handle_fatal_signal (sig=11) in sql_digest_storage::copy

[Roel Van de Paar]

2017-01-31T21:58:09.430634Z 0 [Note] /sda/PS300117-percona-server-5.7.17-11-linux-x86_64-debug/bin/mysqld: ready for connections.
Version: '5.7.17-11-debug' socket: '/sda/PS300117-percona-server-5.7.17-11-linux-x86_64-debug/socket.sock' port: 12438 MySQL Community Server (GPL)
21:58:13 UTC - mysqld got signal 11 ;

Core was generated by `/sda/PS300117-percona-server-5.7.17-11-linux-x86_64-debug/bin/mysqld --no-defau'.
Program terminated with signal 11, Segmentation fault.
#0 0x00007fd808000741 in __pthread_kill (threadid=<optimized out>, signo=11) at ../nptl/sysdeps/unix/sysv/linux/pthread_kill.c:61
61 val = INTERNAL_SYSCALL (tgkill, err, 3, THREAD_GETMEM (THREAD_SELF, pid),
(gdb) bt
#0 0x00007fd808000741 in __pthread_kill (threadid=<optimized out>, signo=11) at ../nptl/sysdeps/unix/sysv/linux/pthread_kill.c:61
#1 0x000000000185027e in my_write_core (sig=11) at /git/PS-5.7.17_dbg/mysys/stacktrace.c:249
#2 0x0000000000e81bc1 in handle_fatal_signal (sig=11) at /git/PS-5.7.17_dbg/sql/signal_handler.cc:223
#3 <signal handler called>
#4 __memcpy_ssse3_back () at ../sysdeps/x86_64/multiarch/memcpy-ssse3-back.S:1809
#5 0x000000000188a298 in sql_digest_storage::copy (this=0x7fd8044326d8, from=0x7fd7c381afd8) at /git/PS-5.7.17_dbg/sql/sql_digest.h:94
#6 0x000000000188ebd4 in find_or_create_digest (thread=0x7fd801420480, digest_storage=0x7fd7c381afd8, schema_name=0x7fd7c381b0c0 "test", schema_name_length=4)
    at /git/PS-5.7.17_dbg/storage/perfschema/pfs_digest.cc:285
#7 0x0000000001885de7 in pfs_end_statement_v1 (locker=0x7fd7c381b018, stmt_da=0x7fd7c381c0b8) at /git/PS-5.7.17_dbg/storage/perfschema/pfs.cc:5407
#8 0x00000000015159a9 in inline_mysql_end_statement (locker=0x7fd7c381b018, stmt_da=0x7fd7c381c0b8) at /git/PS-5.7.17_dbg/include/mysql/psi/mysql_statement.h:228
#9 0x000000000151b031 in dispatch_command (thd=0x7fd7c3819000, com_data=0x7fd8085ecc90, command=COM_QUERY) at /git/PS-5.7.17_dbg/sql/sql_parse.cc:1952
#10 0x0000000001518849 in do_command (thd=0x7fd7c3819000) at /git/PS-5.7.17_dbg/sql/sql_parse.cc:1023
#11 0x0000000001656dfe in handle_connection (arg=0x7fd7ee01f6a0) at /git/PS-5.7.17_dbg/sql/conn_handler/connection_handler_per_thread.cc:312
#12 0x000000000187ee4d in pfs_spawn_thread (arg=0x7fd7fd791420) at /git/PS-5.7.17_dbg/storage/perfschema/pfs.cc:2188
#13 0x00007fd807ffbdc5 in start_thread (arg=0x7fd8085ed700) at pthread_create.c:308
#14 0x00007fd80645673d in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:113

Start server with --no-defaults --performance-schema-digests-size=1
Then execute any query like SELECT 1;
This testcase does NOT crash MS 5.7.17
Related to TokuDB PFS perhaps?

[Laurynas Biveinis (laurynas-biveinis)]

TokuDB PFS is not merged. Roel, can you please test PS and MySQL under Valgrind and/or ASan with this testcase?

[Roel Van de Paar (roel11)]

Tested with Valgrind. Discussing w/ Laurynas

[Roel Van de Paar (roel11)]

Reproduced Valgrind SIGSEGV on MS also
Now dup of https://bugs.mysql.com/bug.php?id=84786

[Laurynas Biveinis (laurynas-biveinis)]

https://github.com/percona/percona-server/pull/1678

[Laurynas Biveinis (laurynas-biveinis)]

https://github.com/percona/percona-server/pull/1695

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PS-1056