I am a Linux newbie (sort of) meaning I've known about Linux and used it occasionally for years but have only started switching to it from Windows. I currently have 2 systems running SuSE 10.1 64bit. One is a Dell Poweredge server that serves as a Samba server and VMWare host. It runs flawlessly. The second is my workstation. It runs really good except for one irritating issue: Sometimes in KDE when I click something (close a window, empty the recycle bin, etc.) it just hangs. And not just KDE, the whole system becomes unresponsive. I can't even ping from the server, all I can do is a reset. When it reboots it comes back fine with no problems and a few days later the same thing happens.

What I need to know is where can I look for any place that this might be logged? After it reboots I look at dmesg but it seems to only show data from the recent reboot (not prior). I do need to point out there are no crashes if I boot to the Windows partition so I don;t think this is a hardware problem.

System Specifics:
SuSE 10.1 x86_64 kernel 2.6.16.21-0.13-smp
Asus P5WD2-Premium Motherboard
Pentium D 920 Dual Core
4GB RAM
ATI x850 PCIe (with 3d drivers fully loaded, 7900fps on GLXGears)
Dual monitor configuration
3x Maxtor 300GB SATA2 drives

Non SuSE Software:
Xine
Mplayer (with Win32 codecs)
Kino
TVTime (Not working)

I *think* it has something to do with the video based on this part of the system log I did find, but here I am too much of a newbie to figure out what to do.

****************************************************
Aug 8 13:24:09 hydra kernel: fglrx: module not supported by Novell, setting U taint flag.
Aug 8 13:24:09 hydra kernel: fglrx: module license 'Proprietary. (C) 2002 - ATI Technologies, Starnberg, GERMANY' taints kernel.
Aug 8 13:24:09 hydra kernel: [fglrx] Maximum main memory to use for locked dma buffers: 3788 MBytes.
Aug 8 13:24:09 hydra kernel: [fglrx] module loaded - fglrx 8.26.18 [Jun 22 2006] on minor 0
Aug 8 13:24:09 hydra kernel: ACPI: PCI Interrupt 0000:03:00.0[A] -> GSI 16 (level, low) -> IRQ 169
Aug 8 13:24:10 hydra zmd: NetworkManagerModule (WARN): Failed to connect to NetworkManager
Aug 8 13:24:11 hydra /usr/sbin/cron[3947]: (CRON) STARTUP (V5.0)
Aug 8 13:24:11 hydra SuSEfirewall2: Warning: ip6tables does not support state matching. Extended IPv6 support disabled.
Aug 8 13:24:11 hydra SuSEfirewall2: Setting up rules from /etc/sysconfig/SuSEfirewall2 ...
Aug 8 13:24:12 hydra SuSEfirewall2: batch committing...
Aug 8 13:24:12 hydra SuSEfirewall2: Firewall rules successfully set
Aug 8 13:24:12 hydra kernel: bootsplash: status on console 0 changed to on
Aug 8 13:24:15 hydra kernel: eth1: no IPv6 routers present
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e0000000,8000000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: [fglrx:firegl_addmap] *ERROR* mtrr allocation failed (-22)
Aug 8 13:24:16 hydra kernel: [fglrx:drm_parse_option] *ERROR* "agplock" is not a valid option
Aug 8 13:24:16 hydra kernel: [fglrx] Maximum main memory to use for locked dma buffers: 3788 MBytes.
Aug 8 13:24:16 hydra kernel: [fglrx] total GART = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] free GART = 118226944
Aug 8 13:24:16 hydra kernel: [fglrx] max single GART = 118226944
Aug 8 13:24:16 hydra kernel: [fglrx] total LFB = 124305408
Aug 8 13:24:16 hydra kernel: [fglrx] free LFB = 108843008
Aug 8 13:24:16 hydra kernel: [fglrx] max single LFB = 108843008
Aug 8 13:24:16 hydra kernel: [fglrx] total Inv = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] free Inv = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] max single Inv = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] total TIM = 0
Aug 8 13:24:16 hydra kernel: [fglrx] total GART = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] free GART = 118226944
Aug 8 13:24:16 hydra kernel: [fglrx] max single GART = 118226944
Aug 8 13:24:16 hydra kernel: [fglrx] total LFB = 124305408
Aug 8 13:24:16 hydra kernel: [fglrx] free LFB = 90996736
Aug 8 13:24:16 hydra kernel: [fglrx] max single LFB = 90996736
Aug 8 13:24:16 hydra kernel: [fglrx] total Inv = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] free Inv = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] max single Inv = 134217728
Aug 8 13:24:16 hydra kernel: [fglrx] total TIM = 0
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e17a8000,4000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e17a0000,8000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e1780000,20000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e1700000,80000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e1600000,100000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e1400000,200000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e1200000,200000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e1100000,100000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e10c0000,40000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e10b0000,10000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e10ac000,4000 old: uncachable new: write-combining
Aug 8 13:24:16 hydra kernel: mtrr: type mismatch for e10ab000,1000 old: uncachable new: write-combining
*************************************************************

Thanks for any help. I can continue using the system which is (other than this) very stable but it is an annoyance I don;t want to deal with much longer.

The output of dmesg won't tell you this. Look into the system logs - /var/log/messages and look at the time before the reboot. There should be an entry indicating why your system froze.

-twantrd

I thought that was the right place but if that's the case it does me no good. For almost an hour prior to the lockup nothing was written to this log, I am posting the last entry before lockup and the first after reboot. System lockup was at about 0935.

Aug 9 08:48:46 hydra syslog-ng[2572]: STATS: dropped 0
Aug 9 09:43:47 hydra syslog-ng[2534]: syslog-ng version 1.6.8 starting

Where else can I look?

This was somehow being caused by the powersave daemon, although I never did figure out exactly what. I ended up rebuilding the system to change the way my RAID was set up (and see if I could remember how I did everything) and since the rebuild this issue has not shown up.

Don't you hate it when you can't reproduce some behaviour consistently? :}

I mean, in this case it's obviously a desired effect, but in general ...

Congratulations!



Cheers,
Tink

Yeah, I hate that and in this line of work that is pretty much the norm. Oh well as you say this time it was to my benefit.