In your remarks and in your code you've used "getuid" a few times where you should have used "getpid". I'll assume that was just a persistent typo and nothing more.
None of this new complexity is worth the trouble, nor was the old complexity. The `getpid' function simply returns the value of a global variable (_hurd_pid) and so is very cheap. I've changed both filesystems to just always call getpid in read_inode/read_disknode. Note that both diskfs_set_statfs implementations already do this. In yet another example of why you should separate semantically distinct changes into separate patches, I would be happy to take your comment changes and removal of superfluous initializers if I got that patch alone. _______________________________________________ Bug-hurd mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/bug-hurd
