Instability of pg_hint_plan.sql test in PG16 + master on a slow box (or when testing with valgrind).

[svglukhov]

Hi,

There is an instability of pg_hint_plan.sql test in PG16 + master
on a slow box (or when testing with valgrind).

How to repeat: run make installcheck with the postgres under valgrind.

cat regression.diffs
diff -U3 contrib/pg_hint_plan/expected/pg_hint_plan.out contrib/pg_hint_plan/results/pg_hint_plan.out
--- contrib/pg_hint_plan/expected/pg_hint_plan.out
+++ contrib/pg_hint_plan/results/pg_hint_plan.out
@@ -9145,7 +9145,7 @@

relname | seq_scan | idx_scan
---------+----------+----------

• t1 | f | f

• t1 | f | t
  (1 row)

-- hint error level

Suggested fix is in the attachment.

Regards,
Gluh

fix-test-instability.txt

[michaelpq]

It is a very bad idea to penalize the tests with some hardcoded pg_sleep() calls that would make the tests even slower for everybody. IMO, we should redesign the tests so as the existing sleeps are avoided as much as possible, as this is already the slowest part of them.