[PATCH v2 0/7] Fix some problems with reflog expiration

This is v2 of the patch series. Thanks to Eric Sunshine, Stefan Beller, Peff, and Junio for their comments about v1 [1].

The two main changes since v1:

* Don't change the locking policy for changing symbolic references.
  Even though the existing policy is partly broken, the change
  proposed in v1 would have introduced an incompatibility with older
  versions of Git and with libgit2 and possibly other Git clients. I
  will try to implement the transition plan described in [2], but I
  will do it in a separate patch series.

* I rewrote the "git reflog" documentation more completely, following
  Junio's suggestion.

Other changes since v1:

* Rebased onto a more recent version of upstream master (because
  master now includes mh/reflog-expire and sb/atomic-push). The rebase
  was conflict-free.

* Tweaked some commit messages and added some Reviewed-by lines.

* Appended a patch by Stefan Beller to delete the "force_write" member
  of "struct ref_lock".

[1] http://thread.gmane.org/gmane.comp.version-control.git/263552 [2] http://article.gmane.org/gmane.comp.version-control.git/263811

Michael Haggerty (6):
  write_ref_sha1(): remove check for lock == NULL
  write_ref_sha1(): Move write elision test to callers
  lock_ref_sha1_basic(): do not set force_write for missing references
  reflog: improve and update documentation
  reflog_expire(): ignore --updateref for symbolic references
  reflog_expire(): never update a reference to null_sha1

Stefan Beller (1):
  struct ref_lock: delete the force_write member

 Documentation/git-reflog.txt | 139 +++++++++++++++++++++++++------------------
 builtin/reflog.c             |   4 +-
 refs.c                       |  63 +++++++++++---------
 3 files changed, 117 insertions(+), 89 deletions(-)