Hi folks, Eric pointed out that this morning that if xfs_fsr didn't fsync the target file, the swap extents ioctl deadlocked. Dumb bug, simple fix. I also made sure that swap extent treats both inodes the same in terms of flushing, page cache pages and being mmap()d so that if anyone other than xfs_fsr uses it it doesn't go boom when the second inode is not a newly created empty file. Thoughts? -Dave. _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
