Eric Blake wrote: > Pádraig Brady <P <at> draigBrady.com> writes: > >> ln: creating hard link `hardlink' => `symlink': Invalid argument >> >> `man linkat` says that AT_SYMLINK_FOLLOW is only supported since 2.6.18 >> and my FC5 system is 2.6.17 > > Bingo. For FC5, I need to implement rpl_linkat in gnulib, which mimics the > link_follow fallback used on systems that lack linkat altogether, if linkat > (,AT_SYMLINK_FOLLOW) fails with EINVAL. And probably enable rpl_linkat > unconditionally at configure time for Linux, for the benefit of people that > compiled against newer headers but run against an older kernel; the > replacement > code would need a static variable that learns whether we expect success or > failure, similar to how we've wrapped O_CLOEXEC support in recent modules when > compiling against newer headers and older kernel. > > Jim, do you want to wait for me to get that done in gnulib, or do we just > state > that since 8.0 will be beta that it is a known bug which will be fixed for > 8.1?
I'm ready to release today, and that failure is far enough out on the fringe that I have no qualms. This makes two known bugs. The fix for the other will be always-required l?getfilecon wrappers to avoid segfaulting on old-kernel-munged disks and/or old versions of libselinux.