Hi! Some new information about the MySQL-4.0.10 compilation problem on Red Hat 8.0 (x86), but unfortunately no solution found yet, except running a MySQL official binary you can download from www.mysql.com. A 4 day stress test of an official binary ran fine.
One of MySQL AB employees said that on his Red Hat 8.0 computer mysqld sometimes crashes, but only if he compiles with -O3. But, in my 8.0 computer all versions, whether compiled with gcc-3.2 or gcc-2.95.3, with -O or -O3, always crash in the first SQL statement. Thus, the bug is apparently connected to hardware, and it is also non-deterministic. It might be some library bug which only appears when the code is physically located in a certain way. As many of our users have reported the problem from Red Hat 8.0 it is certainly a very common problem. Since a stable MySQL can be compiled on all other Unix platforms except Red Hat 8.0, that suggests the bug is in Red Hat 8.0. When I run mysqld inside gdb, the debugger seems to get confused when mysqld creates a new pthread for a new connection. Probably the stack gets badly corrupt in that phase. I solved the problem in my computer by installing Red Hat 7.3. When MySQL is compiled with gcc-2.96 there, it is stable. Best regards, Heikki Innobase Oy http://www.innodb.com InnoDB - transactions, foreign keys, and a hot backup tool for MySQL sql query --------------------------------------------------------------------- Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail <[EMAIL PROTECTED]> To unsubscribe, e-mail <[EMAIL PROTECTED]> Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
