Hi folks, The extent swap code is not symmetrical, so correct behaviour is dependent on userspace doing the right thing. If we try to swap extents with the temporary inode in the wrong state we can have interesting failures. These two patches fix those issues and treat the two files identically in terms of the state of the data contained in the files being swapped. Failure to write or remove all cached data from either file will now cause the swap extents operation to fail. Comments, thoughts? Cheers, Dave. _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
