Permanent solution to FGLRX/Kernel 2.6.34.7-0.3.1 issue?

I’ve been reading about a temporary solution to the kernel/Fglrx issue caused by the last kernel update, but I haven’t read about any permanent solution to this issue. Anyone know anything about a permanent solution (new kernel update?, new Fglrx driver?) to this compilation error?

my guess is that when Catalyst 10.10 comes out they would have fixed the issue but until that you will need to stick to using the temporary fix

I think the ‘permanent’ solution would be

  • path the fglrx module sources rather than the kernel
  • have a service which would check if the kernel has been updated and recompile the module if needed (the same way VirtualBox does).

It’s not too complicated to patch the sources, neither to write such a service… and I think a German guy did it already. He did provide (IMHO too many) explanations on his site and in german language (which doesn’t bother me, but I think the story he made around this patch is a bit exagerated).

Here is the link that please_try_again is talking about:

openSUSE 11.3 – proprietären Grafik-Treiber ATI Catalyst 10.9 als RPM installieren

@ ah7013

Thank you. I would have posted it, but didn’t have the link anymore. BTW I do not criticize his work: He did a good job. It’s just that by the time it would take me to find out where to download his patch, I would have written my own. When I’m trying to solve a problem, I don’t have time to play hide and seek on people’s websites. You have to scroll down 11 times to see the first direct link to download a shell script … and even a newbie webmaster should know that people don’t like to scroll down more than 7 times. What will be the next step? Search for kernel hacks on FaceBook?!

Yes it took me a while to find the patch on his website as well. He has done a good job on the patch but I agree that he he needs to make it easier to find

I love his patch. I applied it and all is well with the kernel and Fglrx. I’m just a bit disappointed that something like this could happen… if only ATI had waited 2 weeks to release 10.9, because then they might have known that a function within the Linux kernel was changing.