Hi! I released a new dpkg yesterday, with new support to try to print the executable name of the contending process holding the database lock, but it looks like I had not really tested the changes on GNU/Hurd (more on that on another mail).
This should now be fixed in dpkg git with: https://git.dpkg.org/cgit/dpkg/dpkg.git/commit/?id=127d25ed1220221f03d9bf11b906f89a065fcfad so there is no need to file a report about this. I'm going to wait a bit in case there are other regressions, but otherwise plan to do another release before this next weekend at most. While testing this I then realized that the fcntl F_GETLK call always returns with -1 in the l_pid member, and investigating a bit I found in hurd.git/libfshelp/rlock-tweak.c the following code: /* XXX: This will be fixed when the proc_{user,server}_identify RPCs are implemented */ lock->l_pid = -1; I was quickly skimming over the bug-hurd mailing list, and I think there have been some references to those RPCs in the discussions around the locking patches by Neal and then Svante, but I'm not sure what's the current status. It would be nice though if some one would get this implemented! :) Thanks, Guillem
