Re: Old, unnoticed bug in OSS output plugin.

Could you please send a patch?

On Mon, 2005-03-28 at 11:27 -0800, Sean Meiners wrote:
> While using XMMS's OSS output plugin to help me test an OSS emulation library 
> that uses bio2jack I found a sneaky little bug that managed to go unnoticed 
> for quite some time.
> 
> In write_all in audio.c it goes into a loop until all data is written.  The 
> problem is that while it reduces the number of bytes to write based on how 
> many were written in the previous loop it forgets to advance the buffer, 
> causing some very odd sound effects.  I suspect this has gone unnoticed 
> because the OSS drivers themselves don't seem to do short-writes.  However, 
> my emulation library did.  When I modified my code to prevent short-writes 
> the problem went away.  It's not a high-priority issue since non-emulated OSS 
> devices don't seem to exhibit this and emulated ones probably shouldn't do 
> partial-frame writes, but it's a bug nonetheless and easy to fix at that.
> 
> PS: I tried filing a bug in the XMMS bugzilla system, but I'd rather not open 
> a bugzilla account just for this, especially since I don't expect to be doing 
> any future work on XMMS.  Oh, and I'm not subscribed to this list either, so 
> keep that in mind if you reply and want me to see it.

_______________________________________________
xmms-devel mailing list
[email protected]
http://lists.xmms.org/mailman/listinfo/xmms-devel

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQBCUYs9OpjtAl+gMRURAvzTAJ0WHcXyWM3oeuZ2CyfTTmA14sz0qQCgx8l+
z4p6CV5zy/rN2kg+s62jlb0=
=/9h3
-----END PGP SIGNATURE-----