Euler Taveira de Oliveira <eulerto@yahoo.com.br> writes:
> Sorry for that. I didn't have the data yet. Some data to test it is in:
> http://www.timbira.com/tst2.sql.gz
Thanks.
> In 8.0 and previous I didn't reproduce it.
Which versions did you try exactly? I've reproduced it in 8.0.6 and HEAD.
> Is it something related to UPDATE ... FROM patch that I submitted?
No, it looks to me like it's a problem with this patch:
2005-08-19 20:39 tgl
* src/: backend/access/heap/heapam.c, backend/commands/async.c,
backend/commands/trigger.c, backend/commands/vacuum.c,
backend/executor/execMain.c, backend/utils/time/tqual.c,
include/access/heapam.h, include/access/htup.h,
include/executor/executor.h, include/utils/tqual.h: Repair problems
with VACUUM destroying t_ctid chains too soon, and with
insufficient paranoia in code that follows t_ctid links.
I think EvalPlanQual is deciding that updated tuples are valid in
some cases where it shouldn't. Unfortunately, if that's correct it
means that all the branches are broken since last August :-(
regards, tom lane