| Bug #111156 | main.subquery_sj_all_bka_nobnl and main.subquery_sj_mat_bka_nobnl fail on s390x | ||
|---|---|---|---|
| Submitted: | 25 May 2023 16:34 | Modified: | 7 Aug 14:25 |
| Reporter: | Lena Voytek | Email Updates: | |
| Status: | Unsupported | Impact on me: | |
| Category: | Tests: Server | Severity: | S7 (Test Cases) |
| Version: | 8.0.33 | OS: | Debian (Fails on Debian sid and Ubuntu 20.04-23.04) |
| Assigned to: | CPU Architecture: | Other (s390x) | |
[25 May 2023 16:34]
Lena Voytek
[26 May 2023 13:11]
MySQL Verification Team
Hi Mr. Voytek, Thank you for your bug report. However, IBM's mainframes are NOT supported to all. We might have attempted to see what is going wrong, but we do NOT have access to ANY of these systems. Some of us have worked on the 43x series and OS/400 series, but we do not have even access to those. Ten to fifteen years ago , we did manage to make MySQL fully functional on AS/400 and on VM that ran on s390x , but never on the metal itself. We can not repeat those experiences, since we do not have access to those systems, nor do we have any plans to support MySQL on those systems. Unsupported. We do not even have ambitions for porting MySQL to any of those platforms.
[7 Aug 14:25]
Lena Voytek
Hi, Mrs. Voytek here, As stated before, I understand s390x is unsupported, but these testsuite failures may be useful to know about. Just as an update, with the release of 8.4.2, main.subquery_sj_all_bka_nobnl and main.subquery_sj_mat_bka_nobnl are passing again. However, main.archive and main.func_compress now fail with the following: CURRENT_TEST: main.archive 767s --- /usr/lib/mysql-test/r/archive.result 2024-07-12 22:20:22.000000000 +0300 767s +++ /tmp/tmp.jZHfOwlPLE/var/2/log/archive.reject 2024-08-05 06:27:53.397196328 +0300 767s @@ -12732,7 +12732,7 @@ 767s SELECT DATA_LENGTH, AVG_ROW_LENGTH FROM 767s INFORMATION_SCHEMA.TABLES WHERE TABLE_NAME='t1' AND TABLE_SCHEMA='test'; 767s DATA_LENGTH AVG_ROW_LENGTH 767s -122 61 767s +121 60 767s DROP TABLE t1; 767s SET @save_join_buffer_size= @@join_buffer_size; 767s SET @@join_buffer_size= 8192; 767s 767s mysqltest: Result content mismatch 767s 767s 767s The result from queries just before the failure was: 767s 3 3 767s SELECT * FROM innodb_table ORDER BY id; 767s id x 767s 1 3 767s 3 5 767s 5 6 767s # Delete is allowed if the archive table is only read from. 767s DELETE t2 FROM archive_table AS t1, innodb_table AS t2 WHERE t1.id = t2.id; 767s SELECT * FROM archive_table ORDER BY id; 767s id x 767s 1 1 767s 2 2 767s 3 3 767s SELECT * FROM innodb_table ORDER BY id; 767s id x 767s 5 6 767s DROP VIEW v; 767s DROP TABLE archive_table, innodb_table; 767s Warnings: 767s Warning 1287 '@@binlog_format' is deprecated and will be removed in a future release. 767s safe_process[13310]: Child process: 13311, exit: 1 CURRENT_TEST: main.func_compress 1242s --- /usr/lib/mysql-test/r/func_compress.result 2024-07-12 22:20:22.000000000 +0300 1242s +++ /tmp/tmp.jZHfOwlPLE/var/1/log/func_compress.reject 2024-08-05 06:35:48.637478848 +0300 1242s @@ -58,7 +58,7 @@ 1242s length(a) 1242s NULL 1242s 12 1242s -76 1242s +454 1242s 50000 1242s select length(uncompress(a)) from t1; 1242s length(uncompress(a)) 1242s 1242s mysqltest: Result content mismatch 1242s 1242s 1242s The result from queries just before the failure was: 1242s Warning 1259 ZLIB: Input data corrupted 1242s drop table t1; 1242s CREATE TABLE t1 (c1 INT); 1242s INSERT INTO t1 VALUES (1), (1111), (11111); 1242s SELECT UNCOMPRESS(c1), UNCOMPRESSED_LENGTH(c1) FROM t1; 1242s UNCOMPRESS(c1) UNCOMPRESSED_LENGTH(c1) 1242s NULL 0 1242s NULL 0 1242s NULL 825307441 1242s EXPLAIN SELECT * FROM (SELECT UNCOMPRESSED_LENGTH(c1) FROM t1) AS s; 1242s DROP TABLE t1; 1242s End of 5.0 tests 1242s # 1242s # Bug#18693654 VALGRIND WARNINGS IN INFLATE ON UNCOMPRESS 1242s # 1242s SELECT UNCOMPRESS( CAST( 0 AS BINARY(5) ) ); 1242s UNCOMPRESS( CAST( 0 AS BINARY(5) ) ) 1242s NULL 1242s Warnings: 1242s Warning 1259 ZLIB: Input data corrupted 1242s safe_process[26323]: Child process: 26324, exit: 1 Thanks
[7 Aug 14:49]
MySQL Verification Team
Dear Mrs. Voytek, We are so happy to hear from you again. Some of us here are still nostalgic about the times when we worked on 4381 systems and used that fantastic 3278 IBM terminal. But, you probably do not know what we are talking about here....... Now, about the tests ....... Regarding the archive test, forget about it ....... It is about ARCHIVE storage engine, which is not maintained any more and nobody uses it. Regarding the other test, well, we might guess what the problem is ...... You should consider rebuilding the zlib library that you are using ........ You should have this one: zlib-1.2.13 ....... It could be also a problem with how you build it ....... that is all ...... ......................................................................................................... One more unrelated question ...... Are you running MySQL on Linux within the VM machine or on the metal ....... If it is first, you do not have to respond ........ that is what we presume ....... We wish you all the best with porting MySQL .......
