Gerd Hoffmann <[email protected]> writes: > This way the tests run without sound hardware being present > on the build machine. Even with sound hardware it IMO isn't > very useful to use it in regression testing. Once the sound > card tests are advanced enougth that they try to actually > play sound we probably want the guests sound output written > to a file (via QEMU_AUDIO_DRV=wav) rather than played on the > build machines sound hardware. > > Signed-off-by: Gerd Hoffmann <[email protected]>
Bonus: I no longer get "audio: Could not init `oss' audio driver" noise. Reviewed-by: Markus Armbruster <[email protected]>
