[cairo] Render EXTEND_REPEAT bug is back...

Behdad Esfahbod behdad at behdad.org
Sun Jan 27 19:45:28 PST 2008

On Sun, 2008-01-27 at 14:34 -0500, Owen Taylor wrote:
> > > Are you sure it's really *the* old buggy repeat bug and not another  
> > > bug
> > > that the same work-around code works around?
> > 
> > I'm not -- the symptoms look identical, but I don't know if the  
> > underlying cause is the same.  The mozilla bug report got updated; it  
> > doesn't seem to be limited to the radeon driver, people using sis and  
> > i810 at least are seeing it.  Do we have a testcase for this?  I can't  
> > remember if it's in the testsuite since it's platform-specific, but it  
> > would be good to be able to reproduce it without firefox.  If we  
> > don't, I can put one together tomorrow.
> The original bug report on the X server:
>  https://bugs.freedesktop.org/show_bug.cgi?id=3566
> The patch attachment is no longer available (all
> bugzilla.freedesktop.org attachments got scrambled some years ago), but
> as you can see from the description, that bug had to do with a very
> specific logic error in the core RENDER framebuffer code. I haven't
> looked at the current version of that file, but it seems improbable to
> me that that particular logic error was reintroduced.

That patch is here:

The bug doesn't exist in pixman these days as it's been switched over to
bit_image_t type that doesn't have x/y.  It's probably in the xserver
code calling pixman now.

> So, I think the first step here is to track down and figure out what the
> current X server bug you are seeing is, and what exactly is the scope of
> it. Before that, trying to figure out how to work around it seems like a
> risky and uncertain proposition.


> - Owen


"Those who would give up Essential Liberty to purchase a little
 Temporary Safety, deserve neither Liberty nor Safety."
        -- Benjamin Franklin, 1759

More information about the cairo mailing list