Re: gnome-vfs freeze break request

[Alexander Larsson <alexl redhat com>]

On Wed, 2005-08-31 at 13:37 +0200, Murray Cumming wrote:
> > While debugging a nautilis bug i noticed a strange crash in the
> > gnome-vfs xdgmime system with valgrind. It seems likely that this was
> > caused by missed locks similar to those fixed in bug 170947. I've
> > repopened that bug and added a new patch that wraps the remaining
> > xdgmime calls with locks.
> >
> > Is it ok to commit this to HEAD? (Patch is attached.)
> 
> It's hard for me to know how safe this is. How much have you tested this ?

I haven't tested it much really, apart from just running nautilus a bit
with it. But I think its pretty safe. 

The patch itself contains all references to the lock (due to the
rename), and the patch (I'm sure you agree) doesn't ever leave the lock
locked after a critical region, nor do the addition of the locking
change the semantics of the code (ignoring threads). So, the risk would
be the addition of the locking causing a deadlock somehow. I don't think
that can happen, because all the locking are around calls to the
xdg_mime subsystem, which in itself never calls back to the normal
gnome-vfs code.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 Alexander Larsson                                            Red Hat, Inc 
                   alexl redhat com    alla lysator liu se 
He's an ungodly dishevelled assassin haunted by an iconic dead American 
confidante She's a sharp-shooting African-American magician's assistant from 
beyond the grave. They fight crime!