[Intel-gfx] xf86-video-intel 2.5 release planning

[Brice Goglin]

Jesse Barnes wrote:
> It's my turn to handle a 2D driver release again, so I've been thinking about 
> what we should include (and remove!) in the 2.5 release.  I've created a 
> blocker bug for the release, 16926, and I'm interested in hearing about bugs 
> you think should be added

Here's some stuff from Debian.

For the record, we are freezing "Lenny" and we currently plan to ship
Xserver 1.4.2 with Intel 2.3.2 (with several patches already
backported). Xserver 1.4.99 and Intel 2.4.0 are in Debian experimental
but not used by many people so far.

Looking at the Debian BTS, I see 3 types of bugs that many people have seen:

1) ExaComposite rendering problems (mostly fonts) on i965. Switching to
XAA fixes the bug for everybody. Adding EXANoComposite seems to help
lots of people as well (some didn't try yet).
This is http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=451791 and
friends. It is supposed to be
https://bugs.freedesktop.org/show_bug.cgi?id=15278 and 15012 in
bugzilla. Your intel friends didn't reproduce these exact symptoms iirc,
but I am not sure that they tried with Xserver 1.4. We have Xserver
1.4.99 in Debian experimental now, so we might have some users try to
reproduce with it soon.

2) EXA is slower than XAA. We need to have all users switch to Xserver
1.5 before commenting on this one since EXA should be much better there.

3) Lots of lockups with pretty much no way to reproduce them, some are
related to EXA as well, some are not. We can't do much about them
without a way to reproduce them so...

By the way, it looks like some people get X failures after rebooting
after a crash/lockup, with the log saying "Existing errors found in
hardware state" or so. Is there a way to cleanup this hardware state at
startup so that users do not get confused by this message and ignore the
actual bug that caused it?


More recently, we got:
3) Force Enable Pipe A bug in 2.4.0 on i855 reported at
https://bugs.freedesktop.org/show_bug.cgi?id=16957
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=493096

4) Corruption when virtual > 2048 reported at
https://bugs.freedesktop.org/show_bug.cgi?id=16534

5) We have one occurence of
http://bugs.freedesktop.org/show_bug.cgi?id=16169 that our user would
like to see fixed in Debian Lenny. But we don't feel like backporting
this dsparb fix into 2.3.2. Any idea how to handle this?
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=484087

thanks,
Brice