Michael Fuhr <mike(at)fuhr(dot)org> writes:
> Do you have a way to revert to the old installation to check whether
> the checks fail again? It might be useful to track down exactly
> what happened. It seems wrong that a currently-installed version
> should have adverse effects on a just-built version's regression
> tests.
We've seen that happen many times with shared library dependencies
... but libpq hasn't changed in the last few days AFAIR.
regards, tom lane