roachtest: unoptimized-query-oracle: incorrect oid typing in row-by-row merge joiner

[cockroach-teamcity]

roachtest.unoptimized-query-oracle/disable-rules=half/rand-tables failed with artifacts on release-24.1 @ 0d7d1ecbe9ba4abcb8fa97d521c3f9448f4dd590:

(query_comparison_util.go:389).runOneRoundQueryComparison: . 2869 statements run: expected unoptimized and optimized results to be equal
  []string{
- 	"-,0",
+ 	"0,0",
- 	"-,0",
+ 	"0,0",
- 	"-,0",
+ 	"0,0",
- 	"-,0",
+ 	"0,0",
  	"0,0",
  	"0,0",
  	... // 10 identical elements
  }
sql: SELECT
	tab_2688."col5""_1" AS col_6495, 0:::OID AS col_6496
FROM
	defaultdb.public."tabl%ve_5" AS tab_2687
	JOIN defaultdb.public."tabl%ve_5" AS tab_2688 ON (tab_2687."col😇5	_0") = (tab_2688."col5""_1")
ORDER BY
	tab_2687."col😇5	_0", col_6495 NULLS FIRST, col_6496 DESC NULLS FIRST, tab_2688."col5""_1"
LIMIT
	63:::INT8
test artifacts and logs in: /artifacts/unoptimized-query-oracle/disable-rules=half/rand-tables/run_1

Parameters:

• ROACHTEST_arch=amd64
• ROACHTEST_cloud=gce
• ROACHTEST_coverageBuild=false
• ROACHTEST_cpu=4
• ROACHTEST_encrypted=false
• ROACHTEST_metamorphicBuild=false
• ROACHTEST_ssd=0

Help

See: roachtest README

See: How To Investigate (internal)

See: Grafana

This test on roachdash | Improve this report!

Jira issue: CRDB-38366

[yuzefovich]

Hm, I'd think that it's the same as #118273 the fix for which is included in the SHA.

[yuzefovich]

Reduced repro on 0d7d1ec (and on master 270e1e3)

CREATE TABLE t (
  col_0 REGROLE, col_1 OID, col_2 INT,
  INDEX (col_1 DESC) STORING (col_2),
  INDEX (col_0) STORING (col_1)
);

INSERT INTO t (col_0, col_1, col_2) VALUES (0, 0, 0);

SET testing_optimizer_random_seed = 6047211422050928467;
SET testing_optimizer_disable_rule_probability = 0.500000;
SET vectorize = off;

SELECT
	t2.col_1
FROM
	t AS t1
	JOIN t AS t2 ON (t1.col_0) = (t2.col_1)
ORDER BY
	t1.col_0;

RESET testing_optimizer_random_seed;
RESET testing_optimizer_disable_rule_probability;
RESET vectorize;

SELECT
	t2.col_1
FROM
	t AS t1
	JOIN t AS t2 ON (t1.col_0) = (t2.col_1)
ORDER BY
	t1.col_0;