CentOS - Aug 2006 - cpuspeed causing hangs on warm boots

We are using CentOS-4.3 (x86_64) with Opteron servers, and noticed a 
problem with cpuspeed:

If we reboot by doing  "# reboot" or similar from a shell, the boot
will
hang not long after cpuspeed starts. The machine freezes: ctrl-alt-del 
doesn't work, and only pressing the reset button or a power cycle will 
restart the machine.

If we reboot from power-up or by pressing the reset button, there is no 
hang, and the boot proceeds normally.

Another strange thing is that only some boxes have this problem. Others, 
seemingly identical, reboot just fine. We've compared the BIOS settings, 
the CPU steppings, etc., and they seem to be identical. It's very odd. 
Maybe we're missing a BIOS setting, but we couldn't find a difference.

We've fixed this just by turning off the cpuspeed service.

It took some trial and error to figure out this was caused by cpuspeed.  
Misleadingly, the hang doesn't occur immediately after "Starting 
cpuspeed: [OK]" prints. Instead it occurs when trying to bring up a 
network interface, which is almost right after cpuspeed starting. With 
the original 2.6.9-34.ELsmp, it's at "Bringing up interface eth0".
With
the 2.6.9-34.0.2.ELsmp (and all other rpms updated), it hangs at 
"Bringing up loopback interface". This may just be a timing thing.

We can just forget about cpuspeed, but if any of you have any advice, 
we'd appreciate it. Otherwise I am basically posting this to save 
diagnostic time for those who encounter these symptoms in the future.

These are dual-core dual Opteron 265/270 servers with Tyan K8SD Pro 
motherboards.

Thanks,
Dan

On Tue, 2006-08-08 at 18:10 -0400, Dan Halbert wrote:
> We are using CentOS-4.3 (x86_64) with Opteron servers, and noticed a 
> problem with cpuspeed:
> 
> If we reboot by doing  "# reboot" or similar from a shell, the
boot will
> hang not long after cpuspeed starts. The machine freezes: ctrl-alt-del 
> doesn't work, and only pressing the reset button or a power cycle will 
> restart the machine.
> 
> If we reboot from power-up or by pressing the reset button, there is no 
> hang, and the boot proceeds normally.
> 
> Another strange thing is that only some boxes have this problem. Others, 
> seemingly identical, reboot just fine. We've compared the BIOS
settings,
> the CPU steppings, etc., and they seem to be identical. It's very odd. 
> Maybe we're missing a BIOS setting, but we couldn't find a
difference.
> 
> We've fixed this just by turning off the cpuspeed service.
> 
> It took some trial and error to figure out this was caused by cpuspeed.  
> Misleadingly, the hang doesn't occur immediately after "Starting 
> cpuspeed: [OK]" prints. Instead it occurs when trying to bring up a 
> network interface, which is almost right after cpuspeed starting. With 
> the original 2.6.9-34.ELsmp, it's at "Bringing up interface
eth0". With
> the 2.6.9-34.0.2.ELsmp (and all other rpms updated), it hangs at 
> "Bringing up loopback interface". This may just be a timing
thing.
> 
> We can just forget about cpuspeed, but if any of you have any advice, 
> we'd appreciate it. Otherwise I am basically posting this to save 
> diagnostic time for those who encounter these symptoms in the future.
> 
> These are dual-core dual Opteron 265/270 servers with Tyan K8SD Pro 
> motherboards.
> 
> Thanks,
> Dan
Personally, I don't like cpuspeed ... and I turn it off all the time
except on laptops.

I have also had issues, especially on AMD machines, with cpuspeed.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL:
<http://lists.centos.org/pipermail/centos/attachments/20060812/a28bf00d/attachment-0002.sig>