MDEV-25683 Atomic DDL: With innodb_force_recovery=3 InnoDB: Trying to load index but the index tree has been freed

The purpose of the parameter innodb_force_recovery is to allow some
data to be dumped from a corrupted database. Its values used to be
as follows:

innodb_force_recovery=0: normal (default)

innodb_force_recovery=1: ignore (skip log for) corrupted pages or
missing data files when applying the redo log

innodb_force_recovery=2: additionally, disable background tasks
(such as the purge of committed undo logs)

innodb_force_recovery=3: additionally, disable the rollback of
recovered incomplete (not committed or XA PREPARE) transactions

innodb_force_recovery=4: same as 3 (since MDEV-19514 in MariaDB 10.5)

innodb_force_recovery=5: additionally, do not process any undo log,
disallow any writes, and force READ UNCOMMITTED isolation level

innodb_force_recovery=6: additionally, pretend that ib_logfile0 does
not exist (prevent any recovery). Never use this!

The bad thing that happens with innodb_force_recovery=3 and
innodb_force_recovery=4 is that also the rollback of any recovered
DDL transaction will be skipped. This would break the DDL log recovery
that was introduced in MDEV-17567.

For one data directory sample, the DDL log recovery would hangs due to
a conflict on the InnoDB SYS_TABLES table, because the lock holder
transaction was not rolled back due to innodb_force_recovery=3.

Fix: Make innodb_force_recovery=3 skip the DML transaction rollback only,
and make innodb_force_recovery=4 (renamed to SRV_FORCE_NO_DDL_UNDO)
behave like innodb_force_recovery=3 used to (skip the rollback of all
recovered transactions, both DML and DDL).

Startup with innodb_force_recovery=4 will be unaffected by this change.
(There may be hangs, possibly preceded by messages about failing to
load an index.)

Side note: With innodb_force_recovery=5, any DDL log for InnoDB tables
will be essentially ignored by InnoDB, but the server will start up.

5 files changed, 11 insertions, 12 deletions

diff --git a/storage/innobase/dict/dict0stats.cc b/storage/innobase/dict/dict0stats.cc
index f92beac0f67..0af5982806f 100644
--- a/storage/innobase/dict/dict0stats.cc
+++ b/storage/innobase/dict/dict0stats.cc
@@ -3812,7 +3812,7 @@ dict_stats_update(
 
 	if (!table->is_readable()) {
 		return (dict_stats_report_error(table));
-	} else if (srv_force_recovery > SRV_FORCE_NO_IBUF_MERGE) {
+	} else if (srv_force_recovery >= SRV_FORCE_NO_UNDO_LOG_SCAN) {
 		/* If we have set a high innodb_force_recovery level, do
 		not calculate statistics, as a badly corrupted index can
 		cause a crash in it. */
diff --git a/storage/innobase/handler/ha_innodb.cc b/storage/innobase/handler/ha_innodb.cc
index b508d0246e7..64ecc7ae957 100644
--- a/storage/innobase/handler/ha_innodb.cc
+++ b/storage/innobase/handler/ha_innodb.cc
@@ -5845,7 +5845,7 @@ initialize_auto_increment(dict_table_t* table, const Field* field)
 		table->persistent_autoinc without
 		autoinc_mutex protection, and there might be multiple
 		ha_innobase::open() executing concurrently. */
-	} else if (srv_force_recovery > SRV_FORCE_NO_IBUF_MERGE) {
+	} else if (srv_force_recovery >= SRV_FORCE_NO_UNDO_LOG_SCAN) {
 		/* If the recovery level is set so high that writes
 		are disabled we force the AUTOINC counter to 0
 		value effectively disabling writes to the table.
@@ -14872,7 +14872,7 @@ ha_innobase::info_low(
 		}
 	}
 
-	if (srv_force_recovery > SRV_FORCE_NO_IBUF_MERGE) {
+	if (srv_force_recovery >= SRV_FORCE_NO_UNDO_LOG_SCAN) {
 
 		goto func_exit;
 
diff --git a/storage/innobase/include/srv0srv.h b/storage/innobase/include/srv0srv.h
index 0c32d5d686a..be25bc86519 100644
--- a/storage/innobase/include/srv0srv.h
+++ b/storage/innobase/include/srv0srv.h
@@ -562,11 +562,9 @@ enum {
 	SRV_FORCE_NO_BACKGROUND	= 2,	/*!< prevent the main thread from
 					running: if a crash would occur
 					in purge, this prevents it */
-	SRV_FORCE_NO_TRX_UNDO = 3,	/*!< do not run trx rollback after
+	SRV_FORCE_NO_TRX_UNDO = 3,	/*!< do not run DML rollback after
 					recovery */
-	SRV_FORCE_NO_IBUF_MERGE = 4,	/*!< prevent also ibuf operations:
-					if they would cause a crash, better
-					not do them */
+	SRV_FORCE_NO_DDL_UNDO = 4,	/*!< prevent also DDL rollback */
 	SRV_FORCE_NO_UNDO_LOG_SCAN = 5,	/*!< do not look at undo logs when
 					starting the database: InnoDB will
 					treat even incomplete transactions
diff --git a/storage/innobase/srv/srv0start.cc b/storage/innobase/srv/srv0start.cc
index 3f0ccf44b1e..1fc5905ce26 100644
--- a/storage/innobase/srv/srv0start.cc
+++ b/storage/innobase/srv/srv0start.cc
@@ -1065,7 +1065,7 @@ dberr_t srv_start(bool create_new_db)
 	}
 
 	high_level_read_only = srv_read_only_mode
-		|| srv_force_recovery > SRV_FORCE_NO_IBUF_MERGE
+		|| srv_force_recovery >= SRV_FORCE_NO_UNDO_LOG_SCAN
 		|| srv_sys_space.created_new_raw();
 
 	srv_started_redo = false;
@@ -1704,7 +1704,7 @@ file_checked:
 
 	if (!create_new_db) {
 		ut_ad(high_level_read_only
-		      || srv_force_recovery <= SRV_FORCE_NO_IBUF_MERGE);
+		      || srv_force_recovery < SRV_FORCE_NO_UNDO_LOG_SCAN);
 
 		/* Validate a few system page types that were left
 		uninitialized before MySQL or MariaDB 5.5. */
@@ -1745,7 +1745,7 @@ file_checked:
 		should guarantee that there is at most one data
 		dictionary transaction active at a time. */
 		if (!high_level_read_only
-		    && srv_force_recovery < SRV_FORCE_NO_TRX_UNDO) {
+		    && srv_force_recovery <= SRV_FORCE_NO_TRX_UNDO) {
 			/* If the following call is ever removed, the
 			first-time ha_innobase::open() must hold (or
 			acquire and release) a table lock that
@@ -1759,7 +1759,7 @@ file_checked:
 			trx_rollback_recovered(false);
 		}
 
-		if (srv_force_recovery <= SRV_FORCE_NO_IBUF_MERGE) {
+		if (srv_force_recovery < SRV_FORCE_NO_UNDO_LOG_SCAN) {
 			/* The following call is necessary for the insert
 			buffer to work with multiple tablespaces. We must
 			know the mapping between space id's and .ibd file
diff --git a/storage/innobase/trx/trx0roll.cc b/storage/innobase/trx/trx0roll.cc
index a0582413d07..34e623b6b1a 100644
--- a/storage/innobase/trx/trx0roll.cc
+++ b/storage/innobase/trx/trx0roll.cc
@@ -709,7 +709,8 @@ void trx_rollback_recovered(bool all)
 {
   std::vector<trx_t*> trx_list;
 
-  ut_a(srv_force_recovery < SRV_FORCE_NO_TRX_UNDO);
+  ut_a(srv_force_recovery <
+       (all ? SRV_FORCE_NO_TRX_UNDO : SRV_FORCE_NO_DDL_UNDO));
 
   /*
     Collect list of recovered ACTIVE transaction ids first. Once collected, no