On Fri, Jun 26, 2020 at 12:31:19AM -0700, Junio C Hamano wrote:

Carlo Marcelo Arenas Belón [off-list ref] writes:

quoted

On Thu, Jun 25, 2020 at 06:12:40PM -0700, Junio C Hamano wrote:

quoted

* cb/is-descendant-of (2020-06-23) 1 commit
  (merged to 'next' on 2020-06-25 at ca0eaddf5e)
 + commit-reach: avoid is_descendant_of() shim

 Code clean-up.

 Will merge to 'master'.

apologize for the most likely unnecessary clarification (probably implied
by the merged comment indicating ca0eaddf5e), but this branch uses both
rs/pull-leakfix and rs/commit-reach-leakfix, that are currently merged
earlier into next and also scheduled for master.

Hmph, Does it break if it is merged alone to 'master' without these
other two topics?  I somehow thought I disentangled the topic from
unnecessary dependencies before queuing it, but it is possible that
I misremember...

Doesn't break and applies perfectly on its own, or in any order with
rs/pull-leakfix, but will conflict with rs/commit-reach-leakfix with
the obvious conflict resolution needed:

--- >8 ---

diff --cc commit-reach.c
index 1761217663,2d85265a35..f846d30f22

--- a/commit-reach.c
+++ b/commit-reach.c

@@@ -433,9 -427,8 +428,10 @@@ int ref_newer(const struct object_id *n
  		return 0;
  
  	commit_list_insert(old_commit, &old_commit_list);
- 	ret = is_descendant_of(new_commit, old_commit_list);
 -	return repo_is_descendant_of(the_repository,
++	ret = repo_is_descendant_of(the_repository,
+ 				    new_commit, old_commit_list);
 +	free_commit_list(old_commit_list);
 +	return ret;
  }
  
  /*