Bug #12536 rpl_multi_delete causes core in slave process for rbr
Submitted: 11 Aug 2005 22:11 Modified: 15 Sep 2005 15:33
Reporter: Jonathan Miller Email Updates:
Status: Closed Impact on me:
None 
Category:MySQL Server: Replication Severity:S1 (Critical)
Version:5.0-wl1012 OS:Linux (Linux)
Assigned to: Lars Thalmann CPU Architecture:Any

[11 Aug 2005 22:11] Jonathan Miller
Description:
This may be the same as bug #12326, but there was enough different in the back trace that I decided to open a new bug report.

BT:
#0  0x40250da1 in kill () from /lib/libc.so.6
(gdb) bt
#0  0x40250da1 in kill () from /lib/libc.so.6
#1  0x40181f4a in pthread_kill () from /lib/libpthread.so.0
#2  0x0828fd2d in write_core (sig=11) at stacktrace.c:220
#3  0x0818b4ed in handle_segfault (sig=11) at mysqld.cc:2064
#4  0x401855cd in __pthread_sighandler () from /lib/libpthread.so.0
#5  <signal handler called>
#6  0x40297662 in _int_malloc () from /lib/libc.so.6
#7  0x402968a0 in malloc () from /lib/libc.so.6
#8  0x084b8ea4 in my_malloc (size=128, my_flags=24) at my_malloc.c:35
#9  0x08182d5a in prepare_pending<Delete_rows_log_event> (thd=0x8771568,
    table=0x872b058, server_id=1, cols=0xbedff7fc, colcnt=0, needed=5,
    is_transactional=false) at log_event.h:614
#10 0x08181231 in THD::binlog_delete_row(st_table*, bool, st_bitmap const*, unsigned, char const*) (this=0x8771568, table=0x872b058, is_trans=false,
    cols=0xbedff7fc, colcnt=0, record=0x8746228 "�\001") at sql_class.cc:2234
#11 0x08236c55 in binlog_log_row<Delete_rows_log_event> (table=0x872b058,
    before_record=0x40189bdc "�\n\001", after_record=0x0) at log_event.h:2047
#12 0x08236821 in handler::ha_delete_row(char const*) (this=0x86c4e01,
    buf=0xbedff7fc "��߾\001") at handler.cc:2579
#13 0x0820c570 in Delete_rows_log_event::do_exec_row(st_table*, st_relay_log_info*) (this=0x40189bdc, table=0x8771568, rli=0x873d6dc) at log_event.cc:5871
#14 0x0820b59c in Rows_log_event::exec_event(st_relay_log_info*) (
    this=0x872b910, rli=0x873d6dc) at log_event.cc:5096
---Type <return> to continue, or q <return> to quit---
#15 0x08285c3f in exec_relay_log_event (thd=0x8771568, rli=0x873d6dc)
    at slave.cc:2960
#16 0x082839a8 in handle_slave_sql (arg=0x87374f8) at slave.cc:3508
#17 0x4017f6de in pthread_start_thread () from /lib/libpthread.so.0

How to repeat:
run rpl_multi_delete.test against rbr
[11 Sep 2005 23:17] Lars Thalmann
Not failing when running:
./mysql-test-run --mysqld="--binlog-format=row" --skip-ndb --do-test=rpl_multi_delete
[12 Sep 2005 11:53] Jonathan Miller
Still causes core on slave.
[15 Sep 2005 15:33] Lars Thalmann
Fixed