On Wed, 04 Jun 2008, Shai Berger wrote: > I checked. The problem reproduces, as described, when the nvidia blob is not > tainting anything.
okay great! > The dmesg lines when alsaconf is run are > > [ 43.386828] ACPI: PCI interrupt for device 0000:00:1b.0 disabled > [ 55.508800] ACPI: PCI Interrupt 0000:00:1b.0[A] -> GSI 22 (level, low) -> > IRQ 22 > [ 55.508823] PCI: Setting latency timer of device 0000:00:1b.0 to 64 > > (middle line got broken here, but these are the last three lines of the > kernel > log in the attached info). sure this just module forced unloading + loading. > > > > hmm > > still not buying the kernel error yet. > > > > I wouldn't either, if I were you :-). Actually, I suspect a subtle bug > higher-up exposed by the new kernel. However, since alsa and artsd weren't > updated when the problem appeared, and returning to the old kernel makes the > problem go away, I think the new kernel is guilty until proven otherwise. unconvincing believe ;) > > > > what sound application are you using? > > Amarok, KLettres, Iceweasel with Adobe flash, KDE system notifications > (login, > error, logout). ok one of those might be the problem, maybe one of this is still oss based? > I'm attaching four files: output from this script under kernel 2.6.24 (where > there's no problem), under kernel 2.6.25 when sound works (that is, right > after an alsaconf run), and under kernel 2.6.25 when sound has failed; plus > the reportbug header and footer generated without the nvidia binary driver. > > If you have any more ideas for data that can help, just shoot. I'd really > rather see this bug solved. thanks according to your output in the "not working" 2.6.25 case it looks like some app is using the sound card. so please retry with 2.6.25 and give us the following output (working, + non working case) ps aux lsof /dev/snd/* best regards -- maks -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
