Hi,

this patch series addresses a regression reported by Mike in Git v2.44
where remote helpers cannot access the Git repository anymore when
running git-clone(1).

The root cause of this is that we have started to initialize the refdb
at a later point, after the helper is spawned. This is required such
that we can initialize it with the correct object format, which fixes
clones of SHA256 repositories with the reftable format and when using
bundles.

The proposed fix here is to partially initialize the refdb with just
enough data such that it can be discovered. The fix isn't pretty, but
addresses the issue. I also couldn't come up with a better idea than
this.

Patrick

Patrick Steinhardt (2):
  refs/reftable: don't fail empty transactions in repo without HEAD
  builtin/clone: allow remote helpers to detect repo

 builtin/clone.c            | 46 ++++++++++++++++++++++++++++++++++++++
 refs/reftable-backend.c    |  1 +
 setup.c                    |  9 +++++++-
 t/t0610-reftable-basics.sh | 13 +++++++++++
 t/t5801/git-remote-testgit |  5 +++++
 5 files changed, 73 insertions(+), 1 deletion(-)


base-commit: a2082dbdd315aa4dd3f315545e5b3ab3b3e2d894
-- 
2.44.0