Radeon 8500 + 2.4.19

[civ1492]

Hi,

anyone here who has random lockups with his radeon card and the new
2.4.19 kernels ?

well I need some specs and some driver info before the problem can be pinpointed, but my Radeon 7500 seems fine using the daily drivers of dri.

[civ1492]

what specs do you need ?

2.4.19 stock kernel
latest dri drivers
xfree 4.20

the problem is that the monitor is going into standby mode randomly - but im using no power management at all :/

well one guess, the only thing I know that can cause random lockups for AGP devices is the AMD bug. But are you sure you installed the drivers correctly as modules ?

[civ1492]

hmm what amd bug ? .. and what modules should i install as drivers ?
tia,
chb

okay, when you compile your kernel make sure that agpgart is a modules and that you hardmount the correct AGPgart subdriver, in my case VIA
and then you need to have XFree86 DRI drivers, radeon as module (at least this works for the 7500 I dont know the specifics of the 8500 driver.

then reboot your new kernel, modprobe agpgart and modprobe radeon and compile the new drivers of the dri website - they will redo the modprobing - now make sure that your PC runs the modprobes on each reboot, in gentoo you can put this in /etc/modules.autoload - otherwise write a init script for it.

[civ1492]

ah .. i will try that thx

[civ1492]

great , now its working like it should. Thanks for you help Lc the problem was the agpgart driver which was compiled into the kernel now i use the modules supplied with dri and everything is working 8)

[quote author=Lovechild link=board=3;threadid=4489;start=0#44861 date=1028748532]
well one guess, the only thing I know that can cause random lockups for AGP devices is the AMD bug.
[/quote]

Bah FUD (gee i am picking on LC again i guess that will be another spite for me). It is not a bug but a design difference and the Linux developer's failure to realize that the design for the AMD and Intel chips are different.

[Schotty]

Just to be an asshole, I have to state that an unwanted side effect in a program is a bug. So technically LC is correct, that the AMD design oversight is a bug.

Glad to be of anal assistance