2013/7/19 Stephen Frost <sfrost@snowman.net>:
> * Greg Smith (greg@2ndQuadrant.com) wrote:
>> On 7/18/13 7:57 PM, Karol Trzcionka wrote:
>> >Current head 4cbe3ac3e86790d05c569de4585e5075a62a9b41 -> patch applies
>> >correct (only change needed in parallel_schedule).
>> >However it fails on own regression tests (other tests pass).
>>
>> I got a rejected hunk in src/backend/nodes/nodeFuncs.c as well as
>> that parallel_schedule issue. Maybe you didn't get the nodeFuncs
>> change but didn't notice that? That might explain why the tests
>> didn't work for you either.
>
> The nodeFuncs.c hunk seems likely to have been impacted by the patch I
> committed today (WITH CHECK OPTION), so I doubt that was the issue.
>
>> Attached is an updated patch where I tried to only fix the two small
>> hunks of bit rot. I get "All 140 tests passed" here, on a Mac no
>> less.
>
> Thanks for updating the patch, I ran into the failed hunks too and
> expected to have to deal with them. :)
>
Thanks for pointing out this problem. I synchronized my local master
with the upstream one, then adjusted the row-security branch.
I'll submit the patch soon, with fixing-up the portion that Tom pointed
out.
Thanks,