On Tue, Jan 8, 2013 at 04:08:42PM -0500, Tom Lane wrote:
> Bruce Momjian <bruce@momjian.us> writes:
> > On Tue, Jan 8, 2013 at 01:08:44PM -0500, Tom Lane wrote:
> >> In a tree in which I previously ran "make check" in contrib/pg_upgrade:
> >>
> >> $ make -s distclean
> >> $ git status
> >> # On branch master
> >> # Untracked files:
> >> # (use "git add <file>..." to include in what will be committed)
> >> #
> >> # contrib/pg_upgrade/pg_upgrade_dump_1.log
> >> # contrib/pg_upgrade/pg_upgrade_dump_12912.log
> >> # contrib/pg_upgrade/pg_upgrade_dump_16384.log
> >> nothing added to commit but untracked files present (use "git add" to track)
> >>
> >> Not sure how long this has been happening.
>
> > Those look like files left over from a failed upgrade, or you used
> > --retain. Does that make sense?
>
> It's possible that I had one or more failed regression test runs on that
> machine ... don't recall for sure. In any case the point here is that
> "make clean" ought to get rid of anything that might be left over from a
> test run, successful or otherwise.
That seems like something more for the regression script (test.sh) to
delete. Those are output by _running_ the program, and I never expected
people to be running it in the git tree.
-- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB
http://enterprisedb.com
+ It's impossible for everything to be true. +