I think waiting a little bit is the sensible and least worrysome fo me.
> Seth is correct, this problem is fixed in later kernels, you can use the cvs
> tree - which I admit does have other problems right now on highmem systems,
> or you can wait a little bit and I will update the patches on the ftp
> site in /projects/xfs/download/patches - the code which is there now does
> not contain the final version of the fix (and I need to go through a few
> hoops to create a new patch at this os level).