refint (& others?) on current snapshot - Mailing list pgsql-hackers

From Nat Howard
Subject refint (& others?) on current snapshot
Date
Msg-id 000001bea891$22249660$1242abc7@dummy.cello-us.com
Whole thread Raw
Responses Re: [HACKERS] refint (& others?) on current snapshot  (Bruce Momjian <maillist@candle.pha.pa.us>)
List pgsql-hackers
Snapshot of a few hours ago on 3.2 FreeBSD.  The trigger regression test
(and a few others) fail.  From looking at it, the trigger regression test
fails because refint fails.

If one simply tries to use the stuff in contrib/spi, the failure is pretty
easy to
see -- the example in contrib/spi/refint.example fails:
CREATE TRIGGER CT BEFORE INSERT OR UPDATE ON C FOR EACH ROW
EXECUTE PROCEDURE
check_primary_key ('REFC', 'A', 'ID');
CREATE

-- Now try

INSERT INTO A VALUES (10);
INSERT 18567 1
INSERT INTO A VALUES (20);
INSERT 18568 1
INSERT INTO A VALUES (30);
INSERT 18569 1
INSERT INTO A VALUES (40);
INSERT 18570 1
INSERT INTO A VALUES (50);
INSERT 18571 1

INSERT INTO B VALUES (1);       -- invalid reference
pqReadData() -- backend closed the channel unexpectedly.       This probably means the backend terminated abnormally
  before or while processing the request.
 
We have lost the connection to the backend, so further processing is
impossible.  Terminating.





pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Ye olde "relation doesn't quite exist" problem
Next
From: Pablo Funes
Date:
Subject: nonblocking lock?