Failure of EXA/DRM/xf86-video-ati/xf86-video-radeonhd in FreeBSD/amd64 -CURRENT
Hi,
I have been tracking a DRI crash-on-startup bug with the Xorg server at:
* https://bugs.freedesktop.org/show_bug.cgi?id=16200
The bug was originally testing on the xf86-video-ati driver. I recently
pulled down the quick-and-dirty-2d branch of radeonhd and tried running
Xorg with EXA+DRI, and got a similar server busy-loop at startup. I did
some more testing, and also I still don't get and DRI rendering from the
card in the Accel-disabled mode. The recent changes to remove some
hard-locks from the driver (by glisse@) have the drm kmod now producing
the following info message, in an infinite loop:
info: [drm] wait for fifo failed status : 0x9001C100 0x00080000
Before these changes went in, I was able to get intermittent success
using EXA+DRI with the xf86-video-ati driver. However, lately, I cannot
get the server to start anymore.
Using the radeonhd driver, I was able to get it to start with DRI when I
turned acceleration off, but the DRI windows just remain black, with no
output. I get those info: messages listed above dumped to the console in
this case too, as well as when I run radeonhd with EXA+DRI and the
xserver locks up (busy-loops) on startup.
Curious if you have any instructions in particular that I can look at
regarding this. I am suspicious of, perhaps, EXA and DRI stepping on
each other, but I don't really know. Maybe someone else can take a look
and get a better idea...
In all cases I can SSH to the laptop, and reboot it from the shell.
Output of lspci -s 01:05 -vv -xxxx -i /usr/local/share/pciids/pci.ids
-nn:
01:05.0 VGA compatible controller [0300]: ATI Technologies Inc RS690M
[Radeon X1200 Series] [1002:791f] (prog-if 00 [VGA controller])
Subsystem: Hewlett-Packard Company Unknown device [103c:30c2]
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop-
ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort-
On Fri, Jun 27, 2008 at 9:46 AM, Coleman Kane
Hi,
I have been tracking a DRI crash-on-startup bug with the Xorg server at: * https://bugs.freedesktop.org/show_bug.cgi?id=16200
The bug was originally testing on the xf86-video-ati driver. I recently pulled down the quick-and-dirty-2d branch of radeonhd and tried running Xorg with EXA+DRI, and got a similar server busy-loop at startup. I did some more testing, and also I still don't get and DRI rendering from the card in the Accel-disabled mode. The recent changes to remove some hard-locks from the driver (by glisse@) have the drm kmod now producing the following info message, in an infinite loop:
info: [drm] wait for fifo failed status : 0x9001C100 0x00080000
Before these changes went in, I was able to get intermittent success using EXA+DRI with the xf86-video-ati driver. However, lately, I cannot get the server to start anymore.
Using the radeonhd driver, I was able to get it to start with DRI when I turned acceleration off, but the DRI windows just remain black, with no output. I get those info: messages listed above dumped to the console in this case too, as well as when I run radeonhd with EXA+DRI and the xserver locks up (busy-loops) on startup.
Curious if you have any instructions in particular that I can look at regarding this. I am suspicious of, perhaps, EXA and DRI stepping on each other, but I don't really know. Maybe someone else can take a look and get a better idea...
I think the drm still has issues on freebsd. I'm note sure what the current status is there. Alex -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
participants (2)
-
Alex Deucher
-
Coleman Kane