[alsa-devel] Problem with snd_hda_intel from 05-27-2010 snapshot
Hi. I had been mostly successfully using the snd_hda_intel driver from the stable kernel 2.6.34 for HDMI audio with my Nvidia GT220 video card. In an attempt to see if any recent changes to snd_hda_intel would fix a problem I've been having with "pcm_hw.c: snd_pcm_hw_delay() SNDRV_PCM_IOCTL_DELAY failed" I decided to download the latest snapshot. Unfortunately with these drivers I have no audio. All my settings, and everything have remained the same. The only noticable difference I've found is /sys/module/snd_hda_intel/parameters/probe_only is 0,0,0,0,0,0,0,0 in the snapshot drivers, and N,N,N,N,N,N,N,N in the 2.6.34 kernel drivers (which are compiled into the kernel).
Any help?
Thanks, Derek
Hi. I had been mostly successfully using the snd_hda_intel driver from the stable kernel 2.6.34 for HDMI audio with my Nvidia GT220 video card. In an attempt to see if any recent changes to snd_hda_intel would fix a problem I've been having with "pcm_hw.c: snd_pcm_hw_delay() SNDRV_PCM_IOCTL_DELAY failed" I decided to download the latest snapshot. Unfortunately with these drivers I have no audio. All my settings, and everything have remained the same. The only noticable difference I've found is /sys/module/snd_hda_intel/parameters/probe_only is 0,0,0,0,0,0,0,0 in the snapshot drivers, and N,N,N,N,N,N,N,N in the 2.6.34 kernel drivers (which are compiled into the kernel).
As a followup, I tried each snapshot going backwards from 05-27-2010 to discover where the breakage occured. The last working snapshot is 05-11-2010, it breaks at 05-12-2010. After making a diff of the sound/pci/hda/ dir, I've discovered that the offending patch is:
author Jaroslav Kysela perex@perex.cz Tue, 11 May 2010 08:21:46 +0000 (10:21 +0200) committer Jaroslav Kysela perex@perex.cz Tue, 11 May 2010 08:25:50 +0000 (10:25 +0200) commit 432f57d106e6d5c69fce74c888de32c6df347494 tree 536120322429edb82d672547595d8dba7ac10b5e tree | snapshot parent 533538fe6243d87b68db6942b1c7a1acb9fe504d commit | diff
[ALSA] snd-hda-intel: use WALLCLK register to check for early irqs
Use 24Mhz WALLCLK register to ignore too early interrupts and wrong interrupt status. The bad timing confuses the higher ALSA layer and causes audio skipping. More information about behaviour and debugging can be found in kernel bz#15912.
https://bugzilla.kernel.org/show_bug.cgi?id=15912
Signed-off-by: Jaroslav Kysela perex@perex.cz
sound/pci/hda/hda_intel.c
On Wed, 26 May 2010, VDR User wrote:
Hi. I had been mostly successfully using the snd_hda_intel driver from the stable kernel 2.6.34 for HDMI audio with my Nvidia GT220 video card. In an attempt to see if any recent changes to snd_hda_intel would fix a problem I've been having with "pcm_hw.c: snd_pcm_hw_delay() SNDRV_PCM_IOCTL_DELAY failed" I decided to download the latest snapshot. Unfortunately with these drivers I have no audio. All my settings, and everything have remained the same. The only noticable difference I've found is /sys/module/snd_hda_intel/parameters/probe_only is 0,0,0,0,0,0,0,0 in the snapshot drivers, and N,N,N,N,N,N,N,N in the 2.6.34 kernel drivers (which are compiled into the kernel).
As a followup, I tried each snapshot going backwards from 05-27-2010 to discover where the breakage occured. The last working snapshot is 05-11-2010, it breaks at 05-12-2010. After making a diff of the sound/pci/hda/ dir, I've discovered that the offending patch is:
author Jaroslav Kysela perex@perex.cz Tue, 11 May 2010 08:21:46 +0000 (10:21 +0200) committer Jaroslav Kysela perex@perex.cz Tue, 11 May 2010 08:25:50 +0000 (10:25 +0200) commit 432f57d106e6d5c69fce74c888de32c6df347494 tree 536120322429edb82d672547595d8dba7ac10b5e tree | snapshot parent 533538fe6243d87b68db6942b1c7a1acb9fe504d commit | diff
[ALSA] snd-hda-intel: use WALLCLK register to check for early irqs
Use 24Mhz WALLCLK register to ignore too early interrupts and wrong interrupt status. The bad timing confuses the higher ALSA layer and causes audio skipping. More information about behaviour and debugging can be found in kernel bz#15912.
Check any suspicious lines in /var/log/messages and follow proc xrun_debug settings noted in bug above. I need at least 20 first lines after the playback is started.
Also, use 'aplay -Dplughw:0' from the alsa-utils package to trigger the playback.
Jaroslav
----- Jaroslav Kysela perex@perex.cz Linux Kernel Sound Maintainer ALSA Project, Red Hat, Inc.
participants (2)
-
Jaroslav Kysela -
VDR User