Re: [PATCH 3/3] gc: do not return error for prior errors in daemonized mode

From: Junio C Hamano <gitster@pobox.com>
Date: Jul 18, 2018, 16:21 UTC
Message-ID: <xmqqk1psldkx.fsf@gitster-ct.c.googlers.com>
In-Reply-To: <20180717201348.GD26218@sigill.intra.peff.net>

Jeff King <peff@peff.net> writes:

Show 18 quoted lines

>> There's nothing for the calling program to act on on the basis of that
>> error.  Use status 0 consistently instead, to indicate that we decided
>> not to run a gc (just like if no housekeeping was required).  This
>> way, repo and similar tools can get the benefit of the same behavior
>> as tools like "git fetch" that ignore the exit status from gc --auto.
>
> I think this is a good change.
>
> In theory it might be useful to propagate the original exit code (_not_
> "did we see a warning or an error", but the true original exit code. But
> as you note, it's not deterministic anyway (we'd miss that exit code on
> the first run, or even any simultaneous runs that exit early due to lock
> contention). So it's clear that callers can't really do anything robust
> based on the exit code of a daemonized "gc --auto".
>
> I still think that "repo" should probably stop respecting the exit code.
> But that's no excuse for Git not to have a sensible exit code in the
> first place.

I am not yet convinced that this last step to exit with 0 is a good change, even though I can understand that it would be more convenient, as there currently is no easy way for the calling script to tell two error cases apart.

I think the "sensible exit code" you mention would be something like "1 for hard error, 2 for 'I am punting as I see there were previous errors---you may want to examine your repository'".

If we did that from day one and documented that behaviour, nobody would have complained, but adopting that suddenly is of course a breaking change.

Perhaps we should exit with 2 (not 0) in that "previous error" case by default, and then have a configuration knob to turn that 2 into 0 for those who cannot easily modify the calling script? That way, we by default will *not* break those who have been paying attention to zero-ness of the exit status, we allow those who want to treat this "prior error" case as if there were no error with just a knob, and then those who are willing to update their script can tell two cases by the exit status and act differently.