ATI drivers and kernel 2.6.7
This is a discussion about ATI drivers and kernel 2.6.7 in the Everything Linux category; Greetings I just had some sort of odd experience while trying to get ATI's 3. 7. 0-drivers to work under kernel 2. 6. 7. Here are the bothersome and significant lines that /lib/modules/fglrx/make. sh produce.
Greetings
I just had some sort of odd experience while trying to get ATI's 3.7.0-drivers to work under kernel 2.6.7. Here are the bothersome and significant lines that "/lib/modules/fglrx/make.sh" produce ...
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c: In function `agp_generic_alloc_page':
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c:1405: structure has no member named `count'
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c: In function `ali_alloc_page':
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c:4416: structure has no member named `count'
Looks like the 3.7-driver doesn't go too well with kernel version 2.6.7. I went back to the standard setup afterwards (ATI3.7 under kernel2.6.5) which works like a dream.
If any of you guys/girls are just as curious as me as it goes for kernel-testing then you might want to check out this link that is said to solve the driver-issue ...
http://kerneltrap.org/node/view/3298
Be informed, though, that I hadn't have the time yet to verify the patch featured on this page. Also, as soon as time will allow, I will check the new kernel against the nVidia-drivers on some other boxes. Til then ..
have a luverly day and weekend
I just had some sort of odd experience while trying to get ATI's 3.7.0-drivers to work under kernel 2.6.7. Here are the bothersome and significant lines that "/lib/modules/fglrx/make.sh" produce ...
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c: In function `agp_generic_alloc_page':
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c:1405: structure has no member named `count'
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c: In function `ali_alloc_page':
/lib/modules/fglrx/build_mod/2.6.x/agpgart_be.c:4416: structure has no member named `count'
Looks like the 3.7-driver doesn't go too well with kernel version 2.6.7. I went back to the standard setup afterwards (ATI3.7 under kernel2.6.5) which works like a dream.
If any of you guys/girls are just as curious as me as it goes for kernel-testing then you might want to check out this link that is said to solve the driver-issue ...
http://kerneltrap.org/node/view/3298
Be informed, though, that I hadn't have the time yet to verify the patch featured on this page. Also, as soon as time will allow, I will check the new kernel against the nVidia-drivers on some other boxes. Til then ..
have a luverly day and weekend
Participate in our website and join the conversation
This subject has been archived. New comments and votes cannot be submitted.
Jun 18
Jun 20
0
3 minutes
Responses to this topic
Wish I could help. The Nvidia cards seem to work well with the new kernel, but we're having problems with 3-D and opengl support with X.org. Is your distro using X.org instead of XFree86? I'm sure we'll all get everything sorted out shortly.
OP
Thanx DepperDan,
fortunately I was only peeking into the whole thing (ATI+k2.6.7) out of curiosity and to find out if there might be any probs with those drivers/kernel versions in case a system-upgrade is immiment. So these incompatabilities do not bother me too much at the moment, especially as the standard setup (ATI3.7+k2.6.5) works perfectly.
As far as it goes for the fundamental windowing server: The test-boxes used were running under Mandrake 10 and Suse 9 (see detailed specs below) which both came with the XFree-server (recent Suse already comes with X.Org, me thinks).
But another box for XOrg-testing is already set up, so - even if it sounds peculiar - I'm sort of "hoping" and expecting to encounter some problems that can be solved in some joint effort After all, XFree will be history in many future distros and so it's a good idea to peek into XOrg right away.
by
specs/test machines:
basic: AMD Athlon XP 2800, Asus A7N8X (box 1), Asus A7V600, VIA chipset with GBit-LAN (box 2), 512MB RAM, DAC960 SCSI-RAID, Adaptec 2940UW SCSI, Radeon 9600pro (box 1), Geforce 4MX (box 2)
storage subsystem: 4xSeagate SCSI 10k (30gig) drives as RAID 0+1 array as boot/system drive, Seagate 120GB IDE (data storage)
operating systems:
box 1 - Mandrake 10 Community running kernel 2.6.5 and KDE 3.2.2
box 2 - Suse 9 running kernel 2.6.5 and KDE 3.x (dunno exactly)
fortunately I was only peeking into the whole thing (ATI+k2.6.7) out of curiosity and to find out if there might be any probs with those drivers/kernel versions in case a system-upgrade is immiment. So these incompatabilities do not bother me too much at the moment, especially as the standard setup (ATI3.7+k2.6.5) works perfectly.
As far as it goes for the fundamental windowing server: The test-boxes used were running under Mandrake 10 and Suse 9 (see detailed specs below) which both came with the XFree-server (recent Suse already comes with X.Org, me thinks).
But another box for XOrg-testing is already set up, so - even if it sounds peculiar - I'm sort of "hoping" and expecting to encounter some problems that can be solved in some joint effort After all, XFree will be history in many future distros and so it's a good idea to peek into XOrg right away.
by
specs/test machines:
basic: AMD Athlon XP 2800, Asus A7N8X (box 1), Asus A7V600, VIA chipset with GBit-LAN (box 2), 512MB RAM, DAC960 SCSI-RAID, Adaptec 2940UW SCSI, Radeon 9600pro (box 1), Geforce 4MX (box 2)
storage subsystem: 4xSeagate SCSI 10k (30gig) drives as RAID 0+1 array as boot/system drive, Seagate 120GB IDE (data storage)
operating systems:
box 1 - Mandrake 10 Community running kernel 2.6.5 and KDE 3.2.2
box 2 - Suse 9 running kernel 2.6.5 and KDE 3.x (dunno exactly)
