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?  Because they are tracked by oid, it
is possible a later successful upgrade would not remove all those files,
bit it should remove contrib/pg_upgrade/pg_upgrade_dump_1.log because it
is "1".

-- 
  Bruce Momjian  <bruce(at)momjian(dot)us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + It's impossible for everything to be true. +

In response to

• pg_upgrade regression test litters the source tree with log files at 2013-01-08 18:08:44 from Tom Lane

Responses

• Re: pg_upgrade regression test litters the source tree with log files at 2013-01-08 21:08:42 from Tom Lane
• Re: pg_upgrade regression test litters the source tree with log files at 2013-01-08 21:11:41 from Peter Eisentraut

pgsql-hackers by date

Next:	From: Tom Lane	Date: 2013-01-08 21:08:42
	Subject: Re: pg_upgrade regression test litters the source tree with log files
Previous:	From: Heikki Linnakangas	Date: 2013-01-08 21:02:15
	Subject: Re: [PATCH 2/5] Make relpathbackend return a statically result instead of palloc()'ing it