Re: [PATCH 1/3] t7900: fix flaky test due to leaking background job

From: Patrick Steinhardt <ps@pks.im>
Date: Aug 19, 2024, 08:55 UTC
Message-ID: <ZsMIa_5jCFs4OWYx@tanuki>
In-Reply-To: <20240819084943.GB2955268@coredump.intra.peff.net>

On Mon, Aug 19, 2024 at 04:49:43AM -0400, Jeff King wrote:

Show 25 quoted lines

> On Mon, Aug 19, 2024 at 09:47:59AM +0200, Patrick Steinhardt wrote:
> 
> > diff --git a/t/t7900-maintenance.sh b/t/t7900-maintenance.sh
> > index 06ab43cfb5..074eadcd1c 100755
> > --- a/t/t7900-maintenance.sh
> > +++ b/t/t7900-maintenance.sh
> > @@ -967,8 +967,13 @@ test_expect_success '--detach causes maintenance to run in background' '
> >  		git config set maintenance.loose-objects.auto 1 &&
> >  		git config set maintenance.incremental-repack.enabled true &&
> >  
> > -		git maintenance run --detach >out 2>&1 &&
> > -		test_must_be_empty out
> > +		# The extra file descriptor gets inherited to the child
> > +		# process, and by reading stdout we thus essentially wait for
> > +		# that descriptor to get closed, which indicates that the child
> > +		# is done, too.
> > +		output=$(git maintenance run --detach 2>&1 9>&1) &&
> > +		printf "%s" "$output" >output &&
> > +		test_must_be_empty output
> >  	)
> >  '
> 
> This looks correct, but should we be doing it for all of the "git
> maintenance" runs in that script? They're all going to kick off detached
> gc jobs, I think.

Only those that use `--detach` run in the background.

Patrick