| Bug #32421 | mysqlbinlog crashes server when sending row-based binlog events | ||
|---|---|---|---|
| Submitted: | 15 Nov 2007 20:55 | Modified: | 28 Jan 2008 8:53 |
| Reporter: | Guilhem Bichot | Email Updates: | |
| Status: | Duplicate | Impact on me: | |
| Category: | MySQL Server: Row Based Replication ( RBR ) | Severity: | S3 (Non-critical) |
| Version: | 5.1-bk | OS: | Linux |
| Assigned to: | Assigned Account | CPU Architecture: | Any |
[22 Nov 2007 11:11]
Sveta Smirnova
Thank you for the report. Verified as described. bt on Linux: Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1220482128 (LWP 2715)] 0x08293fad in unpack_row () (gdb) bt #0 0x08293fad in unpack_row () #1 0x082860a9 in Rows_log_event::write_row () #2 0x08286301 in Write_rows_log_event::do_exec_row () #3 0x082873d2 in Rows_log_event::do_apply_event () #4 0x0832218c in mysql_client_binlog_statement () #5 0x081ed712 in mysql_execute_command () #6 0x081f4d7f in mysql_parse () #7 0x081f5323 in dispatch_command () #8 0x081f6424 in do_command () #9 0x081e780f in handle_one_connection () #10 0x0045fbd4 in start_thread () from /lib/libpthread.so.0 #11 0x003b74fe in clone () from /lib/libc.so.6 bt on Mac: Program received signal EXC_BAD_ACCESS, Could not access memory. Reason: KERN_PROTECTION_FAILURE at address: 0x00000000 [Switching to process 25486 thread 0x2013] 0x00629d38 in table_def::field_metadata (this=0x0, index=0) at rpl_utility.h:195 195 DBUG_ASSERT(index < m_size); (gdb) bt #0 0x00629d38 in table_def::field_metadata (this=0x0, index=0) at rpl_utility.h:195 #1 0x001a5a28 in unpack_row (rli=0x5870200, table=0x5813818, colcnt=1, row_data=0x5873a18 "?\006", cols=0x4505450, row_end=0x45054b4, master_reclength=0x450547c) at rpl_record.cc:249 #2 0x0062961e in Rows_log_event::unpack_current_row (this=0x4505418, rli=0x5870200) at log_event.h:2320 #3 0x00195253 in Rows_log_event::write_row (this=0x4505418, rli=0x5870200, overwrite=true) at log_event.cc:7167 #4 0x00195c4f in Write_rows_log_event::do_exec_row (this=0x4505418, rli=0x5870200) at log_event.cc:7339 #5 0x00199b5f in Rows_log_event::do_apply_event (this=0x4505418, rli=0x5870200) at log_event.cc:6144 #6 0x0062fe57 in Log_event::apply_event (this=0x4505418, rli=0x5870200) at log_event.h:843 #7 0x00285c05 in mysql_client_binlog_statement (thd=0x5825818) at sql_binlog.cc:176 #8 0x000d5315 in mysql_execute_command (thd=0x5825818) at sql_parse.cc:4410 #9 0x000d5996 in mysql_parse (thd=0x5825818, inBuf=0x586c428 "BINLOG '\n1WJFRxcBAAAADwQAAHkEAAAAAB", 'A' <repeats 11 times>, "f/+BgAAAP4HAAAA/gYAAAD+BwAAAP4HAAAA/ggA\nAAD+BQAAAP4GAAAA/gYAAAD+BwAAAP4GAAAA/gcAAAD+BwAAAP4IAAAA/gYAAAD+BwAAAP4HAAAA\n/ggAAAD+BwAAAP4IAAAA/ggAAAD+CQAAAP4FA"..., length=2187, found_semicolon=0xb0061dcc) at sql_parse.cc:5487 #10 0x000d6676 in dispatch_command (command=COM_QUERY, thd=0x5825818, packet=0x444b019 "BINLOG '\n1WJFRxcBAAAADwQAAHkEAAAAAB", 'A' <repeats 11 times>, "f/+BgAAAP4HAAAA/gYAAAD+BwAAAP4HAAAA/ggA\nAAD+BQAAAP4GAAAA/gYAAAD+BwAAAP4GAAAA/gcAAAD+BwAAAP4IAAAA/gYAAAD+BwAAAP4HAAAA\n/ggAAAD+BwAAAP4IAAAA/ggAAAD+CQAAAP4FA"..., packet_length=2187) at sql_parse.cc:991 #11 0x000d78e6 in do_command (thd=0x5825818) at sql_parse.cc:725 #12 0x000c42d9 in handle_one_connection (arg=0x5825818) at sql_connect.cc:1107 #13 0x90024227 in _pthread_body ()
[28 Jan 2008 8:52]
Zhenxing He
duplicate of bug#27779, and already been fixed
Description: My tree is a few weeks old. Under certain circumstances, I see repeatable segmentation fault of mysqld: Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -1219826800 (LWP 11543)] 0x082f83ae in unpack_row (rli=0x8902c08, table=0x890dba8, colcnt=1, row_data=0x89051f8 "รพ\006", cols=0x890e9ec, row_end=0x890ea50, master_reclength=0x890ea18) at rpl_utility.h:195 195 DBUG_ASSERT(index < m_size); when piping into it BINLOG statements produced by mysqlbinlog How to repeat: [INS 21:54 /m/mysql-5.1-maint/mysql-test $] cat t/bug.test -- source include/have_binlog_format_row.inc create table t(a int) engine=myisam; insert into t values(1); insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; insert into t select a+1 from t; [INS 21:57 /m/mysql-5.1-maint/mysql-test $] cat t/bug-master.opt -O max_binlog_size=4096 then ./mtr --record --mem --mysqld=--binlog-format=row bug --gdb This runs the test and creates six binlogs. Transform these binlogs into plain text: ../client/mysqlbinlog var/log/master-bin.00000* > a Now connect to mysqld (still alive as we used --gdb): ../client/mysql -uroot -S var/tmp/master.soc -e "drop table test.t" (this drop is needed because the binlog contains a CREATE) and then ../client/mysql -uroot -S var/tmp/master.sock < ./a I see the crash during execution of one BINLOG statement.