Auto-rebaseline for r393557

Auto-rebaseline for r393557

https://chromium.googlesource.com/chromium/src/+/928d2159e

BUG=487344
TBR=liberato@chromium.org

Committed: https://crrev.com/8ae35b9b0e3a2e37d4d61b84db47ceef5c10e3a1
Cr-Commit-Position: refs/heads/master@{#393582}

[blink-rebaseline-bot, 2016-05-13 18:14:39]

Description was changed from

==========
Auto-rebaseline for r393557

https://chromium.googlesource.com/chromium/src/+/928d2159e

BUG=487344
TBR=liberato@chromium.org
==========

to

==========
Auto-rebaseline for r393557

https://chromium.googlesource.com/chromium/src/+/928d2159e

BUG=487344
TBR=liberato@chromium.org

Committed:
https://chromium.googlesource.com/chromium/src/+/8ae35b9b0e3a2e37d4d61b84db47...
==========

[commit-bot: I haz the power, 2016-05-13 18:14:39]

Description was changed from

==========
Auto-rebaseline for r393557

https://chromium.googlesource.com/chromium/src/+/928d2159e

BUG=487344
TBR=liberato@chromium.org
==========

to

==========
Auto-rebaseline for r393557

https://chromium.googlesource.com/chromium/src/+/928d2159e

BUG=487344
TBR=liberato@chromium.org

Committed: https://crrev.com/8ae35b9b0e3a2e37d4d61b84db47ceef5c10e3a1
Cr-Commit-Position: refs/heads/master@{#393582}
==========

[commit-bot: I haz the power, 2016-05-13 18:14:41]

Patchset 1 (id:??) landed as
https://crrev.com/8ae35b9b0e3a2e37d4d61b84db47ceef5c10e3a1
Cr-Commit-Position: refs/heads/master@{#393582}

[blink-rebaseline-bot, 2016-05-13 18:14:42]

Committed patchset #1 (id:1) manually as
8ae35b9b0e3a2e37d4d61b84db47ceef5c10e3a1 (presubmit successful).

[Ken Rockot(use gerrit already), 2016-05-15 00:56:37]

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

[Ken Rockot(use gerrit already), 2016-05-15 00:56:38]

This quite suspiciously coincides with new webkit_tests failures introduced
below:

https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7%20%28dbg%...
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9/builds...

I assume it would not be safe to revert this entire CL. Any ideas?

[chromium-reviews, 2016-05-15 17:37:45]

win7: it looks like a font rendering difference.  confusingly, the
'expected' image looks wrong; the font is too squished.  'actual' looks
right.  i remember that the rebaseline bots were having some issues lately,
don't remember the OS, and it had to do with fonts.  These just might need
a re-rebaseline for win7 to see if the 'expected' becomes sane again.

mac: not sure about the timeout.  will need to look further.

i'm not in a spot where i can do either of those immediately, but i can
take a look this afternoon.

thanks
-fl

On Sat, May 14, 2016 at 5:56 PM, <rockot@chromium.org> wrote:

> This quite suspiciously coincides with new webkit_tests failures introduced
> below:
>
>
>
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7%20%28dbg%...
>
>
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9/builds...
>
> I assume it would not be safe to revert this entire CL. Any ideas?
>
> https://codereview.chromium.org/1974373002/
>

-- 
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.

[blink-reviews, 2016-05-15 17:37:45]

win7: it looks like a font rendering difference.  confusingly, the
'expected' image looks wrong; the font is too squished.  'actual' looks
right.  i remember that the rebaseline bots were having some issues lately,
don't remember the OS, and it had to do with fonts.  These just might need
a re-rebaseline for win7 to see if the 'expected' becomes sane again.

mac: not sure about the timeout.  will need to look further.

i'm not in a spot where i can do either of those immediately, but i can
take a look this afternoon.

thanks
-fl

On Sat, May 14, 2016 at 5:56 PM, <rockot@chromium.org> wrote:

> This quite suspiciously coincides with new webkit_tests failures introduced
> below:
>
>
>
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7%20%28dbg%...
>
>
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9/builds...
>
> I assume it would not be safe to revert this entire CL. Any ideas?
>
> https://codereview.chromium.org/1974373002/
>

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

[wkorman, 2016-05-15 20:27:20]

The recent issue that I'm aware of involved Mac OS X 10.9.5 specific fonts.
Please take a look at the involved test expectations and the revision log for
the ones you feel have incorrect baselines. If you think re-rebaselining is the
right thing to do, go for it. If issue persists please file a bug on me and I
will investigate.

On 2016/05/15 at 17:37:45, blink-reviews wrote:
> win7: it looks like a font rendering difference.  confusingly, the
> 'expected' image looks wrong; the font is too squished.  'actual' looks
> right.  i remember that the rebaseline bots were having some issues lately,
> don't remember the OS, and it had to do with fonts.  These just might need
> a re-rebaseline for win7 to see if the 'expected' becomes sane again.
> 
> mac: not sure about the timeout.  will need to look further.
> 
> i'm not in a spot where i can do either of those immediately, but i can
> take a look this afternoon.
> 
> thanks
> -fl
> 
> On Sat, May 14, 2016 at 5:56 PM, <rockot@chromium.org> wrote:
> 
> > This quite suspiciously coincides with new webkit_tests failures introduced
> > below:
> >
> >
> >
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7%20%28dbg%...
> >
> >
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Mac10.9/builds...
> >
> > I assume it would not be safe to revert this entire CL. Any ideas?
> >
> > https://codereview.chromium.org/1974373002/
> >
> 
> -- 
> You received this message because you are subscribed to the Google Groups
"Blink Reviews" group.
> To unsubscribe from this group and stop receiving emails from it, send an
email to blink-reviews+unsubscribe@chromium.org.
>

[liberato (no reviews please), 2016-05-15 22:18:32]

On 2016/05/15 20:27:20, wkorman wrote:
> The recent issue that I'm aware of involved Mac OS X 10.9.5 specific fonts.
> Please take a look at the involved test expectations and the revision log for
> the ones you feel have incorrect baselines. If you think re-rebaselining is
the
> right thing to do, go for it. If issue persists please file a bug on me and I
> will investigate.

thanks for the info.  assuming that the revision log shows nothing interesting,
i'm going to try manually replacing the expectations with those from the build
bots.  those all look right after manual inspection.

is there any situation in which the rebaselined images are not expected to pass
on the build bots?  my initial reaction is no, but it's also a complicated
system that i don't understand.  my initial reaction might be an
oversimplification of the actual problem.

thanks
-fl

[hiroshige, 2016-05-17 13:54:59]

On 2016/05/15 22:18:32, liberato wrote:
> On 2016/05/15 20:27:20, wkorman wrote:
> > The recent issue that I'm aware of involved Mac OS X 10.9.5 specific fonts.
> > Please take a look at the involved test expectations and the revision log
for
> > the ones you feel have incorrect baselines. If you think re-rebaselining is
> the
> > right thing to do, go for it. If issue persists please file a bug on me and
I
> > will investigate.
> 
> thanks for the info.  assuming that the revision log shows nothing
interesting,
> i'm going to try manually replacing the expectations with those from the build
> bots.  those all look right after manual inspection.
> 
> is there any situation in which the rebaselined images are not expected to
pass
> on the build bots?  my initial reaction is no, but it's also a complicated
> system that i don't understand.  my initial reaction might be an
> oversimplification of the actual problem.
> 
> thanks
> -fl

Filed https://crbug.com/612442.