On Mon, Aug 10, 2015 at 09:33:11PM -0700, Tim Kientzle wrote: > I have a guess, though: I noticed that the old bufwait/dirhash LOR is being triggered before this. Could the backtrace from that be teaching Witness a bogus ufs -> kld ordering? I do not see how it could. The (directory) buffer lock is after the directory vnode lock, and both of them should be after the kld_sx. If some buffer lock appears before kld_sx, this is the same problem as with the vnode lock.Received on Tue Aug 11 2015 - 02:43:33 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:59 UTC