<div dir="ltr">I'm afraid I'm not going to help much with the tests. I had a short look and couldn't figure it out, but I'm going mostly offline for the next two weeks, so I'll not be able to spend more time on it soon. So the best option is to disable the functionality for now, as Uli suggested.<div><br></div><div>Lukas</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Sep 16, 2014 at 9:14 AM, Uli Schlachter <span dir="ltr"><<a href="mailto:psychon@znc.in" target="_blank">psychon@znc.in</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi & thanks for running these weekly tests,<br>
<br>
Am 16.09.2014 um 08:49 schrieb Bryce Harrington:<br>
> git git git git git git<br>
> 05/03 06/09 07/21 08/04 08/29 09/12<br>
> -------------------------------------------------------------------<br>
> Tests run: 24838 24813 24839 24838 24838 24838<br>
> Passed: 15774 16015 16024 15889 15989 15808<br>
> Failed: 6476 6226 6228 6372 6272 6451<br>
> Expected Failed: 332 315 332 332 332 332<br>
> Error: 1 1 0 0 0 0<br>
> Crashed: 32 33 32 22 22 24<br>
> Untested: 2223 2223 2223 2223 2223 2223<br>
> Total: 24838 24813 24839 24838 24838 24838<br>
><br>
> Here's the tests that changed results since the previous run:<br>
><br>
> <a href="http://www.bryceharrington.org/Cairo/changes-20140829-to-20140912-trent.txt" target="_blank">http://www.bryceharrington.org/Cairo/changes-20140829-to-20140912-trent.txt</a><br>
><br>
> 102 xcb and 70 xlib tests failed here, including a bunch of<br>
> *twin-antialias-* tests. Anyone have ideas on these?<br>
<br>
There weren't many changes to cairo-xcb and git bisect confirms:<br>
<br>
e691d242d592a8556e25659fb091a2031abee4c9 is the first bad commit<br>
commit e691d242d592a8556e25659fb091a2031abee4c9<br>
Author: Lukáš Lalinský <<a href="mailto:lukas@oxygene.sk">lukas@oxygene.sk</a>><br>
Date: Wed Sep 3 22:53:55 2014 +0200<br>
<br>
xcb: Initialize font options from Xft resources<br>
<br>
There is a similar code in the Xlib backend. The logic here is the same, but<br>
XCB doesn't support X resources directly, so there is some custom code<br>
to get and parse the resources from the root window.<br>
<br>
Signed-off-by: Lukáš Lalinský <<a href="mailto:lukas@oxygene.sk">lukas@oxygene.sk</a>><br>
Reviewed-by: Uli Schlachter <<a href="mailto:psychon@znc.in">psychon@znc.in</a>><br>
<br>
The attached patch makes these tests pass again. Both hunks are required. The<br>
first one fixes a shameful oversight (whoops), the second one effectively<br>
reverts the above patch, which is not what we want.<br>
<br>
Something is merging font options wrongly? Why does this code work for<br>
cairo-xlib? I don't know...<br>
<br>
For the xlib backend: Are you running with --enable-xlib-xcb? Aka "are you<br>
testing xlib at all?".<br>
<br>
Not much time right now (or in the near future), so can't look too closely into this<br>
<br>
Cheers,<br>
Uli<br>
<span class="HOEnZb"><font color="#888888">--<br>
“I’m Olaf and I like warm hugs.”<br>
</font></span></blockquote></div><br></div>