Hi,
On 2025-03-22 16:14:11 -0400, Andres Freund wrote:
> On 2025-03-22 22:00:00 +0200, Alexander Lakhin wrote:
> > @@ -24,14 +24,14 @@
> > SELECT count(*) FROM bmscantest WHERE a = 1 AND b = 1;
> > count
> > -------
> > - 23
> > + 18
> > (1 row)
>
> Is it possible that this is the bug reported here:
> https://postgr.es/m/873c33c5-ef9e-41f6-80b2-2f5e11869f1c%40garret.ru
>
> I.e. that the all-visible logic in bitmap heapscans is fundamentally broken?
>
> I can reproduce different results on a fast machien by just putting a VACUUM
> FREEZE bmscantest after the CREATE INDEXes in bitmapops.sql. After I disable
> the all-visible logic in bitmapheap_stream_read_next(), I can't observe such a
> difference anymore.
Hm, it's clearly related to the all-visible path, but I think the bug is
actually independent of what was reported there. The bug you reported is
perfectly reproducible, without any concurrency, after all.
Here's a smaller reproducer:
CREATE TABLE bmscantest (a int, t text);
INSERT INTO bmscantest
SELECT (r%53),
'foooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo'
FROM generate_series(1,70000) r;
CREATE INDEX i_bmtest_a ON bmscantest(a);
set enable_indexscan=false;
set enable_seqscan=false;
-- Lower work_mem to trigger use of lossy bitmaps
set work_mem = 64;
SELECT count(*) FROM bmscantest WHERE a = 1;
vacuum freeze bmscantest;
SELECT count(*) FROM bmscantest WHERE a = 1;
-- clean up
DROP TABLE bmscantest;
The first SELECT reports 1321, the second 572 tuples.
Greetings,
Andres Freund
