Hello,
Thank you for testing, and for all the others who have tested.
I am not sure why cpuspeed it isn't failling on a CentOS machine, but
there have been several tests on pure SL machine's.
I also was able to find a machine to do some tests over the weekend and
gave it a wide battery of tests.
I think the most important test was that installing the newer cpuspeed
doesn't break anything on the older kernels. That was my biggest
concern. So I will be pushing out the newer cpuspeed out today into the
security errata. It is marked as a security errata because a real
security errata has a dependancy on it.
Thanks
Troy
J S Jayakumar wrote:
>
> ------------------------------------------------------------------------
>
> Subject:
> Re: kernel 2.6.18-128.1.1.el5
> From:
> J S Jayakumar <[log in to unmask]>
> Date:
> Mon, 2 Mar 2009 03:58:54 -0600
> To:
> Troy J Dawson <[log in to unmask]>
>
> To:
> Troy J Dawson <[log in to unmask]>
>
>
> Dear Troy,
>
> I hv tested the cpuspeed rpm cpuspeed-1.2.1-5.el5.x86_64 as you
> suggested for kernel 2.6.18-128.1.1.el5 and IS WORKING FINE.
>
> Thanks.
>
> J S Jayakumar
>
> Troy Dawson wrote:
>> I *think* what has to happen is we need to push out the new cpuspeed
>> that is in SL 5.3 to everyone
>>
>> https://bugzilla.redhat.com/show_bug.cgi?id=459441
>> http://rhn.redhat.com/errata/RHEA-2009-0098.html
>>
>> Basically, it looks like what happened was the powernow-k8 changes
>> from being built into the kernel to a kernel module. And the cpuspeed
>> needed to be changed to understand this.
>>
>> Can someone test this out and see if it works for them.
>>
>> i386:
>> http://ftp.scientificlinux.org/linux/scientific/5rolling/i386/SL/cpuspeed-1.2.1-5.el5.i386.rpm
>>
>> x86_64:
>> http://ftp.scientificlinux.org/linux/scientific/5rolling/x86_64/SL/cpuspeed-1.2.1-5.el5.x86_64.rpm
>>
>>
>> Troy
>>
>> Dusan Bruncko wrote:
>>> Dear Troy,
>>>
>>> it is true also for i386 ...
>>>
>>> Dusan
>>>
>>> J S Jayakumar wrote:
>>>> Dear Troy,
>>>>
>>>> unfortunately kernel 2.6.18-128.1.1.el5 does not support cpu
>>>> frequency scaling on AMD64 cpus. We were using this feature on all
>>>> of our machines with the previous kernels. Any solution to this
>>>> problem?
>>>>
>>>> J S Jayakumar
>>
>
>
--
__________________________________________________
Troy Dawson [log in to unmask] (630)840-6468
Fermilab ComputingDivision/LCSI/CSI DSS Group
__________________________________________________
