On Sun, Mar 24, 2019 at 12:19:20PM +0200, Niko Tyni wrote: > While the current test data files from stretch indeed cannot be read > on sid, I cannot make new test data files on stretch that reproduce the > behaviour. I've created 100 such databases on each affected architecture > and all of those are readable on sid. > > I note that the test data files we're shipping for stretch versions of > ppc64el and arm64 are identical, which seems improbable. My best guess > is that I've mixed up the test data files from different architectures > when copying them around and committing them to git. How embarrassing.
Trying to fix things for a new upload I finally understood what went wrong here. The stretch version of the GDBM "compat library" created a hard link between the .dir and .pag files. The buster version checks for this and handles it as a special case. https://sources.debian.org/src/gdbm/1.8.3-14/dbminit.c/#L99 https://sources.debian.org/src/gdbm/1.18.1-4/compat/dbmopen.c/#L78 As git doesn't store hard links, this broke when I copied the files around and committed them to git. What a mess. At least I now know I wasn't *that* careless :) Possibly I need to tar the files, or just recreate the hard links in the autopkgtest check. Will try to come up with something tomorrow. -- Niko
