"Phillip Wood via GitGitGadget" <gitgitgadget@xxxxxxxxx> writes:
> From: Phillip Wood <phillip.wood@xxxxxxxxxxxxx>
>
> If ORIG_HEAD is not set by passing RESET_ORIG_HEAD then there is no
> need to pass anything for reflog_orig_head. In addition to the callers
> fixed in this commit move_to_original_branch() also passes
> reflog_orig_head without setting ORIG_HEAD. That caller is mistakenly
> passing the message it wants to put in the branch reflog which is not
> currently possible so we delay fixing that caller until we can pass
> the message as the branch reflog.
As I hinted elsewhere, these rules should be spelled out in a
comment before "int reset_head(...)" either in reset.[ch].
For this particular one, I wonder if
(A) we can lose RESET_ORIG_HEAD bit and use the presence of
reflog_orig_head to mean what that bit currently means, or
(B) we keep the current code strucure, but make it a BUG() if
a non-NULL reflog_orig_head is given without RESET_ORIG_HEAD
bit set.
> Signed-off-by: Phillip Wood <phillip.wood@xxxxxxxxxxxxx>
> ---
> builtin/rebase.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/builtin/rebase.c b/builtin/rebase.c
> index 3d78b5c8bef..fdd822c470f 100644
> --- a/builtin/rebase.c
> +++ b/builtin/rebase.c
> @@ -675,7 +675,7 @@ static int run_am(struct rebase_options *opts)
>
> reset_head(the_repository, &opts->orig_head,
> opts->head_name, 0,
> - "HEAD", NULL, DEFAULT_REFLOG_ACTION);
> + NULL, NULL, DEFAULT_REFLOG_ACTION);
> error(_("\ngit encountered an error while preparing the "
> "patches to replay\n"
> "these revisions:\n"
> @@ -1777,7 +1777,7 @@ int cmd_rebase(int argc, const char **argv, const char *prefix)
> options.head_name ? options.head_name : "detached HEAD",
> oid_to_hex(&options.onto->object.oid));
> reset_head(the_repository, NULL, options.head_name,
> - RESET_HEAD_REFS_ONLY, "HEAD", msg.buf, NULL);
> + RESET_HEAD_REFS_ONLY, NULL, msg.buf, NULL);
> strbuf_release(&msg);
> ret = finish_rebase(&options);
> goto cleanup;
