On Wed, Apr 06, 2011 at 09:38:17AM -0400, Christoph Hellwig wrote: > On Wed, Apr 06, 2011 at 04:19:10PM +1000, Dave Chinner wrote: > > + ext_buffer = kmem_alloc(XFS_IFORK_SIZE(ip, whichfork), > > + KM_SLEEP | KM_NOFS); > > As mentioned before the KM_NOFS is a change from the previous version > and really should not be needed. Ok. If you feel strongly enough to point it out a second time, I'll fix it ;) > Also the feedback that the new helper needs to be under the > !XFS_NATIVE_ENDIAN ifdef wasn't picked up either. The function is needed regardless of XFS_NATIVE_HOST due to the fact it is needed to pick the real extents out of the n-core data fork when delayed allocation extents are present. This is endian independent. Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs