Thanks,

I found another answer that works for me; I'll try the patch also.

With many thanks to Alex Deucher (see attached),
the answer I found was to use the following script to set the bus-mastering
on:

#!/bin/bash
# set bus mastering for ATI card (4 bit on in low-order nibble; e.g. 0x0007)
dev=00:12.0   # change as appropriate
echo enabling bus mastering on device $dev
setpci -s $dev 4.w=$(printf %x $((0x$(setpci -s $dev 4.w)|4)))

# check that 4 bit is now on
setpci -s $dev 4.w

On a somewhat related topic:
When looking for the device specification to use, get the bus:slot.function 
tuple from scanpci, not from cat /proc/pci. Not sure what the "device"
number in
/proc/pci is, but it definitely isn't the same as "slot"!

-----Original Message-----
From: Michel D�nzer
To: Peter Johnson
Cc: '[EMAIL PROTECTED] '
Sent: 9/3/2002 7:27 AM
Subject: RE: [Xpert]Ant success with PCI/DRI???

On Die, 2002-09-03 at 16:12, Peter Johnson wrote:

> Figured that one out shortly after sending.  Still having what amounts
to a
> system hang, when I try to run the DRI extension.  This happens with
1.3 GHz
> Athlon, running either kernel 2.4.20-pre5 or 2.4.18 and ATI Rage 128
Pro PP rev 0
> (Xpert/PCI).

One possible cause is that bus mastering is only enabled explicitly for
AGP cards. The attached patch fixes that as well as related hangs when
VT switching back to the server.


-- 
Earthling Michel D�nzer (MrCooper)/ Debian GNU/Linux (powerpc) developer
XFree86 and DRI project member   /  CS student, Free Software enthusiast
 <<r128-busmaster.diff>> 

Attachment: atibusmaster5658.html
Description: Binary data

Reply via email to