On Sat, Oct 8, 2011 at 7:43 AM, Thomas Bächler <thomas@archlinux.org> wrote:

Am 07.10.2011 22:16, schrieb Matthew Gyurgyik:

...

On 10/07/2011 08:26 AM, Tobias Powalowski wrote:

...

Latest kernel is in testing, - fixed archiso support - revert to performance governor

Looked in the tracker and didn't see anything. Just curious, why are we reverting back to the performance governor?

There might be trouble, as the ondemand governor does not work with all CPUs.

Could you point me to some more info about this (I couldn't find any bug reports)? I'm interested in following up on it so we can one day move to ondemand by default. For what it's worth, fedora and opensuse have been using ondemand for some time. Is it known why we are seeing problems and they are not? My understanding was that ondemand is recommended by upstream (http://www.codon.org.uk/~mjg59/power/good_practices.html), and that it should automatically fall back to performance if the HW latencies are too high. Please correct me if I'm wrong. Cheers, Tom

Am 09.10.2011 07:43, schrieb Tobias Powalowski:

...

Could you point me to some more info about this (I couldn't find any bug reports)? I'm interested in following up on it so we can one day move to ondemand by default. For what it's worth, fedora and opensuse have been using ondemand for some time. Is it known why we are seeing problems and they are not?

Thats why I changed it, Thomas comes up with the concerns. You can always use cpufreq to change to whatever you need though. I have no issues with changing it again to ondemand or let it on performance. Shall we vote about it?

greetings tpowa

It does not matter that much. The cpufreq modules are not loaded by default. But that might change some day. On my old Opteron cpu I noticed an increased response time when using the ondemand governor. On the other hand I have seen cheap notebooks which will overheat easily if you don't clock down the cpu most of the time. I think Thomas was concerned about p4-clockmod. But this isn't a real clock scaling driver anyway; so shouldn't be used with cpufreq. In short: the user has to decide what is best for them. -- Pierre Schmitz, https://users.archlinux.de/~pierre

There are also differences in the two configs caused by http://projects.archlinux.org/svntogit/packages.git/commit/trunk?h=packages/linux&id=d7ad85513496578dc9e8dbfa87fac8e0db9e490e

Thanks corrected this. greetings tpowa -- Tobias Powalowski Archlinux Developer & Package Maintainer (tpowa) http://www.archlinux.org tpowa@archlinux.org

11.10.2011 21:59, David C. Rankin пишет:

On 10/07/2011 07:26 AM, Tobias Powalowski wrote:

...

Latest kernel is in testing, - fixed archiso support - revert to performance governor

please signoff both arches, greetings tpowa

I get the following error on one x86_64 box, others have updated fine - don't know why?

(65/65) checking for file conflicts [#######################################] 100% error: failed to commit transaction (conflicting files) linux-headers: /usr/src/linux-3.0-ARCH/include/config/cpu/freq/default/gov/performance.h exists in filesystem linux-headers: /usr/src/linux-3.0-ARCH/include/config/mtd/phram.h exists in filesystem Errors occurred, no packages were upgraded.

Try update this package with "--force" key, probably your previous update was not properly finished. Anyway it will not be a big problem because kernel source doesn't make any influence on working system. -- Cheers, Roman.