[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: DRM failure with ATI radeon 9200 and Kernel 2.6



Daniel Johnson <teknotus@gmail.com> wrote:
> On 6/26/05, Terry Hancock <hancock@anansispaceworks.com> wrote:
>> Hardware acceleration (DRI) is not working with my ATI Radeon 9200 card
>> on a fresh install of Debian Sarge (3.1), using the AMD 64 build of Kernel 2.6.
>
> I have a radeon 8500, and it works with  a fresh install of  sarge with a  2.6.8 kernel. 
> That is a  similar  card  I  think.  I don't remember if I had to manually uncomment the
> dri, and glx lines in the config file. 
>
>>    From /var/log/XFree86.0.log I'm seeing this message:
>>    """
>>    (II) RADEON(0): [drm] drmSetBusid failed (5, PCI:1:0:0), Invalid argument
>>    (EE) RADEON(0): [dri] DRIScreenInit failed.  Disabling DRI.
>>     """
>
> Could you have the busid wrong?

Well, I suppose, but this is all automatic.  To my knowledge, I haven't
specified it anywhere.  One thing bothers me a little --- this is an AGP card,
why is the bus id a PCI specification?  Is that part normal?

There is a series of lines in the XFree86.0.log file which appears to show the
autodetection process:
"""
drmOpenDevice: minor is 0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is -1, (Unknown error 999)
drmOpenDevice: open result is -1, (Unknown error 999)
drmOpenDevice: Open failed
drmOpenDevice: minor is 0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is -1, (Unknown error 999)
drmOpenDevice: open result is -1, (Unknown error 999)
drmOpenDevice: Open failed
drmOpenDevice: minor is 0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 5, (OK)
drmOpenDevice: minor is 1
drmOpenDevice: node name is /dev/dri/card1
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: open result is -1, (No such device)
drmOpenDevice: Open failed
"""
(I've included some failures for context -- there's a dozen or so failed
combinations, and only the one device at /dev/dri/card0 found.  One
odd thing -- what is the difference between the first 3 attempts? They
all look like they are using the same parameters to reach /dev/dri/card0,
but only the last one fails.  Is that normal?  Why "Unknown error"
instead of "No such device"?)

> Did you check it against lspci?
I have done so now. ;-)  
Here's the relevant lines:
"""
0000:01:00.0 VGA compatible controller: ATI Technologies Inc RV280 [Radeon 9200] (rev 01)
0000:01:00.1 Display controller: ATI Technologies Inc RV280 [Radeon 9200] (Secondary) (rev 01)
"""

I'm a little uncertain about how to read the bus ids here, but they look plausible.  Maybe
it should read PCI:1:0:1 or PCI:0:1:0?  Even if I want to change this, though, where do I
put the information?  Is there a line in XF86Config-4 for this?  I suppose I can try to
set these with "dpkg-reconfigure xserver-xfree86" -- I'm pretty sure there is an option to
specify the bus.  But what should I tell it, if this isn't right?

> Or maybe just the syntax of that line wrong?

Um, which line?  Again, I don't know where this would be specified.  A syntax error
sounds promising, but I don't know where to look for it.  ISTM, whatever syntax is
being used is being created by the driver software?

I'm willing to go to a new driver or whatever if that's required, but this looks like a
very minor problem --- it seems like there *ought* to be something I can specify
somewhere to make this behave.

I'm also looking at the Gatos project and the Debugging DRI page:
http://gatos.sourceforge.net/dri-debug.php

but it appears to me that this would involve compiling and installing new drivers,
which means getting kernel-headers and so on.  I'd like to avoid that if there's a
simple configuration solution (which I'd likely have to solve anyway, even for new
drivers).

--
Terry Hancock ( hancock at anansispaceworks.com )
Anansi Spaceworks  http://www.anansispaceworks.com



Reply to: