On Thu, Jul 7, 2016 at 12:34 AM, Noah Misch <noah@leadboat.com> wrote:
>> How do you feel about adding testing to tuplesort.c not limited to
>> hitting this bug (when Valgrind memcheck is used)?
>
> Sounds great, but again, not in the patch fixing this bug.
I'll work on a minimal CLUSTER testcase within the next day or two,
and post a revision. Separately, I'll propose a patch that further
expands tuplesort test coverage. This will add coverage for hash
tuplesorts, as well as coverage of external sorts that require
multiple passes.
>> Are you satisfied that I have adequately described steps to reproduce?
>
> I can confirm that (after 62 minutes) your test procedure reached SIGSEGV
> today and then completed successfully with your patch.
Thanks for going to the trouble of confirming that the test procedure
causes a segmentation fault, and that my patch appears to fix the
issue.
--
Peter Geoghegan