Horton

Re: GNU/Hurd portability patches

Alexander Lakhin <exclusion@gmail.com>

View thread
From: Alexander Lakhin <exclusion@gmail.com>
To: Michael Banck <mbanck@gmx.net>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Michael Paquier <michael@paquier.xyz>, pgsql-hackers@lists.postgresql.org, Thomas Munro <thomas.munro@gmail.com>
Date: 2025-10-28T05:00:01Z
Lists: pgsql-hackers 
Hello Michael,

25.09.2025 08:00, Alexander Lakhin wrote:
>
>> I saw those issues frequently on the initial 32bit Hurd VM I started to
>> run the buildfarm code on, before I switched it to HPET timers. Since
>> then, I don't think I saw that particular error again, but 4 out 1000 is
>> not a lot of course.
>
> There is also contrib/pg_stat_statements/entry_timestamp, which fails for
> me when running in a loop:
> for i in `seq 100`; do echo "ITERATION $i"; NO_TEMP_INSTALL=1 make -s check -C contrib/pg_stat_statements || break; done
>
> on iterations 42, 60, 12, 5, 28:
> ITERATION 28
> ...
> ok 8         - wal                                        14 ms
> not ok 9     - entry_timestamp                            14 ms
> ok 10        - privileges                                 16 ms
> ...
> 1..15
> # 1 of 15 tests failed.

One month later, fruitcrow has generated this failure too:
https://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=fruitcrow&dt=2025-10-25%2007%3A45%3A03

pgsql.build/contrib/pg_stat_statements/regression.diffs
diff -U3 
/home/demo/client-code-REL_19_1/buildroot/HEAD/pgsql.build/contrib/pg_stat_statements/expected/entry_timestamp.out 
/home/demo/client-code-REL_19_1/buildroot/HEAD/pgsql.build/contrib/pg_stat_statements/results/entry_timestamp.out
--- /home/demo/client-code-REL_19_1/buildroot/HEAD/pgsql.build/contrib/pg_stat_statements/expected/entry_timestamp.out 
2025-10-25 08:45:03.000000000 +0100
+++ /home/demo/client-code-REL_19_1/buildroot/HEAD/pgsql.build/contrib/pg_stat_statements/results/entry_timestamp.out 
2025-10-25 08:57:31.000000000 +0100
@@ -147,7 +147,7 @@
  WHERE query LIKE '%STMTTS%';
   total | minmax_exec_zero | minmax_ts_after_ref | stats_since_after_ref
  -------+------------------+---------------------+-----------------------
-     2 |                1 |                   2 |                     0
+     2 |                2 |                   2 |                     0
  (1 row)

  -- Cleanup

Thus, the "zero time difference" issue in general still exists.

Best regards,
Alexander