Re: 4.5 - strange performance issue

2009-05-12 Thread Josh
Andrei GUDIU wrote: Try to enable EXA and play with Option "MigrationHeuristic" "greedy" I can confirm this solved my X problem. And it was really really a slow X. I added Option "AccelMethod" "EXA" Option "MigrationHeuristic" "greedy" in Section "Device". This also

Re: 4.5 - strange performance issue

2009-05-06 Thread The Wraith
I can confirm the problem, but it was not an X problem only...everything was slow. The problem was that my interrupts were up to 82.9%. Disabled acpiprt and acpimadt in the kernel and it all works ok. On Wed, May 6, 2009 at 11:35 AM, Andrei GUDIU wrote: > > Try to enable EXA and play with Option

Re: 4.5 - strange performance issue

2009-05-06 Thread Andrei GUDIU
> Try to enable EXA and play with Option "MigrationHeuristic" "greedy" > I can confirm this solved my X problem. And it was really really a slow X. I added Option "AccelMethod" "EXA" Option "MigrationHeuristic" "greedy" in Section "Device".

Re: 4.5 - strange performance issue

2009-05-05 Thread Gregory Edigarov
Tobias Ulmer wrote: Try to enable EXA and play with Option "MigrationHeuristic" "greedy" Thanks, I will try that later today. -- With best regards, Gregory Edigarov

Re: 4.5 - strange performance issue

2009-05-05 Thread Gregory Edigarov
Gregory Edigarov wrote: Hello, Does anybody else notice strange X performance degradation? It takes forever for X to start, and after it start it takes forever for them to bring up firefox... and after all it is really slw Well I turned off the acpi completely, that seems to solve

Re: 4.5 - strange performance issue

2009-05-04 Thread frantisek holop
hmm, on Mon, May 04, 2009 at 12:24:20PM +0200, Otto Moerbeek said that > Check top(1) without a runing X first. You might have problems with > interrupts. If that's the case, top should show prettu high interrupt %'s. > Espcially some nvidia chipsets have these problems. > > If you are suffering f

Re: 4.5 - strange performance issue

2009-05-04 Thread Tobias Ulmer
Try to enable EXA and play with Option "MigrationHeuristic" "greedy"

Re: 4.5 - strange performance issue

2009-05-04 Thread Gregory Edigarov
It's not an interrupt load. top seems normal, even running X, 0.0 - <0.9 % interrupt. but everything is very slow. Otto Moerbeek wrote: Check top(1) without a runing X first. You might have problems with interrupts. If that's the case, top should show prettu high interrupt %'s. Espcially some

Re: 4.5 - strange performance issue

2009-05-04 Thread Otto Moerbeek
Check top(1) without a runing X first. You might have problems with interrupts. If that's the case, top should show prettu high interrupt %'s. Espcially some nvidia chipsets have these problems. If you are suffering from high interrupt load, try disabling acpirt(4) or if that does not work, acpi(4