Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
5.5(EOL), 10.0(EOL), 10.1(EOL), 10.2(EOL), 10.3(EOL)
-
None
-
10.1.29, 10.1.30, 5.5.59
Description
In the following example (which is quite common) only mysqld frames were resolved properly. Frames from shared objects are not resolved.
This is because addr2line attempts to resolve shared object frames while it is being called against mysqld binary.
To resolve shared object frames addr2line should be called individually for these objects.
mysys/stacktrace.c:269(my_print_stacktrace)[0x56491efeceff]
|
sql/signal_handler.cc:168(handle_fatal_signal)[0x56491e8126b5]
|
/lib/x86_64-linux-gnu/libpthread.so.0(+0x11390)[0x7fcaf2825390]
|
/lib/x86_64-linux-gnu/libc.so.6(gsignal+0x38)[0x7fcaf1bde428]
|
/lib/x86_64-linux-gnu/libc.so.6(abort+0x16a)[0x7fcaf1be002a]
|
/lib/x86_64-linux-gnu/libc.so.6(+0x2dbd7)[0x7fcaf1bd6bd7]
|
/lib/x86_64-linux-gnu/libc.so.6(+0x2dc82)[0x7fcaf1bd6c82]
|
/home/svoj/devel/maria/debug/mysql-test/var/plugins/proc_info.so(+0xd1f)[0x7fcaec38ad1f]
|
sql/sql_show.cc:8350(get_schema_tables_result(JOIN*, enum_schema_table_state))[0x56491e600e92]
|
sql/sql_select.cc:3511(JOIN::exec_inner())[0x56491e59ce16]
|
sql/sql_select.cc:3344(JOIN::exec())[0x56491e59c4bc]
|
sql/sql_select.cc:3745(mysql_select(THD*, TABLE_LIST*, unsigned int, List<Item>&, Item*, unsigned int, st_order*, st_order*, Item*, st_order*, unsigned long long, select_result*, st_select_lex_unit*, st_select_lex*))[0x56491e59d694]
|
sql/sql_select.cc:378(handle_select(THD*, LEX*, select_result*, unsigned long))[0x56491e591c08]
|
sql/sql_parse.cc:6467(execute_sqlcom_select(THD*, TABLE_LIST*))[0x56491e55cf79]
|
sql/sql_parse.cc:3731(mysql_execute_command(THD*))[0x56491e55312c]
|
sql/sql_parse.cc:7921(mysql_parse(THD*, char*, unsigned int, Parser_state*, bool, bool))[0x56491e5608a8]
|
sql/sql_parse.cc:1821(dispatch_command(enum_server_command, THD*, char*, unsigned int, bool, bool))[0x56491e54db6b]
|
sql/sql_parse.cc:1370(do_command(THD*))[0x56491e54c51a]
|
sql/sql_connect.cc:1418(do_handle_one_connection(CONNECT*))[0x56491e6a325a]
|
sql/sql_connect.cc:1325(handle_one_connection)[0x56491e6a2fad]
|
perfschema/pfs.cc:1864(pfs_spawn_thread)[0x56491ea46791]
|
/lib/x86_64-linux-gnu/libpthread.so.0(+0x76ba)[0x7fcaf281b6ba]
|
/lib/x86_64-linux-gnu/libc.so.6(clone+0x6d)[0x7fcaf1cb03dd]
|
Attachments
Issue Links
- causes
-
-
- Closed
-
- relates to
-
-
- Open
-
-
-
- Closed
-
Hi Sergey!
Can you please review the patch for this issue? The code is at:
http://lists.askmonty.org/pipermail/commits/2018-January/011826.html
The main worry I have is with the LIBDL change in mysys/CMakeLists.txt and if we need any checks for the extra used functions.
The rest is tested and produces the following output, with a purposefully introduced NULL pointer dereference:
Attempting backtrace. You can use the following information to find outwhere mysqld died. If you see no messages after this, something wentterribly wrong...stack_bottom = 0x0 thread_stack 0x48000mysys/stacktrace.c:246(my_print_stacktrace)[0x55d6b69c1ba4]sql/signal_handler.cc:155(handle_fatal_signal)[0x55d6b6484763]sigaction.c:0(__restore_rt)[0x7efcd9187da0]dict/dict0boot.c:456(dict_boot)[0x7efcd1759a34]dict/dict0boot.c:489(dict_create)[0x7efcd1759ac5]srv/srv0start.c:1681(innobase_start_or_create_for_mysql)[0x7efcd1893dcf]handler/ha_innodb.cc:2611(innobase_init(void*))[0x7efcd1814577]sql/handler.cc:470(ha_initialize_handlerton(st_plugin_int*))[0x55d6b6486632]sql/sql_plugin.cc:1376(plugin_initialize(st_mem_root*, st_plugin_int*, int*, char**, bool))[0x55d6b62ce4bd]sql/sql_plugin.cc:1650(plugin_init(int*, char**, int))[0x55d6b62cefb3]sql/mysqld.cc:4294(init_server_components())[0x55d6b6205e58]sql/mysqld.cc:4891(mysqld_main(int, char**))[0x55d6b6206f21]sql/main.cc:26(main)[0x55d6b61fee0a]/usr/lib/libc.so.6(__libc_start_main+0xea)[0x7efcd781ff4a]/home/vicentiu/Workspace/MariaDB-5.5/sql/mysqld(_start+0x2a)[0x55d6b61fed0a]