Bug#2454: mount-2.5-5 hangs when type not specified
> Here's the fix. Because of the block_signals (SIG_BLOCK) at line
> 482, the SEGV from strcasecmp(NULL,"auto") is blocked, and strcasecmp
> spins (probably retrying at that address? not sure) with signals
> blocked until you hit it with a kill -9. Please propagate this back to
> the upstream maintainer. (Simply leaving *type == NULL is enough,
> since that's equivalent to explicitly specifying "auto"...)
Thanks. The upstream maintainer has already in effect fixed this; I'll be
releasing a new (temporary?) mount package to test all of the changes to date.
Fwiw, a future version of `umount' will also restore the -n option.