[cairo] xcb is passing, xlib? Re: What does it take to get a make check to pass with the xcb target, using CAIRO_REF_DIR?

[Adrian Johnson]

On 23/07/16 08:29, darxus at chaosreigns.com wrote:
> Adrian, did this work with the xlib target when you committed it?

I only tested with image, pdf, and ps.

> record-neg-extents is what's currently preventing the xlib target from
> passing, with regenerated references.

I just ran that test with xcb. The output looks fine. There are just a
few pixel differences so probably just needs a ref image refresh.

> This test currently fails for the commit where it was added (with
> references regenerated from the same commit):

Adding a new test will often expose latent bugs.

> commit 9fff6f0be3ed1a7af474ddfd8e384eb0376093c2
> Author:     Adrian Johnson <ajohnson at redneon.com>
> AuthorDate: Sat Jun 4 14:43:43 2016 +0930
> Commit:     Adrian Johnson <ajohnson at redneon.com>
> CommitDate: Sun Jun 5 20:43:36 2016 +0930
> 
>     test: add record-neg-extents
>     
>     Test case for bug 89232 - painting a recording surface to a
>     pdf/ps surface omits objects on the recording surface with negative
>     coordinates even though the pattern matrix has transformed the objects
>     to within the page extents.
>     
>     The image surface also fails when the recording surface is bounded.
> 
> 
> xlib (argb32): 1 failed - record-neg-extents-bounded
> xlib (rgb24): 1 failed - record-neg-extents-bounded
> 
> 
> The referenced bug: https://bugs.freedesktop.org/show_bug.cgi?id=89232
> 
> On 07/21, darxus at chaosreigns.com wrote:
>> Yay, xcb target passes (at least with regenerated references):
>> https://bugs.freedesktop.org/show_bug.cgi?id=84330
>>
>> image16 and image also appear to be passing.
>>
>> The one I'm most curious about now is xlib:
>>
>>    1062 xlib PASS
>>       4 xlib FAIL
>>       2 xlib XFAIL
>>
>> $ cat *.log | grep ^TEST: | grep 'xlib ' | grep ' FAIL'
>> TEST: record-neg-extents-bounded TARGET: xlib FORMAT: argb32 OFFSET: 0 SCALE: 1 SIMILAR: 0 RESULT: FAIL
>> TEST: record-neg-extents-bounded TARGET: xlib FORMAT: rgb24 OFFSET: 0 SCALE: 1 SIMILAR: 0 RESULT: FAIL
>> TEST: record-neg-extents-unbounded TARGET: xlib FORMAT: argb32 OFFSET: 0 SCALE: 1 SIMILAR: 0 RESULT: FAIL
>> TEST: record-neg-extents-unbounded TARGET: xlib FORMAT: rgb24 OFFSET: 0 SCALE: 1 SIMILAR: 0 RESULT: FAIL
>>
>> http://www.chaosreigns.com/tmp/cairo/xlib/record-neg-extents-bounded.ref.png
>> http://www.chaosreigns.com/tmp/cairo/xlib/record-neg-extents-bounded.xlib.argb32.out.png
>> http://www.chaosreigns.com/tmp/cairo/xlib/record-neg-extents-bounded.xlib.rgb24.out.png
>>
>> http://www.chaosreigns.com/tmp/cairo/xlib/record-neg-extents-unbounded.ref.png
>> http://www.chaosreigns.com/tmp/cairo/xlib/record-neg-extents-unbounded.xlib.argb32.out.png
>> http://www.chaosreigns.com/tmp/cairo/xlib/record-neg-extents-unbounded.xlib.rgb24.out.png
>>
>> Yup, something pretty wrong is happening there.
>> -- 
>> cairo mailing list
>> cairo at cairographics.org
>> https://lists.cairographics.org/mailman/listinfo/cairo
>