After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 562745 - no sound using oss plugin on ezx
no sound using oss plugin on ezx
Status: RESOLVED INCOMPLETE
Product: GStreamer
Classification: Platform
Component: gst-plugins-good
0.10.17
Other All
: Normal major
: git master
Assigned To: GStreamer Maintainers
GStreamer Maintainers
Depends on:
Blocks:
 
 
Reported: 2008-11-30 10:33 UTC by Ilya Petrov
Modified: 2011-05-19 09:10 UTC
See Also:
GNOME target: ---
GNOME version: Unversioned Enhancement



Description Ilya Petrov 2008-11-30 10:33:05 UTC
Please describe the problem:
"gst-launch audiotestsrc ! osssink" opens device ("/dev/dsp") and writes a lot of 0x00 to it unstead of sound data. 

"gst-launch audiotestsrc ! filesink location=/dev/dsp" writes data (not 0x00) but sound is ugly. 

any other sources also produce a lot of 0x00.


Steps to reproduce:
1. run "gst-launch audiotestsrc ! osssink" on motorola ezx



Actual results:
no sound

Expected results:


Does this happen every time?
yes

Other information:
kernel 2.4.20, motorola a1200 (and other ezx). 

other oss software (mpd, qtopia) works well.
Comment 1 Jan Schmidt 2009-01-09 15:47:06 UTC
Did you make any progress with this? I'd think the most likely cause is that you're flooding the CPU because audiotestsrc uses float operations to generate the sine wave it outputs.

Try audiotestsrc wave=7 ! osssink 
Comment 2 Ilya Petrov 2009-01-09 20:00:18 UTC
same behavior with wave=7: zeros to dsp, sound to file.

cpu load is near 20% in both cases.


>you're flooding the CPU because audiotestsrc uses float operations to generate
the sine wave it outputs.

ezx uses patched glibc with iwmmx-specific softfloat (not via illegal instruction faults in kernel).

and i have same problem with real sources, not only with  audiotestsrc
Comment 3 Jan Schmidt 2009-02-13 17:39:32 UTC
Dropping the blocker flag from this bug - it's not a blocker for us unless it's a regression.

The only reason I can think of why osssink writes 0 is still that the writer thread is not receiving data quickly enough. It's hard to say more without either a debug log or access to the platform.
Comment 4 Sebastian Dröge (slomo) 2011-05-19 09:10:42 UTC
Let's close this as incomplete then