https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=218512
--- Comment #10 from Frank <fhri...@gmail.com> --- (In reply to Alexander Motin from comment #9) It's not paranoia at all. If you don't specify -i, the iterations are not set to -1, they are computed and passed in. When setting passphrases on a set of disks, freenas serializes that process, but what if I wanted to parallelize it? geli is happily computing iterations on the cpus when another task happens to come in. One or more of the iterations end up getting computed differently that what it is currently set to and subsequently causing one or more passphrase changes to fail. This is but one of many scenarios of how this could really happen. Of course, I thought of all this when I made my changes, and I added tests to test those changes. But I suppose it's the state of software these days to make the easy fix for the specific use case I care about. This will be my last comment on this as I am not a committer to freebsd, but if I were, I would label the commit as incomplete. -- You are receiving this mail because: You are the assignee for the bug. _______________________________________________ freebsd-geom@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-geom To unsubscribe, send any mail to "freebsd-geom-unsubscr...@freebsd.org"
