<html>
<head>
<base href="https://bugs.freedesktop.org/" />
</head>
<body>
<p>
<div>
<b><a class="bz_bug_link
bz_status_NEW "
title="NEW --- - broken on big-endian systems"
href="https://bugs.freedesktop.org/show_bug.cgi?id=66663#c3">Comment # 3</a>
on <a class="bz_bug_link
bz_status_NEW "
title="NEW --- - broken on big-endian systems"
href="https://bugs.freedesktop.org/show_bug.cgi?id=66663">bug 66663</a>
from <span class="vcard"><a class="email" href="mailto:kettenis@openbsd.org" title="Mark Kettenis <kettenis@openbsd.org>"> <span class="fn">Mark Kettenis</span></a>
</span></b>
<pre>Looks like I jumped to conclusions here. For the (EXA) accelerated case, the
problem is that UTS/DFS get bypassed for the "screen pixmap". That was the
right
thing to do when the front buffer had a surface reg allocated to it that made
sure
the required byte swapping was done in hardware. But now that this isn't the
case anymore, all pixmaps over 8bpp need byte swapping. See the attached
patch.
Of course this will leave the unaccelerated case broken. In that case the
RADEON_TILING_SURFACE flag really has to be set. But there should be no
interference with page flipping in that case.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are the assignee for the bug.</li>
</ul>
</body>
</html>