On Mon, Jan 25, 2021 at 12:06 AM Mahendra Singh Thalor <mahi6...@gmail.com> wrote: > > On Sun, 24 Jan 2021 at 11:48, Julien Rouhaud <rjuju...@gmail.com> wrote: > > > > Hi, > > > > I'm not really familiar with this part of the code, so it's not exactly > > clear > > to me if some logic is missing in compute_new_xmax_infomask() / > > heap_prepare_insert(), or if this should actually be an allowed combination > > of > > hint bit. > > Thanks Juliean for reporting this. I am also able to reproduce this assert.
Thanks for looking at it! > > Small test case to reproduce: >> >> DROP TABLE IF EXISTS t1; >> CREATE TABLE t1(id integer, val text); >> INSERT INTO t1 SELECT i, 'val' FROM generate_series(1, 2) i; >> >> BEGIN; >> SAVEPOINT s1; >> SELECT 1 FROM t1 WHERE id = 2 FOR UPDATE; >> UPDATE t1 SET val = 'hoho' WHERE id = 2; >> release s1; >> SELECT 1 FROM t1 WHERE id = 2 FOR UPDATE; >> UPDATE t1 SET val = 'hoho' WHERE id = 2; > > > If we remove the "release s1;" step from the test case, then we are not > getting this assert failure. Yes, this is the smallest reproducer that could trigger the problem, and the release is required.
