Bugzilla Bug 214134

I have a machine with two radeon (rv370) cards, each of which is dual head.
This setup works fine with the following versions:
 * x11-base/xorg-server-1.4.0.90-r3
 * x11-drivers/xf86-video-ati-6.6.3
 * linux 2.6.24 (vanilla kernel)

It does not work with linux 2.6.25-rc2 . I will attach the Xorg log file.

Created an attachment (id=146753) [details]
Xorg log file with linux 2.6.25-rc2 (fails)

Created an attachment (id=146755) [details]
Xorg log file with linux 2.6.24 (succeeds)

Could you test with the 6.8.0 driver? Please reopen when you reply.

Ok, I tested it with 6.8.0 (but please note that I have other issues with that
driver version -- see #214136).

New log files to follow.

Created an attachment (id=146868) [details]
xf86-video-ati-6.8.0 Xorg log file with linux 2.6.25-rc2 (fails)

Created an attachment (id=146870) [details]
xf86-video-ati-6.8.0 Xorg log file with linux 2.6.24 (succeeds)

https://bugs.freedesktop.org/show_bug.cgi?id=15523

Just planning to stabilize 6.8 to fix this..

Following on from the bug on fdo, booting with kernel parameter pci=rom solves
the problem. However the 6.8.0 driver still doesn't work for me (see
https://bugs.freedesktop.org/show_bug.cgi?id=15524 ).

Well, for me, on r300 card
 * x11-base/xorg-server-1.4.0.90-r3
 * x11-drivers/xf86-video-ati-6.8.0
 * gentoo-sources 2.6.25-r2
it seems to work without any tweaking,
though those dri+composite issues are annoying.

Friendly reminder: we'll probably push for 2.6.25 stable in the next week or
so, so it would be great to see this fixed

Arches, please stabilize x11-drivers/xf86-video-ati-6.8.0-r1.

(In reply to comment #12)
> Arches, please stabilize x11-drivers/xf86-video-ati-6.8.0-r1.

 Should we ignore bug 216196?

Alpha AT Report: compiles and runs cleanly :)

(In reply to comment #13)
> (In reply to comment #12)
> > Arches, please stabilize x11-drivers/xf86-video-ati-6.8.0-r1.
> 
>  Should we ignore bug 216196?

Yes, that's a corner case that nobody should care about because it means you
can't do direct rendering. I should just force on DRI all the time, but I
haven't gotten around to it.

x86 stable

ppc stable

alpha/ia64 stable, thanks Matt for testing

Created an attachment (id=160397) [details]
Xorg.0.log, startup failure on sparc64(mach64)

On sparc with an ATI 3D Rage XL(SUN PGX64) card I get:
(EE) MACH64(0): Unable to register the following bus resources:
        [0] 0   1       0x000003b0 - 0x000003bb (0xc) IS[B]
        [1] 0   1       0x000003c0 - 0x000003df (0x20) IS[B]
(II) UnloadModule: "mach64"
(EE) Screen(s) found, but none have a usable configuration.

I did a little research and found the release anouncement[0] which says that
the mach64, r128 and radeon driver have been ported to libpciaccess. Are they
supposed to work with a non-libpciaccess based server(1.3.0.0-r6)? If yes, any
ideas why these resource conflicts occur? After switching back to 6.6.3 it
works fine.

[0] http://lists.freedesktop.org/archives/xorg/2008-February/032992.html

(In reply to comment #19)
> Created an attachment (id=160397) [edit] [details]
> Xorg.0.log, startup failure on sparc64(mach64)
> 
> On sparc with an ATI 3D Rage XL(SUN PGX64) card I get:
> (EE) MACH64(0): Unable to register the following bus resources:
>         [0] 0   1       0x000003b0 - 0x000003bb (0xc) IS[B]
>         [1] 0   1       0x000003c0 - 0x000003df (0x20) IS[B]
> (II) UnloadModule: "mach64"
> (EE) Screen(s) found, but none have a usable configuration.
> 
> I did a little research and found the release anouncement[0] which says that
> the mach64, r128 and radeon driver have been ported to libpciaccess. Are they
> supposed to work with a non-libpciaccess based server(1.3.0.0-r6)? If yes, any
> ideas why these resource conflicts occur? After switching back to 6.6.3 it
> works fine.

That split happened for ati version 6.9, so it shouldn't be relevant here with
6.8. Yes, ati 6.8 is supposed to work without pciaccess. I think there might be
some kind of sparc bug open or marked upstream on our Bugzilla right now.

(In reply to comment #20)
> That split happened for ati version 6.9, so it shouldn't be relevant here with
> 6.8. Yes, ati 6.8 is supposed to work without pciaccess. I think there might be
> some kind of sparc bug open or marked upstream on our Bugzilla right now.

Are you referring to bug 230953? That one's related to mesa 7.1(6.5 and 7.0 are
fine).

I talked to another user who experience the same problems as I do(and I found
noone who got it working). To sum it up: =xf86-video-ati-6.6.3 works with
=xorg-server-1.3.* and =xorg-server-1.4.*. >=xf86-video-ati-6.7.197 works with
=xorg-server-1.4.99.* but not with 1.3 or 1.4.

Therefore sparc will pass but please invite us again for
xf86-video-{ati,mach64} when xorg-server 1.4.99(1.5) is going stable :)

ppc64 stable

arm/sh stable

@amd64, ping?

amd64 stable, all arches done.