You are not logged in.

#26 2008-05-28 21:39:35

Xelinis
Member
Registered: 2008-05-28
Posts: 10

Re: Ati ... worse and worse

Same problem with an nVidia driver.  I'm using 2.6.24-ARCH.

Just out of curiousity, has anyone tried this with a patched kernel?  From my understanding, Arch's CFS is a bit buggy.

Offline

#27 2008-05-31 21:41:23

nirvanix
Member
From: Saskatoon
Registered: 2005-01-31
Posts: 193

Re: Ati ... worse and worse

I have the same problem with an nvidia card.


I've got a lovely bunch of coconuts...

Offline

#28 2008-05-31 22:02:38

shazeal
Member
From: New Zealand
Registered: 2007-06-05
Posts: 341

Re: Ati ... worse and worse

Works fine for me, using custom 2.6.25.4 kernel + nvidia 173.14.05 drivers

Offline

#29 2008-05-31 22:11:41

shazeal
Member
From: New Zealand
Registered: 2007-06-05
Posts: 341

Re: Ati ... worse and worse

Xelinis wrote:

Just out of curiousity, has anyone tried this with a patched kernel?  From my understanding, Arch's CFS is a bit buggy.

That is correct, for some reason 2.6.24 had FAIR_USER_SCHED (Which is actually really bad, since root processes steal all your IO time slices) enabled, now 2.6.25 has no SCHED options at all enabled (Which is basically as bad as having FAIR_USER_SCHED, but root gets a flat 2x more IO slices than the user), Ive raise a Feature request in the bug tracker already. No idea if the two things are related however as I have a great deal of other things enabled/disabled in my kernel.

http://bugs.archlinux.org/task/10512

Offline

Board footer

Powered by FluxBB