Re: DoS with POSIX file locks?

[Trond Myklebust]

On Wed, 2006-03-22 at 07:21 +0100, Miklos Szeredi wrote:
> > i concur with Trond, there's no sane way to get rid of it w/out
> > formalizing CLONE_FILES and locks on exec
> 
> Probably there is.  It would involve allocating a separate
> lock-owner-ID stored in files_struct but separate from it.  But it's
> more complicated than simply not propagating locks on exec in the
> CLONE_FILES case.

That doesn't solve the fundamental problem.

You would still have to be able to tell a remote server that some locks
which previously belonged to one owner are being reallocated to several
owners. That is not something that NFS, CIFS, AFS,... support, and I'd
be surprised if any of the clustered filesystems support it either.

Cheers,
  Trond

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/