MDEV-32444 Data from orphaned XA transaction is lost after online alter

XA support for online alter was totally missing.

Tying on binlog_hton made this hardly visible: simply having binlog_commit
called from xa_commit made an impression that it will automagically work
for online alter, which turns out wrong: all binlog does is writes
"XA END" into trx cache and flushes it to a real binlog.

In comparison, online alter can't do the same, since online replication
happens in a single transaction.

Solution: make a dedicated XA support.
* Extend struct xid_t with a pointer to Online_alter_cache_list
* On prepare: move online alter cache from THD::ha_data to XID passed
* On XA commit/rollback: use the online alter cache stored in this XID.
  This makes us pass xid_cache_element->xid to xa_commit/xa_rollback
  instead of lex->xid
* Use manual memory management for online alter cache list, instead of
  mem_root allocation, since we don't have mem_root connected to the XA
  transaction.

mysql-test/main/alter_table_online_debug.result

@@ -1563,6 +1563,116 @@ connection default;
 drop table t1, t2;
 set @@binlog_format=default;
 set debug_sync= reset;
+# MDEV-32444 Data from orphaned XA transaction is lost after online alter
+create table t (a int primary key) engine=innodb;
+insert into t values (1);
+# XA commit
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+alter table t force, algorithm=copy, lock=none;
+connection con1;
+set debug_sync= 'now wait_for downgraded';
+xa begin 'x1';
+update t set a = 2 where a = 1;
+xa end 'x1';
+xa prepare 'x1';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+disconnect con1;
+connection con2;
+set debug_sync= 'now wait_for xa_detach';
+xa commit 'x1';
+set debug_sync= 'now signal go';
+connection default;
+select * from t;
+a
+2
+# XA rollback
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+alter table t force, algorithm=copy, lock=none;
+connect con1, localhost, root,,;
+set debug_sync= 'now wait_for downgraded';
+xa begin 'x2';
+insert into t values (53);
+xa end 'x2';
+xa prepare 'x2';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+disconnect con1;
+connection con2;
+set debug_sync= 'now wait_for xa_detach';
+xa rollback 'x2';
+set debug_sync= 'now signal go';
+connection default;
+select * from t;
+a
+2
+# XA transaction is left uncommitted
+# end then is rollbacked after alter fails
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+set statement innodb_lock_wait_timeout=0, lock_wait_timeout= 0
+for alter table t force, algorithm=copy, lock=none;
+connect con1, localhost, root,,;
+set debug_sync= 'now wait_for downgraded';
+xa begin 'xuncommitted';
+insert into t values (3);
+xa end 'xuncommitted';
+xa prepare 'xuncommitted';
+set debug_sync= 'now signal go';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+disconnect con1;
+connection default;
+ERROR HY000: Lock wait timeout exceeded; try restarting transaction
+set debug_sync= 'now wait_for xa_detach';
+xa rollback 'xuncommitted';
+select * from t;
+a
+2
+# Same, but commit
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+set statement innodb_lock_wait_timeout=0, lock_wait_timeout= 0
+for alter table t force, algorithm=copy, lock=none;
+connect con1, localhost, root,,;
+set debug_sync= 'now wait_for downgraded';
+xa begin 'committed_later';
+insert into t values (3);
+xa end 'committed_later';
+xa prepare 'committed_later';
+set debug_sync= 'now signal go';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+disconnect con1;
+connection default;
+ERROR HY000: Lock wait timeout exceeded; try restarting transaction
+set debug_sync= 'now wait_for xa_detach';
+xa commit 'committed_later';
+select * from t;
+a
+2
+3
+# Commit, but error in statement, and there is some stmt data to rollback
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+alter table t force, algorithm=copy, lock=none;
+connect con1, localhost, root,,;
+set debug_sync= 'now wait_for downgraded';
+xa begin 'x1';
+insert into t values (4), (3);
+ERROR 23000: Duplicate entry '3' for key 'PRIMARY'
+insert into t values (5);
+xa end 'x1';
+xa prepare 'x1';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+disconnect con1;
+connection con2;
+set debug_sync= 'now wait_for xa_detach';
+xa commit 'x1';
+set debug_sync= 'now signal go';
+connection default;
+select * from t;
+a
+2
+3
+5
+connect con1, localhost, root,,;
+connection default;
+drop table t;
+set debug_sync= reset;
 disconnect con1;
 disconnect con2;
 #

mysql-test/main/alter_table_online_debug.test

@@ -1791,6 +1791,135 @@ set @@binlog_format=default;
 set debug_sync= reset;
 
 
+--echo # MDEV-32444 Data from orphaned XA transaction is lost after online alter
+
+create table t (a int primary key) engine=innodb;
+insert into t values (1);
+
+--echo # XA commit
+
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+send alter table t force, algorithm=copy, lock=none;
+
+--connection con1
+set debug_sync= 'now wait_for downgraded';
+xa begin 'x1';
+update t set a = 2 where a = 1;
+xa end 'x1';
+xa prepare 'x1';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+--disconnect con1
+
+--connection con2
+set debug_sync= 'now wait_for xa_detach';
+xa commit 'x1';
+set debug_sync= 'now signal go';
+--connection default
+--reap # alter table
+
+select * from t;
+
+--echo # XA rollback
+
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+send alter table t force, algorithm=copy, lock=none;
+--connect(con1, localhost, root,,)
+set debug_sync= 'now wait_for downgraded';
+xa begin 'x2';
+insert into t values (53);
+xa end 'x2';
+xa prepare 'x2';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+--disconnect con1
+
+--connection con2
+set debug_sync= 'now wait_for xa_detach';
+xa rollback 'x2';
+set debug_sync= 'now signal go';
+--connection default
+--reap # alter table
+
+select * from t;
+
+--echo # XA transaction is left uncommitted
+--echo # end then is rollbacked after alter fails
+
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+send set statement innodb_lock_wait_timeout=0, lock_wait_timeout= 0
+     for alter table t force, algorithm=copy, lock=none;
+
+--connect(con1, localhost, root,,)
+set debug_sync= 'now wait_for downgraded';
+xa begin 'xuncommitted';
+insert into t values (3);
+xa end 'xuncommitted';
+xa prepare 'xuncommitted';
+set debug_sync= 'now signal go';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+--disconnect con1
+
+--connection default
+--error ER_LOCK_WAIT_TIMEOUT
+--reap # alter table
+set debug_sync= 'now wait_for xa_detach';
+xa rollback 'xuncommitted';
+
+select * from t;
+
+--echo # Same, but commit
+
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+send set statement innodb_lock_wait_timeout=0, lock_wait_timeout= 0
+     for alter table t force, algorithm=copy, lock=none;
+
+--connect(con1, localhost, root,,)
+set debug_sync= 'now wait_for downgraded';
+xa begin 'committed_later';
+insert into t values (3);
+xa end 'committed_later';
+xa prepare 'committed_later';
+set debug_sync= 'now signal go';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+--disconnect con1
+
+--connection default
+--error ER_LOCK_WAIT_TIMEOUT
+--reap # alter table
+set debug_sync= 'now wait_for xa_detach';
+xa commit 'committed_later';
+
+
+select * from t;
+
+--echo # Commit, but error in statement, and there is some stmt data to rollback
+
+set debug_sync= 'alter_table_online_downgraded signal downgraded wait_for go';
+send alter table t force, algorithm=copy, lock=none;
+--connect(con1, localhost, root,,)
+set debug_sync= 'now wait_for downgraded';
+xa begin 'x1';
+--error ER_DUP_ENTRY
+insert into t values (4), (3);
+insert into t values (5);
+xa end 'x1';
+xa prepare 'x1';
+set debug_sync= 'THD_cleanup_after_trans_cleanup signal xa_detach';
+--disconnect con1
+
+--connection con2
+set debug_sync= 'now wait_for xa_detach';
+xa commit 'x1';
+set debug_sync= 'now signal go';
+--connection default
+--reap # alter table
+
+select * from t;
+
+--connect(con1, localhost, root,,)
+--connection default
+drop table t;
+set debug_sync= reset;
+
 --disconnect con1
 --disconnect con2
 --echo #

sql/handler.cc

@@ -2596,7 +2596,7 @@ static bool xarecover_decide_to_commit(xid_recovery_member* member,
 static void xarecover_do_commit_or_rollback(handlerton *hton,
                                             xarecover_complete_arg *arg)
 {
-  xid_t x;
+  XA_data x;
   my_bool rc;
   xid_recovery_member *member= arg->member;
   Binlog_offset *ptr_commit_max= arg->binlog_coord;
@@ -2605,7 +2605,7 @@ static void xarecover_do_commit_or_rollback(handlerton *hton,
     // Populate xid using the server_id from original transaction
     x.set(member->xid, member->server_id);
   else
-    x= *member->full_xid;
+    (XID)x= *member->full_xid;
 
   rc= xarecover_decide_to_commit(member, ptr_commit_max) ?
     hton->commit_by_xid(hton, &x) : hton->rollback_by_xid(hton, &x);

sql/handler.h

@@ -890,6 +890,7 @@ typedef ulonglong my_xid; // this line is the same as in log_event.h
 #define COMPATIBLE_DATA_YES 0
 #define COMPATIBLE_DATA_NO  1
 
+
 /**
   struct xid_t is binary compatible with the XID structure as
   in the X/Open CAE Specification, Distributed Transaction Processing:
@@ -973,6 +974,12 @@ struct xid_t {
 };
 typedef struct xid_t XID;
 
+struct Online_alter_cache_list;
+struct XA_data: XID
+{
+  Online_alter_cache_list *online_alter_cache= NULL;
+};
+
 /*
   Enumerates a sequence in the order of
   their creation that is in the top-down order of the index file.