The tarball appears to have been re-uploaded, because both
Profpatsch[1] and r-rmcgibbo[2] got the previous hash, but by the time
it got to Hydra, the hash had changed[3].
By getting Profpatsch to send me his cached tarball, and comparing
each to the s6-rc git repo, I've determined that the difference
between them is a fast-forward of 7cadbf1..3d1af07 (summarised below).
So I think we're fine to bump the hash — it looks like some commits
were just mistakenly excluded the first time round.
Laurent Bercot (2):
Add lock-fd support
version: 0.5.2.3
NEWS | 1 +
src/libs6rc/s6rc_servicedir_internal.c | 1 +
src/s6-rc/s6-rc-compile.c | 9 +++++++++
3 files changed, 11 insertions(+)
[1]: https://github.com/NixOS/nixpkgs/pull/139544
[2]: https://github.com/NixOS/nixpkgs/pull/139544#issuecomment-927332319
[3]: https://hydra.nixos.org/build/154845872
fc24bb88c3