When we pick another commit's message, we die() immediately
if we find that it's empty and we are not going to run an
editor (i.e., when running "-C" instead of "-c").  However,
this check is redundant and harmful.

It's redundant because we will already notice the empty
message later, after we would have run the editor, and die
there (just as we would for a regular, not "-C" case, where
the user provided an empty message in the editor).

It's harmful for a few reasons:

  1. It does not respect --allow-empty-message. As a result,
     a "git rebase -i" cannot "pick" such a commit. So you
     cannot even go back in time to fix it with a "reword"
     or "edit" instruction.

  2. It does not take into account other ways besides the
     editor to modify the message. For example, "git commit
     -C empty-commit -m foo" could take the author
     information from empty-commit, but add a message to it.
     There's more to do to make that work correctly (and
     right now we explicitly forbid "-C with -m"), but this
     removes one roadblock.

  3. The existing check is not enough to prevent segfaults.
     We try to find the "\n\n" header/body boundary in the
     commit. If it is at the end of the string (i.e., no
     body), _or_ if we cannot find it at all (i.e., a
     truncated commit object), we consider the message
     empty. With "-C", that's OK; we die in either case. But
     with "-c", we continue on, and in the case of a
     truncated commit may end up dereferencing NULL+2.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

Signed-off-by: Junio C Hamano <gitster@pobox.com>

"git merge-base --octopus" used to leave cleaning up suboptimal
result to the caller, but now it does the clean-up itself.

* bm/merge-base-octopus-dedup:
  merge-base --octopus: reduce the result from get_octopus_merge_bases()
  merge-base: separate "--independent" codepath into its own helper

"git log --left-right A...B" lost the "leftness" of commits
reachable from A when A is a tag as a side effect of a recent
bugfix.  This is a regression in 1.8.4.x series.

* jc/revision-range-unpeel:
  revision: propagate flag bits from tags to pointees
  revision: mark contents of an uninteresting tree uninteresting

"git clone" would fail to clone from a repository that has a ref
directly under "refs/", e.g. "refs/stash", because different
validation paths do different things on such a refname.  Loosen the
client side's validation to allow such a ref.

* jk/allow-fetch-onelevel-refname:
  fetch-pack: do not filter out one-level refs

A handful of bugs around interpreting $branch@{upstream} notation
and its lookalike, when $branch part has interesting characters,
e.g. "@", and ":", have been fixed.

* jk/interpret-branch-name-fix:
  interpret_branch_name: find all possible @-marks
  interpret_branch_name: avoid @{upstream} past colon
  interpret_branch_name: always respect "namelen" parameter
  interpret_branch_name: rename "cp" variable to "at"
  interpret_branch_name: factor out upstream handling

A recent update to "git send-email" broke platforms where
/etc/ssl/certs/ directory exists but cannot be used as SSL_ca_path
(e.g. Fedora rawhide).

* rk/send-email-ssl-cert:
  send-email: /etc/ssl/certs/ directory may not be usable as ca_path

"git repack --max-pack-size=8g" stopped being parsed correctly when
the command was reimplemented in C.

* sb/repack-in-c:
  repack: propagate pack-objects options as strings
  repack: make parsed string options const-correct
  repack: fix typo in max-pack-size option

The pathspec matching code, while comparing two trees (e.g. "git
diff A B -- path1 path2") was too aggressive and failed to match
some paths when multiple pathspecs were involved.

* as/tree-walk-fix-aggressive-short-cut:
  tree_entry_interesting: match against all pathspecs

Signed-off-by: Junio C Hamano <gitster@pobox.com>