APM messes up sound on dell dimension 5000

[heinjan]

After my computer isn't used for a while, every sound played will repeat
forever. This can only be stopped by rebooting, even a /etc/init.d/alsa restart
isn't sufficient.

In /var/log/messages the following lines are relevant:

Feb 3 12:55:18 localhost -- MARK --
Feb 3 12:57:25 localhost kernel: apm: BIOS version 1.2 Flags 0x03 (Driver
version 1.16ac)
Feb 3 12:57:25 localhost kernel: mtrr: base(0xc0020000) is not aligned on a
size(0x180000) boundary
Feb 3 12:57:25 localhost kernel: drivers/usb/input/hid-input.c: event field not
found
Feb 3 12:57:25 localhost kernel: [__report_bad_irq+42/143]
__report_bad_irq+0x2a/0x8f
Feb 3 12:57:25 localhost kernel: [handle_IRQ_event+46/100]
handle_IRQ_event+0x2e/0x64
Feb 3 12:57:25 localhost kernel: [note_interrupt+124/223] note_interrupt+0x7c/0xdf
Feb 3 12:57:25 localhost kernel: [__do_IRQ+318/330] __do_IRQ+0x13e/0x14a
Feb 3 12:57:25 localhost kernel: [do_IRQ+25/36] do_IRQ+0x19/0x24
Feb 3 13:02:09 localhost shutdown[10130]: shutting down for system reboot

The reboot is initiated by myself because the system becomes unworkable. I use
hoary with kernel image 686 version 2.6.10-3. I boot with acpi=off, because I
have a dell dimension 5000 which won't boot with acpi on. (No idea if this is a
known bug, maybe should file another one?) My soundcard uses the ice1712 alsa
driver.

I will try booting with noapm to disable apm as workaround. Haven't tried it yet
though.

[Fabio Massimo Di Nitto (fabbione)]

Please try a more recent version of the kernel. We are up to 2.6.10-13
and try to boot with the option irqpoll.

Fabio

[heinjan (heinjan)]

My fault: I'm running the latest kernel (2.6.10-13); the 2.6.10-3 is the
dependancy-package.

I will boot with irqpoll to see if there's some more info.

[Chuck Short (zulcss)]

Could you tell us the results of your test?

Thanks
chuck

[heinjan (heinjan)]

After booting with irqpoll none of these hangs occured. Also, a new kernel is
installed on the machine. So maybe it is fixed... I'm still booting with irqpoll
though, but if you want you can close this bug. I will reopen it if there is
another hang.

[Matt Zimmerman (mdz)]

Would you please remove the irqpoll parameter and confirm that the problem is
fixed in that case as well?

[heinjan (heinjan)]

Without irqpoll the problems start again. If I kill polypaudio, the hanging
sound stops. If I start polypaudio again, the stuttering sound comes back if I
play a sound. The same with esound.

The current kernel used is 2.6.10-31. The last interesting messages in
/var/log/messages are:

Mar 28 19:12:59 localhost kernel: [__report_bad_irq+42/160]
__report_bad_irq+0x2a/0xa0
Mar 28 19:12:59 localhost kernel: [handle_IRQ_event+48/112]
handle_IRQ_event+0x30/0x70
Mar 28 19:12:59 localhost kernel: [note_interrupt+128/240] note_interrupt+0x80/0xf0
Mar 28 19:12:59 localhost kernel: [__do_IRQ+318/336] __do_IRQ+0x13e/0x150
Mar 28 19:12:59 localhost kernel: [do_IRQ+25/48] do_IRQ+0x19/0x30
Mar 28 19:12:59 localhost kernel: [common_interrupt+26/32]
common_interrupt+0x1a/0x20
Mar 28 19:12:59 localhost kernel: [mwait_idle+51/80] mwait_idle+0x33/0x50
Mar 28 19:12:59 localhost kernel: [cpu_idle+58/96] cpu_idle+0x3a/0x60
Mar 28 19:12:59 localhost kernel: [start_kernel+386/448] start_kernel+0x182/0x1c0
Mar 28 19:12:59 localhost kernel: [unknown_bootoption+0/480]
unknown_bootoption+0x0/0x1e0
Mar 28 19:33:17 localhost -- MARK --

I will boot again with the irqpoll option.

I'm not using the onboard sound of the dell, but a pci soundcard with the
snd_ice1712 alsa driver. The lspci output of this card is:

0000:04:00.0 Multimedia audio controller: VIA Technologies Inc. ICE1712 [Envy24]
PCI Multi-Channel I/O Controller (rev 02)

[heinjan (heinjan)]

Extra info: if I play a sound with mplayer with the alsa output the same
stuttering happens. However, using the oss output there is still heavy
stuttering, but less than with the alsa output.

[Matt Zimmerman (mdz)]

This is similar to the problem I have in bug #11671

[Ben Collins (ben-collins)]

This bug has been marked as a duplicate of bug 11671.