Hello, 2008 m. January 12 d., Saturday, David Brownell rašė: > What happens if you don't include firewire_ohci? Yeah, deleting firewire_ohci fixed the problem (with "noapic" specified).
> Must you use "noapic"? That can cause IRQ routing problems... Ok, I retested without "noapic" and, fortunately, everything worked just fine. I was using "noapic" because: 1) early kernels (up to 2.6.15, iirc) had a problem with time ticking twice normal speed with APIC enabled (my laptop has a famous ATI X200 Express chipset). 2) when 1) problem was fixed, I wasn't able to use suspend-to-mem (S3) with APIC enabled (the laptop wouldn't simply wake up). I've tested this again with 2.6.24-rc7 and the problem seems to be gone so, finally, there is no reason to use "noapic" anymore. In addition, /proc/interrupts looks much better with APIC enabled. Therefore I'll stick with "don't use noapic" solution. > The reason your audio craps out *too* is that it's one of the > zillion devices sharing IRQ #11. Bugs in any one of their > drivers could cause problems like this. If you are 100% sure that it was not usb-ohci fault (could be firewire_ohci's then), then let it be. However, reverting that patch _helped_ for me previously (with "noapic"). Hopefully, other people won't get bitten by the same problem. -- Modestas Vainius <[EMAIL PROTECTED]>
signature.asc
Description: This is a digitally signed message part.
