Bug #68939 | mysqld crashes by signal 5 | ||
---|---|---|---|
Submitted: | 12 Apr 2013 8:06 | Modified: | 12 Apr 2013 14:45 |
Reporter: | Tsubasa Tanaka (OCA) | Email Updates: | |
Status: | Not a Bug | Impact on me: | |
Category: | MySQL Server: General | Severity: | S3 (Non-critical) |
Version: | 5.6.10 | OS: | Linux (CentOS 6.3) |
Assigned to: | CPU Architecture: | Any | |
Tags: | alter_table_manage_keys, foreign_key |
[12 Apr 2013 8:06]
Tsubasa Tanaka
[12 Apr 2013 8:13]
Tsubasa Tanaka
Uploaded to ftp server. core-68939.gz schema-68939.sql(table definitions without data)
[12 Apr 2013 9:35]
MySQL Verification Team
I think your schema-68939.sql is wrong file. Can you upload the full schema again, output of: mysqladmin -uroot --no-data --all-databases > schema2-68939.sql ? Also, please start server with --thread-stack=5M and let me know if that prevents a crash?
[12 Apr 2013 9:42]
MySQL Verification Team
Also upload your mysqld file (compressed and named 68939-mysqld.gz (for example) and output of "rpm -qa|grep -i glibc" so we can examine the corefile
[12 Apr 2013 9:58]
Tsubasa Tanaka
I uploaded full schema including datas schema2-68939.sql.gz and mysqld-68939.gz. And it doesn't crash with --thread-stack=5M and "ALTER TABLE", it returns errno 150 correctly. mysql56> show tables; +----------------+ | Tables_in_test | +----------------+ | addresses | | conditions | | hosts | | scripts | +----------------+ 4 rows in set (0.00 sec) mysql56> SHOW GLOBAL VARIABLES LIKE 'thread_stack'; +---------------+---------+ | Variable_name | Value | +---------------+---------+ | thread_stack | 5242880 | +---------------+---------+ 1 row in set (0.00 sec) mysql56> ALTER TABLE scripts DROP PRIMARY KEY; ERROR 1025 (HY000): Error on rename of './logs/#sql-4690_5' to './logs/scripts' (errno: 150 - Foreign key constraint is incorrectly formed) Thanks.
[12 Apr 2013 10:41]
MySQL Verification Team
Hi! There is no crash in the corefile. There is a breakpoint, see int3 : (gdb) disassemble alter_table_manage_keys Dump of assembler code for function alter_table_manage_keys(TABLE*, int, Alter_info::enum_enable_or_disable): 0x0000000000712a70 <+0>: push %rbp 0x0000000000712a71 <+1>: cmp $0x1,%edx 0x0000000000712a74 <+4>: mov %rsp,%rbp 0x0000000000712a77 <+7>: push %r12 0x0000000000712a79 <+9>: push %rbx 0x0000000000712a7a <+10>: mov %rdi,%rbx 0x0000000000712a7d <+13>: je 0x712ac8 <alter_table_manage_keys(TABLE*, int, Alter_info::enum_enable_or_disable)+88> 0x0000000000712a7f <+15>: cmp $0x2,%edx 0x0000000000712a82 <+18>: je 0x712a94 <alter_table_manage_keys(TABLE*, int, Alter_info::enum_enable_or_disable)+36> 0x0000000000712a84 <+20>: test %edx,%edx 0x0000000000712a86 <+22>: je 0x712a90 <alter_table_manage_keys(TABLE*, int, Alter_info::enum_enable_or_disable)+32> 0x0000000000712a88 <+24>: xor %eax,%eax 0x0000000000712a8a <+26>: pop %rbx 0x0000000000712a8b <+27>: pop %r12 0x0000000000712a8d <+29>: leaveq 0x0000000000712a8e <+30>: retq 0x0000000000712a8f <+31>: nop 0x0000000000712a90 <+32>: int3 => 0x0000000000712a91 <+33>: divb 0x48(%rsp,%rsi,8) 0x0000000000712a95 <+37>: mov 0x8(%rbx),%edi 0x0000000000712a98 <+40>: mov $0x2,%esi Please check your source code, and your builds... Or try our official builds?
[12 Apr 2013 14:43]
Tsubasa Tanaka
Hi, > There is a breakpoint, Maybe I mistaked when I used gdb some days ago. I'm sorry for taking your time. Would you like to mark "Not a Bug" this case? > Or try our official builds? Yes, we take them in production environment, I'm testing in not-production environment in this time. Regards, tsubasa