xvfb: set dpi to 96 (xvfb defaults to 100).

xvfb: set dpi to 96 (xvfb defaults to 100).

Tests assume that the DPI is effectively 96. Up until recently, this was
the case because chromium only supported 100% (corresponding to 96 dpi)
or 200% (corresponding to 192 dpi) scaling in its UI.

With https://codereview.chromium.org/1078423002/, we want to change that
to better reflect the hardware that many people have today. Applying
that CL without also changing the DPI results in tests failing because
of off-by-a-few-pixel errors.

BUG=477716, 143619
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=294979

[stapelberg, 2015-04-24 14:18:57]

stapelberg@google.com changed reviewers:
+ glider@chromium.org

[Alexander Potapenko, 2015-04-24 14:20:28]

On 2015/04/24 14:18:58, stapelberg wrote:

rubberstamp LGTM

[stapelberg, 2015-04-24 14:21:58]

The CQ bit was checked by stapelberg@google.com

[commit-bot: I haz the power, 2015-04-24 14:22:04]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1105903002/1

[commit-bot: I haz the power, 2015-04-24 14:26:10]

Committed patchset #1 (id:1) as
http://src.chromium.org/viewvc/chrome?view=rev&revision=294979

[pfeifer, 2015-04-24 14:59:51]

pfeifer@gmail.com changed reviewers:
+ pfeifer@gmail.com

[pfeifer, 2015-04-24 14:59:52]

(apologies in advance if the comment is not pertinent, I don't have the full
picture)

https://codereview.chromium.org/1105903002/diff/1/scripts/slave/xvfb.py
File scripts/slave/xvfb.py (right):

https://codereview.chromium.org/1105903002/diff/1/scripts/slave/xvfb.py#newco...
scripts/slave/xvfb.py:76: '-dpi', '96'],
With HiDPI becoming more popular, it's conceivable that one would want to write
tests for different DPI settings.

Is it worth making dpi an extra parameter to StartVirtualX which defaults to 96
?

[stapelberg, 2015-04-24 15:01:15]

On 2015/04/24 14:59:52, pfeifer wrote:
> (apologies in advance if the comment is not pertinent, I don't have the full
> picture)
> 
> https://codereview.chromium.org/1105903002/diff/1/scripts/slave/xvfb.py
> File scripts/slave/xvfb.py (right):
> 
>
https://codereview.chromium.org/1105903002/diff/1/scripts/slave/xvfb.py#newco...
> scripts/slave/xvfb.py:76: '-dpi', '96'],
> With HiDPI becoming more popular, it's conceivable that one would want to
write
> tests for different DPI settings.
> 
> Is it worth making dpi an extra parameter to StartVirtualX which defaults to
96
> ?

Possibly, but that’s for another time. I’ll leave that up to people who actually
want to port the existing hidpi tests that we have for OS X to Linux.