On Wed, 17 May 2000, Richard Gooch wrote:
> Alexander Viro writes:
> > To fix:
> > race in devfs_open() (bogus use of ->d_count; waiting for comments from
> > Richard),
>
> Yep. I'm busy with conferences at the moment, so please just hang
> on. There was a good reason for it.
OK with me, just tell what that reason was (when you'll have spare time,
that is). As it is, code can be easily tricked into missing the first
open() - if you have something else holding the dentry while you are
opening the thing (e.g. stat() with buffer mmaped on a file on floppy).
If it's "do it once, just make sure that open() will not return until
that's done" - fine, we'll need some flag instead of ->d_count. If it
actually depends on number of openers - well, so we'll need a counter...
So far I'm leaving it as-is, but I hope you'll be back in some reasonable
time - it's racey right now, regardless of any dcache rewrites.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed May 17 2000 - 23:05:01 EDT