Hi! I'm encountering what I believe to be a bug in xfs_db with some code that worked previously for me. I have some code that uses xfs_db to copy some specific data out of an XFS disk image based on an inode number. Basically, it does similar to: inode [inodenum] dblock 0 p dblock 1 p dblock 2 p [etc] This worked on older versions of xfs_db but is now resulting in corrupted output. I believe I've traced the issue to the code introduced in https://git.kernel.org/pub/scm/fs/xfs/xfsprogs-dev.git/commit/?id=b05a31722f5d4c5e071238cbedf491d5b109f278 to support realtime files. Specifically, the use of a `dblock` command when the previous command was not an `inode` command looks to lead to the data in iocur_top->data to no longer contain inode data as expected in the line if (is_rtfile(iocur_top->data)) I don't know the code well enough to submit a fix, but some quick experimentation suggested it may be sufficient to move the check for an rtfile to inside the push/pop context above after the call to set_cur_inode(iocur_top->ino). Hopefully this describes the issue sufficiently but please let me know if you need anything else from me. -Tom Samstag NetRise.io
