ati catalyst 10.6 released

[psychok9]

You are using an unsupported kernel. Your best bet is to downgrade to the stock 2.6.32 kernel. Someone around here might have a patch for you to enable 2.6.34.. Good luck.

Yeah, I hope this.
The "stable" 2.6.32 of Canonical, it's unstable. It have an huge bug with my platform that cause random and frequently freeze with the hdds. From 2 months of 10.04, not fix yet...

[phoen1x]

phoen1x, is that ticket #1700 you're talking about on ati.cchtml.com ? If so, that ticket was reported as debian-only and the debian ticket containing the info was subsequently closed because the debian devs felt it was a kernel issue (mucked up MTRRs).

Any chance you could confirm the problem also occurs on a supported distro (maybe try Ubuntu 9.10) and update the ticket ?

I can confirm that this is not debian only bug. It doesn't depend on distribution, tested on ubuntu 9.04/9.10, archlinux and gentoo. The problem now is i got only gentoo and yesterday i tried cat 10.6 and still blank screen as usual xorg.log shows nothing suspicious. I tested cat 9.4-10.6, all versions higher than 9.8 just dont work and kernel version really doesn't matter, i tried 2.6.28-2.6.33 (even 2.6.34rc). Should i fill bug report? Bug #1794 perfectly describes problem.

[bridgman]

The mtrr output looks funny, doesn't it ? That sounds like it might be a system BIOS issue; do you know if you're running the most recent BIOS image available for your mobo ?

[spoilerhead]

btw, i wanted to note that i seem to ahve fixed my issues with a clean reinstall, speed is very fast :O

[phoen1x]

The mtrr output looks funny, doesn't it ? That sounds like it might be a system BIOS issue; do you know if you're running the most recent BIOS image available for your mobo ?

Could You please specify which mtrr output Your are referring to?
Mine mtrr output:
reg00: base=0x000000000 ( 0MB), size= 2048MB, count=1: write-back
reg01: base=0x080000000 ( 2048MB), size= 1024MB, count=1: write-back
reg02: base=0x0c0000000 ( 3072MB), size= 512MB, count=1: write-back
reg03: base=0x100000000 ( 4096MB), size= 512MB, count=1: write-back
Does this looks bad?

[mar1o]

Ati did it again!Worst tearing ever with 10.6.In 2d and 3d (Ati HD3870)Going back to 10.5
Sorry for bad english

[devius]

I have to say I was pleasantly surprised by this new driver! I'm using a supported OS (openSUSE 11.2) but I've installed some extra libraries that aren't standard in the distro, namely Kernel 2.6.33, libdrm 2.4.21 and a few others I can't remember right now, although I'm still using xorg-server 7.4. I just build a package from the list of distros (SuSE/SUSE 11.2), but I wasn't hoping for much due to the upgraded kernel and not being sure if SUSE was referring to openSUSE or the commercial variant. The package build fine and I installed it, again with no errors to my big surprise. I just assumed that it would probably break my system on reboot since so far it had been a pleasant process. But once again my expectations weren't met and the damn thing worked fine and booted nicely.

By this time I was thinking that something HAD to be wrong, since I haven't had such a trouble free graphics driver installation process since the days I had a geforce4. I tried a few 3D apps and desktop effects and all worked great. So I have to say this is the best catalyst release yet for me and it is working perfectly... almost! After I changed the screen mode in the CCC everything also worked fine, until the next time I booted the system. This time it was more like a real catalyst driver and X wouldn't start. Some complaint about a ModeLine that was expecting H"something". Turns out that in each ModeLine line the first number after the mode name ("1600x1200") was like this -> 217,0 and xorg was expecting only integers so I had to manually delete the ",0" from every line to be able to boot into X. Why did this happen? I did an aticonfig --initial as was stated and everything was working perfectly until this. Why?? I suppose that CCC added those unnecessary ",0" to each number, I just don't understand why...

Anyway, it's not news that there are problems installing the catalyst drivers, and this was an easy one to solve, but it's a shame this still isn't a flawless victory for the driver. Maybe next month

[j0v4nn1]

Could You please specify which mtrr output Your are referring to?
Mine mtrr output:
reg00: base=0x000000000 ( 0MB), size= 2048MB, count=1: write-back
reg01: base=0x080000000 ( 2048MB), size= 1024MB, count=1: write-back
reg02: base=0x0c0000000 ( 3072MB), size= 512MB, count=1: write-back
reg03: base=0x100000000 ( 4096MB), size= 512MB, count=1: write-back
Does this looks bad?

Just to add to the discussion, very similar setup (HD5770, Nforce 430, 64-bit Ubuntu, mobo: Asus M2NPV-VM, btw just upgraded to the last BIOS available, as per Bridgman suggestion) same negative experience (black screen / hanging). I'll append my mtrr too, who knows..

reg00: base=0x000000000 ( 0MB), size= 2048MB, count=1: write-back
reg01: base=0x080000000 ( 2048MB), size= 1024MB, count=1: write-back
reg02: base=0x0c0000000 ( 3072MB), size= 256MB, count=1: write-back
reg03: base=0x100000000 ( 4096MB), size= 1024MB, count=1: write-back
reg04: base=0x0cff00000 ( 3327MB), size= 1MB, count=1: uncachable

And no, the MTRR ranges didn't change upgrading the BIOS.

I've tried w/ all recent Catalyst versions, now I'm going to try the 10.6 release but I have little hope..

[Pfanne]

oh btw.
with fglrx 10.6 you can finally record your desktop without everything slowing to a crawl.
runs smooth as silk!

[Sarvatt]

What I put in x-updates is working with xserver 1.8 here outside of the crashes unless you specify a busID in xorg.conf (aticonfig --initial works)

http://sarvatt.com/downloads/Xorg.0.log

backtrace with jockey's xorg.conf

http://sarvatt.com/downloads/xserver.gdb.txt