OSS playback via xine is broken.

[adamk]

All four methods work fine with snd_emu10kx, including no problems when positioning further down the stream.

Adam

[adamk]

Any more thoughts on the matter?

Adam

[cesium]

There has been a report of problems with xine and vmix on Linux (here), not sure if this is the same issue, since you have problems without vmix too. Another (possibly unrelated) thing I've noticed: If you Ctrl-C osstest while it's playing, it will stop only after a bit more is played. On Linux this is instantaneous.

[adamk]

Are you guys considering this a bug in xine or a bug in the oss drivers?

Adam

[cesium]

I'm not quite sure yet (testing on FreeBSD is quite difficult for me as it doesn't support my network card. VBox works sometimes, so I'll try it).

[adamk]

Any luck yet?

[adamk]

Seriously, any update on this? So much multimedia software out there uses the xine engine, it makes OSS4 practically unusable on FreeBSD.

Adam

[cesium]

I've been very busy lately. Is this the same machine that has the mplayer related crash?

[adamk]

The crash happens only on one machine. The xine issue happens on the two FreeBSD machines that I have SBLive! cards in.

EDIT: One of those machines also has a ICH5 ac97, which has the same problem with the xine engine.

Adam

[adamk]

And, in fact, while testing xine (specifically, moving the progress bar around to skip to various points in the song), that machine locked up again. So mplayer is not the only trigger for the crash.

I think I'll switch to the ICH sound chip to see if I can get a kernel crash from that.

Adam

EDIT: It also happens on the intel hda sound chip on one of the two machines (the one that doesn't have the ich, obviously).