On Tue, 30 Apr 2024 at 14:14, Tom Lane <t...@sss.pgh.pa.us> wrote: > Yeah, the affected buildfarm members seem still not happy.
Isn't your latest fix adding COLLATE "C" to the wrong test? Crake is failing with [1]: --- /home/andrew/bf/root/HEAD/pgsql/src/test/modules/test_pg_dump/expected/test_pg_dump.out 2024-04-29 20:27:02.792272385 -0400 +++ /home/andrew/bf/root/HEAD/pgsql.build/testrun/test_pg_dump/regress/results/test_pg_dump.out 2024-04-29 20:31:43.662916513 -0400 @@ -145,8 +145,8 @@ ----------------------------------------------------+-----------------------------+--------- column c1 of foreign table ft1 | role regress_dump_test_role | a column c1 of table test_pg_dump_t1 | role regress_dump_test_role | a - foreign table ft1 | role regress_dump_test_role | a foreign-data wrapper dummy | role regress_dump_test_role | a + foreign table ft1 | role regress_dump_test_role | a function regress_pg_dump_schema.test_agg(smallint) | role regress_dump_test_role | a function regress_pg_dump_schema.test_agg(smallint) | role regress_dump_test_role | i function regress_pg_dump_schema.test_func() | role regress_dump_test_role | a Whereas your fix in [2] adjusts some other tests at line 70 and 207 rather than around line 145. The test at [3] should have the COLLATE "C". David [1] https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=crake&dt=2024-04-30%2000%3A27%3A04 [2] https://git.postgresql.org/gitweb/?p=postgresql.git;a=blobdiff;f=src/test/modules/test_pg_dump/expected/test_pg_dump.out;h=dd1a2389644ea7d8a5f8ae34a974525bde4cc817;hp=dc493e5be26e9ea97a07a2bd84fba48e979198ac;hb=900d1144256a63250a2e326567b636ee3220b731;hpb=9d9ece4c16dbbaf3b9d60c2fe201b8e99a407be3 [3] https://git.postgresql.org/gitweb/?p=postgresql.git;a=blob;f=src/test/modules/test_pg_dump/expected/test_pg_dump.out;h=dd1a2389644ea7d8a5f8ae34a974525bde4cc817;hb=900d1144256a63250a2e326567b636ee3220b731#l138
