From: Jonathan Nieder <jrnieder@gmail.com>
Date: Tue, 16 Sep 2014 22:28:14 GMT
Subject: Re: [PATCH v5 22/35] git_config_set_multivar_in_file(): avoid call to rollback_lock_file()
Message-ID: <20140916222814.GV29050@google.com>
In-Reply-To: <1410896036-12750-23-git-send-email-mhagger@alum.mit.edu>

Michael Haggerty wrote:

> After commit_lock_file() is called, then the lock_file object is
> necessarily either committed or rolled back.  So there is no need to
> call rollback_lock_file() again in either of these cases.
>
> Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>

This seems to involve an unadvertised behavior change: before
it wouldn't call git_config_clear() on failure, and after the
patch it would.  Intended?

I think it would be clearer with the goto for exception handling
maintained (and just a second 'lock = NULL' assignment).