Fails to umount

[David Zeuthen]

On Thu, 2005-01-27 at 19:41 +0100, Pierre Ossman wrote:
> >It's probably interesting to see the hald output of why unmounting
> >fails in this case; does hald attempt to lazy unmount?
> >  
> >
> Can it do any other? I could only find one in the code and that always 
> has the '-l' included.

Right.

> And as I stated before, it doesn't attempt to do even that in the 
> failure case since the condition 'volume.is_mounted' isn't satisfied. I 
> haven't had the time to study the design of hal enough to determine how 
> this can happen. No other tool tries to unmount the device, so it is hal 
> that screws up.
>

Uhm, unmount -l could fail. Theoretically at least. Another thing is
that you shouldn't just yank out devices without unmounting them, but,
yeah, I agree that a system should survice this. So does all the
interaction designers I've spoken with.

> I'll send you a complete dump when I have time to run the tests again. 
> Be warned though, it will probably be huge.
> 

Looking forward to that. Just cut out the relevant parts and it
shouldn't be too large. Thanks for looking into this.

Cheers,
David


_______________________________________________
hal mailing list
hal at lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/hal