Re: [PATCH v5 09/16] bisect run: fix the error message

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Aug 29, 2022, 12:20 UTC
Message-ID: <06q285oq-np8q-qpo1-s470-74q5s3s3478o@tzk.qr>
In-Reply-To: <xmqq8rn8voh2.fsf@gitster.g>

Hi Junio,

On Sun, 28 Aug 2022, Junio C Hamano wrote:

Show 40 quoted lines

> "Johannes Schindelin via GitGitGadget" <gitgitgadget@gmail.com>
> writes:
>
> > From: Johannes Schindelin <johannes.schindelin@gmx.de>
> >
> > In d1bbbe45df8 (bisect--helper: reimplement `bisect_run` shell function
> > in C, 2021-09-13), we ported the `bisect run` subcommand to C, including
> > the part that prints out an error message when the implicit `git bisect
> > bad` or `git bisect good` failed.
> >
> > However, the error message was supposed to print out whether the state
> > was "good" or "bad", but used a bogus (because non-populated) `args`
> > variable for it.
> >
> > Helped-by: Elijah Newren <newren@gmail.com>
> > Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
> > ---
> >  builtin/bisect--helper.c    |  2 +-
> >  t/t6030-bisect-porcelain.sh | 10 ++++++++++
> >  2 files changed, 11 insertions(+), 1 deletion(-)
> >
> > diff --git a/builtin/bisect--helper.c b/builtin/bisect--helper.c
> > index d797cd1cef8..a00167f1373 100644
> > --- a/builtin/bisect--helper.c
> > +++ b/builtin/bisect--helper.c
> > @@ -1430,7 +1430,7 @@ static int cmd_bisect_run(int argc, const char **argv, const char *prefix)
> >  			printf(_("bisect found first bad commit"));
> >  			res = BISECT_OK;
> >  		} else if (res) {
> > -			error(_("bisect run failed: 'git bisect--helper --bisect-state"
> > +			error(_("bisect run failed: 'git bisect"
> >  			" %s' exited with error code %d"), new_state, res);
> >  		} else {
> >  			continue;
>
> The change to retire "bisect--helper" from the end-user facing error
> message makes tons of sense, but I am not quite sure if the proposed
> log message describes the change correctly.  Or is this a fallout of
> some "rebase -i" gotcha that a log message meant for one commit was
> applied to a different commit, or something?

This is indeed a fall-out from some interactive rebase. Thank you for catching it.

You can actually see it in the thread history: v2 still had a diff that was aligned with the commit message, but <46fe0774-66e7-8947-cd79-d35229eec25a@web.de> graduated between v2 and v3 and partially addressed the issue fixed by this here patch.

I'm glad that we caught this in time for v6. This is the commit message I plan on using:

-- snip --
    bisect run: fix the error message

    In d1bbbe45df8 (bisect--helper: reimplement `bisect_run` shell function
    in C, 2021-09-13), we ported the `bisect run` subcommand to C, including
    the part that prints out an error message when the implicit `git bisect
    bad` or `git bisect good` failed.

    However, the error message was supposed to print out whether the state
    was "good" or "bad", but used a bogus (because non-populated) `args`
    variable for it. This was fixed in 80c2e9657f2 (bisect--helper: report
    actual bisect_state() argument on error, 2022-01-18), but the error
    message still talks about `bisect--helper`, which is an implementation
    detail that should not concern end users.

    Fix that, and add a regression test to ensure that the intended form of
    the error message.
-- snap --

Ciao, Dscho