Re: t5607 fail with GIT_TEST_FAIL_PREREQS enabled

From: Jeff King <peff@peff.net>
Date: Aug 11, 2021, 14:03 UTC
Message-ID: <YRPYrXYwmwVnligD@coredump.intra.peff.net>
In-Reply-To: <CAL3xRKcMAkfaoAWo5VfQ_Q6nYkR94spSN4Ko6xfpSRKa1cEuAg@mail.gmail.com>

On Wed, Aug 11, 2021 at 03:02:52PM +0200, Son Luong Ngoc wrote:

Show 6 quoted lines

>     git/t% GIT_TEST_FAIL_PREREQS=1 ./t5607-clone-bundle.sh
> [...]
>     #               if ! test_have_prereq SHA1
>     #               then
>     #                       echo "@object-format=sha256"
>     #               fi >expect &&

The problem is presumably here. If test_have_prereq lies and say "no, we are using sha256" then we cannot expect what the built binary does to match that lie.

Perhaps that is a sign that test_have_prereq is not the right tool to check "which hash format are we using", but I don't think we have another convenient mechanism to do so currently.

I also think that the FAIL_PREREQS system may be mis-designed a bit. We had a similar problem a few months ago, and I think Junio's response here points in a good direction:

  https://lore.kernel.org/git/xmqqblbgrwkg.fsf@gitster.g/

-Peff