Kernel bug #394392

[Jeff Licquia <licquia@debian.org>]

Per your blog post this morning
(http://www.itp.tuwien.ac.at/~mattems/blog/2006/11/07#2.6.18-status),
would you mind looking into the status of bug #394392?

I haven't seen any response from the kernel team, and as it's a LSB
issue, I'm very concerned that the fix gets into the kernel ASAP.

I've attached a patch to the bug (straight from the Fedora Core kernel
package, and originally derived from the Linus tree for 2.6.19), but if
that patch is deemed to be too much for the current state of the kernel,
I can create an alternate patch which reverts msync() to the behavior in
2.6.16 and earlier.

Thanks!