Re: [HACKERS] Rewriting the test of pg_upgrade as a TAP test

On Fri, Apr 14, 2017 at 8:03 PM, Stephen Frost <sfrost@snowman.net> wrote:
> Some of those were specifically left around to test those code paths.
> I'm not sure if these were those or not though, Andrew was the one who
> reviewed the various pg_dumpall calls to add --no-sync in places.

Well, Andrew has pushed the patch I have written, and the calls of
pg_dumpall in pg_upgrade use --no-sync. The ones intentionally left
are in src/bin/pg_dump/t/002_pg_dump.pl.
-- 
Michael