Bugzilla – Bug 116336
powersaved causes system to freeze on bootup
Last modified: 2006-01-09 16:19:37 UTC
I ran a fresh install of SuSE 10.0 RC1 and all went fine. The first time I booted the PC it worked fine. Then I booted the PC from the second time, and soon after the desktop appeared, the system froze - hard reboot was required. This happened every time I booted. I tried booting to runlevel 3 - same problem. I then booted to runlevel 1 which worked fine. I looked at the /var/log/boot.omsg file which had the following at the end: Starting powersaved (accessing ACPI events over acpid) <notice>startproc: exec done Master Resource Control: runlevel 3 has been reached Skipped services in runlevel 3: nfs smbfs <notice>killproc: kill(5146,3) So I rebooted to runlevel 1, ran YaST and disabled powersaved from booting on startup. I then switched to runlevel 5 and everything ran fine. Every time subsequently that I have booted with powersaved disabled, the system runs great.
What happens if you start powersave with 'rcpowersaved start' after boot?
this is probably one of the systems where loading the cpufreq modules crashes the kernel because of bad ACPI tables in a substandard BIOS.
In case it's any use to you, what may be relevant to the cpufreq and powersave issue is that I have a motherboard with AMD's Cool 'n' Quiet technology enabled.
So I expect that: a) ACPI modules or b) cpufreq freezes your system. Do you get any suspicious logs in /var/log/messages, could you post them (the freeze should have been logged at least one time - compare the time of logs). Is this a dual-core machine, then it's probably cpufreq? To find out above you could: modprobe button;modprobe ac;modprobe battery;modprobe fan Still running? -> then it should be cpufreq. Try: modprobe powernow-k8 modprobe cpufreq_ondemand echo ondemand > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor Still not freezing? Increase powersave daemon output (/etc/sysconfig/powersave/common -> DEBUG="15" start powersave daemon. In any case where it freezes: Reboot and send/attach /var/log/message output, there should be logged some errors. If it's not cpufreq then please also attach the output of acpidmp. Thanks.
Mark, can you please try what Thomas wrote in the previous post? If not, we have to close the bug as INVALID...
Please test Open Suse 10.1 and reopen if the problem still exists.