-05-16 18:12:39]

The CQ bit was checked by estade@chromium.org

[commit-bot: I haz the power, 2016-05-16 18:13:24]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1976113003/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1976113003/20001

[commit-bot: I haz the power, 2016-05-16 19:12:02]

Description was changed from

==========
Mark kasko as a hard dependency since it depends on kasko_features which
generates a header file.

BUG=611849
==========

to

==========
Mark kasko as a hard dependency since it depends on kasko_features which
generates a header file.

BUG=611849
==========

[commit-bot: I haz the power, 2016-05-16 19:12:03]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-05-16 19:13:26]

Description was changed from

==========
Mark kasko as a hard dependency since it depends on kasko_features which
generates a header file.

BUG=611849
==========

to

==========
Mark kasko as a hard dependency since it depends on kasko_features which
generates a header file.

BUG=611849

Committed: https://crrev.com/a81d8c56784867ada51a97a9dbac4ffd0200eb74
Cr-Commit-Position: refs/heads/master@{#393891}
==========

[commit-bot: I haz the power, 2016-05-16 19:13:27]

Patchset 2 (id:??) landed as
https://crrev.com/a81d8c56784867ada51a97a9dbac4ffd0200eb74
Cr-Commit-Position: refs/heads/master@{#393891}

[blundell, 2016-05-17 08:16:38]

https://codereview.chromium.org/1976113003/diff/20001/third_party/kasko/kasko...
File third_party/kasko/kasko.gyp (right):

https://codereview.chromium.org/1976113003/diff/20001/third_party/kasko/kasko...
third_party/kasko/kasko.gyp:46: 'hard_dependency': 1,
I don't think that this will fix the problem because lines 66-68 already export
the fact that kasko_features is a hard dependency, so the problem is somewhere
else in the dependency chain.

I think that the crash_component target should depend directly on kasko_features
and do export_dependent_settings on it, since it includes kasko_features.h
directly in crashpad.h. I'm not sure that our (meta)-build systems make any
guarantees when there are missing dependencies in the dependency graph (e.g.,
the fact that crash_component depends directly on kasko_features).

As a sanity check, I assume that you're sure that the failing builds are GYP and
not GN builds?

[Evan Stade, 2016-05-17 17:25:28]

On 2016/05/17 08:16:38, blundell wrote:
>
https://codereview.chromium.org/1976113003/diff/20001/third_party/kasko/kasko...
> File third_party/kasko/kasko.gyp (right):
> 
>
https://codereview.chromium.org/1976113003/diff/20001/third_party/kasko/kasko...
> third_party/kasko/kasko.gyp:46: 'hard_dependency': 1,
> I don't think that this will fix the problem because lines 66-68 already
export
> the fact that kasko_features is a hard dependency, so the problem is somewhere
> else in the dependency chain.
> 
> I think that the crash_component target should depend directly on
kasko_features
> and do export_dependent_settings on it, since it includes kasko_features.h
> directly in crashpad.h. I'm not sure that our (meta)-build systems make any
> guarantees when there are missing dependencies in the dependency graph (e.g.,
> the fact that crash_component depends directly on kasko_features).

As I stated, stab in the dark :) Your argument sounds plausible.

> 
> As a sanity check, I assume that you're sure that the failing builds are GYP
and
> not GN builds?

the bot linked in the bug has these lines in the update logs:

python build\gyp_chromium -G output_dir=out
Enabled Psyco JIT.
Updating projects from gyp files...