Re: [Alsa-devel] Slab corruptions & Re: 2.6.17-rc1: Oops in sound applications

[Jan Niehusmann <jan@gondor.com>]

On Wed, Apr 05, 2006 at 01:19:11AM +0200, Jan Niehusmann wrote:
> ...which happens to be the commit which contains the bug you already
> mentioned. I wonder if there is a second one hidden somewhere in that
> commit, or if git-bisect led me to that bug while the second one is
> hidden in a different commit...

Well, I did some additional debugging on that, working on git version
3bf75f9b90c981f18f27a0d35a44f488ab68c8ea:

In snd_pcm_oss_release() I added 
snd_assert(substream != NULL, return -ENXIO); 
in front of the first access to substream->pcm, which leads to

Apr  5 02:13:13 knautsch kernel: [17180638.784000] BUG? (substream != ((void *)0))

when opening /dev/dsp for write.

If I now add the patch you suggested, correcting the check in
snd_pcm_oss_open_file(), accessing /dev/dsp instead leads to EINVAL.

So I guess git bisect really lead me to this already known bug.

Jan