On Friday 27 June 2008 02:36, Johnny Hughes wrote:
Johnny Hughes wrote:
Wojtek Pilorz wrote:
After yum upgrade, the new CentOS5 kernel [ kernel-2.6.18-92.1.1.el5.i686 ] crashes each time early in boot process on Compaq ProLiant 3000 with dual Pentium3 CPU (600 MHz, CPU0: Intel Pentium III (Katmai) stepping 03, CPU1 the same).
I get kernel panic at powernowk8_init+0x05e/0x1c3.
The previous kernel [ kernel-2.6.18-53.1.21.el5.i686 ] seems to work fine.
I am wondering if anyone has met this on PIII CPU ?
I have found 99.9% identical problem on http://www.nabble.com/net-5501-and-RHEL5.2-td18024549.html
(only esp and task slightly different from what I have seen on screen after crash, all other registers, call trace, code bytes identical to my case) , but that was on a different CPU.
On the other hand, I run CentOS upgraded to 5.2 on a home PC with PIII (single CPU, Abit motherboard with i440BX chipset) with no problem.
Is it normal, that powernowk8_init gets called on PIII CPU?
We have created a new test kernel that has an upstream patch to fix this issue, please test this and see if it fixes your issues:
I have heard off list that this kernel does solve the problem with powernow-k8.
As always though, this kernel will be available to people with the problem (at the above link), however we won't officially release it as a replacement until it is in an upstream kernel and released into the tree by them.
This is even worse because it renders the install media worthless on these machines, and I am sorry, but upstream needs to initiate things in the main tree.
I have an Athlon 64 X2 4200+, and that hit the powernow-k8 problem too. I installed your test kernel which has cured it. Thanks
Anne