[MDEV-30883] Assertion `err != DB_DUPLICATE_KEY' failed in row_rename_table_for_mysql on ALTER TABLE Created: 2023-03-20  Updated: 2023-11-28

Status: Confirmed
Project: MariaDB Server
Component/s: Replication, Storage Engine - InnoDB, XA
Affects Version/s: 10.6, 10.9, 10.10, 10.11, 11.0, 11.1, 11.2, 11.3, 11.4
Fix Version/s: 10.6, 10.11, 11.0, 11.1, 11.2, 11.3

Type: Bug Priority: Major
Reporter: Angelique Sklavounos (Inactive) Assignee: Nikita Malyavin
Resolution: Unresolved Votes: 0
Labels: upstream-8.0

Issue Links:
Relates
relates to MDEV-14193 innodb.log_file_name failed in buildb... Open
relates to MDEV-30105 ER_TABLE_EXISTS_ERROR or Assertion `e... Open

 Description   

10.6 535a38c4559160c7a4

 
Version: '10.6.13-MariaDB-debug-log'  socket: '/dev/shm/alter/mysql.sock'  port: 19300  Source distribution
 
mysqld: /home/angelique/dev/server/storage/innobase/row/row0mysql.cc:2691: dberr_t row_rename_table_for_mysql(const char*, const char*, trx_t*, bool): Assertion `err != DB_DUPLICATE_KEY' failed.
 
230320 14:19:51 [ERROR] mysqld got signal 6 ;
 
This could be because you hit a bug. It is also possible that this binary
 
or one of the libraries it was linked against is corrupt, improperly built,
 
or misconfigured. This error can also be caused by malfunctioning hardware.
 
 
 
To report this bug, see https://mariadb.com/kb/en/reporting-bugs
 
 
 
We will try our best to scrape up some info that will hopefully help
 
diagnose the problem, but since we have already crashed,
 
something is definitely wrong and this may fail.
 
 
 
Server version: 10.6.13-MariaDB-debug-log source revision: 535a38c4559160c7a46f0d9ece52495b83718d45
 
key_buffer_size=134217728
 
read_buffer_size=131072
 
max_used_connections=3
 
max_threads=153
 
thread_count=3
 
It is possible that mysqld could use up to
 
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 468107 K  bytes of memory
 
Hope that's ok; if not, decrease some variables in the equation.
 
 
 
Thread pointer: 0xffff34000dd8
 
Attempting backtrace. You can use the following information to find out
 
where mysqld died. If you see no messages after this, something went
 
terribly wrong...
 
stack_bottom = 0xffff98543508 thread_stack 0x49000
 
mysys/stacktrace.c:215(my_print_stacktrace)[0xaaaaebdcf9e4]
 
sql/signal_handler.cc:238(handle_fatal_signal)[0xaaaaeb48c8c4]
 
addr2line: 'linux-vdso.so.1': No such file
 
linux-vdso.so.1(__kernel_rt_sigreturn+0x0)[0xffffb14af7dc]
 
nptl/pthread_kill.c:44(__pthread_kill_implementation)[0xffffb082f200]
 
posix/raise.c:27(__GI_raise)[0xffffb07ea67c]
 
stdlib/abort.c:81(__GI_abort)[0xffffb07d7130]
 
assert/assert.c:89(__assert_fail_base)[0xffffb07e3fd0]
 
assert/assert-perr.c:31(__GI___assert_perror_fail)[0xffffb07e4040]
 
row/row0mysql.cc:2695(row_rename_table_for_mysql(char const*, char const*, trx_t*, bool))[0xaaaaebabd9bc]
 
handler/ha_innodb.cc:13763(innobase_rename_table(trx_t*, char const*, char const*, bool))[0xaaaaeb8e9130]
 
handler/ha_innodb.cc:14184(ha_innobase::rename_table(char const*, char const*))[0xaaaaeb8eaa44]
 
sql/handler.cc:5376(handler::ha_rename_table(char const*, char const*))[0xaaaaeb49f860]
 
sql/sql_table.cc:5007(mysql_rename_table(handlerton*, st_mysql_const_lex_string const*, st_mysql_const_lex_string const*, st_mysql_const_lex_string const*, st_mysql_const_lex_string const*, st_mysql_const_unsigned_lex_string*, unsigned int))[0xaaaaeb1f5f34]
 
sql/sql_table.cc:9241(simple_rename_or_index_change(THD*, TABLE_LIST*, Alter_info::enum_enable_or_disable, TRIGGER_RENAME_PARAM*, Alter_table_ctx*))[0xaaaaeb2014d8]
 
sql/sql_table.cc:9945(mysql_alter_table(THD*, st_mysql_const_lex_string const*, st_mysql_const_lex_string const*, HA_CREATE_INFO*, TABLE_LIST*, Recreate_info*, Alter_info*, unsigned int, st_order*, bool, bool))[0xaaaaeb2030f0]
 
sql/sql_alter.cc:553(Sql_cmd_alter_table::execute(THD*))[0xaaaaeb2bb9ec]
 
sql/sql_parse.cc:6002(mysql_execute_command(THD*, bool))[0xaaaaeb10b9c0]
 
sql/sql_parse.cc:8021(mysql_parse(THD*, char*, unsigned int, Parser_state*))[0xaaaaeb1115f8]
 
sql/sql_parse.cc:1898(dispatch_command(enum_server_command, THD*, char*, unsigned int, bool))[0xaaaaeb0ff080]
 
sql/sql_parse.cc:1409(do_command(THD*, bool))[0xaaaaeb0fdce4]
 
sql/sql_connect.cc:1416(do_handle_one_connection(CONNECT*, bool))[0xaaaaeb2b0f18]
 
sql/sql_connect.cc:1320(handle_one_connection)[0xaaaaeb2b0c5c]
 
perfschema/pfs.cc:2203(pfs_spawn_thread)[0xaaaaeb7f2504]
 
nptl/pthread_create.c:442(start_thread)[0xffffb082d5c8]
 
aarch64/clone.S:82(thread_start)[0xffffb0895d1c]
 
 
 
Trying to get some variables.
 
Some pointers may be invalid and cause the dump to abort.
 
Query (0xffff340141f0): ALTER TABLE t1 /*!100301 WAIT 6 */ RENAME TO t6 /* QNO 5029 CON_ID 10 */


# 2023-03-20T14:20:03 [2792643] Core was generated by `/home/angelique/dev/build-10.6/build1/bin/mysqld --no-defaults --basedir=/home/'.
 
# 2023-03-20T14:20:03 [2792643] Program terminated with signal SIGABRT, Aborted.
 
# 2023-03-20T14:20:03 [2792643] #0  __pthread_kill_implementation (threadid=281473237401216, signo=6, no_tid=<optimized out>) at ./nptl/pthread_kill.c:44
 
# 2023-03-20T14:20:03 [2792643] [Current thread is 1 (Thread 0xffff98543e80 (LWP 2792715))]
 
# 2023-03-20T14:20:03 [2792643] #0  __pthread_kill_implementation (threadid=281473237401216, signo=6, no_tid=<optimized out>) at ./nptl/pthread_kill.c:44
 
# 2023-03-20T14:20:03 [2792643] #1  0x0000aaaaebdcfaa0 in my_write_core (sig=6) at /home/angelique/dev/server/mysys/stacktrace.c:424
 
# 2023-03-20T14:20:03 [2792643] #2  0x0000aaaaeb48cca4 in handle_fatal_signal (sig=6) at /home/angelique/dev/server/sql/signal_handler.cc:357
 
# 2023-03-20T14:20:03 [2792643] #3  <signal handler called>
 
# 2023-03-20T14:20:03 [2792643] #4  __pthread_kill_implementation (threadid=281473237401216, signo=signo@entry=6, no_tid=no_tid@entry=0) at ./nptl/pthread_kill.c:44
 
# 2023-03-20T14:20:03 [2792643] #5  0x0000ffffb082f254 in __pthread_kill_internal (signo=6, threadid=<optimized out>) at ./nptl/pthread_kill.c:78
 
# 2023-03-20T14:20:03 [2792643] #6  0x0000ffffb07ea67c in __GI_raise (sig=sig@entry=6) at ../sysdeps/posix/raise.c:26
 
# 2023-03-20T14:20:03 [2792643] #7  0x0000ffffb07d7130 in __GI_abort () at ./stdlib/abort.c:79
 
# 2023-03-20T14:20:03 [2792643] #8  0x0000ffffb07e3fd0 in __assert_fail_base (fmt=0xffffb08fd3f8 "%s%s%s:%u: %s%sAssertion `%s' failed.\n%n", assertion=assertion@entry=0xaaaaec286118 "err != DB_DUPLICATE_KEY", file=file@entry=0xaaaaec2847f8 "/home/angelique/dev/server/storage/innobase/row/row0mysql.cc", line=line@entry=2691, function=function@entry=0xaaaaec285fe0 "dberr_t row_rename_table_for_mysql(const char*, const char*, trx_t*, bool)") at ./assert/assert.c:92
 
# 2023-03-20T14:20:03 [2792643] #9  0x0000ffffb07e4040 in __GI___assert_fail (assertion=0xaaaaec286118 "err != DB_DUPLICATE_KEY", file=0xaaaaec2847f8 "/home/angelique/dev/server/storage/innobase/row/row0mysql.cc", line=2691, function=0xaaaaec285fe0 "dberr_t row_rename_table_for_mysql(const char*, const char*, trx_t*, bool)") at ./assert/assert.c:101
 
# 2023-03-20T14:20:03 [2792643] #10 0x0000aaaaebabd9bc in row_rename_table_for_mysql (old_name=0xffff9853e088 "test/t1", new_name=0xffff9853de88 "test/t6", trx=0xffffacf4a500, use_fk=true) at /home/angelique/dev/server/storage/innobase/row/row0mysql.cc:2691
 
# 2023-03-20T14:20:03 [2792643] #11 0x0000aaaaeb8e9130 in innobase_rename_table (trx=0xffffacf4a500, from=0xffff9853ec08 "./test/t1", to=0xffff9853ee08 "./test/t6", use_fk=true) at /home/angelique/dev/server/storage/innobase/handler/ha_innodb.cc:13763
 
# 2023-03-20T14:20:03 [2792643] #12 0x0000aaaaeb8eaa44 in ha_innobase::rename_table (this=0xffff34014de0, from=0xffff9853ec08 "./test/t1", to=0xffff9853ee08 "./test/t6") at /home/angelique/dev/server/storage/innobase/handler/ha_innodb.cc:14184
 
# 2023-03-20T14:20:03 [2792643] #13 0x0000aaaaeb49f860 in handler::ha_rename_table (this=0xffff34014de0, from=0xffff9853ec08 "./test/t1", to=0xffff9853ee08 "./test/t6") at /home/angelique/dev/server/sql/handler.cc:5375
 
# 2023-03-20T14:20:03 [2792643] #14 0x0000aaaaeb1f5f34 in mysql_rename_table (base=0xaaaaf6674398, old_db=0xffff98540cf8, old_name=0xffff98540d08, new_db=0xffff98540d38, new_name=0xffff98540d58, id=0xffff9853f4a0, flags=0) at /home/angelique/dev/server/sql/sql_table.cc:5007
 
# 2023-03-20T14:20:03 [2792643] #15 0x0000aaaaeb2014d8 in simple_rename_or_index_change (thd=0xffff34000dd8, table_list=0xffff340144c0, keys_onoff=Alter_info::LEAVE_AS_IS, trigger_param=0xffff9853fd28, alter_ctx=0xffff98540ce8) at /home/angelique/dev/server/sql/sql_table.cc:9241
 
# 2023-03-20T14:20:03 [2792643] #16 0x0000aaaaeb2030f0 in mysql_alter_table (thd=0xffff34000dd8, new_db=0xffff34005a40, new_name=0xffff34005e90, create_info=0xffff98541af0, table_list=0xffff340144c0, recreate_info=0xffff985419d0, alter_info=0xffff98541a08, order_num=0, order=0x0, ignore=false, if_exists=false) at /home/angelique/dev/server/sql/sql_table.cc:9945
 
# 2023-03-20T14:20:03 [2792643] #17 0x0000aaaaeb2bb9ec in Sql_cmd_alter_table::execute (this=0xffff34014c08, thd=0xffff34000dd8) at /home/angelique/dev/server/sql/sql_alter.cc:553
 
# 2023-03-20T14:20:03 [2792643] #18 0x0000aaaaeb10b9c0 in mysql_execute_command (thd=0xffff34000dd8, is_called_from_prepared_stmt=false) at /home/angelique/dev/server/sql/sql_parse.cc:6002
 
# 2023-03-20T14:20:03 [2792643] #19 0x0000aaaaeb1115f8 in mysql_parse (thd=0xffff34000dd8, rawbuf=0xffff340141f0 "ALTER TABLE t1 /*!100301 WAIT 6 */ RENAME TO t6 /* QNO 5029 CON_ID 10 */", length=72, parser_state=0xffff98542bf8) at /home/angelique/dev/server/sql/sql_parse.cc:8021
 
# 2023-03-20T14:20:03 [2792643] #20 0x0000aaaaeb0ff080 in dispatch_command (command=COM_QUERY, thd=0xffff34000dd8, packet=0xffff3400b849 "ALTER TABLE t1 /*!100301 WAIT 6 */ RENAME TO t6 /* QNO 5029 CON_ID 10 */ ", packet_length=73, blocking=true) at /home/angelique/dev/server/sql/sql_parse.cc:1896
 
# 2023-03-20T14:20:03 [2792643] #21 0x0000aaaaeb0fdce4 in do_command (thd=0xffff34000dd8, blocking=true) at /home/angelique/dev/server/sql/sql_parse.cc:1409
 
# 2023-03-20T14:20:03 [2792643] #22 0x0000aaaaeb2b0f18 in do_handle_one_connection (connect=0xaaaaf6c53658, put_in_cache=true) at /home/angelique/dev/server/sql/sql_connect.cc:1416
 
# 2023-03-20T14:20:03 [2792643] #23 0x0000aaaaeb2b0c5c in handle_one_connection (arg=0xaaaaf6c53658) at /home/angelique/dev/server/sql/sql_connect.cc:1318
 
# 2023-03-20T14:20:03 [2792643] #24 0x0000aaaaeb7f2504 in pfs_spawn_thread (arg=0xaaaaf6bc84a8) at /home/angelique/dev/server/storage/perfschema/pfs.cc:2201
 
# 2023-03-20T14:20:03 [2792643] #25 0x0000ffffb082d5c8 in start_thread (arg=0x0) at ./nptl/pthread_create.c:442
 
# 2023-03-20T14:20:03 [2792643] #26 0x0000ffffb0895d1c in thread_start () at ../sysdeps/unix/sysv/linux/aarch64/clone.S:79

The crash only seems to happen with more than 1 thread. The actual ALTER TABLE command it fails on will differ among runs.

To recreate the crash

git clone https://github.com/MariaDB/randgen --branch mdev30883 mdev30883
 
cd mdev30883
 
./mdev30883.cmd <your basedir>



 Comments   
Comment by Roel Van de Paar [ 2023-11-22 ] 

--source include/have_innodb.inc
 
--source include/have_binlog_format_mixed.inc
 
CREATE TABLE t1 (a tinyBLOB) ENGINE=InnoDB;
 
SET SESSION pseudo_slave_mode=ON;
 
SET default_storage_engine=InnoDB,default_storage_engine='HEAP',GLOBAL default_storage_engine='MERGE';
 
XA START 'a';
 
INSERT INTO t1 VALUES (1);
 
XA END 'a';
 
XA PREPARE 'a';
 
ALTER TABLE t1 ENGINE=MERGE UNION (t_not_exists,t2);
 
DROP TABLE t1;
 
CREATE TABLE t1 (c1 SMALLINT,c2 INT,c3 BINARY (1));
 
ALTER TABLE t1 ENGINE=InnoDB;

Leads to:

10.6.16 6f091434f313996436b4f01d7e7d62a9f3786958 (Debug)

 
mariadbd: /test/10.6_dbg/storage/innobase/row/row0mysql.cc:2695: dberr_t row_rename_table_for_mysql(const char*, const char*, trx_t*, bool): Assertion `err != DB_DUPLICATE_KEY' failed.

10.6.16 6f091434f313996436b4f01d7e7d62a9f3786958 (Debug)

 
Core was generated by `/test/MD311023-mariadb-10.6.16-linux-x86_64-dbg/bin/mariadbd --no-defaults --co'.
 
Program terminated with signal SIGABRT, Aborted.
 
#0  __pthread_kill_implementation (no_tid=0, signo=6, threadid=22694742533696)
 
    at ./nptl/pthread_kill.c:44
 
[Current thread is 1 (Thread 0x14a408112640 (LWP 3954235))]
 
(gdb) bt
 
#0  __pthread_kill_implementation (no_tid=0, signo=6, threadid=22694742533696) at ./nptl/pthread_kill.c:44
 
#1  __pthread_kill_internal (signo=6, threadid=22694742533696) at ./nptl/pthread_kill.c:78
 
#2  __GI___pthread_kill (threadid=22694742533696, signo=signo@entry=6) at ./nptl/pthread_kill.c:89
 
#3  0x000014a41ce42476 in __GI_raise (sig=sig@entry=6) at ../sysdeps/posix/raise.c:26
 
#4  0x000014a41ce287f3 in __GI_abort () at ./stdlib/abort.c:79
 
#5  0x000014a41ce2871b in __assert_fail_base (fmt=0x14a41cfdd150 "%s%s%s:%u: %s%sAssertion `%s' failed.\n%n", assertion=0x55a2c6c07a94 "err != DB_DUPLICATE_KEY", file=0x55a2c6c05bc8 "/test/10.6_dbg/storage/innobase/row/row0mysql.cc", line=2695, function=<optimized out>) at ./assert/assert.c:92
 
#6  0x000014a41ce39e96 in __GI___assert_fail (assertion=0x55a2c6c07a94 "err != DB_DUPLICATE_KEY", file=0x55a2c6c05bc8 "/test/10.6_dbg/storage/innobase/row/row0mysql.cc", line=2695, function=0x55a2c6c06ab0 "dberr_t row_rename_table_for_mysql(const char*, const char*, trx_t*, bool)") at ./assert/assert.c:101
 
#7  0x000055a2c6561ed4 in row_rename_table_for_mysql (old_name=<optimized out>, old_name@entry=0x14a40810c6d0 "test/#sql-alter-3c52c8-5", new_name=new_name@entry=0x14a40810c4d0 "test/t1", trx=trx@entry=0x14a403c01c80, use_fk=use_fk@entry=true) at /test/10.6_dbg/storage/innobase/row/row0mysql.cc:2695
 
#8  0x000055a2c63e9990 in innobase_rename_table (trx=trx@entry=0x14a403c01c80, from=from@entry=0x14a40810d1e0 "./test/#sql-alter-3c52c8-5", to=to@entry=0x14a40810d3e0 "./test/t1", use_fk=use_fk@entry=true) at /test/10.6_dbg/storage/innobase/handler/ha_innodb.cc:13825
 
#9  0x000055a2c63f2ce9 in ha_innobase::rename_table (this=<optimized out>, from=<optimized out>, to=<optimized out>) at /test/10.6_dbg/storage/innobase/handler/ha_innodb.cc:14247
 
#10 0x000055a2c606f45c in handler::ha_rename_table (this=this@entry=0x14a3d00152a0, from=from@entry=0x14a40810d1e0 "./test/#sql-alter-3c52c8-5", to=to@entry=0x14a40810d3e0 "./test/t1") at /test/10.6_dbg/sql/handler.cc:5394
 
#11 0x000055a2c5e672f4 in mysql_rename_table (base=base@entry=0x55a2c963acc8, old_db=old_db@entry=0x14a40810f4a0, old_name=old_name@entry=0x14a40810f4d0, new_db=new_db@entry=0x14a40810f4a0, new_name=new_name@entry=0x14a40810f4c0, id=id@entry=0x14a40810f4f0, flags=1) at /test/10.6_dbg/sql/sql_table.cc:5007
 
#12 0x000055a2c5e7c999 in mysql_alter_table (thd=thd@entry=0x14a3d0000d58, new_db=new_db@entry=0x14a3d00059e0, new_name=new_name@entry=0x14a3d0005e38, create_info=create_info@entry=0x14a4081102e0, table_list=<optimized out>, table_list@entry=0x14a3d00134d0, recreate_info=recreate_info@entry=0x14a408110150, alter_info=<optimized out>, order_num=<optimized out>, order=<optimized out>, ignore=<optimized out>, if_exists=<optimized out>) at /test/10.6_dbg/sql/sql_table.cc:10791
 
#13 0x000055a2c5efe30e in Sql_cmd_alter_table::execute (this=<optimized out>, thd=0x14a3d0000d58) at /test/10.6_dbg/sql/sql_alter.cc:672
 
#14 0x000055a2c5dad762 in mysql_execute_command (thd=thd@entry=0x14a3d0000d58, is_called_from_prepared_stmt=is_called_from_prepared_stmt@entry=false) at /test/10.6_dbg/sql/sql_parse.cc:6026
 
#15 0x000055a2c5daf30a in mysql_parse (thd=thd@entry=0x14a3d0000d58, rawbuf=<optimized out>, length=<optimized out>, parser_state=parser_state@entry=0x14a4081111f0) at /test/10.6_dbg/sql/sql_parse.cc:8050
 
#16 0x000055a2c5db1680 in dispatch_command (command=command@entry=COM_QUERY, thd=thd@entry=0x14a3d0000d58, packet=packet@entry=0x14a3d000b049 "", packet_length=packet_length@entry=28, blocking=blocking@entry=true) at /test/10.6_dbg/sql/sql_class.h:241
 
#17 0x000055a2c5db379d in do_command (thd=0x14a3d0000d58, blocking=blocking@entry=true) at /test/10.6_dbg/sql/sql_parse.cc:1409
 
#18 0x000055a2c5ef71d9 in do_handle_one_connection (connect=<optimized out>, connect@entry=0x55a2c9a859a8, put_in_cache=put_in_cache@entry=true) at /test/10.6_dbg/sql/sql_connect.cc:1416
 
#19 0x000055a2c5ef74ce in handle_one_connection (arg=arg@entry=0x55a2c9a859a8) at /test/10.6_dbg/sql/sql_connect.cc:1318
 
#20 0x000055a2c6344e16 in pfs_spawn_thread (arg=0x55a2c9b12a08) at /test/10.6_dbg/storage/perfschema/pfs.cc:2201
 
#21 0x000014a41ce94ac3 in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:442
 
#22 0x000014a41cf26a40 in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81

Bug confirmed present in:
MariaDB: 10.6.16 (dbg), 10.9.8 (dbg), 10.10.7 (dbg), 10.11.6 (dbg), 11.0.4 (dbg), 11.1.3 (dbg), 11.2.2 (dbg), 11.3.0 (dbg), 11.4.0 (dbg)

Bug (or feature/syntax) confirmed not present in:
MariaDB: 10.4.32 (dbg), 10.4.32 (opt), 10.5.23 (dbg), 10.5.23 (opt), 10.6.16 (opt), 10.9.8 (opt), 10.10.7 (opt), 10.11.6 (opt), 11.0.4 (opt), 11.1.3 (opt), 11.2.2 (opt), 11.3.0 (opt), 11.4.0 (opt)
MySQL: 5.5.62 (dbg), 5.5.62 (opt), 5.6.51 (dbg), 5.6.51 (opt), 5.7.40 (dbg), 8.0.33 (dbg), 8.0.33 (opt)

Additionally, the ALTER TABLE takes a long time (~50 sec) to complete:

10.6.16 6f091434f313996436b4f01d7e7d62a9f3786958 (Debug)

 
10.6.16-dbg>ALTER TABLE t1 ENGINE=MERGE UNION (t_not_exists,t2);
 
Query OK, 0 rows affected, 1 warning (50.013 sec)
 
Records: 0  Duplicates: 0  Warnings: 1

It should also be noted that MySQL 8.0 hang on the query even longer:

MySQL 8.0.33 ea7087d885006918ad54458e7aad215b1650312c (Optimized and Debug)

 
Command: Query
 
   Time: 614
 
  State: Waiting for table metadata lock
 
   Info: ALTER TABLE t1 ENGINE=MERGE UNION (t_not_exists,t2)

Comment by Roel Van de Paar [ 2023-11-22 ]

Also see possibly related MDEV-14193 (ref stack in first comment), MDEV-30992 (but no concurrency is involved here), MDEV-30105 (which is 10.3-10.4 only), MDEV-30883, and MDEV-25119 (which is 10.5+).

Comment by Marko Mäkelä [ 2023-11-22 ]

I think that the bug here is that the DDL statements are not being rejected while the current connection is attached to a transaction in XA PREPARE state. I do not think that this can be fixed in InnoDB.

Generated at Thu Feb 08 10:19:36 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.