Sound distortion, white noise

[fer1578]

Hi, I'm using an Audigy 4 on Gentoo linux 64 bits, after upgrading the kernel to 2.6.33 I started to get terrible distortions, white noise sometimes when playing a video or listening to a mp3. It's not happening all the time, for example if I open a video with a mplayer and get the white noise, I can close it and open it again and the sound plays normally, other times it continues to get white noise no matter how many times I close and open it. When that is happening it also happens in all other apps that uses sound (sound players, other video players, gnome sounds) and the white noise remains even if I restart oss. The only solutions without having to restart the pc is to wait a little and everything returns to normal.
After upgrading to kernel 2.6.34 it's even worse as it's happening almost all the time.
For now I reverted back to kernel 2.6.32 and everything works as expected but I would to get this problem fixed and be able to upgrade the kernel.
I tried running the apps from the terminal and see if something useful appears when the problem arise but all the output was the normal one. I also reinstalled oss several times without luck.
I'll appreciate any suggestions to resolve this issue.

Code: Select all

Version info: OSS 4.2 (b 2002/201005081836) (0x00040100) OSS_HG
Hg revision: changeset: 891:cac211efd60d, tag: tip, date: Tue Apr 20 22:49:24 2010 +0300, summary: Haiku port: reserve PCI devices to avoid conflicts with native drivers
Platform: Linux/x86_64 2.6.32-gentoo-r7 #6 SMP Wed Jun 16 23:29:21 ART 2010 (phenom)

Number of audio devices:   8
Number of audio engines:   17
Number of MIDI devices:      1
Number of mixer devices:   1


Device objects
0: osscore0 OSS core services
1: oss_sblive0 SB Audigy4 interrupts=4419804 (4419804)

MIDI devices (/dev/midi*)
0: Audigy UART (MIDI port 1 of device object 1)

Mixer devices
0: SB Audigy4 (STAC9750) (Mixer 0 of device object 1)

Audio devices
SB Audigy4 main                   /dev/oss/oss_sblive0/pcm0  (device index 0)
SB Audigy4 front out              /dev/oss/oss_sblive0/pcm1  (device index 1)
SB Audigy4 side out               /dev/oss/oss_sblive0/pcm2  (device index 2)
SB Audigy4 center/lfe out         /dev/oss/oss_sblive0/pcm3  (device index 3)
SB Audigy4 rear out               /dev/oss/oss_sblive0/pcm4  (device index 4)
SB Audigy4 extra out              /dev/oss/oss_sblive0/pcm5  (device index 5)
SB Audigy4 raw S/PDIF (output only)  /dev/oss/oss_sblive0/pcm6  (device index 6)
SB Audigy4 5.1 output device      /dev/oss/oss_sblive0/pcm7  (device index 7)

Nodes
  /dev/dsp -> /dev/oss/oss_sblive0/pcm5
  /dev/dsp_in -> /dev/oss/oss_sblive0/pcm0
  /dev/dsp_out -> /dev/oss/oss_sblive0/pcm0
  /dev/dsp_ac3 -> /dev/oss/oss_sblive0/pcm0
  /dev/dsp_mmap -> /dev/oss/oss_sblive0/pcm0
  /dev/dsp_multich -> /dev/oss/oss_sblive0/pcm7


[dev]

sounds like the memory may not be aligned - just a guess.

[cesium]

That's not the latest OSS version, but the difference isn't too great. Can you check with 2.6.33 if b2003/2.6.34 fails? Maybe by sifting via kernel git logs we'll be able to find what's going on... Also, you may wish to try this diff.

[fer1578]

Thanks for the replies. I updated to b2003 and it seems to be fixed, or at least I can't make it happen again. I'll continue to test it a bit more and let you know how it goes.

[fer1578]

Some update on the issue. After rebooting the system a couple of days later the problem returned, I tried reinstalling but didn't fix it. After that I applied the patch cesium suggested and it worked again, rebooted the system and it continued to work. I hope this solves it definitively.

[gott_im_himmel]

hello,
I have the same problem. I'm using oss from mercurial(changeset 917). Removing /usr/lib/oss or restarting oss solves the issue temporarily. Is patch posted by cesium(link's not working by the way) applied to oss? If not, would you mind, cesium, uploading it again?
thanks

[cesium]

Try this:

Code: Select all

diff -r e300d16046f7 kernel/drv/oss_sblive/oss_sblive.c
--- a/kernel/drv/oss_sblive/oss_sblive.c   Wed Oct 20 04:14:01 2010 +0300
+++ b/kernel/drv/oss_sblive/oss_sblive.c   Mon Nov 01 22:54:50 2010 +0200
@@ -2515,7 +2515,7 @@

   if (devc->feature_mask & SB_AUDIGY)
     {
-      memlimit=MEMLIMIT_32BITS;
+      memlimit=MEMLIMIT_31BITS;
       sblive_write_reg (devc, 0x5e, 0, 0xf00);   /* ?? */
       sblive_write_reg (devc, 0x5f, 0, 0x3);   /* ?? */
     }

[gott_im_himmel]

I can't apply this patch
"Hunk failed at #1."

[cesium]

Not sure what went wrong... Just do it manually. Or even "sed -i -e s/32BITS/31BITS/ kernel/drv/oss_sblive/oss_sblive.c"

[gott_im_himmel]

I managed to apply the patch, but the issue is still not solved. OSS works better now, noise appears rarely, but it still appears.What do I do?

[nexx]

Same here. Applied the patch, rebuilt & reinstalled, but the noise is still there
Audigy 1 on source from 05-jan-11 with squeeze and 2.6.32-5-amd64
And no sound on flash anymore, oh greatness

Edit: got it to work. see attachment. content: libflashplayer.so for 64bit, needs to go to /usr/lib/flashplugin-nonfree/

[igorzwx]

Same here. Applied the patch, rebuilt & reinstalled, but the noise is still there
Audigy 1 on source from 05-jan-11 with squeeze and 2.6.32-5-amd64
And no sound on flash anymore, oh greatness

Edit: got it to work. see attachment. content: libflashplayer.so for 64bit, needs to go to /usr/lib/flashplugin-nonfree/

Do you think that "necrobumping" is a good idea?
https://wiki.archlinux.org/index.php/Fo ... Bumping.27

[cesium]

At the end, a different change helped gott_in_himmel. IIRC, it was this:

Code: Select all

diff -r 09a210f84f55 setup/Linux/oss/build/osscore.c
--- a/setup/Linux/oss/build/osscore.c   Thu Jan 06 07:55:36 2011 +0200
+++ b/setup/Linux/oss/build/osscore.c   Sat Jan 08 17:42:37 2011 +0200
@@ -808,7 +808,7 @@
     flags |= GFP_DMA32;
#endif

-  if (memlimit < 0x00000000ffffffffLL)
+  if (memlimit <= 0x00000000ffffffffLL)
     flags |= GFP_DMA;

   start_addr = NULL;

[nexx]

Do you think that "necrobumping" is a good idea?
https://wiki.archlinux.org/index.php/Fo ... Bumping.27

Not really, but i just didnt look at the date of the last post, sorry 'bout this.
@cesium: well, i think that made it. no noise for a couple of hours and many pause/play, thanks!