Re: [PATCH 3/3] read-cache: don't write index twice if we can't write shared index

From: Jeff King <peff@peff.net>
Date: Jan 18, 2018, 21:37 UTC
Message-ID: <20180118213732.GA7403@sigill.intra.peff.net>
In-Reply-To: <20180118150014.GA2097@ash>

On Thu, Jan 18, 2018 at 10:00:14PM +0700, Duy Nguyen wrote:

Show 7 quoted lines

> The test suite was run as root, no wonder why my removing write access
> has no effect. I got the test to pass with this, but then it fails
> with
> 
>     Can't write .prove (Permission denied) at /usr/share/perl/5.22/App/Prove.pm line 542.
> 
> Some more chown'ing or chmod'ing is required....

Ah, right. I agree that we probably ought to run the ci as non-root. However, if the test requires non-root, it probably needs to be marked with the SANITY prereq.

I also ran into one funny thing: if you run the script with "-i", then we do not run the test_when_finished block. And therefore the "ro" directory is left without its write bit, and the next test run fails, as it cannot "rm -rf" the old trash directory out of the way.

I'm not sure there's a good solution, though. Skipping the test_when_finished block on a "-i" run is intentional, to let you inspect the broken state.

Show 6 quoted lines

> Subject: [PATCH] ci: don't accidentally run the test suite as root
> 
> This script assigns and adds a user named "ci" in a subshell so the
> outer CI_USER is not affected. For some reason, CI_USER is actually
> empty on Travis linux32 builds. This makes the following "su" useless
> and the test suite is run as root.

Are we sure this was the problem on Travis, and it wasn't just an issue with how I reproduced via docker?

-Peff