Hi! I finally got the chance to take a look at the dpkg FTBFS on Hurd (after recovering my VM! :), and it appears as if it was a fakeroot problem?
Building dpkg w/o fakeroot works fine. And running the specific test w/ and w/o fakeroot shows the problem: ,--- $ apt-get source dpkg $ cd dpkg-*/ $ ./configure $ make -C scripts check <success> $ cd scripts/t.tmp/Dpkg_Source_Patch/indent-header-tree $ ls dummy-file symlink $ ls ../indent-header-out/ <empty> $ LC_ALL=C LANG=C PATCH_GET=0 patch -t -F0 -N -p1 -u -Vnever -b \ -z .dpkg-orig <../../../t/Dpkg_Source_Patch/indent-header.patch (Patch is indented 2 spaces.) Invalid file name symlink/index-file -- skipping patch The next patch would create the file dummy-file, which already exists! Skipping patch. 1 out of 1 hunk ignored $ ls ../indent-header-out/ <empty> $ LC_ALL=C LANG=C PATCH_GET=0 fakeroot patch -t -F0 -N -p1 -u -Vnever \ -b -z .dpkg-orig <../../../t/Dpkg_Source_Patch/indent-header.patch (Patch is indented 2 spaces.) patching file symlink/index-file The next patch would create the file dummy-file, which already exists! Skipping patch. 1 out of 1 hunk ignored /bin/fakeauth: Error 1 for child 8118 /bin/settrans: Error 1 for child 8117 $ ls ../indent-header-out/ index-file index-file.dpkg-orig `--- I'm not sure though why it would fail only for this test and not the others, because they are all similar in structure. Anyway, I'm not sure I'll have time to dig further, but I'd really like to get this sorted out, otherwise dpkg will be out-of-sync on the Hurd. :( I've not filed a bug yet, because I'm not entirely sure whether this is a bug in the Hurd or not. Thanks, Guillem