Search:

If it only occurred on kernel 2.4 it would be a thread, because 2.4 gives each thread its own PID. That it happens under 2.6 proves it is a fork.

No, Microsoft gets LESS stable with each major release. It takes 2-3 service packs to get stable again.

If the driver can't handle the old cards anymore, they need to fork the driver for each...

I see 2 problems and a possible fix:


(EE) module ABI major version (0) doesn't match the server's version (1)


dlopen: /usr/lib/xorg/modules/drivers//fglrx_drv.so: undefined symbol:...

glxgears and fgl_glxgears slowdown (50% drop) as of 8.10, still present in 8.11
however, CPU isn't buried anymore
-CPU usage is under 10% instead of 100% like with 8.10
-however CPU usage...

It has long been said:
ATI = good hardware, sucky drivers
nVidia = sucky hardware, good drivers
and if you put the two together, you'd get nVidia's hardware and ATI's drivers ;)

nobody tell...

$ uudecode < temp
$ patch -p1 < firegl_public.patch

another work around:
run program in windowed mode, without border, at desired resolution

with wine, this requires an additional step:
enable 'emulate desktop' at the desired resolution
set...

That does narrow things down.

If it works with a monitor and svideo at the same time, then it doesn't like not having a monitor. If it doesn't, then it doesn't like svideo.

That would seem to...

It seems phoronix corrupts patches when you post them. :confused:

md5sum
f9c15acbf28f780a15930b400cb07cc1 firegl_public.c (original)
aa259a13effc4b0a941f707818b8474a firegl_public.patch...

htop includes wait in idle :/
try using CPU detail, or use regular top, please

I hear you on compiling kernels. Must be getting old :P

Are you able to boot with a regular monitor and no svidio?

Odur, Does it still lock if you disable composite? Or if you enable it? Some people have reported <i>enabling</i> composite can fix it.

Also, try disabling DRI in xorg.conf.

Even so, we are agreed that DRI initialization triggers the fault.

Possibilities:
1 bug in fglrx binary blob
-not much we can do here.
2 bug/miscompile fglrx kernel module.
-gcc version...

Booting to runlevel 3 was to prevent X from loading. This is standard across most distros.

Deleting /etc/ati only works if X is not running because it gets rebuilt when X closes. I think...

I'll post it when I get home from work.

Just in case, please make sure you are patching an unmodified source tree. If Kano's patch partially applied, mine won't work anymore, so you would need to...

try changing:
src/sysdeps/linux-orig.c
to:
src/sysdeps/linux.c

Also, are you in the source tree itself? It looks like you are 1 folder up, unless you unpacked into your Downloads folder.

2 things:
do you have your display manager set to restart xorg on logout? (default is no. turn this option on)
is atieventsd running? (default is yes. turn it off)

aljaz, it may still be a misconfigured APIC. If that is the case, updating to a recent kernel and updating your BIOS may correct it.

Linuxhippy, are you still having problems? if so, please...

That definitely sounds like it isn't a GPU problem. Was that 75% user with 25% idle, 25% system, or 25% wait?
idle - CPU isn't loaded either, could be IRQ or DMA problem
system - likely driver,...

I stand corrected. Thank-you.

This, ladies and gentlemen, was the definitive straight answer.