Bug #26067 | Falcon: SELECT fails on a database after many changes | ||
---|---|---|---|
Submitted: | 5 Feb 2007 2:56 | Modified: | 6 Sep 2007 21:11 |
Reporter: | Peter Gulutzan | Email Updates: | |
Status: | Can't repeat | Impact on me: | |
Category: | MySQL Server: Falcon storage engine | Severity: | S3 (Non-critical) |
Version: | 5.2.2-falcon-alpha-debug-log | OS: | Linux (SUSE 10.0 / 64-bit) |
Assigned to: | Christopher Powers | CPU Architecture: | Any |
[5 Feb 2007 2:56]
Peter Gulutzan
[5 Feb 2007 8:40]
Hakan Küçükyılmaz
The description about how to populate the the table is in BUG#26055 Regards, Hakan
[5 Feb 2007 20:09]
Hakan Küçükyılmaz
After populating the table as described in BUG#26055 I could reproduce the error. I had to use following setting: SET names utf8; SET sql_mode=''; SET autocommit=1; mysql> select * from tu where cp932 like 'a%'; +------+-------+------+ | s1 | cp932 | otto | +------+-------+------+ | A | AA A | 66 | +------+-------+------+ 1 row in set (0.01 sec) mysql> insert into tu (cp932) values (0x4141204120); Query OK, 1 row affected (0.01 sec) mysql> insert into tu (cp932) values (0x41412041); Query OK, 1 row affected (0.00 sec) mysql> select * from tu where cp932 like 'a%'; +------+-------+------+ | s1 | cp932 | otto | +------+-------+------+ | A | AA A | 66 | +------+-------+------+ 1 row in set (0.00 sec) Best regards, Hakan
[31 Aug 2007 20:12]
Christopher Powers
Reproduced in post alpha IV (release 6.0.2) build, changeset 1.2730.
[6 Sep 2007 21:08]
Christopher Powers
Unable to reproduce in the post-5.1 merge codebase, changeset 1.2592 and later. NOTE: The fix for Bug#27405 introduced a regression that crashes the server when Falcon tables containing autoincrement fields are partitioned (see Bug#30878). The test scenario for this bug will only work if these lines in ha_partition::write_row() are commented out: //autoincrement_lock= true; //pthread_mutex_lock(&table_share->mutex);