catalyst 10.4 vs tuxonice

[haplo602]

Hi all

This is a bit late, since 10.5 will arrive shortly I guess, but since I installed 10.4, hibernation no longer works for me.

10.3, kernel 2.6.33 and X 1.6.5 suspended in about 50% of cases (I guess my extra page allowance was not high enough). However since I got 10.4 and 1.7 X I cannot hibernate the system at all while X is running. it hangs in first atomic copy.

Anybody has working hibernation with 10.4 catalyst ? my GFX card is a HD3650 AGP.

[Qaridarium]

Hi all

This is a bit late, since 10.5 will arrive shortly I guess, but since I installed 10.4, hibernation no longer works for me.

10.3, kernel 2.6.33 and X 1.6.5 suspended in about 50% of cases (I guess my extra page allowance was not high enough). However since I got 10.4 and 1.7 X I cannot hibernate the system at all while X is running. it hangs in first atomic copy.

Anybody has working hibernation with 10.4 catalyst ? my GFX card is a HD3650 AGP.

catalyst 10-5 chancelog:
1.* SLED/SLES 11 SP1 support (early look)
2.* Fixed the RedHat installer so that building packages in a directory path that contains spaces works
3.* New OpenGL component version
4.* Works on Ubuntu 10.04 again
5.* Fixed Xserver 1.7 segfault during startup on some systems with a backtrace pointing to xf86HandleInt10Options
6.* Disabled consumer card support for Eyefinity until QA verifies it
7.* Fixed failure reported when trying to turn off discrete card on some PowerXpress AMD+AMD systems
8.* Fixed CPU lockup on resume from S3/S4 on some systems
9.* Fixed Xserver startup failure in certain Crossfire configurations
10.* Assorted OpenGL fixes
11.* Fixed incorrect display arrangement in big-desktop mode on RandR 1.1 OSes
12.* OpenGL fixes for Unigine


just wait 1-2 days

[Kirurgs]

I have fglrx + tuxonice working on archlinux perfectly. No jokes, this is first fglrx which I've ever used so long (since the day it was released).
So far resume id 100% success either from S3 or S4.
I had 2.6.33.2 kernel, now it is 2.6.33.4, xorg 1.7.6, ati 3650 discrete card in Lenovo T500 which has intel + amd switchable graphics.
Kernel is custom compiled (added just TOI patch).
So it IS working.

[sovok_kpss]

I have fglrx + tuxonice working on archlinux perfectly. No jokes, this is first fglrx which I've ever used so long (since the day it was released).
So far resume id 100% success either from S3 or S4.
I had 2.6.33.2 kernel, now it is 2.6.33.4, xorg 1.7.6, ati 3650 discrete card in Lenovo T500 which has intel + amd switchable graphics.
Kernel is custom compiled (added just TOI patch).
So it IS working.

I have ArchLinux on the laptop but my story is not that good than yours
It is about 20% success resumes, rest resumes are not so successful. It resumes fine but shows me a cursor on the black screen(actually there isn't only the black screen and cursor, because I can unlock screen(still the black) type my password(it resumes to gnome lockscreen and doing some weird things)) and I need to switch to tty1-5 and restart gdm. This happens on both 2.6.33 and 2.6.34 kernels, both TOI git and 3.1 versions, xorg server 1.7 and 1.8. catalyst 10.4.

Could you post your config files(/etc/hibernate/ and related) ?

[Kirurgs]

I'm not using hibernate script at all, pm-utils integrates wtih TOI very well... pm-utils integrates with gnome very well and it seems I have very good configuration
Is that only catalyst gives you this error or opensource drivers as well?
I remember when I had nvidia, it was real nightmare with hibernate

Edit: All configs are default, no custom changes added...

[Kirurgs]

Please post Xorg.0.log, everything.log file after unsucessful resume, there might be some clue in there.
Have you customized xorg.conf?

P.S. Damned one minute