Merge 10.10 into 10.11

.github/CODEOWNERS

@@ -1,2 +0,0 @@
-/debian @ottok
-

.gitlab-ci.yml

@@ -42,7 +42,7 @@ variables:
   CMAKE_FLAGS: "-DWITH_SSL=system -DPLUGIN_COLUMNSTORE=NO -DPLUGIN_ROCKSDB=NO -DPLUGIN_S3=NO -DPLUGIN_MROONGA=NO -DPLUGIN_CONNECT=NO -DPLUGIN_MROONGA=NO -DPLUGIN_TOKUDB=NO -DPLUGIN_PERFSCHEMA=NO -DWITH_WSREP=OFF"
   # Major version dictates which branches share the same ccache. E.g. 10.6-abc
   # and 10.6-xyz will have the same cache.
-  MARIADB_MAJOR_VERSION: "10.8"
+  MARIADB_MAJOR_VERSION: "10.9"
   # NOTE! Currently ccache is only used on the Centos8 build. As each job has
   # sufficiently different environments they are unable to benefit from each
   # other's ccaches. As each build generates about 1 GB of ccache, having

cmake/cpack_rpm.cmake

@@ -297,7 +297,7 @@ ELSEIF(RPM MATCHES "sles")
 ENDIF()
 
 # MDEV-24629, we need it outside of ELSIFs
-IF(RPM MATCHES "fedora3[234]")
+IF(RPM MATCHES "fedora")
   ALTERNATIVE_NAME("common" "mariadb-connector-c-config" ${MARIADB_CONNECTOR_C_VERSION}-1)
 ENDIF()
 

debian/additions/innotop/innotop

@@ -20,6 +20,9 @@
 # Street, Fifth Floor, Boston, MA 02110-1335 USA
 
 use strict;
+use warnings;
+use utf8;
+use feature ':5.16';
 use warnings FATAL => 'all';
 
 our $VERSION = '1.11.4';
@@ -265,7 +268,7 @@ sub get_dbh {
             $dbh->do($sql);
             MKDEBUG && _d('Enabling charset for STDOUT');
             if ( $charset eq 'utf8' ) {
-               binmode(STDOUT, ':utf8')
+               binmode(STDOUT, ':encoding(UTF-8)')
                   or die "Can't binmode(STDOUT, ':utf8'): $OS_ERROR";
             }
             else {
@@ -612,6 +615,9 @@ sub ts_to_string {
 
 sub parse_innodb_timestamp {
    my $text = shift;
+   if ( ! defined $text ) {
+        return (0, 0, 0, 0, 0, 0);
+   }
    my ( $y, $m, $d, $h, $i, $s )
       = $text =~ m/^(\d\d)(\d\d)(\d\d) +(\d+):(\d+):(\d+)$/;
    die("Can't get timestamp from $text\n") unless $y;
@@ -803,7 +809,8 @@ sub parse_fk_transaction_error {
 # TODO: write some tests for this
 sub parse_innodb_record_dump {
    my ( $dump, $complete, $debug ) = @_;
-   return undef unless $dump;
+   # Use bare return as recommend in page 199 of PBP
+   return unless $dump;
 
    my $result = {};
 
@@ -6769,6 +6776,9 @@ sub set_precision {
    my ( $num, $precision ) = @_;
    $num = 0 unless defined $num;
    $precision = $config{num_digits}->{val} if !defined $precision;
+   if ( $num eq "" ) {
+       $num = int(0);
+   }
    sprintf("%.${precision}f", $num);
 }
 
@@ -6777,6 +6787,9 @@ sub set_precision {
 sub percent {
    my ( $num ) = @_;
    $num = 0 unless defined $num;
+   if ( $num eq "" ) {
+       $num = int(0);
+   }
    my $digits = $config{num_digits}->{val};
    return sprintf("%.${digits}f", $num * 100)
       . ($config{show_percent}->{val} ? '%' : '');
@@ -6841,7 +6854,7 @@ sub make_color_func {
       push @criteria,
          "( defined \$set->{$spec->{col}} && \$set->{$spec->{col}} $spec->{op} $val ) { return '$spec->{color}'; }";
    }
-   return undef unless @criteria;
+   return unless @criteria;
    my $sub = eval 'sub { my ( $set ) = @_; if ' . join(" elsif ", @criteria) . '}';
    die if $EVAL_ERROR;
    return $sub;
@@ -7521,10 +7534,10 @@ sub choose_connections {
 sub do_stmt {
    my ( $cxn, $stmt_name, @args ) = @_;
 
-   return undef if $file;
+   return if $file;
 
    # Test if the cxn should not even be tried
-   return undef if $dbhs{$cxn}
+   return if $dbhs{$cxn}
       && $dbhs{$cxn}->{failed}
       && ( !$dbhs{$cxn}->{dbh} || !$dbhs{$cxn}->{dbh}->{Active} || $dbhs{$cxn}->{mode} eq $config{mode}->{val} );
 
@@ -7596,10 +7609,10 @@ sub handle_cxn_error {
 sub do_query {
    my ( $cxn, $query ) = @_;
 
-   return undef if $file;
+   return if $file;
 
    # Test if the cxn should not even be tried
-   return undef if $dbhs{$cxn}
+   return if $dbhs{$cxn}
       && $dbhs{$cxn}->{failed}
       && ( !$dbhs{$cxn}->{dbh} || !$dbhs{$cxn}->{dbh}->{Active} || $dbhs{$cxn}->{mode} eq $config{mode}->{val} );
 
@@ -7781,7 +7794,7 @@ sub compile_select_stmt {
 sub compile_filter {
    my ( $text ) = @_;
    my ( $sub, $err );
-   eval "\$sub = sub { my \$set = shift; $text }";
+   eval { $sub = sub { my $set = shift; $text } };
    if ( $EVAL_ERROR ) {
       $EVAL_ERROR =~ s/at \(eval.*$//;
       $sub = sub { return $EVAL_ERROR };
@@ -8013,7 +8026,7 @@ sub load_config_plugins {
 
    # First, find a list of all plugins that exist on disk, and get information about them.
    my $dir = $config{plugin_dir}->{val};
-   foreach my $p_file ( <$dir/*.pm> ) {
+   foreach my $p_file (glob($dir."/*.pm")) {
       my ($package, $desc);
       eval {
          open my $p_in, "<", $p_file or die $OS_ERROR;
@@ -9192,7 +9205,7 @@ sub switch_var_set {
 # edit_stmt_sleep_times {{{3
 sub edit_stmt_sleep_times {
    $clear_screen_sub->();
-   my $stmt = prompt_list('Specify a statement', '', sub { return sort keys %stmt_maker_for });
+   my $stmt = prompt_list('Specify a statement', '', sub { my @tmparray = sort keys %stmt_maker_for; return @tmparray });
    return unless $stmt && exists $stmt_maker_for{$stmt};
    $clear_screen_sub->();
    my $curr_val = $stmt_sleep_time_for{$stmt} || 0;
@@ -9843,7 +9856,7 @@ sub get_slave_status {
 sub is_func {
    my ( $word ) = @_;
    return defined(&$word)
-      || eval "my \$x= sub { $word  }; 1"
+      || eval { my $x = sub { $word  }; 1 }
       || $EVAL_ERROR !~ m/^Bareword/;
 }
 

debian/mariadb-server.mariadb.init

@@ -86,7 +86,7 @@ sanity_checks() {
   datadir=`mariadbd_get_param datadir`
   # As preset blocksize of GNU df is 1024 then available bytes is $df_available_blocks * 1024
   # 4096 blocks is then lower than 4 MB
-  df_available_blocks=`LC_ALL=C BLOCKSIZE= df --output=avail "$datadir" | tail -n 1`
+  df_available_blocks="$(LC_ALL=C BLOCKSIZE='' df --output=avail "$datadir" | tail -n 1)"
   if [ "$df_available_blocks" -lt "4096" ]; then
     log_failure_msg "$0: ERROR: The partition with $datadir is too full!"
     echo                "ERROR: The partition with $datadir is too full!" | $ERR_LOGGER

debian/mariadb-server.preinst

@@ -223,15 +223,24 @@ then
   mkdir -Z $mysql_datadir
 fi
 
+# Check if MariaDB datadir is available if not fails.
+# There should be symlink or directory available or something will fail.
+if [ -d "$mysql_datadir" ] || [ -L "$mysql_datadir" ]
+then
     # As preset blocksize of GNU df is 1024 then available bytes is $df_available_blocks * 1024
     # 4096 blocks is then lower than 4 MB
-df_available_blocks=`LC_ALL=C BLOCKSIZE= df --output=avail "$datadir" | tail -n 1`
+    df_available_blocks="$(LC_ALL=C BLOCKSIZE='' df --output=avail "$mysql_datadir" | tail -n 1)"
     if [ "$df_available_blocks" -lt "4096" ]
     then
       echo "ERROR: There's not enough space in $mysql_datadir/" 1>&2
       db_stop
       exit 1
     fi
+else
+    echo "ERROR: There's no directory or symlink available: $mysql_datadir/" 1>&2
+    db_stop
+    exit 1    
+fi
 
 # Since the home directory was created before putting the user into
 # the mysql group and moreover we cannot guarantee that the

extra/mariabackup/xtrabackup.cc

@@ -847,26 +847,48 @@ void mdl_lock_all()
 
 
 // Convert non-null terminated filename to space name
+// Note that in 10.6 the filename may be an undo file name
 static std::string filename_to_spacename(const void *filename, size_t len)
 {
-	// null- terminate filename
+  char f[FN_REFLEN];
-	char *f = (char *)malloc(len + 1);
+  char *p= 0, *table, *db;
-	ut_a(f);
+  DBUG_ASSERT(len < FN_REFLEN);
-	memcpy(f, filename, len);
+
-	f[len] = 0;
+  strmake(f, (const char*) filename, len);
+
+#ifdef _WIN32
   for (size_t i = 0; i < len; i++)
+  {
     if (f[i] == '\\')
       f[i] = '/';
-	char *p = strrchr(f, '.');
+  }
-	ut_a(p);
+#endif
+
+  /* Remove extension, if exists */
+  if (!(p= strrchr(f, '.')))
+    goto err;
   *p= 0;
-	char *table = strrchr(f, '/');
+
-	ut_a(table);
+  /* Find table name */
+  if (!(table= strrchr(f, '/')))
+    goto err;
   *table = 0;
-	char *db = strrchr(f, '/');
+
+  /* Find database name */
+  db= strrchr(f, '/');
   *table = '/';
-	std::string s(db ? db+1 : f);
+  if (!db)
-	free(f);
+    goto err;
+  {
+    std::string s(db+1);
+    return s;
+  }
+
+err:
+  /* Not a database/table. Return original (converted) name */
+  if (p)
+    *p= '.';                                    // Restore removed extension
+  std::string s(f);
   return s;
 }
 
@@ -3146,7 +3168,7 @@ static bool xtrabackup_copy_logfile()
       if (log_sys.buf[recv_sys.offset] <= 1)
         break;
 
-      if (recv_sys.parse_mtr(STORE_NO) == recv_sys_t::OK)
+      if (recv_sys.parse_mtr<false>(false) == recv_sys_t::OK)
       {
         do
         {
@@ -3156,7 +3178,7 @@ static bool xtrabackup_copy_logfile()
                                                  sequence_offset));
           *seq= 1;
         }
-        while ((r= recv_sys.parse_mtr(STORE_NO)) == recv_sys_t::OK);
+        while ((r= recv_sys.parse_mtr<false>(false)) == recv_sys_t::OK);
 
         if (ds_write(dst_log_file, log_sys.buf + start_offset,
                      recv_sys.offset - start_offset))

include/m_string.h

@@ -249,14 +249,15 @@ static inline void lex_string_set3(LEX_CSTRING *lex_str, const char *c_str,
 */
 static inline int safe_strcpy(char *dst, size_t dst_size, const char *src)
 {
-  memset(dst, '\0', dst_size);
+  DBUG_ASSERT(dst_size > 0);
-  strncpy(dst, src, dst_size - 1);
+  /* Note, strncpy will zerofill end of dst if src shorter than dst_size */
-  /*
+  strncpy(dst, src, dst_size);
-     If the first condition is true, we are guaranteed to have src length
+  if (dst[dst_size-1])
-     >= (dst_size - 1), hence safe to access src[dst_size - 1].
+  {
-  */
+    /* Ensure string is zero terminated */
-  if (dst[dst_size - 2] != '\0' && src[dst_size - 1] != '\0')
+    dst[dst_size-1]= 0;
-    return 1; /* Truncation of src. */
+    return 1;
+  }
   return 0;
 }
 

include/myisammrg.h

@@ -71,6 +71,7 @@ typedef struct st_myrg_info
   ulong  cache_size;
   uint	 merge_insert_method;
   uint	 tables,options,reclength,keys;
+  uint   key_parts;
   my_bool cache_in_use;
   /* If MERGE children attached to parent. See top comment in ha_myisammrg.cc */
   my_bool children_attached;

include/mysql/service_wsrep.h

@@ -57,6 +57,7 @@ extern struct wsrep_service_st {
   my_bool                     (*wsrep_on_func)(const MYSQL_THD thd);
   bool                        (*wsrep_prepare_key_for_innodb_func)(MYSQL_THD thd, const unsigned char*, size_t, const unsigned char*, size_t, struct wsrep_buf*, size_t*);
   void                        (*wsrep_thd_LOCK_func)(const MYSQL_THD thd);
+  int                         (*wsrep_thd_TRYLOCK_func)(const MYSQL_THD thd);
   void                        (*wsrep_thd_UNLOCK_func)(const MYSQL_THD thd);
   const char *                (*wsrep_thd_query_func)(const MYSQL_THD thd);
   int                         (*wsrep_thd_retry_counter_func)(const MYSQL_THD thd);
@@ -89,7 +90,6 @@ extern struct wsrep_service_st {
   ulong                       (*wsrep_OSU_method_get_func)(const MYSQL_THD thd);
   my_bool                     (*wsrep_thd_has_ignored_error_func)(const MYSQL_THD thd);
   void                        (*wsrep_thd_set_ignored_error_func)(MYSQL_THD thd, my_bool val);
-  bool                        (*wsrep_thd_set_wsrep_aborter_func)(MYSQL_THD bf_thd, MYSQL_THD thd);
   void                        (*wsrep_report_bf_lock_wait_func)(const MYSQL_THD thd,
                                                                 unsigned long long trx_id);
   void                        (*wsrep_thd_kill_LOCK_func)(const MYSQL_THD thd);
@@ -111,6 +111,7 @@ extern struct wsrep_service_st {
 #define wsrep_on(thd) (thd) && WSREP_ON && wsrep_service->wsrep_on_func(thd)
 #define wsrep_prepare_key_for_innodb(A,B,C,D,E,F,G) wsrep_service->wsrep_prepare_key_for_innodb_func(A,B,C,D,E,F,G)
 #define wsrep_thd_LOCK(T) wsrep_service->wsrep_thd_LOCK_func(T)
+#define wsrep_thd_TRYLOCK(T) wsrep_service->wsrep_thd_TRYLOCK_func(T)
 #define wsrep_thd_UNLOCK(T) wsrep_service->wsrep_thd_UNLOCK_func(T)
 #define wsrep_thd_kill_LOCK(T) wsrep_service->wsrep_thd_kill_LOCK_func(T)
 #define wsrep_thd_kill_UNLOCK(T) wsrep_service->wsrep_thd_kill_UNLOCK_func(T)
@@ -141,7 +142,6 @@ extern struct wsrep_service_st {
 #define wsrep_OSU_method_get(T) wsrep_service->wsrep_OSU_method_get_func(T)
 #define wsrep_thd_has_ignored_error(T) wsrep_service->wsrep_thd_has_ignored_error_func(T)
 #define wsrep_thd_set_ignored_error(T,V) wsrep_service->wsrep_thd_set_ignored_error_func(T,V)
-#define wsrep_thd_set_wsrep_aborter(T) wsrep_service->wsrep_thd_set_wsrep_aborter_func(T1, T2)
 #define wsrep_report_bf_lock_wait(T,I) wsrep_service->wsrep_report_bf_lock_wait(T,I)
 #define wsrep_thd_set_PA_unsafe(T) wsrep_service->wsrep_thd_set_PA_unsafe_func(T)
 #else
@@ -175,6 +175,8 @@ void wsrep_set_data_home_dir(const char *data_dir);
 extern "C" my_bool wsrep_on(const MYSQL_THD thd);
 /* Lock thd wsrep lock */
 extern "C" void wsrep_thd_LOCK(const MYSQL_THD thd);
+/* Try thd wsrep lock. Return non-zero if lock could not be taken. */
+extern "C" int wsrep_thd_TRYLOCK(const MYSQL_THD thd);
 /* Unlock thd wsrep lock */
 extern "C" void wsrep_thd_UNLOCK(const MYSQL_THD thd);
 
@@ -197,8 +199,6 @@ extern "C" my_bool wsrep_thd_is_local(const MYSQL_THD thd);
 /* Return true if thd is in high priority mode */
 /* todo: rename to is_high_priority() */
 extern "C" my_bool wsrep_thd_is_applying(const MYSQL_THD thd);
-/* set wsrep_aborter for the target THD */
-extern "C" bool wsrep_thd_set_wsrep_aborter(MYSQL_THD bf_thd, MYSQL_THD victim_thd);
 /* Return true if thd is in TOI mode */
 extern "C" my_bool wsrep_thd_is_toi(const MYSQL_THD thd);
 /* Return true if thd is in replicating TOI mode */
@@ -249,7 +249,6 @@ extern "C" my_bool wsrep_thd_is_applying(const MYSQL_THD thd);
 extern "C" ulong wsrep_OSU_method_get(const MYSQL_THD thd);
 extern "C" my_bool wsrep_thd_has_ignored_error(const MYSQL_THD thd);
 extern "C" void wsrep_thd_set_ignored_error(MYSQL_THD thd, my_bool val);
-extern "C" bool wsrep_thd_set_wsrep_aborter(MYSQL_THD bf_thd, MYSQL_THD victim_thd);
 extern "C" void wsrep_report_bf_lock_wait(const THD *thd,
                                           unsigned long long trx_id);
 /* declare parallel applying unsafety for the THD */

mysql-test/lib/My/File/Path.pm

@@ -34,7 +34,7 @@ use strict;
 
 use Exporter;
 use base "Exporter";
-our @EXPORT= qw /rmtree mkpath copytree/;
+our @EXPORT= qw /rmtree mkpath copytree make_readonly/;
 
 use File::Find;
 use File::Copy;
@@ -184,6 +184,10 @@ sub copytree {
     # Only copy plain files
     next unless -f "$from_dir/$_";
     copy("$from_dir/$_", "$to_dir/$_");
+    if (!$use_umask)
+    {
+      chmod(0666, "$to_dir/$_");
+    }
   }
   closedir(DIR);
 
@@ -193,4 +197,29 @@ sub copytree {
   }
 }
 
+
+sub make_readonly {
+  my ($dir) = @_;
+
+  die "Usage: make_readonly(<dir>])"
+    unless @_ == 1;
+
+  opendir(DIR, "$dir")
+    or croak("Can't find $dir$!");
+  for(readdir(DIR)) {
+
+    next if "$_" eq "." or "$_" eq "..";
+
+    if ( -d "$dir/$_" )
+    {
+      make_readonly("$dir/$_");
+      next;
+    }
+
+    # Only copy plain files
+    next unless -f "$dir/$_";
+    chmod 0444, "$dir/$_";
+  }
+  closedir(DIR);
+}
 1;

mysql-test/lib/My/SafeProcess/Base.pm

@@ -40,7 +40,7 @@ our @EXPORT= qw(create_process);
 #   Retry a couple of times if fork returns EAGAIN
 #
 sub _safe_fork {
-  my $retries= 5;
+  my $retries= 100;
   my $pid;
 
  FORK:

mysql-test/main/distinct.result

@@ -1157,3 +1157,28 @@ explain select * from t1 limit 0 offset 10;
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	SIMPLE	NULL	NULL	NULL	NULL	NULL	NULL	NULL	Zero limit
 drop table t1, t2;
+#
+# MDEV-28285 Unexpected result when combining DISTINCT, subselect
+# and LIMIT
+#
+create table t1 (a int primary key);
+create table t2 (a int primary key, b int not null);
+insert into t1 select seq from seq_1_to_10;
+insert into t2 select seq,seq from seq_1_to_10;
+select distinct a from t1 where t1.a=1 and t1.a in (select a from t2 where t2.b in (1,2));
+a
+1
+explain select distinct a from t1 where t1.a=1 and t1.a in (select a+0 from t2 where t2.b in (1,2)) limit 10,10;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t1	const	PRIMARY	PRIMARY	4	const	1	Using index; Using temporary
+1	PRIMARY	<subquery2>	eq_ref	distinct_key	distinct_key	8	func	1	
+2	MATERIALIZED	t2	ALL	NULL	NULL	NULL	NULL	10	Using where
+select distinct a from t1 where t1.a=1 and t1.a in (select a+0 from t2 where t2.b in (1,2)) limit 10,10;
+a
+select distinct a from t1 where t1.a=1 and t1.a in (select a+0 from t2 where t2.b in (1,2)) limit 0,1;
+a
+1
+drop table t1,t2;
+#
+# end of 10.5 tests
+#

mysql-test/main/distinct.test

@@ -892,3 +892,24 @@ explain select * from t1 limit 0;
 explain select * from t1 limit 0 offset 10;
 
 drop table t1, t2;
+
+--echo #
+--echo # MDEV-28285 Unexpected result when combining DISTINCT, subselect
+--echo # and LIMIT
+--echo #
+
+create table t1 (a int primary key);
+create table t2 (a int primary key, b int not null);
+
+insert into t1 select seq from seq_1_to_10;
+insert into t2 select seq,seq from seq_1_to_10;
+
+select distinct a from t1 where t1.a=1 and t1.a in (select a from t2 where t2.b in (1,2));
+explain select distinct a from t1 where t1.a=1 and t1.a in (select a+0 from t2 where t2.b in (1,2)) limit 10,10;
+select distinct a from t1 where t1.a=1 and t1.a in (select a+0 from t2 where t2.b in (1,2)) limit 10,10;
+select distinct a from t1 where t1.a=1 and t1.a in (select a+0 from t2 where t2.b in (1,2)) limit 0,1;
+drop table t1,t2;
+
+--echo #
+--echo # end of 10.5 tests
+--echo #

mysql-test/main/func_json.result

@@ -2578,5 +2578,29 @@ SELECT JSON_EXTRACT('{ "my-key": 1 }', '$.my-key');
 JSON_EXTRACT('{ "my-key": 1 }', '$.my-key')
 1
 #
+# MDEV-23187: Assorted assertion failures in json_find_path with certain collations
+#
+SET @save_collation_connection= @@collation_connection;
+SET @json='{ "A": [ [{"k":"v"},[1]],true],"B": {"C": 1} }';
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+JSON_VALUE(@json,'$.A[last-1][last-1].key1')
+NULL
+SET @json='{ "A": [ [{"k":"v"},[1]],true],"B": {"C": 1} }';
+SET collation_connection='ucs2_bin';
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+JSON_VALUE(@json,'$.A[last-1][last-1].key1')
+NULL
+SET @json='{ "A": [ [{"k":"v"},[15]],true],"B": {"C": 1} }';
+SET sql_mode=0,character_set_connection=utf32;
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+JSON_VALUE(@json,'$.A[last-1][last-1].key1')
+NULL
+SET @json='{ "A": [ [{"k":"v"},[15]],true],"B": {"C": 1} }';
+SET sql_mode=0,character_set_connection=utf32;
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+JSON_VALUE(@json,'$.A[last-1][last-1].key1')
+NULL
+SET @@collation_connection= @save_collation_connection;
+#
 # End of 10.9 Test
 #

mysql-test/main/func_json.test

@@ -1772,6 +1772,32 @@ DROP TABLE t1;
 SELECT JSON_EXTRACT('{ "my-key": 1 }', '$."my-key"');
 SELECT JSON_EXTRACT('{ "my-key": 1 }', '$.my-key');
 
+--echo #
+--echo # MDEV-23187: Assorted assertion failures in json_find_path with certain collations
+--echo #
+
+
+SET @save_collation_connection= @@collation_connection;
+
+SET @json='{ "A": [ [{"k":"v"},[1]],true],"B": {"C": 1} }';
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+
+SET @json='{ "A": [ [{"k":"v"},[1]],true],"B": {"C": 1} }';
+SET collation_connection='ucs2_bin';
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+
+SET @json='{ "A": [ [{"k":"v"},[15]],true],"B": {"C": 1} }';
+SET sql_mode=0,character_set_connection=utf32;
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+
+
+SET @json='{ "A": [ [{"k":"v"},[15]],true],"B": {"C": 1} }';
+SET sql_mode=0,character_set_connection=utf32;
+SELECT JSON_VALUE(@json,'$.A[last-1][last-1].key1');
+
+SET @@collation_connection= @save_collation_connection;
+
+
 --echo #
 --echo # End of 10.9 Test
 --echo #

mysql-test/main/group_min_max.result

@@ -4095,6 +4095,116 @@ MIN(pk)	a
 5	10
 DROP TABLE t1;
 #
+# MDEV-6768 Wrong result with agregate with join with no resultset
+#
+create table t1
+(
+PARENT_ID int(10) unsigned NOT NULL AUTO_INCREMENT,
+PARENT_FIELD VARCHAR(10),
+PRIMARY KEY (PARENT_ID)
+) engine=innodb;
+create table t2
+(
+CHILD_ID INT NOT NULL AUTO_INCREMENT,
+PARENT_ID INT NOT NULL,
+CHILD_FIELD varchar(10),
+PRIMARY KEY (CHILD_ID)
+)engine=innodb;
+INSERT INTO t1 (PARENT_FIELD)
+SELECT 'AAAA';
+INSERT INTO t2 (PARENT_ID, CHILD_FIELD)
+SELECT 1, 'BBBB';
+explain select
+t1.PARENT_ID,
+min(CHILD_FIELD)
+from t1 straight_join t2
+where t1.PARENT_ID = 1
+and t1.PARENT_ID = t2.PARENT_ID
+and t2.CHILD_FIELD = "ZZZZ";
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	SIMPLE	t1	const	PRIMARY	PRIMARY	4	const	1	Using index
+1	SIMPLE	t2	ALL	NULL	NULL	NULL	NULL	1	Using where
+select
+t1.PARENT_ID,
+min(CHILD_FIELD)
+from t1 straight_join t2
+where t1.PARENT_ID = 1
+and t1.PARENT_ID = t2.PARENT_ID
+and t2.CHILD_FIELD = "ZZZZ";
+PARENT_ID	min(CHILD_FIELD)
+NULL	NULL
+select
+1,
+min(CHILD_FIELD)
+from t1 straight_join t2
+where t1.PARENT_ID = 1
+and t1.PARENT_ID = t2.PARENT_ID
+and t2.CHILD_FIELD = "ZZZZ";
+1	min(CHILD_FIELD)
+1	NULL
+select
+IFNULL(t1.PARENT_ID,1),
+min(CHILD_FIELD)
+from t1 straight_join t2
+where t1.PARENT_ID = 1
+and t1.PARENT_ID = t2.PARENT_ID
+and t2.CHILD_FIELD = "ZZZZ";
+IFNULL(t1.PARENT_ID,1)	min(CHILD_FIELD)
+1	NULL
+# Check that things works with MyISAM (which has different explain)
+alter table t1 engine=myisam;
+alter table t2 engine=myisam;
+explain select
+t1.PARENT_ID,
+min(CHILD_FIELD)
+from t1 straight_join t2
+where t1.PARENT_ID = 1
+and t1.PARENT_ID = t2.PARENT_ID
+and t2.CHILD_FIELD = "ZZZZ";
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	SIMPLE	NULL	NULL	NULL	NULL	NULL	NULL	NULL	Impossible WHERE noticed after reading const tables
+select
+t1.PARENT_ID,
+min(CHILD_FIELD)
+from t1 straight_join t2
+where t1.PARENT_ID = 1
+and t1.PARENT_ID = t2.PARENT_ID
+and t2.CHILD_FIELD = "ZZZZ";
+PARENT_ID	min(CHILD_FIELD)
+NULL	NULL
+drop table t1,t2;
+# Check that things works if sub queries are re-executed
+create table t1 (a int primary key, b int);
+create table t2 (a int primary key, b int);
+create table t3 (a int primary key, b int);
+insert into t1 values (1,1),(2,2),(3,3);
+insert into t2 values (1,1),(2,2),(3,3);
+insert into t3 values (1,1),(3,3);
+explain
+select *,
+(select
+CONCAT('t2:',       IFNULL(t2.a,      't2a-null'), ';',
+'min_t3_b:', IFNULL(min(t3.b), 't3b-null'))
+from t2,t3
+where t2.a=1 and t1.b = t3.a) as s1
+from t1;
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	3	
+2	DEPENDENT SUBQUERY	t2	const	PRIMARY	PRIMARY	4	const	1	Using index
+2	DEPENDENT SUBQUERY	t3	eq_ref	PRIMARY	PRIMARY	4	test.t1.b	1	
+select *,
+(select
+CONCAT('t2:',       IFNULL(t2.a,      't2a-null'), ';',
+'min_t3_b:', IFNULL(min(t3.b), 't3b-null'))
+from t2,t3
+where t2.a=1 and t1.b = t3.a) as s1
+from t1;
+a	b	s1
+1	1	t2:1;min_t3_b:1
+2	2	t2:t2a-null;min_t3_b:t3b-null
+3	3	t2:1;min_t3_b:3
+drop table t1,t2,t3;
+#
 # End of 10.5 tests
 #
 #

mysql-test/main/group_min_max.test

@@ -1749,6 +1749,116 @@ SELECT MIN(pk), a FROM t1 WHERE pk <> 1 GROUP BY a;
 
 DROP TABLE t1;
 
+--echo #
+--echo # MDEV-6768 Wrong result with agregate with join with no resultset
+--echo #
+
+create table t1
+(
+	PARENT_ID int(10) unsigned NOT NULL AUTO_INCREMENT,
+	PARENT_FIELD VARCHAR(10),
+	PRIMARY KEY (PARENT_ID)
+) engine=innodb;
+
+create table t2
+(
+	CHILD_ID INT NOT NULL AUTO_INCREMENT,
+	PARENT_ID INT NOT NULL,
+	CHILD_FIELD varchar(10),
+	PRIMARY KEY (CHILD_ID)
+)engine=innodb;
+
+INSERT INTO t1 (PARENT_FIELD)
+SELECT 'AAAA';
+
+INSERT INTO t2 (PARENT_ID, CHILD_FIELD)
+SELECT 1, 'BBBB';
+
+explain select
+ t1.PARENT_ID,
+ min(CHILD_FIELD)
+ from t1 straight_join t2
+ where t1.PARENT_ID = 1
+ and t1.PARENT_ID = t2.PARENT_ID
+ and t2.CHILD_FIELD = "ZZZZ";
+
+select
+ t1.PARENT_ID,
+ min(CHILD_FIELD)
+ from t1 straight_join t2
+ where t1.PARENT_ID = 1
+ and t1.PARENT_ID = t2.PARENT_ID
+ and t2.CHILD_FIELD = "ZZZZ";
+
+select
+ 1,
+ min(CHILD_FIELD)
+ from t1 straight_join t2
+ where t1.PARENT_ID = 1
+ and t1.PARENT_ID = t2.PARENT_ID
+ and t2.CHILD_FIELD = "ZZZZ";
+
+select
+ IFNULL(t1.PARENT_ID,1),
+ min(CHILD_FIELD)
+ from t1 straight_join t2
+ where t1.PARENT_ID = 1
+ and t1.PARENT_ID = t2.PARENT_ID
+ and t2.CHILD_FIELD = "ZZZZ";
+
+
+--echo # Check that things works with MyISAM (which has different explain)
+
+alter table t1 engine=myisam;
+alter table t2 engine=myisam;
+
+explain select
+ t1.PARENT_ID,
+ min(CHILD_FIELD)
+ from t1 straight_join t2
+ where t1.PARENT_ID = 1
+ and t1.PARENT_ID = t2.PARENT_ID
+ and t2.CHILD_FIELD = "ZZZZ";
+
+select
+ t1.PARENT_ID,
+ min(CHILD_FIELD)
+ from t1 straight_join t2
+ where t1.PARENT_ID = 1
+ and t1.PARENT_ID = t2.PARENT_ID
+ and t2.CHILD_FIELD = "ZZZZ";
+
+drop table t1,t2;
+
+--echo # Check that things works if sub queries are re-executed
+
+create table t1 (a int primary key, b int);
+create table t2 (a int primary key, b int);
+create table t3 (a int primary key, b int);
+
+insert into t1 values (1,1),(2,2),(3,3);
+insert into t2 values (1,1),(2,2),(3,3);
+insert into t3 values (1,1),(3,3);
+
+explain
+select *,
+   (select
+        CONCAT('t2:',       IFNULL(t2.a,      't2a-null'), ';',
+               'min_t3_b:', IFNULL(min(t3.b), 't3b-null'))
+    from t2,t3
+    where t2.a=1 and t1.b = t3.a) as s1
+from t1;
+
+select *,
+   (select
+        CONCAT('t2:',       IFNULL(t2.a,      't2a-null'), ';',
+               'min_t3_b:', IFNULL(min(t3.b), 't3b-null'))
+    from t2,t3
+    where t2.a=1 and t1.b = t3.a) as s1
+from t1;
+
+drop table t1,t2,t3;
+
 --echo #
 --echo # End of 10.5 tests
 --echo #

mysql-test/main/merge.result

@@ -3902,6 +3902,18 @@ DROP TABLE t1;
 DROP TABLE m1;
 set global default_storage_engine=@save_default_storage_engine;
 #
+# MDEV-31083 ASAN use-after-poison in myrg_attach_children
+#
+CREATE TABLE t1 (f TEXT, FULLTEXT (f)) ENGINE=MyISAM;
+INSERT INTO t1 VALUES ('foo'),('bar');
+CREATE TABLE mrg (f TEXT) ENGINE=MERGE, UNION(t1);
+SELECT * FROM mrg;
+f
+foo
+bar
+DROP TABLE mrg, t1;
+End of 10.5 tests
+#
 # End of 10.0 tests
 #
 #

mysql-test/main/merge.test

@@ -2859,6 +2859,18 @@ set global default_storage_engine=@save_default_storage_engine;
 # gone so execution of other tests won't be affected by their presence.
 --source include/wait_until_count_sessions.inc
 
+--echo #
+--echo # MDEV-31083 ASAN use-after-poison in myrg_attach_children
+--echo #
+
+CREATE TABLE t1 (f TEXT, FULLTEXT (f)) ENGINE=MyISAM;
+INSERT INTO t1 VALUES ('foo'),('bar');
+CREATE TABLE mrg (f TEXT) ENGINE=MERGE, UNION(t1);
+SELECT * FROM mrg;
+DROP TABLE mrg, t1;
+
+--echo End of 10.5 tests
+
 --echo #
 --echo # End of 10.0 tests
 --echo #

mysql-test/main/selectivity.result

@@ -1824,7 +1824,6 @@ test.t1	analyze	status	Table is already up to date
 test.t2	analyze	status	Engine-independent statistics collected
 test.t2	analyze	status	Table is already up to date
 set optimizer_switch='exists_to_in=off';
-set optimizer_use_condition_selectivity=2;
 SELECT * FROM t1
 WHERE
 EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
@@ -1849,18 +1848,39 @@ id	a
 17	17
 18	18
 19	19
-explain SELECT * FROM t1
+set statement optimizer_use_condition_selectivity=2 for explain SELECT * FROM t1
 WHERE
 EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
 WHERE A.a=t1.a AND t2.b < 20);
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	100	Using where
-2	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	
-2	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (10%)	Using where; Using rowid filter
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (10%)	Using where; Using rowid filter
-EXPLAIN  SELECT * FROM t1 A, t1 B WHERE A.a = B.a and A.id = 65;
+set statement optimizer_use_condition_selectivity=4 for explain SELECT * FROM t1
+WHERE
+EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
+WHERE A.a=t1.a AND t2.b < 20);
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
-1	SIMPLE	A	const	PRIMARY,a	PRIMARY	4	const	1	
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	100	Using where
-1	SIMPLE	B	ref	a	a	5	const	1	
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (10%)	Using where; Using rowid filter
+set @query="EXPLAIN  SELECT * FROM t1 A, t1 B WHERE A.a = B.a and A.id = 65";
+set statement optimizer_use_condition_selectivity=2 for explain SELECT * FROM t1
+WHERE
+EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
+WHERE A.a=t1.a AND t2.b < 20);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	100	Using where
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (10%)	Using where; Using rowid filter
+set statement optimizer_use_condition_selectivity=4 for explain SELECT * FROM t1
+WHERE
+EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
+WHERE A.a=t1.a AND t2.b < 20);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t1	ALL	NULL	NULL	NULL	NULL	100	Using where
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (10%)	Using where; Using rowid filter
 explain SELECT * FROM t1
 WHERE
 EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
@@ -1870,7 +1890,6 @@ id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 2	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	
 2	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (10%)	Using where; Using rowid filter
 set optimizer_switch= @save_optimizer_switch;
-set optimizer_use_condition_selectivity= @save_optimizer_use_condition_selectivity;
 drop table t1,t2;
 #
 # MDEV-21495: Conditional jump or move depends on uninitialised value in sel_arg_range_seq_next

mysql-test/main/selectivity.test

@@ -1236,13 +1236,10 @@ set optimizer_use_condition_selectivity= @@optimizer_use_condition_selectivity;
 
 drop table t1,t2,t3;
 
-
 --echo #
 --echo # MDEV-20519: Query plan regression with optimizer_use_condition_selectivity=4
 --echo #
 
-
-
 create table t1 (id int, a int, PRIMARY KEY(id), key(a));
 insert into t1 select seq,seq from seq_1_to_100;
 
@@ -1252,7 +1249,6 @@ insert into t2 select seq,seq,seq from seq_1_to_100;
 analyze table t1,t2 persistent for all;
 
 set optimizer_switch='exists_to_in=off';
-set optimizer_use_condition_selectivity=2;
 
 let $query= SELECT * FROM t1
             WHERE
@@ -1260,14 +1256,16 @@ let $query= SELECT * FROM t1
                       WHERE A.a=t1.a AND t2.b < 20);
 
 eval $query;
-eval explain $query;
+eval set statement optimizer_use_condition_selectivity=2 for explain $query;
+eval set statement optimizer_use_condition_selectivity=4 for explain $query;
 
-EXPLAIN  SELECT * FROM t1 A, t1 B WHERE A.a = B.a and A.id = 65;
+set @query="EXPLAIN  SELECT * FROM t1 A, t1 B WHERE A.a = B.a and A.id = 65";
+eval set statement optimizer_use_condition_selectivity=2 for explain $query;
+eval set statement optimizer_use_condition_selectivity=4 for explain $query;
 
 eval explain $query;
 
 set optimizer_switch= @save_optimizer_switch;
-set optimizer_use_condition_selectivity= @save_optimizer_use_condition_selectivity;
 drop table t1,t2;
 
 --echo #

mysql-test/main/selectivity_innodb.result

@@ -1836,7 +1836,6 @@ test.t1	analyze	status	OK
 test.t2	analyze	status	Engine-independent statistics collected
 test.t2	analyze	status	OK
 set optimizer_switch='exists_to_in=off';
-set optimizer_use_condition_selectivity=2;
 SELECT * FROM t1
 WHERE
 EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
@@ -1861,18 +1860,39 @@ id	a
 17	17
 18	18
 19	19
-explain SELECT * FROM t1
+set statement optimizer_use_condition_selectivity=2 for explain SELECT * FROM t1
 WHERE
 EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
 WHERE A.a=t1.a AND t2.b < 20);
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 1	PRIMARY	t1	index	NULL	a	5	NULL	100	Using where; Using index
-2	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	Using index
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	Using index
-2	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (19%)	Using where; Using rowid filter
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (19%)	Using where; Using rowid filter
-EXPLAIN  SELECT * FROM t1 A, t1 B WHERE A.a = B.a and A.id = 65;
+set statement optimizer_use_condition_selectivity=4 for explain SELECT * FROM t1
+WHERE
+EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
+WHERE A.a=t1.a AND t2.b < 20);
 id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
-1	SIMPLE	A	const	PRIMARY,a	PRIMARY	4	const	1	
+1	PRIMARY	t1	index	NULL	a	5	NULL	100	Using where; Using index
-1	SIMPLE	B	ref	a	a	5	const	1	Using index
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	Using index
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (19%)	Using where; Using rowid filter
+set @query="EXPLAIN  SELECT * FROM t1 A, t1 B WHERE A.a = B.a and A.id = 65";
+set statement optimizer_use_condition_selectivity=2 for explain SELECT * FROM t1
+WHERE
+EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
+WHERE A.a=t1.a AND t2.b < 20);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t1	index	NULL	a	5	NULL	100	Using where; Using index
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	Using index
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (19%)	Using where; Using rowid filter
+set statement optimizer_use_condition_selectivity=4 for explain SELECT * FROM t1
+WHERE
+EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
+WHERE A.a=t1.a AND t2.b < 20);
+id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
+1	PRIMARY	t1	index	NULL	a	5	NULL	100	Using where; Using index
+3	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	Using index
+3	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (19%)	Using where; Using rowid filter
 explain SELECT * FROM t1
 WHERE
 EXISTS (SELECT * FROM t1 A INNER JOIN t2 ON t2.a = A.id
@@ -1882,7 +1902,6 @@ id	select_type	table	type	possible_keys	key	key_len	ref	rows	Extra
 2	DEPENDENT SUBQUERY	A	ref	PRIMARY,a	a	5	test.t1.a	1	Using index
 2	DEPENDENT SUBQUERY	t2	ref|filter	a,b	a|b	5|5	test.A.id	1 (19%)	Using where; Using rowid filter
 set optimizer_switch= @save_optimizer_switch;
-set optimizer_use_condition_selectivity= @save_optimizer_use_condition_selectivity;
 drop table t1,t2;
 #
 # MDEV-21495: Conditional jump or move depends on uninitialised value in sel_arg_range_seq_next

mysql-test/main/type_timestamp.result

@@ -1230,6 +1230,8 @@ SELECT * FROM t1 HAVING MIN(t1.c1) >= ALL(SELECT 'a' UNION SELECT 'r');
 c1
 Warnings:
 Warning	1292	Truncated incorrect datetime value: 'r'
+SELECT * FROM t1 HAVING MIN(t1.c1) > 0;
+c1
 DROP TABLE t1;
 CREATE TABLE t1 (c1 timestamp);
 INSERT INTO t1 VALUES ('2010-01-01 00:00:00');

mysql-test/main/type_timestamp.test

@@ -810,6 +810,7 @@ DROP TABLE t1;
 CREATE TABLE t1 (c1 timestamp);
 SELECT MIN(t1.c1) AS k1 FROM t1 HAVING (k1 >= ALL(SELECT 'a' UNION SELECT 'r'));
 SELECT * FROM t1 HAVING MIN(t1.c1) >= ALL(SELECT 'a' UNION SELECT 'r');
+SELECT * FROM t1 HAVING MIN(t1.c1) > 0;
 DROP TABLE t1;
 
 CREATE TABLE t1 (c1 timestamp);

mysql-test/mariadb-test-run.pl

@@ -408,8 +408,11 @@ sub main {
 
   mark_time_used('collect');
 
-  mysql_install_db(default_mysqld(), "$opt_vardir/install.db") unless using_extern();
+  if (!using_extern())
-
+  {
+    mysql_install_db(default_mysqld(), "$opt_vardir/install.db");
+    make_readonly("$opt_vardir/install.db");
+  }
   if ($opt_dry_run)
   {
     for (@$tests) {

mysql-test/suite/galera/disabled.def

@@ -27,3 +27,5 @@ galera_bf_kill_debug : timeout after 900 seconds
 galera_ssl_upgrade : [Warning] Failed to load slave replication state from table mysql.gtid_slave_pos: 130: Incorrect file format 'gtid_slave_pos'
 galera_parallel_simple : timeout related to wsrep_sync_wait
 galera_insert_bulk : MDEV-30536 no expected deadlock in galera_insert_bulk test
+MDEV-27713 : test is using get_lock(), which is now rejected in cluster
+galera_bf_abort_group_commit : MDEV-30855 PR to remove the test exists

mysql-test/suite/galera/r/MDEV-29293.result

@@ -0,0 +1,21 @@
+connection node_2;
+connection node_1;
+connect node_1a, 127.0.0.1, root, , test, $NODE_MYPORT_1;
+connect node_1b, 127.0.0.1, root, , test, $NODE_MYPORT_1;
+set wsrep_sync_wait = 0;
+CREATE TABLE t1(a int not null primary key auto_increment, b int) engine=InnoDB;
+INSERT INTO t1 VALUES (1,2);
+connection node_1a;
+BEGIN;
+UPDATE t1 SET b=3 WHERE a=1;
+connection node_1;
+set debug_sync='wsrep_kill_before_awake_no_mutex SIGNAL before_kill WAIT_FOR continue';
+connection node_1b;
+set debug_sync= 'now WAIT_FOR before_kill';
+connection node_2;
+UPDATE t1 SET b=7 WHERE a=1;
+connection node_1b;
+set debug_sync= 'now SIGNAL continue';
+connection node_1;
+DROP TABLE t1;
+SET DEBUG_SYNC= 'RESET';

mysql-test/suite/galera/r/galera_create_table_as_select.result

@@ -82,6 +82,7 @@ connect node_1a, 127.0.0.1, root, , test, $NODE_MYPORT_1;
 LOCK TABLE t2 WRITE;
 connection node_1;
 CREATE TABLE t1 AS SELECT * FROM t2;;
+connection node_1a;
 connection node_2;
 SELECT COUNT(*) = 5 FROM t2;
 COUNT(*) = 5

mysql-test/suite/galera/r/galera_gcache_recover_manytrx.result

@@ -134,6 +134,3 @@ connection node_1;
 call mtr.add_suppression("Error in Log_event::read_log_event():.*");
 CALL mtr.add_suppression("conflict state 7 after post commit");
 CALL mtr.add_suppression("Skipped GCache ring buffer recovery");
-connection node_2;
-call mtr.add_suppression("Error in Log_event::read_log_event():.*");
-CALL mtr.add_suppression("Skipped GCache ring buffer recovery");

mysql-test/suite/galera/r/galera_kill_group_commit.result

@@ -0,0 +1,27 @@
+connection node_2;
+connection node_1;
+connect node_1_kill, 127.0.0.1, root, , test, $NODE_MYPORT_1;
+connect node_1_ctrl, 127.0.0.1, root, , test, $NODE_MYPORT_1;
+SET SESSION wsrep_sync_wait = 0;
+connect node_1_follower, 127.0.0.1, root, , test, $NODE_MYPORT_1;
+SET SESSION wsrep_sync_wait = 0;
+connection node_1;
+CREATE TABLE t1 (f1 INT PRIMARY KEY) ENGINE=InnoDB;
+SET SESSION DEBUG_SYNC = "commit_before_enqueue SIGNAL leader_before_enqueue_reached WAIT_FOR leader_before_enqueue_continue";
+INSERT INTO t1 VALUES (1);
+connection node_1_ctrl;
+SET DEBUG_SYNC = "now WAIT_FOR leader_before_enqueue_reached";
+connection node_1_follower;
+INSERT INTO t1 VALUES (2);;
+connection node_1_ctrl;
+connection node_1_kill;
+# Execute KILL QUERY for group commit follower
+SET DEBUG_SYNC = "now SIGNAL leader_before_enqueue_continue";
+connection node_1_follower;
+connection node_1;
+SELECT * FROM t1;
+f1
+1
+2
+SET DEBUG_SYNC = "RESET";
+DROP TABLE t1;

mysql-test/suite/galera/r/galera_var_retry_autocommit.result

@@ -36,7 +36,10 @@ SET DEBUG_SYNC = 'now SIGNAL wsrep_retry_autocommit_continue';
 connection node_1;
 SELECT COUNT(*) FROM t1;
 COUNT(*)
-1
+connection node_1;
+SELECT COUNT(*) FROM t1;
+COUNT(*)
+0
 SET DEBUG_SYNC = 'RESET';
 SET GLOBAL debug_dbug = NULL;
 DROP TABLE t1;

mysql-test/suite/galera/t/MDEV-29293.test

@@ -0,0 +1,41 @@
+--source include/galera_cluster.inc
+--source include/have_innodb.inc
+--source include/have_debug_sync.inc
+--source include/galera_have_debug_sync.inc
+
+--connect node_1a, 127.0.0.1, root, , test, $NODE_MYPORT_1
+--connect node_1b, 127.0.0.1, root, , test, $NODE_MYPORT_1
+set wsrep_sync_wait = 0;
+
+CREATE TABLE t1(a int not null primary key auto_increment, b int) engine=InnoDB;
+INSERT INTO t1 VALUES (1,2); 
+
+--connection node_1a
+--let $victim_id = `SELECT CONNECTION_ID()`
+BEGIN;
+UPDATE t1 SET b=3 WHERE a=1;
+
+--connection node_1
+set debug_sync='wsrep_kill_before_awake_no_mutex SIGNAL before_kill WAIT_FOR continue';
+--disable_query_log
+--disable_result_log
+--send_eval KILL CONNECTION $victim_id 
+--enable_result_log
+--enable_query_log
+
+--connection node_1b
+set debug_sync= 'now WAIT_FOR before_kill';
+
+--connection node_2
+UPDATE t1 SET b=7 WHERE a=1;
+
+--connection node_1b 
+--let $wait_condition = SELECT COUNT(*) = 1 FROM INFORMATION_SCHEMA.PROCESSLIST WHERE User = 'system user' AND  State LIKE 'Update_rows_log_event%'; 
+--source include/wait_condition.inc
+set debug_sync= 'now SIGNAL continue'; 
+
+--connection node_1
+--reap 
+DROP TABLE t1;
+SET DEBUG_SYNC= 'RESET';
+

mysql-test/suite/galera/t/galera_create_table_as_select.test

@@ -113,6 +113,10 @@ LOCK TABLE t2 WRITE;
 --connection node_1
 --send CREATE TABLE t1 AS SELECT * FROM t2;
 
+--connection node_1a
+--let $wait_condition = SELECT COUNT(*) = 1 FROM information_schema.processlist WHERE STATE LIKE 'Waiting for table metadata lock%'
+--source include/wait_condition.inc
+
 --connection node_2
 SELECT COUNT(*) = 5 FROM t2;
 CREATE TABLE t1 AS SELECT * FROM t2;
@@ -121,7 +125,7 @@ CREATE TABLE t1 AS SELECT * FROM t2;
 UNLOCK TABLES;
 
 --connection node_1
---error ER_TABLE_EXISTS_ERROR,ER_LOCK_DEADLOCK
+--error ER_TABLE_EXISTS_ERROR,ER_QUERY_INTERRUPTED
 --reap
 
 DROP TABLE t1, t2;

mysql-test/suite/galera/t/galera_kill_group_commit.cnf

@@ -0,0 +1,5 @@
+!include ../galera_2nodes.cnf
+
+[mysqld]
+log-bin
+log-slave-updates

mysql-test/suite/galera/t/galera_kill_group_commit.test

@@ -0,0 +1,69 @@
+#
+# Verify that transaction which has reached group commit queue
+# cannot be killed. If the kill succeeds, assertion for
+# wsrep transaction state will fail.
+#
+# If the bug is present, i.e. wsrep transaction gets killed during
+# group commit wait, this test is enough to reproduce the crash
+# most of the time.
+#
+
+--source include/have_innodb.inc
+--source include/have_debug_sync.inc
+--source include/galera_cluster.inc
+
+# Connection for KILL commands
+--connect node_1_kill, 127.0.0.1, root, , test, $NODE_MYPORT_1
+# Connection for sync point control
+--connect node_1_ctrl, 127.0.0.1, root, , test, $NODE_MYPORT_1
+SET SESSION wsrep_sync_wait = 0;
+# Connection for group commit follower
+--connect node_1_follower, 127.0.0.1, root, , test, $NODE_MYPORT_1
+# Need to disable sync wait to reach commit queue when leader
+# is blocked.
+SET SESSION wsrep_sync_wait = 0;
+--let $follower_id = `SELECT CONNECTION_ID()`
+
+--connection node_1
+CREATE TABLE t1 (f1 INT PRIMARY KEY) ENGINE=InnoDB;
+
+SET SESSION DEBUG_SYNC = "commit_before_enqueue SIGNAL leader_before_enqueue_reached WAIT_FOR leader_before_enqueue_continue";
+--send INSERT INTO t1 VALUES (1)
+
+--connection node_1_ctrl
+SET DEBUG_SYNC = "now WAIT_FOR leader_before_enqueue_reached";
+
+--connection node_1_follower
+# SET SESSION DEBUG_SYNC = "group_commit_waiting_for_prior SIGNAL follower_waiting_for_prior_reached WAIT_FOR follower_waiting_for_prior_continue";
+--send INSERT INTO t1 VALUES (2);
+
+--connection node_1_ctrl
+# TODO: Is it possible to use sync points to enforce group commit to happen?
+# The leader will hold commit monitor in commit_before_enqueue sync point,
+# which prevents the follower to reach the group commit wait state.
+# We now sleep and expect the follower to reach group commit, but this
+# may cause false negatives.
+--sleep 1
+
+--connection node_1_kill
+--echo # Execute KILL QUERY for group commit follower
+--disable_query_log
+--disable_result_log
+# Because it is currently impossible to verify that the
+# follower has reached group commit queue, the KILL may
+# sometimes return success.
+--error 0,ER_KILL_DENIED_ERROR
+--eval KILL QUERY $follower_id
+--enable_result_log
+--enable_query_log
+
+SET DEBUG_SYNC = "now SIGNAL leader_before_enqueue_continue";
+--connection node_1_follower
+--reap
+
+--connection node_1
+--reap
+SELECT * FROM t1;
+
+SET DEBUG_SYNC = "RESET";
+DROP TABLE t1;

mysql-test/suite/galera/t/galera_var_retry_autocommit.test

@@ -64,6 +64,7 @@ SELECT COUNT(*) FROM t1;
 SET DEBUG_SYNC = 'now SIGNAL wsrep_retry_autocommit_continue';
 
 --connection node_1
+--error 0,ER_LOCK_DEADLOCK
 --reap
 SELECT COUNT(*) FROM t1;
 

mysql-test/suite/innodb/r/monitor.result

@@ -1,10 +1,9 @@
-set global innodb_monitor_disable = All;
 select name, if(enabled,'enabled','disabled') status
 from information_schema.innodb_metrics;
 name	status
 metadata_table_handles_opened	disabled
-lock_deadlocks	disabled
+lock_deadlocks	enabled
-lock_timeouts	disabled
+lock_timeouts	enabled
 lock_rec_lock_waits	disabled
 lock_table_lock_waits	disabled
 lock_rec_lock_requests	disabled
@@ -14,30 +13,30 @@ lock_rec_locks	disabled
 lock_table_lock_created	disabled
 lock_table_lock_removed	disabled
 lock_table_locks	disabled
-lock_row_lock_current_waits	disabled
+lock_row_lock_current_waits	enabled
-lock_row_lock_time	disabled
+lock_row_lock_time	enabled
-lock_row_lock_time_max	disabled
+lock_row_lock_time_max	enabled
-lock_row_lock_waits	disabled
+lock_row_lock_waits	enabled
-lock_row_lock_time_avg	disabled
+lock_row_lock_time_avg	enabled
-buffer_pool_size	disabled
+buffer_pool_size	enabled
-buffer_pool_reads	disabled
+buffer_pool_reads	enabled
-buffer_pool_read_requests	disabled
+buffer_pool_read_requests	enabled
-buffer_pool_write_requests	disabled
+buffer_pool_write_requests	enabled
-buffer_pool_wait_free	disabled
+buffer_pool_wait_free	enabled
-buffer_pool_read_ahead	disabled
+buffer_pool_read_ahead	enabled
-buffer_pool_read_ahead_evicted	disabled
+buffer_pool_read_ahead_evicted	enabled
-buffer_pool_pages_total	disabled
+buffer_pool_pages_total	enabled
-buffer_pool_pages_misc	disabled
+buffer_pool_pages_misc	enabled
-buffer_pool_pages_data	disabled
+buffer_pool_pages_data	enabled
-buffer_pool_bytes_data	disabled
+buffer_pool_bytes_data	enabled
-buffer_pool_pages_dirty	disabled
+buffer_pool_pages_dirty	enabled
-buffer_pool_bytes_dirty	disabled
+buffer_pool_bytes_dirty	enabled
-buffer_pool_pages_free	disabled
+buffer_pool_pages_free	enabled
-buffer_pages_created	disabled
+buffer_pages_created	enabled
-buffer_pages_written	disabled
+buffer_pages_written	enabled
-buffer_pages_read	disabled
+buffer_pages_read	enabled
-buffer_data_reads	disabled
+buffer_data_reads	enabled
-buffer_data_written	disabled
+buffer_data_written	enabled
 buffer_flush_batch_scanned	disabled
 buffer_flush_batch_num_scan	disabled
 buffer_flush_batch_scanned_per_call	disabled
@@ -70,8 +69,8 @@ buffer_flush_background_pages	disabled
 buffer_LRU_batch_scanned	disabled
 buffer_LRU_batch_num_scan	disabled
 buffer_LRU_batch_scanned_per_call	disabled
-buffer_LRU_batch_flush_total_pages	disabled
+buffer_LRU_batch_flush_total_pages	enabled
-buffer_LRU_batch_evict_total_pages	disabled
+buffer_LRU_batch_evict_total_pages	enabled
 buffer_LRU_single_flush_failure_count	disabled
 buffer_LRU_get_free_search	disabled
 buffer_LRU_search_scanned	disabled
@@ -112,21 +111,21 @@ buffer_page_written_blob	disabled
 buffer_page_written_zblob	disabled
 buffer_page_written_zblob2	disabled
 buffer_page_written_other	disabled
-os_data_reads	disabled
+os_data_reads	enabled
-os_data_writes	disabled
+os_data_writes	enabled
-os_data_fsyncs	disabled
+os_data_fsyncs	enabled
-os_pending_reads	disabled
+os_pending_reads	enabled
-os_pending_writes	disabled
+os_pending_writes	enabled
-os_log_bytes_written	disabled
+os_log_bytes_written	enabled
 trx_rw_commits	disabled
 trx_ro_commits	disabled
 trx_nl_ro_commits	disabled
 trx_commits_insert_update	disabled
 trx_rollbacks	disabled
 trx_rollbacks_savepoint	disabled
-trx_rseg_history_len	disabled
+trx_rseg_history_len	enabled
 trx_undo_slots_used	disabled
-trx_undo_slots_cached	disabled
+trx_undo_slots_cached	enabled
 trx_rseg_current_size	disabled
 purge_del_mark_records	disabled
 purge_upd_exist_or_extern_records	disabled
@@ -142,9 +141,9 @@ log_lsn_current	disabled
 log_lsn_checkpoint_age	disabled
 log_lsn_buf_pool_oldest	disabled
 log_max_modified_age_async	disabled
-log_waits	disabled
+log_waits	enabled
-log_write_requests	disabled
+log_write_requests	enabled
-log_writes	disabled
+log_writes	enabled
 compress_pages_compressed	disabled
 compress_pages_decompressed	disabled
 compression_pad_increments	disabled
@@ -162,34 +161,34 @@ index_page_merge_successful	disabled
 index_page_reorg_attempts	disabled
 index_page_reorg_successful	disabled
 index_page_discards	disabled
-adaptive_hash_searches	disabled
+adaptive_hash_searches	enabled
-adaptive_hash_searches_btree	disabled
+adaptive_hash_searches_btree	enabled
 adaptive_hash_pages_added	disabled
 adaptive_hash_pages_removed	disabled
 adaptive_hash_rows_added	disabled
 adaptive_hash_rows_removed	disabled
 adaptive_hash_rows_deleted_no_hash_entry	disabled
 adaptive_hash_rows_updated	disabled
-file_num_open_files	disabled
+file_num_open_files	enabled
-ibuf_merges_insert	disabled
+ibuf_merges_insert	enabled
-ibuf_merges_delete_mark	disabled
+ibuf_merges_delete_mark	enabled
-ibuf_merges_delete	disabled
+ibuf_merges_delete	enabled
-ibuf_merges_discard_insert	disabled
+ibuf_merges_discard_insert	enabled
-ibuf_merges_discard_delete_mark	disabled
+ibuf_merges_discard_delete_mark	enabled
-ibuf_merges_discard_delete	disabled
+ibuf_merges_discard_delete	enabled
-ibuf_merges	disabled
+ibuf_merges	enabled
-ibuf_size	disabled
+ibuf_size	enabled
 innodb_master_thread_sleeps	disabled
-innodb_activity_count	disabled
+innodb_activity_count	enabled
 innodb_master_active_loops	disabled
 innodb_master_idle_loops	disabled
 innodb_log_flush_usec	disabled
 innodb_dict_lru_usec	disabled
 innodb_dict_lru_count_active	disabled
 innodb_dict_lru_count_idle	disabled
-innodb_dblwr_writes	disabled
+innodb_dblwr_writes	enabled
-innodb_dblwr_pages_written	disabled
+innodb_dblwr_pages_written	enabled
-innodb_page_size	disabled
+innodb_page_size	enabled
 ddl_background_drop_indexes	disabled
 ddl_online_create_index	disabled
 ddl_pending_alter_table	disabled
@@ -199,6 +198,9 @@ icp_attempts	disabled
 icp_no_match	disabled
 icp_out_of_range	disabled
 icp_match	disabled
+set global innodb_monitor_disable = All;
+select name from information_schema.innodb_metrics where enabled;
+name
 set global innodb_monitor_enable = all;
 select name from information_schema.innodb_metrics where not enabled;
 name

mysql-test/suite/innodb/r/stat_tables.result

@@ -26,4 +26,60 @@ UPDATE mysql.innodb_table_stats SET last_update=NULL WHERE table_name='t1';
 XA END 'test';
 XA ROLLBACK 'test';
 DROP TABLE t1;
+#
+# MDEV-30483 After upgrade to 10.6 from Mysql 5.7 seeing "InnoDB: Column last_update in table mysql.innodb_table_stats is BINARY(4) NOT NULL but should be INT UNSIGNED NOT NULL"
+#
+#
+# Testing a non-default format: Field_timestamp0 - UINT4 based
+#
+SET @@global.mysql56_temporal_format=0;
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW COLUMNS FROM mysql.innodb_table_stats LIKE 'last_update';
+Field	Type	Null	Key	Default	Extra
+last_update	timestamp /* mariadb-5.3 */	NO		current_timestamp()	on update current_timestamp()
+SHOW COLUMNS FROM mysql.innodb_index_stats LIKE 'last_update';
+Field	Type	Null	Key	Default	Extra
+last_update	timestamp /* mariadb-5.3 */	NO		current_timestamp()	on update current_timestamp()
+CREATE TABLE t1 (a INT PRIMARY KEY) ENGINE=InnoDB STATS_PERSISTENT=1;
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+TIMESTAMPDIFF(DAY,last_update,now())<=1
+1
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+TIMESTAMPDIFF(DAY,last_update,now())<=1
+1
+DROP TABLE t1;
+#
+# Now as the table t1 is dropped, expect no statistics
+#
+SELECT * FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+database_name	table_name	last_update	n_rows	clustered_index_size	sum_of_other_index_sizes
+SELECT * FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+database_name	table_name	index_name	last_update	stat_name	stat_value	sample_size	stat_description
+#
+# Testing with the default format: Field_timestampf - BINARY(4) based with the UNSIGNED_FLAG
+#
+SET @@global.mysql56_temporal_format=1;
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW COLUMNS FROM mysql.innodb_table_stats LIKE 'last_update';
+Field	Type	Null	Key	Default	Extra
+last_update	timestamp	NO		current_timestamp()	on update current_timestamp()
+SHOW COLUMNS FROM mysql.innodb_index_stats LIKE 'last_update';
+Field	Type	Null	Key	Default	Extra
+last_update	timestamp	NO		current_timestamp()	on update current_timestamp()
+CREATE TABLE t1 (a INT PRIMARY KEY) ENGINE=InnoDB STATS_PERSISTENT=1;
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+TIMESTAMPDIFF(DAY,last_update,now())<=1
+1
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+TIMESTAMPDIFF(DAY,last_update,now())<=1
+1
+DROP TABLE t1;
 # End of 10.6 tests

mysql-test/suite/innodb/t/monitor.test

@@ -5,12 +5,14 @@
 # sys_vars.innodb_monitor_enable_basic
 
 --source include/have_innodb.inc
-set global innodb_monitor_disable = All;
 # Test turn on/off the monitor counter  with "all" option
 # By default, they will be off.
 select name, if(enabled,'enabled','disabled') status
 from information_schema.innodb_metrics;
 
+set global innodb_monitor_disable = All;
+select name from information_schema.innodb_metrics where enabled;
+
 # Turn on all monitor counters
 set global innodb_monitor_enable = all;
 

mysql-test/suite/innodb/t/stat_tables.test

@@ -28,4 +28,57 @@ XA END 'test';
 XA ROLLBACK 'test';
 DROP TABLE t1;
 
+--echo #
+--echo # MDEV-30483 After upgrade to 10.6 from Mysql 5.7 seeing "InnoDB: Column last_update in table mysql.innodb_table_stats is BINARY(4) NOT NULL but should be INT UNSIGNED NOT NULL"
+--echo #
+
+# The following tests demonstrate that these columns:
+# - innodb_table_stats.last_update
+# - innodb_index_stats.last_update
+# have sane values close to NOW(), rather than any garbage,
+# with all TIMESTAMP formats.
+
+--echo #
+--echo # Testing a non-default format: Field_timestamp0 - UINT4 based
+--echo #
+
+SET @@global.mysql56_temporal_format=0;
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW COLUMNS FROM mysql.innodb_table_stats LIKE 'last_update';
+SHOW COLUMNS FROM mysql.innodb_index_stats LIKE 'last_update';
+CREATE TABLE t1 (a INT PRIMARY KEY) ENGINE=InnoDB STATS_PERSISTENT=1;
+
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+DROP TABLE t1;
+
+--echo #
+--echo # Now as the table t1 is dropped, expect no statistics
+--echo #
+
+SELECT * FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+SELECT * FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+
+--echo #
+--echo # Testing with the default format: Field_timestampf - BINARY(4) based with the UNSIGNED_FLAG
+--echo #
+
+SET @@global.mysql56_temporal_format=1;
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW COLUMNS FROM mysql.innodb_table_stats LIKE 'last_update';
+SHOW COLUMNS FROM mysql.innodb_index_stats LIKE 'last_update';
+CREATE TABLE t1 (a INT PRIMARY KEY) ENGINE=InnoDB STATS_PERSISTENT=1;
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+DROP TABLE t1;
+
+
 --echo # End of 10.6 tests

mysql-test/suite/parts/r/partition_purge.result

@@ -0,0 +1,26 @@
+CREATE TABLE t1(f1 INT, f2 INT, INDEX(f1))ENGINE=InnoDB
+PARTITION BY LIST(f1) (
+PARTITION p1 VALUES in (1, 2, 3),
+PARTITION p2 VALUES in (4, 5, 6));
+INSERT INTO t1 VALUES(1, 1), (1, 1), (6, 1);
+connect con1,localhost,root,,,;
+START TRANSACTION WITH CONSISTENT SNAPSHOT;
+connect con2,localhost,root,,,;
+SET DEBUG_SYNC="innodb_rollback_inplace_alter_table SIGNAL default_resume WAIT_FOR alter_resume";
+ALTER TABLE t1 ADD UNIQUE INDEX(f1);
+connection default;
+set DEBUG_SYNC="now WAIT_FOR default_resume";
+SET DEBUG_SYNC="innodb_row_update_for_mysql_begin SIGNAL alter_resume WAIT_FOR alter_finish";
+DELETE FROM t1;
+connection con2;
+ERROR 23000: Duplicate entry '1' for key 'f1_2'
+SET DEBUG_SYNC="now SIGNAL alter_finish";
+connection default;
+connection con1;
+commit;
+connection default;
+disconnect con1;
+disconnect con2;
+InnoDB		0 transactions not purged
+drop table t1;
+SET DEBUG_SYNC=reset;

mysql-test/suite/parts/t/partition_purge.opt

@@ -0,0 +1 @@
+--innodb_purge_threads=1

mysql-test/suite/parts/t/partition_purge.test

@@ -0,0 +1,37 @@
+--source include/have_innodb.inc
+--source include/have_partition.inc
+--source include/have_debug.inc
+--source include/have_debug_sync.inc
+
+CREATE TABLE t1(f1 INT, f2 INT, INDEX(f1))ENGINE=InnoDB
+                PARTITION BY LIST(f1) (
+                        PARTITION p1 VALUES in (1, 2, 3),
+                        PARTITION p2 VALUES in (4, 5, 6));
+INSERT INTO t1 VALUES(1, 1), (1, 1), (6, 1);
+connect(con1,localhost,root,,,);
+START TRANSACTION WITH CONSISTENT SNAPSHOT;
+
+connect(con2,localhost,root,,,);
+SET DEBUG_SYNC="innodb_rollback_inplace_alter_table SIGNAL default_resume WAIT_FOR alter_resume";
+send ALTER TABLE t1 ADD UNIQUE INDEX(f1);
+
+connection default;
+set DEBUG_SYNC="now WAIT_FOR default_resume";
+SET DEBUG_SYNC="innodb_row_update_for_mysql_begin SIGNAL alter_resume WAIT_FOR alter_finish";
+send DELETE FROM t1;
+
+connection con2;
+--error ER_DUP_ENTRY
+reap;
+SET DEBUG_SYNC="now SIGNAL alter_finish";
+
+connection default;
+reap;
+connection con1;
+commit;
+connection default;
+disconnect con1;
+disconnect con2;
+--source ../../innodb/include/wait_all_purged.inc
+drop table t1;
+SET DEBUG_SYNC=reset;

plugin/type_mysql_timestamp/CMakeLists.txt

@@ -0,0 +1,17 @@
+# Copyright (c) 2019, MariaDB corporation
+# 
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; version 2 of the License.
+# 
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+# 
+# You should have received a copy of the GNU General Public License
+# along with this program; if not, write to the Free Software
+# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1335  USA
+
+MYSQL_ADD_PLUGIN(type_mysql_timestamp plugin.cc RECOMPILE_FOR_EMBEDDED
+                 MODULE_ONLY COMPONENT Test)

plugin/type_mysql_timestamp/mysql-test/type_mysql_timestamp/suite.opt

@@ -0,0 +1 @@
+--plugin-load-add=$TYPE_MYSQL_TIMESTAMP_SO

plugin/type_mysql_timestamp/mysql-test/type_mysql_timestamp/suite.pm

@@ -0,0 +1,10 @@
+package My::Suite::Type_test;
+
+@ISA = qw(My::Suite);
+
+return "No TYPE_TEST plugin" unless $ENV{TYPE_MYSQL_TIMESTAMP_SO};
+
+sub is_default { 1 }
+
+bless { };
+

plugin/type_mysql_timestamp/mysql-test/type_mysql_timestamp/type_mysql_timestamp.result

@@ -0,0 +1,45 @@
+#
+# MDEV-30483 After upgrade to 10.6 from Mysql 5.7 seeing "InnoDB: Column last_update in table mysql.innodb_table_stats is BINARY(4) NOT NULL but should be INT UNSIGNED NOT NULL"
+#
+SELECT
+PLUGIN_NAME,
+PLUGIN_VERSION,
+PLUGIN_STATUS,
+PLUGIN_TYPE,
+PLUGIN_AUTHOR,
+PLUGIN_DESCRIPTION,
+PLUGIN_LICENSE,
+PLUGIN_MATURITY,
+PLUGIN_AUTH_VERSION
+FROM INFORMATION_SCHEMA.PLUGINS
+WHERE PLUGIN_TYPE='DATA TYPE'
+    AND PLUGIN_NAME LIKE 'type_mysql_timestamp';
+PLUGIN_NAME	type_mysql_timestamp
+PLUGIN_VERSION	1.0
+PLUGIN_STATUS	ACTIVE
+PLUGIN_TYPE	DATA TYPE
+PLUGIN_AUTHOR	MariaDB Corporation
+PLUGIN_DESCRIPTION	Data type TYPE_MYSQL_TIMESTAMP
+PLUGIN_LICENSE	GPL
+PLUGIN_MATURITY	Experimental
+PLUGIN_AUTH_VERSION	1.0
+CREATE TABLE t1 (a TYPE_MYSQL_TIMESTAMP);
+SHOW CREATE TABLE t1;
+Table	Create Table
+t1	CREATE TABLE `t1` (
+  `a` type_mysql_timestamp NULL DEFAULT NULL
+) ENGINE=MyISAM DEFAULT CHARSET=latin1 COLLATE=latin1_swedish_ci
+DROP TABLE t1;
+CREATE TABLE t1 (a TIMESTAMP);
+SHOW CREATE TABLE t1;
+Table	Create Table
+t1	CREATE TABLE `t1` (
+  `a` timestamp NULL DEFAULT NULL
+) ENGINE=MyISAM DEFAULT CHARSET=latin1 COLLATE=latin1_swedish_ci
+ALTER TABLE t1 MODIFY a TYPE_MYSQL_TIMESTAMP;
+SHOW CREATE TABLE t1;
+Table	Create Table
+t1	CREATE TABLE `t1` (
+  `a` type_mysql_timestamp NULL DEFAULT NULL
+) ENGINE=MyISAM DEFAULT CHARSET=latin1 COLLATE=latin1_swedish_ci
+DROP TABLE t1;

plugin/type_mysql_timestamp/mysql-test/type_mysql_timestamp/type_mysql_timestamp.test

@@ -0,0 +1,31 @@
+--source include/have_innodb.inc
+
+--echo #
+--echo # MDEV-30483 After upgrade to 10.6 from Mysql 5.7 seeing "InnoDB: Column last_update in table mysql.innodb_table_stats is BINARY(4) NOT NULL but should be INT UNSIGNED NOT NULL"
+--echo #
+
+--vertical_results
+SELECT
+  PLUGIN_NAME,
+  PLUGIN_VERSION,
+  PLUGIN_STATUS,
+  PLUGIN_TYPE,
+  PLUGIN_AUTHOR,
+  PLUGIN_DESCRIPTION,
+  PLUGIN_LICENSE,
+  PLUGIN_MATURITY,
+  PLUGIN_AUTH_VERSION
+FROM INFORMATION_SCHEMA.PLUGINS
+  WHERE PLUGIN_TYPE='DATA TYPE'
+    AND PLUGIN_NAME LIKE 'type_mysql_timestamp';
+--horizontal_results
+
+CREATE TABLE t1 (a TYPE_MYSQL_TIMESTAMP);
+SHOW CREATE TABLE t1;
+DROP TABLE t1;
+
+CREATE TABLE t1 (a TIMESTAMP);
+SHOW CREATE TABLE t1;
+ALTER TABLE t1 MODIFY a TYPE_MYSQL_TIMESTAMP;
+SHOW CREATE TABLE t1;
+DROP TABLE t1;

plugin/type_mysql_timestamp/mysql-test/type_mysql_timestamp/type_mysql_timestamp_stat_tables.result

@@ -0,0 +1,108 @@
+#
+# MDEV-30483 After upgrade to 10.6 from Mysql 5.7 seeing "InnoDB: Column last_update in table mysql.innodb_table_stats is BINARY(4) NOT NULL but should be INT UNSIGNED NOT NULL"
+#
+SET @@global.innodb_stats_persistent=0;
+SHOW CREATE TABLE mysql.innodb_table_stats;
+Table	Create Table
+innodb_table_stats	CREATE TABLE `innodb_table_stats` (
+  `database_name` varchar(64) NOT NULL,
+  `table_name` varchar(199) NOT NULL,
+  `last_update` timestamp NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp(),
+  `n_rows` bigint(20) unsigned NOT NULL,
+  `clustered_index_size` bigint(20) unsigned NOT NULL,
+  `sum_of_other_index_sizes` bigint(20) unsigned NOT NULL,
+  PRIMARY KEY (`database_name`,`table_name`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3 COLLATE=utf8mb3_bin STATS_PERSISTENT=0
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_table_stats;
+Table	Create Table
+innodb_table_stats	CREATE TABLE `innodb_table_stats` (
+  `database_name` varchar(64) NOT NULL,
+  `table_name` varchar(199) NOT NULL,
+  `last_update` type_mysql_timestamp NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp(),
+  `n_rows` bigint(20) unsigned NOT NULL,
+  `clustered_index_size` bigint(20) unsigned NOT NULL,
+  `sum_of_other_index_sizes` bigint(20) unsigned NOT NULL,
+  PRIMARY KEY (`database_name`,`table_name`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3 COLLATE=utf8mb3_bin STATS_PERSISTENT=0
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_index_stats;
+Table	Create Table
+innodb_index_stats	CREATE TABLE `innodb_index_stats` (
+  `database_name` varchar(64) NOT NULL,
+  `table_name` varchar(199) NOT NULL,
+  `index_name` varchar(64) NOT NULL,
+  `last_update` type_mysql_timestamp NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp(),
+  `stat_name` varchar(64) NOT NULL,
+  `stat_value` bigint(20) unsigned NOT NULL,
+  `sample_size` bigint(20) unsigned DEFAULT NULL,
+  `stat_description` varchar(1024) NOT NULL,
+  PRIMARY KEY (`database_name`,`table_name`,`index_name`,`stat_name`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3 COLLATE=utf8mb3_bin STATS_PERSISTENT=0
+SET @@global.innodb_stats_persistent=1;
+CREATE TABLE t1 (a INT, KEY(a)) ENGINE=InnoDB;
+INSERT INTO t1 VALUES (10);
+DROP TABLE t1;
+SET @@global.innodb_stats_persistent=0;
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_table_stats;
+Table	Create Table
+innodb_table_stats	CREATE TABLE `innodb_table_stats` (
+  `database_name` varchar(64) NOT NULL,
+  `table_name` varchar(199) NOT NULL,
+  `last_update` timestamp NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp(),
+  `n_rows` bigint(20) unsigned NOT NULL,
+  `clustered_index_size` bigint(20) unsigned NOT NULL,
+  `sum_of_other_index_sizes` bigint(20) unsigned NOT NULL,
+  PRIMARY KEY (`database_name`,`table_name`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3 COLLATE=utf8mb3_bin STATS_PERSISTENT=0
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_index_stats;
+Table	Create Table
+innodb_index_stats	CREATE TABLE `innodb_index_stats` (
+  `database_name` varchar(64) NOT NULL,
+  `table_name` varchar(199) NOT NULL,
+  `index_name` varchar(64) NOT NULL,
+  `last_update` timestamp NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp(),
+  `stat_name` varchar(64) NOT NULL,
+  `stat_value` bigint(20) unsigned NOT NULL,
+  `sample_size` bigint(20) unsigned DEFAULT NULL,
+  `stat_description` varchar(1024) NOT NULL,
+  PRIMARY KEY (`database_name`,`table_name`,`index_name`,`stat_name`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb3 COLLATE=utf8mb3_bin STATS_PERSISTENT=0
+SET @@global.innodb_stats_persistent=1;
+#
+# Testing MySQL-5.6-alike Field_timestampf: BINARY(4) based, without UNSIGNED_FLAG
+#
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW COLUMNS FROM mysql.innodb_table_stats LIKE 'last_update';
+Field	Type	Null	Key	Default	Extra
+last_update	type_mysql_timestamp	NO		current_timestamp()	on update current_timestamp()
+SHOW COLUMNS FROM mysql.innodb_index_stats LIKE 'last_update';
+Field	Type	Null	Key	Default	Extra
+last_update	type_mysql_timestamp	NO		current_timestamp()	on update current_timestamp()
+CREATE TABLE t1 (a INT PRIMARY KEY) ENGINE=InnoDB STATS_PERSISTENT=1;
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+TIMESTAMPDIFF(DAY,last_update,now())<=1
+1
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+TIMESTAMPDIFF(DAY,last_update,now())<=1
+1
+DROP TABLE t1;
+#
+# Now as the table t1 is dropped, expect no statistics
+#
+SELECT * FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+database_name	table_name	last_update	n_rows	clustered_index_size	sum_of_other_index_sizes
+SELECT * FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+database_name	table_name	index_name	last_update	stat_name	stat_value	sample_size	stat_description
+#
+# Restore the structure of the tables
+#
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();

plugin/type_mysql_timestamp/mysql-test/type_mysql_timestamp/type_mysql_timestamp_stat_tables.test

@@ -0,0 +1,62 @@
+--source include/have_innodb.inc
+
+--echo #
+--echo # MDEV-30483 After upgrade to 10.6 from Mysql 5.7 seeing "InnoDB: Column last_update in table mysql.innodb_table_stats is BINARY(4) NOT NULL but should be INT UNSIGNED NOT NULL"
+--echo #
+
+SET @@global.innodb_stats_persistent=0;
+SHOW CREATE TABLE mysql.innodb_table_stats;
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_table_stats;
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_index_stats;
+SET @@global.innodb_stats_persistent=1;
+
+CREATE TABLE t1 (a INT, KEY(a)) ENGINE=InnoDB;
+INSERT INTO t1 VALUES (10);
+DROP TABLE t1;
+
+SET @@global.innodb_stats_persistent=0;
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_table_stats;
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW CREATE TABLE mysql.innodb_index_stats;
+SET @@global.innodb_stats_persistent=1;
+
+
+# The following test demonstrate that these columns:
+# - innodb_table_stats.last_update
+# - innodb_index_stats.last_update
+# have sane values close to NOW(), rather than any garbage,
+# with MySQL-alike Field_timestampf
+
+--echo #
+--echo # Testing MySQL-5.6-alike Field_timestampf: BINARY(4) based, without UNSIGNED_FLAG
+--echo #
+
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TYPE_MYSQL_TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+SHOW COLUMNS FROM mysql.innodb_table_stats LIKE 'last_update';
+SHOW COLUMNS FROM mysql.innodb_index_stats LIKE 'last_update';
+CREATE TABLE t1 (a INT PRIMARY KEY) ENGINE=InnoDB STATS_PERSISTENT=1;
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+SELECT TIMESTAMPDIFF(DAY,last_update,now())<=1 FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+DROP TABLE t1;
+
+--echo #
+--echo # Now as the table t1 is dropped, expect no statistics
+--echo #
+
+SELECT * FROM mysql.innodb_table_stats
+WHERE database_name='test' AND table_name='t1';
+SELECT * FROM mysql.innodb_index_stats
+WHERE database_name='test' AND table_name='t1' AND stat_name='size';
+
+--echo #
+--echo # Restore the structure of the tables
+--echo #
+
+ALTER TABLE mysql.innodb_table_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();
+ALTER TABLE mysql.innodb_index_stats MODIFY last_update TIMESTAMP NOT NULL DEFAULT current_timestamp() ON UPDATE current_timestamp();

plugin/type_mysql_timestamp/plugin.cc

@@ -0,0 +1,177 @@
+/*
+   Copyright (c) 2023, MariaDB Corporation
+
+   This program is free software; you can redistribute it and/or modify
+   it under the terms of the GNU General Public License as published by
+   the Free Software Foundation; version 2 of the License.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU General Public License for more details.
+
+   You should have received a copy of the GNU General Public License
+   along with this program; if not, write to the Free Software
+   Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA */
+
+#include <my_global.h>
+#include <sql_class.h>
+#include <mysql/plugin_data_type.h>
+#include "sql_type.h"
+
+
+class Type_collection_local: public Type_collection
+{
+protected:
+  const Type_handler *aggregate_common(const Type_handler *h1,
+                                       const Type_handler *h2) const;
+public:
+  const Type_handler *handler_by_name(const LEX_CSTRING &name) const override
+  {
+    return NULL;
+  }
+
+  const Type_handler *aggregate_for_result(const Type_handler *h1,
+                                           const Type_handler *h2)
+                                           const override
+  {
+    return aggregate_common(h1, h2);
+  }
+
+  const Type_handler *aggregate_for_comparison(const Type_handler *h1,
+                                               const Type_handler *h2)
+                                               const override
+  {
+    return aggregate_common(h1, h2);
+  }
+
+  const Type_handler *aggregate_for_min_max(const Type_handler *h1,
+                                            const Type_handler *h2)
+                                            const override
+  {
+    return aggregate_common(h1, h2);
+  }
+
+  const Type_handler *aggregate_for_num_op(const Type_handler *h1,
+                                           const Type_handler *h2)
+                                           const override
+  {
+    return aggregate_common(h1, h2);
+  }
+};
+
+
+static Type_collection_local type_collection_local;
+
+
+/*
+  A more MySQL compatible Field:
+  it does not set the UNSIGNED_FLAG.
+  This is how MySQL's Field_timestampf works.
+*/
+class Field_mysql_timestampf :public Field_timestampf
+{
+public:
+  Field_mysql_timestampf(const LEX_CSTRING &name,
+                         const Record_addr &addr,
+                         enum utype unireg_check_arg,
+                         TABLE_SHARE *share, decimal_digits_t dec_arg)
+   :Field_timestampf(addr.ptr(), addr.null_ptr(), addr.null_bit(),
+                     unireg_check_arg, &name, share, dec_arg)
+  {
+    flags&= ~UNSIGNED_FLAG; // MySQL compatibility
+  }
+  void sql_type(String &str) const override
+  {
+    sql_type_opt_dec_comment(str,
+                             Field_mysql_timestampf::type_handler()->name(),
+                             dec, type_version_mysql56());
+  }
+  const Type_handler *type_handler() const override;
+};
+
+
+class Type_handler_mysql_timestamp2: public Type_handler_timestamp2
+{
+public:
+  const Type_collection *type_collection() const override
+  {
+    return &type_collection_local;
+  }
+  Field *make_table_field_from_def(TABLE_SHARE *share, MEM_ROOT *root,
+                                   const LEX_CSTRING *name,
+                                   const Record_addr &rec, const Bit_addr &bit,
+                                   const Column_definition_attributes *attr,
+                                   uint32 flags) const override
+  {
+    return new (root)
+      Field_mysql_timestampf(*name, rec, attr->unireg_check, share,
+                             attr->temporal_dec(MAX_DATETIME_WIDTH));
+  }
+  void Column_definition_implicit_upgrade(Column_definition *c) const override
+  {
+    /*
+      Suppress the automatic upgrade depending on opt_mysql56_temporal_format,
+      derived from Type_handler_timestamp_common.
+    */
+  }
+};
+
+
+static Type_handler_mysql_timestamp2 type_handler_mysql_timestamp2;
+
+
+const Type_handler *Field_mysql_timestampf::type_handler() const
+{
+  return &type_handler_mysql_timestamp2;
+}
+
+
+const Type_handler *
+Type_collection_local::aggregate_common(const Type_handler *h1,
+                                        const Type_handler *h2) const
+{
+  if (h1 == h2)
+    return h1;
+
+  static const Type_aggregator::Pair agg[]=
+  {
+    {
+      &type_handler_timestamp2,
+      &type_handler_mysql_timestamp2,
+      &type_handler_mysql_timestamp2
+    },
+    {NULL,NULL,NULL}
+  };
+
+  return Type_aggregator::find_handler_in_array(agg, h1, h2, true);
+}
+
+
+static struct st_mariadb_data_type plugin_descriptor_type_mysql_timestamp=
+{
+  MariaDB_DATA_TYPE_INTERFACE_VERSION,
+  &type_handler_mysql_timestamp2
+};
+
+
+
+/*************************************************************************/
+
+maria_declare_plugin(type_mysql_timestamp)
+{
+  MariaDB_DATA_TYPE_PLUGIN,     // the plugin type (see include/mysql/plugin.h)
+  &plugin_descriptor_type_mysql_timestamp, // pointer to type-specific plugin descriptor
+  "type_mysql_timestamp",       // plugin name
+  "MariaDB Corporation",        // plugin author
+  "Data type TYPE_MYSQL_TIMESTAMP", // the plugin description
+  PLUGIN_LICENSE_GPL,           // the plugin license (see include/mysql/plugin.h)
+  0,                            // Pointer to plugin initialization function
+  0,                            // Pointer to plugin deinitialization function
+  0x0100,                       // Numeric version 0xAABB means AA.BB version
+  NULL,                         // Status variables
+  NULL,                         // System variables
+  "1.0",                        // String version representation
+  MariaDB_PLUGIN_MATURITY_EXPERIMENTAL // Maturity(see include/mysql/plugin.h)*/
+}
+maria_declare_plugin_end;

sql/field.h

@@ -3354,7 +3354,7 @@ public:
 /**
   TIMESTAMP(0..6) - MySQL56 version
 */
-class Field_timestampf final :public Field_timestamp_with_dec {
+class Field_timestampf :public Field_timestamp_with_dec {
   void store_TIMEVAL(const timeval &tv) override;
 public:
   Field_timestampf(uchar *ptr_arg,

sql/handler.cc

@@ -7972,6 +7972,9 @@ Compare_keys handler::compare_key_parts(const Field &old_field,
   concurrent accesses. And it's an overkill to take LOCK_plugin and
   iterate the whole installed_htons[] array every time.
 
+  @note Object victim_thd is not guaranteed to exist after this
+        function returns.
+
   @param bf_thd       brute force THD asking for the abort
   @param victim_thd   victim THD to be aborted
 
@@ -7985,6 +7988,8 @@ int ha_abort_transaction(THD *bf_thd, THD *victim_thd, my_bool signal)
   if (!WSREP(bf_thd) &&
       !(bf_thd->variables.wsrep_OSU_method == WSREP_OSU_RSU &&
         wsrep_thd_is_toi(bf_thd))) {
+    mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
+    mysql_mutex_unlock(&victim_thd->LOCK_thd_kill);
     DBUG_RETURN(0);
   }
 
@@ -7996,6 +8001,8 @@ int ha_abort_transaction(THD *bf_thd, THD *victim_thd, my_bool signal)
   else
   {
     WSREP_WARN("Cannot abort InnoDB transaction");
+    mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
+    mysql_mutex_unlock(&victim_thd->LOCK_thd_kill);
   }
 
   DBUG_RETURN(0);

sql/handler.h

@@ -45,6 +45,7 @@
 #include "sql_sequence.h"
 #include "mem_root_array.h"
 #include <utility>     // pair
+#include <my_attribute.h> /* __attribute__ */
 
 class Alter_info;
 class Virtual_column_info;
@@ -1460,8 +1461,8 @@ struct handlerton
                             const char *query, uint query_length,
                             const char *db, const char *table_name);
 
-   void (*abort_transaction)(handlerton *hton, THD *bf_thd,
+   void (*abort_transaction)(handlerton *hton, THD *bf_thd, THD *victim_thd,
-			    THD *victim_thd, my_bool signal);
+                             my_bool signal) __attribute__((nonnull));
    int (*set_checkpoint)(handlerton *hton, const XID *xid);
    int (*get_checkpoint)(handlerton *hton, XID* xid);
   /**

sql/item_func.h

@@ -373,7 +373,7 @@ public:
   {
     for (uint i= 0; i < arg_count; i++)
     {
-      args[i]->no_rows_in_result();
+      args[i]->restore_to_before_no_rows_in_result();
     }
   }
   void convert_const_compared_to_int_field(THD *thd);

sql/mysqld.cc

@@ -1747,6 +1747,11 @@ static void close_connections(void)
       (void) unlink(mysqld_unix_port);
     }
   }
+  /*
+    The following is needed to the threads stuck in
+    setup_connection_thread_globals()
+    to continue.
+  */
   listen_sockets.free_memory();
   mysql_mutex_unlock(&LOCK_start_thread);
 
@@ -2032,6 +2037,7 @@ static void clean_up(bool print_message)
   end_ssl();
 #ifndef EMBEDDED_LIBRARY
   vio_end();
+  listen_sockets.free_memory();
 #endif /*!EMBEDDED_LIBRARY*/
 #if defined(ENABLED_DEBUG_SYNC)
   /* End the debug sync facility. See debug_sync.cc. */

sql/service_wsrep.cc

@@ -32,6 +32,11 @@ extern "C" void wsrep_thd_LOCK(const THD *thd)
   mysql_mutex_lock(&thd->LOCK_thd_data);
 }
 
+extern "C" int wsrep_thd_TRYLOCK(const THD *thd)
+{
+  return mysql_mutex_trylock(&thd->LOCK_thd_data);
+}
+
 extern "C" void wsrep_thd_UNLOCK(const THD *thd)
 {
   mysql_mutex_unlock(&thd->LOCK_thd_data);
@@ -196,6 +201,7 @@ extern "C" void wsrep_handle_SR_rollback(THD *bf_thd,
 
   /* Note: do not store/reset globals before wsrep_bf_abort() call
      to avoid losing BF thd context. */
+  mysql_mutex_lock(&victim_thd->LOCK_thd_data);
   if (!(bf_thd && bf_thd != victim_thd))
   {
     DEBUG_SYNC(victim_thd, "wsrep_before_SR_rollback");
@@ -208,6 +214,7 @@ extern "C" void wsrep_handle_SR_rollback(THD *bf_thd,
   {
     wsrep_thd_self_abort(victim_thd);
   }
+  mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
   if (bf_thd)
   {
     wsrep_store_threadvars(bf_thd);
@@ -218,7 +225,7 @@ extern "C" my_bool wsrep_thd_bf_abort(THD *bf_thd, THD *victim_thd,
                                       my_bool signal)
 {
   mysql_mutex_assert_owner(&victim_thd->LOCK_thd_kill);
-  mysql_mutex_assert_not_owner(&victim_thd->LOCK_thd_data);
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_data);
   my_bool ret= wsrep_bf_abort(bf_thd, victim_thd);
   /*
     Send awake signal if victim was BF aborted or does not
@@ -227,19 +234,8 @@ extern "C" my_bool wsrep_thd_bf_abort(THD *bf_thd, THD *victim_thd,
    */
   if ((ret || !wsrep_on(victim_thd)) && signal)
   {
-    mysql_mutex_lock(&victim_thd->LOCK_thd_data);
-
-    if (victim_thd->wsrep_aborter && victim_thd->wsrep_aborter != bf_thd->thread_id)
-    {
-      WSREP_DEBUG("victim is killed already by %llu, skipping awake",
-                  victim_thd->wsrep_aborter);
-      mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
-      return false;
-    }
-
     victim_thd->wsrep_aborter= bf_thd->thread_id;
     victim_thd->awake_no_mutex(KILL_QUERY_HARD);
-    mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
   } else {
     WSREP_DEBUG("wsrep_thd_bf_abort skipped awake, signal %d", signal);
   }
@@ -368,25 +364,6 @@ extern "C" ulong wsrep_OSU_method_get(const MYSQL_THD thd)
     return(global_system_variables.wsrep_OSU_method);
 }
 
-extern "C" bool wsrep_thd_set_wsrep_aborter(THD *bf_thd, THD *victim_thd)
-{
-  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_data);
-  if (!bf_thd)
-  {
-    victim_thd->wsrep_aborter= 0;
-    WSREP_DEBUG("wsrep_thd_set_wsrep_aborter resetting wsrep_aborter");
-    return false;
-  }
-  if (victim_thd->wsrep_aborter && victim_thd->wsrep_aborter != bf_thd->thread_id)
-  {
-    return true;
-  }
-  victim_thd->wsrep_aborter= bf_thd->thread_id;
-  WSREP_DEBUG("wsrep_thd_set_wsrep_aborter setting wsrep_aborter %u",
-              victim_thd->wsrep_aborter);
-  return false;
-}
-
 extern "C" void wsrep_report_bf_lock_wait(const THD *thd,
                                           unsigned long long trx_id)
 {

sql/sql_class.cc

@@ -1293,6 +1293,11 @@ void THD::init()
   wsrep_affected_rows     = 0;
   m_wsrep_next_trx_id     = WSREP_UNDEFINED_TRX_ID;
   wsrep_aborter           = 0;
+  wsrep_abort_by_kill     = NOT_KILLED;
+  wsrep_abort_by_kill_err = 0;
+#ifndef DBUG_OFF
+  wsrep_killed_state      = 0;
+#endif /* DBUG_OFF */
   wsrep_desynced_backup_stage= false;
 #endif /* WITH_WSREP */
 
@@ -1641,6 +1646,13 @@ void THD::reset_for_reuse()
 #endif
 #ifdef WITH_WSREP
   wsrep_free_status(this);
+  wsrep_cs().reset_error();
+  wsrep_aborter= 0;
+  wsrep_abort_by_kill= NOT_KILLED;
+  wsrep_abort_by_kill_err= 0;
+#ifndef DBUG_OFF
+  wsrep_killed_state= 0;
+#endif /* DBUG_OFF */
 #endif /* WITH_WSREP */
 }
 
@@ -1897,7 +1909,9 @@ void THD::awake_no_mutex(killed_state state_to_set)
   }
 
   /* Interrupt target waiting inside a storage engine. */
-  if (state_to_set != NOT_KILLED  && !wsrep_is_bf_aborted(this))
+  if (state_to_set != NOT_KILLED &&
+      IF_WSREP(!wsrep_is_bf_aborted(this) && wsrep_abort_by_kill == NOT_KILLED,
+               true))
     ha_kill_query(this, thd_kill_level(this));
 
   abort_current_cond_wait(false);
@@ -2125,6 +2139,17 @@ void THD::reset_killed()
     mysql_mutex_unlock(&LOCK_thd_kill);
   }
 #ifdef WITH_WSREP
+  if (WSREP_NNULL(this))
+  {
+    if (wsrep_abort_by_kill != NOT_KILLED)
+    {
+      mysql_mutex_assert_not_owner(&LOCK_thd_kill);
+      mysql_mutex_lock(&LOCK_thd_kill);
+      wsrep_abort_by_kill= NOT_KILLED;
+      wsrep_abort_by_kill_err= 0;
+      mysql_mutex_unlock(&LOCK_thd_kill);
+    }
+  }
   mysql_mutex_assert_not_owner(&LOCK_thd_data);
   mysql_mutex_lock(&LOCK_thd_data);
   wsrep_aborter= 0;

sql/sql_class.h

@@ -5382,7 +5382,14 @@ public:
   bool                      wsrep_ignore_table;
   /* thread who has started kill for this THD protected by LOCK_thd_data*/
   my_thread_id              wsrep_aborter;
-
+  /* Kill signal used, if thread was killed by manual KILL. Protected by
+     LOCK_thd_kill. */
+  std::atomic<killed_state> wsrep_abort_by_kill;
+  /* */
+  struct err_info*          wsrep_abort_by_kill_err;
+#ifndef DBUG_OFF
+  int                       wsrep_killed_state;
+#endif /* DBUG_OFF */
   /* true if BF abort is observed in do_command() right after reading
   client's packet, and if the client has sent PS execute command. */
   bool                      wsrep_delayed_BF_abort;

sql/sql_limit.h

@@ -61,6 +61,15 @@ class Select_limit_counters
      with_ties= false;
    }
 
+   /* Send the first row, still honoring offset_limit_cnt */
+   void send_first_row()
+   {
+     /* Guard against overflow */
+     if ((select_limit_cnt= offset_limit_cnt +1 ) == 0)
+       select_limit_cnt= offset_limit_cnt;
+     // with_ties= false;   Remove // on merge to 10.6
+   }
+
    bool is_unlimited() const
    { return select_limit_cnt == HA_POS_ERROR; }
    /*

sql/sql_parse.cc

@@ -9229,23 +9229,20 @@ kill_one_thread(THD *thd, my_thread_id id, killed_state kill_signal, killed_type
         thd->security_ctx->user_matches(tmp->security_ctx))
 #endif /* WITH_WSREP */
     {
-#ifdef WITH_WSREP
-      DEBUG_SYNC(thd, "before_awake_no_mutex");
-      if (tmp->wsrep_aborter && tmp->wsrep_aborter != thd->thread_id)
       {
-        /* victim is in hit list already, bail out */
+#ifdef WITH_WSREP
-	WSREP_DEBUG("victim %lld has wsrep aborter: %lu, skipping awake()",
+        if (WSREP(tmp))
-		    id, tmp->wsrep_aborter);
+        {
-        error= 0;
+          error = wsrep_kill_thd(thd, tmp, kill_signal);
         }
         else
-#endif /* WITH_WSREP */
         {
-        WSREP_DEBUG("kill_one_thread victim: %lld wsrep_aborter %lu"
+#endif /* WITH_WSREP */
-                    " by signal %d",
-                    id, tmp->wsrep_aborter, kill_signal);
         tmp->awake_no_mutex(kill_signal);
         error= 0;
+#ifdef WITH_WSREP
+        }
+#endif /* WITH_WSREP */
       }
     }
     else
@@ -9368,18 +9365,6 @@ static
 void sql_kill(THD *thd, my_thread_id id, killed_state state, killed_type type)
 {
   uint error;
-#ifdef WITH_WSREP
-  if (WSREP(thd))
-  {
-    WSREP_DEBUG("sql_kill called");
-    if (thd->wsrep_applier)
-    {
-      WSREP_DEBUG("KILL in applying, bailing out here");
-      return;
-    }
-    WSREP_TO_ISOLATION_BEGIN(WSREP_MYSQL_DB, NULL, NULL)
-  }
-#endif /* WITH_WSREP */
   if (likely(!(error= kill_one_thread(thd, id, state, type))))
   {
     if (!thd->killed)
@@ -9389,11 +9374,6 @@ void sql_kill(THD *thd, my_thread_id id, killed_state state, killed_type type)
   }
   else
     my_error(error, MYF(0), id);
-#ifdef WITH_WSREP
-  return;
- wsrep_error_label:
-  my_error(ER_KILL_DENIED_ERROR, MYF(0), (long long) thd->thread_id);
-#endif /* WITH_WSREP */
 }
 
 
@@ -9402,18 +9382,6 @@ sql_kill_user(THD *thd, LEX_USER *user, killed_state state)
 {
   uint error;
   ha_rows rows;
-#ifdef WITH_WSREP
-  if (WSREP(thd))
-  {
-    WSREP_DEBUG("sql_kill_user called");
-    if (thd->wsrep_applier)
-    {
-      WSREP_DEBUG("KILL in applying, bailing out here");
-      return;
-    }
-    WSREP_TO_ISOLATION_BEGIN(WSREP_MYSQL_DB, NULL, NULL)
-  }
-#endif /* WITH_WSREP */
   switch (error= kill_threads_for_user(thd, user, state, &rows))
   {
   case 0:
@@ -9429,11 +9397,6 @@ sql_kill_user(THD *thd, LEX_USER *user, killed_state state)
   default:
     my_error(error, MYF(0));
   }
-#ifdef WITH_WSREP
-  return;
- wsrep_error_label:
-  my_error(ER_KILL_DENIED_ERROR, MYF(0), (long long) thd->thread_id);
-#endif /* WITH_WSREP */
 }
 
 

sql/sql_plugin_services.inl

@@ -151,6 +151,7 @@ static struct wsrep_service_st wsrep_handler = {
   wsrep_on,
   wsrep_prepare_key_for_innodb,
   wsrep_thd_LOCK,
+  wsrep_thd_TRYLOCK,
   wsrep_thd_UNLOCK,
   wsrep_thd_query,
   wsrep_thd_retry_counter,
@@ -179,7 +180,6 @@ static struct wsrep_service_st wsrep_handler = {
   wsrep_OSU_method_get,
   wsrep_thd_has_ignored_error,
   wsrep_thd_set_ignored_error,
-  wsrep_thd_set_wsrep_aborter,
   wsrep_report_bf_lock_wait,
   wsrep_thd_kill_LOCK,
   wsrep_thd_kill_UNLOCK,

sql/sql_select.cc

@@ -150,10 +150,10 @@ static void update_depend_map_for_order(JOIN *join, ORDER *order);
 static ORDER *remove_const(JOIN *join,ORDER *first_order,COND *cond,
 			   bool change_list, bool *simple_order);
 static int return_zero_rows(JOIN *join, select_result *res, 
-                            List<TABLE_LIST> &tables,
+                            List<TABLE_LIST> *tables,
-                            List<Item> &fields, bool send_row,
+                            List<Item> *fields, bool send_row,
                             ulonglong select_options, const char *info,
-                            Item *having, List<Item> &all_fields);
+                            Item *having, List<Item> *all_fields);
 static COND *build_equal_items(JOIN *join, COND *cond,
                                COND_EQUAL *inherited,
                                List<TABLE_LIST> *join_list,
@@ -1301,11 +1301,40 @@ int SELECT_LEX::vers_setup_conds(THD *thd, TABLE_LIST *tables)
   DBUG_RETURN(0);
 }
 
+
 /*****************************************************************************
   Check fields, find best join, do the select and output fields.
   mysql_select assumes that all tables are already opened
 *****************************************************************************/
 
+/*
+  Check if we have a field reference. If yes, we have to use
+  mixed_implicit_grouping.
+*/
+
+static bool check_list_for_field(List<Item> *items)
+{
+  List_iterator_fast <Item> select_it(*items);
+  Item *select_el;
+
+  while ((select_el= select_it++))
+  {
+    if (select_el->with_field())
+      return true;
+  }
+  return false;
+}
+
+static bool check_list_for_field(ORDER *order)
+{
+  for (; order; order= order->next)
+  {
+    if (order->item[0]->with_field())
+      return true;
+  }
+  return false;
+}
+
 
 /**
   Prepare of whole select (including sub queries in future).
@@ -1387,52 +1416,43 @@ JOIN::prepare(TABLE_LIST *tables_init, COND *conds_init, uint og_num,
     DBUG_RETURN(-1);
 
   /*
-    TRUE if the SELECT list mixes elements with and without grouping,
+    mixed_implicit_grouping will be set to TRUE if the SELECT list
-    and there is no GROUP BY clause. Mixing non-aggregated fields with
+    mixes elements with and without grouping, and there is no GROUP BY
-    aggregate functions in the SELECT list is a MySQL extenstion that
+    clause.
-    is allowed only if the ONLY_FULL_GROUP_BY sql mode is not set.
+    Mixing non-aggregated fields with aggregate functions in the
+    SELECT list or HAVING is a MySQL extension that is allowed only if
+    the ONLY_FULL_GROUP_BY sql mode is not set.
   */
   mixed_implicit_grouping= false;
   if ((~thd->variables.sql_mode & MODE_ONLY_FULL_GROUP_BY) &&
       select_lex->with_sum_func && !group_list)
   {
-    List_iterator_fast <Item> select_it(fields_list);
+    if (check_list_for_field(&fields_list)  ||
-    Item *select_el; /* Element of the SELECT clause, can be an expression. */
+        check_list_for_field(order))
-    bool found_field_elem= false;
-    bool found_sum_func_elem= false;
-
-    while ((select_el= select_it++))
     {
-      if (select_el->with_sum_func())
-        found_sum_func_elem= true;
-      if (select_el->with_field())
-        found_field_elem= true;
-      if (found_sum_func_elem && found_field_elem)
-      {
-        mixed_implicit_grouping= true;
-        break;
-      }
-    }
-  }
-
-  table_count= select_lex->leaf_tables.elements;
-
-  TABLE_LIST *tbl;
       List_iterator_fast<TABLE_LIST> li(select_lex->leaf_tables);
-  while ((tbl= li++))
+
+      mixed_implicit_grouping= true;            // mark for future
+
+      while (TABLE_LIST *tbl= li++)
       {
         /*
-      If the query uses implicit grouping where the select list contains both
+          If the query uses implicit grouping where the select list
-      aggregate functions and non-aggregate fields, any non-aggregated field
+          contains both aggregate functions and non-aggregate fields,
-      may produce a NULL value. Set all fields of each table as nullable before
+          any non-aggregated field may produce a NULL value. Set all
-      semantic analysis to take into account this change of nullability.
+          fields of each table as nullable before semantic analysis to
+          take into account this change of nullability.
 
-      Note: this loop doesn't touch tables inside merged semi-joins, because
+          Note: this loop doesn't touch tables inside merged
-      subquery-to-semijoin conversion has not been done yet. This is intended.
+          semi-joins, because subquery-to-semijoin conversion has not
+          been done yet. This is intended.
         */
-    if (mixed_implicit_grouping && tbl->table)
+        if (tbl->table)
           tbl->table->maybe_null= 1;
       }
+    }
+  }
+  table_count= select_lex->leaf_tables.elements;
 
   uint real_og_num= og_num;
   if (skip_order_by && 
@@ -1453,7 +1473,7 @@ JOIN::prepare(TABLE_LIST *tables_init, COND *conds_init, uint og_num,
 
   {
     List_iterator_fast<TABLE_LIST> it(select_lex->leaf_tables);
-    while ((tbl= it++))
+    while (TABLE_LIST *tbl= it++)
     {
       if (tbl->table_function &&
           tbl->table_function->setup(thd, tbl, select_lex_arg))
@@ -4074,7 +4094,7 @@ bool JOIN::make_aggr_tables_info()
   set_items_ref_array(items0);
   if (join_tab)
     join_tab[exec_join_tab_cnt() + aggr_tables - 1].next_select=
-      setup_end_select_func(this, NULL);
+      setup_end_select_func(this);
   group= has_group_by;
 
   DBUG_RETURN(false);
@@ -4469,13 +4489,7 @@ JOIN::reinit()
     }
   }
 
-  /* Reset of sum functions */
+  clear_sum_funcs();
-  if (sum_funcs)
-  {
-    Item_sum *func, **func_ptr= sum_funcs;
-    while ((func= *(func_ptr++)))
-      func->clear();
-  }
 
   if (no_rows_in_result_called)
   {
@@ -4758,12 +4772,12 @@ void JOIN::exec_inner()
     }
     else
     {
-      (void) return_zero_rows(this, result, select_lex->leaf_tables,
+      (void) return_zero_rows(this, result, &select_lex->leaf_tables,
-                              *columns_list,
+                              columns_list,
 			      send_row_on_empty_set(),
 			      select_options,
 			      zero_result_cause,
-			      having ? having : tmp_having, all_fields);
+			      having ? having : tmp_having, &all_fields);
       DBUG_VOID_RETURN;
     }
   }
@@ -15641,10 +15655,36 @@ ORDER *simple_remove_const(ORDER *order, COND *where)
 }
 
 
+/*
+  Set all fields in the table to have a null value
+
+  @param tables            Table list
+*/
+
+static void make_tables_null_complemented(List<TABLE_LIST> *tables)
+{
+  List_iterator<TABLE_LIST> ti(*tables);
+  TABLE_LIST *table;
+  while ((table= ti++))
+  {
+    /*
+      Don't touch semi-join materialization tables, as the a join_free()
+      call may have freed them (and HAVING clause can't have references to
+      them anyway).
+    */
+    if (!table->is_jtbm())
+    {
+      TABLE *tbl= table->table;
+      mark_as_null_row(tbl);		// Set fields to NULL
+    }
+  }
+}
+
+
 static int
-return_zero_rows(JOIN *join, select_result *result, List<TABLE_LIST> &tables,
+return_zero_rows(JOIN *join, select_result *result, List<TABLE_LIST> *tables,
-		 List<Item> &fields, bool send_row, ulonglong select_options,
+		 List<Item> *fields, bool send_row, ulonglong select_options,
-		 const char *info, Item *having, List<Item> &all_fields)
+		 const char *info, Item *having, List<Item> *all_fields)
 {
   DBUG_ENTER("return_zero_rows");
 
@@ -15660,24 +15700,15 @@ return_zero_rows(JOIN *join, select_result *result, List<TABLE_LIST> &tables,
       Set all tables to have NULL row. This is needed as we will be evaluating
       HAVING condition.
     */
-    List_iterator<TABLE_LIST> ti(tables);
+    make_tables_null_complemented(tables);
-    TABLE_LIST *table;
+
-    while ((table= ti++))
+    List_iterator_fast<Item> it(*all_fields);
-    {
-      /*
-        Don't touch semi-join materialization tables, as the above join_free()
-        call has freed them (and HAVING clause can't have references to them 
-        anyway).
-      */
-      if (!table->is_jtbm())
-        mark_as_null_row(table->table);		// All fields are NULL
-    }
-    List_iterator_fast<Item> it(all_fields);
     Item *item;
     /*
       Inform all items (especially aggregating) to calculate HAVING correctly,
       also we will need it for sending results.
     */
+    join->no_rows_in_result_called= 1;
     while ((item= it++))
       item->no_rows_in_result();
     if (having && having->val_int() == 0)
@@ -15691,12 +15722,12 @@ return_zero_rows(JOIN *join, select_result *result, List<TABLE_LIST> &tables,
     join->thd->limit_found_rows= 0;
   }
 
-  if (!(result->send_result_set_metadata(fields,
+  if (!(result->send_result_set_metadata(*fields,
                               Protocol::SEND_NUM_ROWS | Protocol::SEND_EOF)))
   {
     bool send_error= FALSE;
     if (send_row)
-      send_error= result->send_data_with_check(fields, join->unit, 0) > 0;
+      send_error= result->send_data_with_check(*fields, join->unit, 0) > 0;
     if (likely(!send_error))
       result->send_eof();				// Should be safe
   }
@@ -15712,31 +15743,25 @@ return_zero_rows(JOIN *join, select_result *result, List<TABLE_LIST> &tables,
 }
 
 /**
-  used only in JOIN::clear (always) and in do_select()
+  Reset table rows to contain a null-complement row (all fields are null)
-  (if there where no matching rows)
+
+  Used only in JOIN::clear() and in do_select() if there where no matching rows.
 
   @param join            JOIN
-  @param cleared_tables  If not null, clear also const tables and mark all
+  @param cleared_tables  Used to mark all cleared tables in the map. Needed for
-                         cleared tables in the map. cleared_tables is only
+                         unclear_tables() to know which tables to restore to
-                         set when called from do_select() when there is a
+                         their original state.
-                         group function and there where no matching rows.
 */
 
 static void clear_tables(JOIN *join, table_map *cleared_tables)
 {
-  /* 
+  DBUG_ASSERT(cleared_tables);
-    must clear only the non-const tables as const tables are not re-calculated.
-  */
   for (uint i= 0 ; i < join->table_count ; i++)
   {
     TABLE *table= join->table[i];
 
     if (table->null_row)
       continue;                                 // Nothing more to do
-    if (!(table->map & join->const_table_map) || cleared_tables)
-    {
-      if (cleared_tables)
-      {
     (*cleared_tables)|= (((table_map) 1) << i);
     if (table->s->null_bytes)
     {
@@ -15746,15 +15771,14 @@ static void clear_tables(JOIN *join, table_map *cleared_tables)
       */
       memcpy(table->record[1], table->null_flags, table->s->null_bytes);
     }
-      }
     mark_as_null_row(table);                  // All fields are NULL
   }
 }
-}
 
 
 /**
    Reverse null marking for tables and restore null bits.
+   This return the tables to the state of before clear_tables().
 
    We have to do this because the tables may be re-used in a sub query
    and the subquery will assume that the const tables contains the original
@@ -21493,9 +21517,9 @@ void set_postjoin_aggr_write_func(JOIN_TAB *tab)
     end_select function to use. This function can't fail.
 */
 
-Next_select_func setup_end_select_func(JOIN *join, JOIN_TAB *tab)
+Next_select_func setup_end_select_func(JOIN *join)
 {
-  TMP_TABLE_PARAM *tmp_tbl= tab ? tab->tmp_table_param : &join->tmp_table_param;
+  TMP_TABLE_PARAM *tmp_tbl= &join->tmp_table_param;
 
   /* 
      Choose method for presenting result to user. Use end_send_group
@@ -21565,7 +21589,7 @@ do_select(JOIN *join, Procedure *procedure)
   join->duplicate_rows= join->send_records=0;
   if (join->only_const_tables() && !join->need_tmp)
   {
-    Next_select_func end_select= setup_end_select_func(join, NULL);
+    Next_select_func end_select= setup_end_select_func(join);
 
     /*
       HAVING will be checked after processing aggregate functions,
@@ -22040,6 +22064,7 @@ sub_select(JOIN *join,JOIN_TAB *join_tab,bool end_of_records)
     }
   }
 
+  /* Restore state if mark_as_null_row() have been called */
   if (join_tab->last_inner)
   {
     JOIN_TAB *last_inner_tab= join_tab->last_inner;
@@ -23452,11 +23477,18 @@ end_send_group(JOIN *join, JOIN_TAB *join_tab, bool end_of_records)
 {
   int idx= -1;
   enum_nested_loop_state ok_code= NESTED_LOOP_OK;
+  /*
+    join_tab can be 0 in the case all tables are const tables and we did not
+    need a temporary table to store the result.
+    In this case we use the original given fields, which is stored in
+    join->fields.
+  */
   List<Item> *fields= join_tab ? (join_tab-1)->fields : join->fields;
   DBUG_ENTER("end_send_group");
 
   if (!join->items3.is_null() && !join->set_group_rpa)
   {
+    /* Move ref_pointer_array to points to items3 */
     join->set_group_rpa= true;
     join->set_items_ref_array(join->items3);
   }
@@ -23464,10 +23496,12 @@ end_send_group(JOIN *join, JOIN_TAB *join_tab, bool end_of_records)
   if (!join->first_record || end_of_records ||
       (idx=test_if_group_changed(join->group_fields)) >= 0)
   {
+
     if (!join->group_sent &&
         (join->first_record ||
          (end_of_records && !join->group && !join->group_optimized_away)))
     {
+      table_map cleared_tables= (table_map) 0;
       if (join->procedure)
 	join->procedure->end_group();
       /* Test if there was a group change. */
@@ -23492,11 +23526,13 @@ end_send_group(JOIN *join, JOIN_TAB *join_tab, bool end_of_records)
           /* Reset all sum functions on group change. */
 	  if (!join->first_record)
 	  {
-            List_iterator_fast<Item> it(*join->fields);
-            Item *item;
             /* No matching rows for group function */
-            join->clear();
 
+            List_iterator_fast<Item> it(*fields);
+            Item *item;
+            join->no_rows_in_result_called= 1;
+
+            join->clear(&cleared_tables);
             while ((item= it++))
               item->no_rows_in_result();
 	  }
@@ -23524,6 +23560,13 @@ end_send_group(JOIN *join, JOIN_TAB *join_tab, bool end_of_records)
 	    if (join->rollup_send_data((uint) (idx+1)))
 	      error= 1;
 	  }
+          if (join->no_rows_in_result_called)
+          {
+            /* Restore null tables to original state */
+            join->no_rows_in_result_called= 0;
+            if (cleared_tables)
+              unclear_tables(join, &cleared_tables);
+          }
         }
 	if (unlikely(error > 0))
           DBUG_RETURN(NESTED_LOOP_ERROR);        /* purecov: inspected */
@@ -23841,6 +23884,7 @@ end_write_group(JOIN *join, JOIN_TAB *join_tab __attribute__((unused)),
   {
     if (join->first_record || (end_of_records && !join->group))
     {
+      table_map cleared_tables= (table_map) 0;
       if (join->procedure)
 	join->procedure->end_group();
       int send_group_parts= join->send_group_parts;
@@ -23849,7 +23893,7 @@ end_write_group(JOIN *join, JOIN_TAB *join_tab __attribute__((unused)),
         if (!join->first_record)
         {
           /* No matching rows for group function */
-          join->clear();
+          join->clear(&cleared_tables);
         }
         copy_sum_funcs(join->sum_funcs,
                        join->sum_funcs_end[send_group_parts]);
@@ -23872,6 +23916,8 @@ end_write_group(JOIN *join, JOIN_TAB *join_tab __attribute__((unused)),
 	    DBUG_RETURN(NESTED_LOOP_ERROR);
           }
 	}
+        if (cleared_tables)
+          unclear_tables(join, &cleared_tables);
 	if (end_of_records)
 	  goto end;
       }
@@ -25603,7 +25649,7 @@ JOIN_TAB::remove_duplicates()
       !(join->select_options & OPTION_FOUND_ROWS))
   {
     // only const items with no OPTION_FOUND_ROWS
-    join->unit->lim.set_single_row();		// Only send first row
+    join->unit->lim.send_first_row();		// Only send first row
     my_free(sortorder);
     DBUG_RETURN(false);
   }
@@ -28007,11 +28053,8 @@ int JOIN::rollup_write_data(uint idx, TMP_TABLE_PARAM *tmp_table_param_arg,
   (end_send_group/end_write_group)
 */
 
-void JOIN::clear()
+void inline JOIN::clear_sum_funcs()
 {
-  clear_tables(this, 0);
-  copy_fields(&tmp_table_param);
-
   if (sum_funcs)
   {
     Item_sum *func, **func_ptr= sum_funcs;
@@ -28021,6 +28064,22 @@ void JOIN::clear()
 }
 
 
+/*
+  Prepare for returning 'empty row' when there is no matching row.
+
+  - Mark all tables with mark_as_null_row()
+  - Make a copy of of all simple SELECT items
+  - Reset all sum functions to NULL or 0.
+*/
+
+void JOIN::clear(table_map *cleared_tables)
+{
+  clear_tables(this, cleared_tables);
+  copy_fields(&tmp_table_param);
+  clear_sum_funcs();
+}
+
+
 /**
   Print an EXPLAIN line with all NULLs and given message in the 'Extra' column
 

sql/sql_select.h

@@ -227,7 +227,7 @@ enum sj_strategy_enum
 
 typedef enum_nested_loop_state
 (*Next_select_func)(JOIN *, struct st_join_table *, bool);
-Next_select_func setup_end_select_func(JOIN *join, JOIN_TAB *tab);
+Next_select_func setup_end_select_func(JOIN *join);
 int rr_sequential(READ_RECORD *info);
 int read_record_func_for_rr_and_unpack(READ_RECORD *info);
 Item *remove_pushed_top_conjuncts(THD *thd, Item *cond);
@@ -1722,7 +1722,8 @@ public:
   void join_free();
   /** Cleanup this JOIN, possibly for reuse */
   void cleanup(bool full);
-  void clear();
+  void clear(table_map *cleared_tables);
+  void inline clear_sum_funcs();
   bool send_row_on_empty_set()
   {
     return (do_send_rows && implicit_grouping && !group_optimized_away &&

sql/sql_type.h

@@ -7574,8 +7574,9 @@ extern Named_type_handler<Type_handler_time>        type_handler_time;
 extern Named_type_handler<Type_handler_time2>       type_handler_time2;
 extern Named_type_handler<Type_handler_datetime>    type_handler_datetime;
 extern Named_type_handler<Type_handler_datetime2>   type_handler_datetime2;
-extern Named_type_handler<Type_handler_timestamp>   type_handler_timestamp;
+
-extern Named_type_handler<Type_handler_timestamp2>  type_handler_timestamp2;
+extern MYSQL_PLUGIN_IMPORT Named_type_handler<Type_handler_timestamp>   type_handler_timestamp;
+extern MYSQL_PLUGIN_IMPORT Named_type_handler<Type_handler_timestamp2>  type_handler_timestamp2;
 
 extern Type_handler_interval_DDhhmmssff type_handler_interval_DDhhmmssff;
 

sql/table.h

@@ -3376,10 +3376,16 @@ inline void mark_as_null_row(TABLE *table)
     bfill(table->null_flags,table->s->null_bytes,255);
 }
 
+/*
+  Restore table to state before mark_as_null_row() call.
+  This assumes that the caller has restored table->null_flags,
+  as is done in unclear_tables().
+*/
+
 inline void unmark_as_null_row(TABLE *table)
 {
   table->null_row= 0;
-  table->status= STATUS_NO_RECORD;
+  table->status&= ~STATUS_NULL_ROW;
 }
 
 bool is_simple_order(ORDER *order);

sql/wsrep_dummy.cc

@@ -56,6 +56,11 @@ my_bool wsrep_on(const THD *)
 void wsrep_thd_LOCK(const THD *)
 { }
 
+int wsrep_thd_TRYLOCK(const THD *)
+{
+  return 0;
+}
+
 void wsrep_thd_UNLOCK(const THD *)
 { }
 
@@ -154,8 +159,6 @@ void wsrep_thd_set_ignored_error(THD*, my_bool)
 { }
 ulong wsrep_OSU_method_get(const THD*)
 { return 0;}
-bool wsrep_thd_set_wsrep_aborter(THD*, THD*)
-{ return 0;}
 
 void wsrep_report_bf_lock_wait(const THD*,
                                unsigned long long)

sql/wsrep_high_priority_service.cc

@@ -510,6 +510,7 @@ int Wsrep_high_priority_service::log_dummy_write_set(const wsrep::ws_handle& ws_
       m_thd->wait_for_prior_commit();
     }
 
+    WSREP_DEBUG("checkpointing dummy write set %lld", ws_meta.seqno().get());
     wsrep_set_SE_checkpoint(ws_meta.gtid(), wsrep_gtid_server.gtid());
 
     if (!WSREP_EMULATE_BINLOG(m_thd))

sql/wsrep_mysqld.cc

@@ -52,6 +52,7 @@
 #include "log_event.h"
 #include "sql_connect.h"
 #include "thread_cache.h"
+#include "debug_sync.h"
 
 #include <sstream>
 
@@ -3107,6 +3108,20 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
                   request_thd, granted_thd);
     ticket->wsrep_report(wsrep_debug);
 
+    DEBUG_SYNC(request_thd, "before_wsrep_thd_abort");
+    DBUG_EXECUTE_IF("sync.before_wsrep_thd_abort", {
+      const char act[]= "now "
+                        "SIGNAL sync.before_wsrep_thd_abort_reached "
+                        "WAIT_FOR signal.before_wsrep_thd_abort";
+      DBUG_ASSERT(!debug_sync_set_action(request_thd, STRING_WITH_LEN(act)));
+    };);
+
+    /* Here we will call wsrep_abort_transaction so we should hold
+    THD::LOCK_thd_data to protect victim from concurrent usage
+    and THD::LOCK_thd_kill to protect from disconnect or delete.
+
+    */
+    mysql_mutex_lock(&granted_thd->LOCK_thd_kill);
     mysql_mutex_lock(&granted_thd->LOCK_thd_data);
 
     if (wsrep_thd_is_toi(granted_thd) ||
@@ -3118,13 +3133,11 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
                     wsrep_thd_query(request_thd));
         THD_STAGE_INFO(request_thd, stage_waiting_isolation);
         ticket->wsrep_report(wsrep_debug);
-        mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
       }
       else if (wsrep_thd_is_SR(granted_thd) && !wsrep_thd_is_SR(request_thd))
       {
         WSREP_MDL_LOG(INFO, "MDL conflict, DDL vs SR",
                       schema, schema_len, request_thd, granted_thd);
-        mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
         WSREP_DEBUG("wsrep_handle_mdl_conflict DDL vs SR for %s",
                     wsrep_thd_query(request_thd));
         THD_STAGE_INFO(request_thd, stage_waiting_isolation);
@@ -3136,6 +3149,7 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
                       request_thd, granted_thd);
         ticket->wsrep_report(true);
         mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
+        mysql_mutex_unlock(&granted_thd->LOCK_thd_kill);
         unireg_abort(1);
       }
     }
@@ -3146,7 +3160,6 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
                   wsrep_thd_query(request_thd));
       THD_STAGE_INFO(request_thd, stage_waiting_ddl);
       ticket->wsrep_report(wsrep_debug);
-      mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
       if (granted_thd->current_backup_stage != BACKUP_FINISHED &&
 	  wsrep_check_mode(WSREP_MODE_BF_MARIABACKUP))
       {
@@ -3160,7 +3173,6 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
                   wsrep_thd_query(request_thd));
       THD_STAGE_INFO(request_thd, stage_waiting_isolation);
       ticket->wsrep_report(wsrep_debug);
-      mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
       wsrep_abort_thd(request_thd, granted_thd, 1);
     }
     else
@@ -3174,7 +3186,6 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
 
       if (granted_thd->wsrep_trx().active())
       {
-        mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
         wsrep_abort_thd(request_thd, granted_thd, 1);
       }
       else
@@ -3183,10 +3194,9 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
           Granted_thd is likely executing with wsrep_on=0. If the requesting
           thd is BF, BF abort and wait.
         */
-        mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
-
         if (wsrep_thd_is_BF(request_thd, FALSE))
         {
+          granted_thd->awake_no_mutex(KILL_QUERY_HARD);
           ha_abort_transaction(request_thd, granted_thd, TRUE);
         }
         else
@@ -3195,10 +3205,14 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
                         schema, schema_len,
                         request_thd, granted_thd);
 	  ticket->wsrep_report(true);
+          mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
+          mysql_mutex_unlock(&granted_thd->LOCK_thd_kill);
 	  unireg_abort(1);
         }
       }
     }
+    mysql_mutex_unlock(&granted_thd->LOCK_thd_data);
+    mysql_mutex_unlock(&granted_thd->LOCK_thd_kill);
   }
   else
   {
@@ -3210,13 +3224,17 @@ void wsrep_handle_mdl_conflict(MDL_context *requestor_ctx,
 static bool abort_replicated(THD *thd)
 {
   bool ret_code= false;
+  mysql_mutex_lock(&thd->LOCK_thd_kill);
+  mysql_mutex_lock(&thd->LOCK_thd_data);
   if (thd->wsrep_trx().state() == wsrep::transaction::s_committing)
   {
     WSREP_DEBUG("aborting replicated trx: %llu", (ulonglong)(thd->real_id));
 
-    (void)wsrep_abort_thd(thd, thd, TRUE);
+    wsrep_abort_thd(thd, thd, TRUE);
     ret_code= true;
   }
+  mysql_mutex_unlock(&thd->LOCK_thd_data);
+  mysql_mutex_unlock(&thd->LOCK_thd_kill);
   return ret_code;
 }
 

sql/wsrep_server_service.cc

@@ -148,9 +148,13 @@ void Wsrep_server_service::release_high_priority_service(wsrep::high_priority_se
   wsrep_delete_threadvars();
 }
 
-void Wsrep_server_service::background_rollback(wsrep::client_state& client_state)
+void Wsrep_server_service::background_rollback(
+    wsrep::unique_lock<wsrep::mutex> &lock WSREP_UNUSED,
+    wsrep::client_state &client_state)
 {
+  DBUG_ASSERT(lock.owns_lock());
   Wsrep_client_state &cs= static_cast<Wsrep_client_state &>(client_state);
+  mysql_mutex_assert_owner(&cs.thd()->LOCK_thd_data);
   wsrep_fire_rollbacker(cs.thd());
 }
 

sql/wsrep_server_service.h

@@ -46,7 +46,8 @@ public:
 
   void release_high_priority_service(wsrep::high_priority_service*);
 
-  void background_rollback(wsrep::client_state&);
+  void background_rollback(wsrep::unique_lock<wsrep::mutex> &,
+                           wsrep::client_state &);
 
   void bootstrap();
   void log_message(enum wsrep::log::level, const char*);

sql/wsrep_thd.cc

@@ -307,48 +307,9 @@ void wsrep_fire_rollbacker(THD *thd)
   }
 }
 
-
+static bool wsrep_bf_abort_low(THD *bf_thd, THD *victim_thd)
-int wsrep_abort_thd(THD *bf_thd,
-                    THD *victim_thd,
-                    my_bool signal)
 {
-  DBUG_ENTER("wsrep_abort_thd");
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_data);
-
-  mysql_mutex_lock(&victim_thd->LOCK_thd_data);
-
-  /* Note that when you use RSU node is desynced from cluster, thus WSREP(thd)
-  might not be true.
-  */
-  if ((WSREP_NNULL(bf_thd) ||
-       ((WSREP_ON || bf_thd->variables.wsrep_OSU_method == WSREP_OSU_RSU) &&
-	 wsrep_thd_is_toi(bf_thd))) &&
-       !wsrep_thd_is_aborting(victim_thd))
-  {
-      WSREP_DEBUG("wsrep_abort_thd, by: %llu, victim: %llu",
-                  (long long)bf_thd->real_id, (long long)victim_thd->real_id);
-      mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
-      ha_abort_transaction(bf_thd, victim_thd, signal);
-      DBUG_RETURN(1);
-  }
-  else
-  {
-    WSREP_DEBUG("wsrep_abort_thd not effective: bf %llu victim %llu "
-                "wsrep %d wsrep_on %d RSU %d TOI %d aborting %d",
-                (long long)bf_thd->real_id, (long long)victim_thd->real_id,
-                WSREP_NNULL(bf_thd), WSREP_ON,
-                bf_thd->variables.wsrep_OSU_method == WSREP_OSU_RSU,
-                wsrep_thd_is_toi(bf_thd),
-                wsrep_thd_is_aborting(victim_thd));
-  }
-
-  mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
-  DBUG_RETURN(1);
-}
-
-bool wsrep_bf_abort(THD* bf_thd, THD* victim_thd)
-{
-  WSREP_LOG_THD(bf_thd, "BF aborter before");
-  WSREP_LOG_THD(victim_thd, "victim before");
 
 #ifdef ENABLED_DEBUG_SYNC
   DBUG_EXECUTE_IF("sync.wsrep_bf_abort",
@@ -362,6 +323,85 @@ bool wsrep_bf_abort(THD* bf_thd, THD* victim_thd)
                   };);
 #endif
 
+  wsrep::seqno bf_seqno(bf_thd->wsrep_trx().ws_meta().seqno());
+  bool ret;
+
+  {
+    /* Adopt the lock, it is being held by the caller. */
+    Wsrep_mutex wsm{&victim_thd->LOCK_thd_data};
+    wsrep::unique_lock<wsrep::mutex> lock{wsm, std::adopt_lock};
+
+    if (wsrep_thd_is_toi(bf_thd))
+    {
+      ret= victim_thd->wsrep_cs().total_order_bf_abort(lock, bf_seqno);
+    }
+    else
+    {
+      DBUG_ASSERT(WSREP(victim_thd) ? victim_thd->wsrep_trx().active() : 1);
+      ret= victim_thd->wsrep_cs().bf_abort(lock, bf_seqno);
+    }
+    if (ret)
+    {
+      /* BF abort should be allowed only once by wsrep-lib.*/
+      DBUG_ASSERT(victim_thd->wsrep_aborter == 0);
+      victim_thd->wsrep_aborter= bf_thd->thread_id;
+      wsrep_bf_aborts_counter++;
+    }
+    lock.release(); /* No unlock at the end of the scope. */
+  }
+
+  /* Sanity check for wsrep-lib calls to return with LOCK_thd_data held. */
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_data);
+
+  return ret;
+}
+
+void wsrep_abort_thd(THD *bf_thd,
+                    THD *victim_thd,
+                    my_bool signal)
+{
+  DBUG_ENTER("wsrep_abort_thd");
+
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_kill);
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_data);
+
+  /* Note that when you use RSU node is desynced from cluster, thus WSREP(thd)
+  might not be true.
+  */
+  if ((WSREP(bf_thd)
+       || ((WSREP_ON || bf_thd->variables.wsrep_OSU_method == WSREP_OSU_RSU)
+           &&  wsrep_thd_is_toi(bf_thd))
+       || bf_thd->lex->sql_command == SQLCOM_KILL)
+      && !wsrep_thd_is_aborting(victim_thd) &&
+      wsrep_bf_abort_low(bf_thd, victim_thd) &&
+      !victim_thd->wsrep_cs().is_rollbacker_active())
+  {
+    WSREP_DEBUG("wsrep_abort_thd, by: %llu, victim: %llu",
+                (long long)bf_thd->real_id, (long long)victim_thd->real_id);
+    victim_thd->awake_no_mutex(KILL_QUERY_HARD);
+    ha_abort_transaction(bf_thd, victim_thd, signal);
+  }
+  else
+  {
+    WSREP_DEBUG("wsrep_abort_thd not effective: bf %llu victim %llu "
+                "wsrep %d wsrep_on %d RSU %d TOI %d aborting %d",
+                (long long)bf_thd->real_id, (long long)victim_thd->real_id,
+                WSREP_NNULL(bf_thd), WSREP_ON,
+                bf_thd->variables.wsrep_OSU_method == WSREP_OSU_RSU,
+                wsrep_thd_is_toi(bf_thd),
+                wsrep_thd_is_aborting(victim_thd));
+  }
+
+  DBUG_VOID_RETURN;
+}
+
+bool wsrep_bf_abort(THD* bf_thd, THD* victim_thd)
+{
+  WSREP_LOG_THD(bf_thd, "BF aborter before");
+  WSREP_LOG_THD(victim_thd, "victim before");
+
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_data);
+
   if (WSREP(victim_thd) && !victim_thd->wsrep_trx().active())
   {
     WSREP_DEBUG("wsrep_bf_abort, BF abort for non active transaction."
@@ -384,30 +424,84 @@ bool wsrep_bf_abort(THD* bf_thd, THD* victim_thd)
         wsrep_check_mode(WSREP_MODE_BF_MARIABACKUP))
     {
       WSREP_DEBUG("killing connection for non wsrep session");
-      mysql_mutex_lock(&victim_thd->LOCK_thd_data);
       victim_thd->awake_no_mutex(KILL_CONNECTION);
-      mysql_mutex_unlock(&victim_thd->LOCK_thd_data);
     }
     return false;
   }
 
-  bool ret;
+  return wsrep_bf_abort_low(bf_thd, victim_thd);
-  wsrep::seqno bf_seqno(bf_thd->wsrep_trx().ws_meta().seqno());
+}
 
-  if (wsrep_thd_is_toi(bf_thd))
+uint wsrep_kill_thd(THD *thd, THD *victim_thd, killed_state kill_signal)
 {
-    ret= victim_thd->wsrep_cs().total_order_bf_abort(bf_seqno);
+  DBUG_ENTER("wsrep_kill_thd");
-  }
+  DBUG_ASSERT(WSREP(victim_thd));
-  else
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_kill);
+  mysql_mutex_assert_owner(&victim_thd->LOCK_thd_data);
+  using trans= wsrep::transaction;
+  auto trx_state= victim_thd->wsrep_trx().state();
+#ifndef DBUG_OFF
+  victim_thd->wsrep_killed_state= trx_state;
+#endif /* DBUG_OFF */
+  /*
+    Already killed or in commit codepath. Mark the victim as killed,
+    the killed status will be restored in wsrep_after_commit() and
+    will be processed after the commit is over. In case of multiple
+    KILLs happened on commit codepath, the last one will be effective.
+  */
+  if (victim_thd->wsrep_abort_by_kill ||
+      trx_state == trans::s_preparing ||
+      trx_state == trans::s_committing ||
+      trx_state == trans::s_ordered_commit)
   {
-    DBUG_ASSERT(WSREP(victim_thd) ? victim_thd->wsrep_trx().active() : 1);
+    victim_thd->wsrep_abort_by_kill= kill_signal;
-    ret= victim_thd->wsrep_cs().bf_abort(bf_seqno);
+    DBUG_RETURN(0);
   }
-  if (ret)
+  /*
+    Mark killed victim_thd with kill_signal so that awake_no_mutex does
+    not dive into storage engine. We use ha_abort_transaction()
+    to do the storage engine part for wsrep THDs.
+  */
+  DEBUG_SYNC(thd, "wsrep_kill_before_awake_no_mutex");
+  victim_thd->wsrep_abort_by_kill= kill_signal;
+  victim_thd->awake_no_mutex(kill_signal);
+  /* ha_abort_transaction() releases tmp->LOCK_thd_kill, so tmp
+     is not safe to access anymore. */
+  ha_abort_transaction(thd, victim_thd, 1);
+  DBUG_RETURN(0);
+}
+
+void wsrep_backup_kill_for_commit(THD *thd)
 {
-    wsrep_bf_aborts_counter++;
+  DBUG_ASSERT(WSREP(thd));
+  mysql_mutex_assert_owner(&thd->LOCK_thd_kill);
+  DBUG_ASSERT(thd->killed != NOT_KILLED);
+  mysql_mutex_lock(&thd->LOCK_thd_data);
+  /* If the transaction will roll back, keep the killed state.
+     For must replay, the replay will happen in different THD context
+     which is high priority and cannot be killed. The owning thread will
+     pick the killed state in after statement processing. */
+  if (thd->wsrep_trx().state() != wsrep::transaction::s_cert_failed &&
+      thd->wsrep_trx().state() != wsrep::transaction::s_must_abort &&
+      thd->wsrep_trx().state() != wsrep::transaction::s_aborting &&
+      thd->wsrep_trx().state() != wsrep::transaction::s_must_replay)
+  {
+    thd->wsrep_abort_by_kill= thd->killed;
+    thd->wsrep_abort_by_kill_err= thd->killed_err;
+    thd->killed= NOT_KILLED;
+    thd->killed_err= 0;
   }
-  return ret;
+  mysql_mutex_unlock(&thd->LOCK_thd_data);
+}
+
+void wsrep_restore_kill_after_commit(THD *thd)
+{
+  DBUG_ASSERT(WSREP(thd));
+  mysql_mutex_assert_owner(&thd->LOCK_thd_kill);
+  thd->killed= thd->wsrep_abort_by_kill;
+  thd->killed_err= thd->wsrep_abort_by_kill_err;
+  thd->wsrep_abort_by_kill= NOT_KILLED;
+  thd->wsrep_abort_by_kill_err= 0;
 }
 
 int wsrep_create_threadvars()

sql/wsrep_thd.h

@@ -88,10 +88,39 @@ bool wsrep_create_appliers(long threads, bool mutex_protected=false);
 void wsrep_create_rollbacker();
 
 bool wsrep_bf_abort(THD* bf_thd, THD* victim_thd);
-int  wsrep_abort_thd(THD *bf_thd,
+/*
+  Abort transaction for victim_thd. This function is called from
+  MDL BF abort codepath.
+*/
+void wsrep_abort_thd(THD *bf_thd,
                      THD *victim_thd,
                      my_bool signal) __attribute__((nonnull(1,2)));
 
+/**
+  Kill wsrep connection with kill_signal. Object thd is not
+  guaranteed to exist anymore when this function returns.
+
+  Asserts that the caller holds victim_thd->LOCK_thd_kill,
+  victim_thd->LOCK_thd_data.
+
+  @param thd THD object for connection that executes the KILL.
+  @param victim_thd THD object for connection to be killed.
+  @param kill_signal Kill signal.
+
+  @return Zero if the kill was successful, otherwise non-zero error code.
+ */
+uint wsrep_kill_thd(THD *thd, THD *victim_thd, killed_state kill_signal);
+
+/*
+  Backup kill status for commit.
+ */
+void wsrep_backup_kill_for_commit(THD *);
+
+/*
+  Restore KILL status after commit.
+ */
+void wsrep_restore_kill_after_commit(THD *);
+
 /*
   Helper methods to deal with thread local storage.
   The purpose of these methods is to hide the details of thread

sql/wsrep_trans_observer.h

@@ -256,6 +256,11 @@ static inline int wsrep_before_prepare(THD* thd, bool all)
                    thd->wsrep_trx().ws_meta().gtid(),
                    wsrep_gtid_server.gtid());
   }
+
+  mysql_mutex_lock(&thd->LOCK_thd_kill);
+  if (thd->killed) wsrep_backup_kill_for_commit(thd);
+  mysql_mutex_unlock(&thd->LOCK_thd_kill);
+
   DBUG_RETURN(ret);
 }
 
@@ -325,6 +330,11 @@ static inline int wsrep_before_commit(THD* thd, bool all)
                    wsrep_gtid_server.gtid());
     wsrep_register_for_group_commit(thd);
   }
+
+  mysql_mutex_lock(&thd->LOCK_thd_kill);
+  if (thd->killed) wsrep_backup_kill_for_commit(thd);
+  mysql_mutex_unlock(&thd->LOCK_thd_kill);
+
   DBUG_RETURN(ret);
 }
 
@@ -343,7 +353,8 @@ static inline int wsrep_before_commit(THD* thd, bool all)
 static inline int wsrep_ordered_commit(THD* thd, bool all)
 {
   DBUG_ENTER("wsrep_ordered_commit");
-  WSREP_DEBUG("wsrep_ordered_commit: %d", wsrep_is_real(thd, all));
+  WSREP_DEBUG("wsrep_ordered_commit: %d %lld", wsrep_is_real(thd, all),
+              (long long) wsrep_thd_trx_seqno(thd));
   DBUG_ASSERT(wsrep_run_commit_hook(thd, all));
   DBUG_RETURN(thd->wsrep_cs().ordered_commit());
 }
@@ -451,10 +462,18 @@ int wsrep_after_statement(THD* thd)
               wsrep::to_c_string(thd->wsrep_cs().state()),
               wsrep::to_c_string(thd->wsrep_cs().mode()),
               wsrep::to_c_string(thd->wsrep_cs().transaction().state()));
-  DBUG_RETURN((thd->wsrep_cs().state() != wsrep::client_state::s_none &&
+  int ret= ((thd->wsrep_cs().state() != wsrep::client_state::s_none &&
                thd->wsrep_cs().mode() == Wsrep_client_state::m_local) &&
               !thd->internal_transaction() ?
               thd->wsrep_cs().after_statement() : 0);
+
+  if (wsrep_is_active(thd))
+  {
+    mysql_mutex_lock(&thd->LOCK_thd_kill);
+    wsrep_restore_kill_after_commit(thd);
+    mysql_mutex_unlock(&thd->LOCK_thd_kill);
+  }
+  DBUG_RETURN(ret);
 }
 
 static inline void wsrep_after_apply(THD* thd)

storage/innobase/buf/buf0buf.cc

@@ -2140,6 +2140,8 @@ void buf_page_free(fil_space_t *space, uint32_t page, mtr_t *mtr)
   }
 
   block->page.lock.x_lock();
+  if (block->page.is_ibuf_exist())
+    ibuf_merge_or_delete_for_page(nullptr, page_id, block->page.zip_size());
 #ifdef BTR_CUR_HASH_ADAPT
   if (block->index)
     btr_search_drop_page_hash_index(block, false);

storage/innobase/buf/buf0dblwr.cc

@@ -372,7 +372,7 @@ void buf_dblwr_t::recover()
     const uint32_t space_id= page_get_space_id(page);
     const page_id_t page_id(space_id, page_no);
 
-    if (recv_sys.lsn < lsn)
+    if (recv_sys.scanned_lsn < lsn)
     {
       ib::info() << "Ignoring a doublewrite copy of page " << page_id
                  << " with future log sequence number " << lsn;

storage/innobase/buf/buf0flu.cc

@@ -966,12 +966,20 @@ uint32_t fil_space_t::flush_freed(bool writable)
   mysql_mutex_assert_not_owner(&buf_pool.flush_list_mutex);
   mysql_mutex_assert_not_owner(&buf_pool.mutex);
 
+  for (;;)
+  {
     freed_range_mutex.lock();
-  if (freed_ranges.empty() || log_sys.get_flushed_lsn() < get_last_freed_lsn())
+    if (freed_ranges.empty())
     {
       freed_range_mutex.unlock();
       return 0;
     }
+    const lsn_t flush_lsn= last_freed_lsn;
+    if (log_sys.get_flushed_lsn() >= flush_lsn)
+      break;
+    freed_range_mutex.unlock();
+    log_write_up_to(flush_lsn, true);
+  }
 
   const unsigned physical{physical_size()};
 
@@ -2430,6 +2438,7 @@ static void buf_flush_page_cleaner()
     else if (buf_pool.ran_out())
     {
       buf_pool.page_cleaner_set_idle(false);
+      buf_pool.get_oldest_modification(0);
       mysql_mutex_unlock(&buf_pool.flush_list_mutex);
       n= srv_max_io_capacity;
       mysql_mutex_lock(&buf_pool.mutex);
@@ -2583,6 +2592,7 @@ ATTRIBUTE_COLD void buf_flush_page_cleaner_init()
 /** Flush the buffer pool on shutdown. */
 ATTRIBUTE_COLD void buf_flush_buffer_pool()
 {
+  ut_ad(!os_aio_pending_reads());
   ut_ad(!buf_page_cleaner_is_active);
   ut_ad(!buf_flush_sync_lsn);
 

storage/innobase/buf/buf0lru.cc

@@ -1086,7 +1086,11 @@ static bool buf_LRU_block_remove_hashed(buf_page_t *bpage, const page_id_t id,
 
 			ut_a(!zip || !bpage->oldest_modification());
 			ut_ad(bpage->zip_size());
-
+			/* Skip consistency checks if the page was freed.
+			In recovery, we could get a sole FREE_PAGE record
+			and nothing else, for a ROW_FORMAT=COMPRESSED page.
+			Its contents would be garbage. */
+			if (!bpage->is_freed())
 			switch (fil_page_get_type(page)) {
 			case FIL_PAGE_TYPE_ALLOCATED:
 			case FIL_PAGE_INODE:
@@ -1217,6 +1221,7 @@ void buf_pool_t::corrupted_evict(buf_page_t *bpage, uint32_t state)
   buf_pool_t::hash_chain &chain= buf_pool.page_hash.cell_get(id.fold());
   page_hash_latch &hash_lock= buf_pool.page_hash.lock_get(chain);
 
+  recv_sys.free_corrupted_page(id);
   mysql_mutex_lock(&mutex);
   hash_lock.lock();
 
@@ -1241,8 +1246,6 @@ void buf_pool_t::corrupted_evict(buf_page_t *bpage, uint32_t state)
     buf_LRU_block_free_hashed_page(reinterpret_cast<buf_block_t*>(bpage));
 
   mysql_mutex_unlock(&mutex);
-
-  recv_sys.free_corrupted_page(id);
 }
 
 /** Update buf_pool.LRU_old_ratio.

storage/innobase/buf/buf0rea.cc

@@ -265,9 +265,6 @@ buf_read_page_low(
 	buf_page_t*	bpage;
 
 	if (buf_dblwr.is_inside(page_id)) {
-		ib::error() << "Trying to read doublewrite buffer page "
-			<< page_id;
-		ut_ad(0);
 		space->release();
 		return DB_PAGE_CORRUPTED;
 	}
@@ -525,7 +522,7 @@ buf_read_ahead_linear(const page_id_t page_id, ulint zip_size, bool ibuf)
 
   /* We will check that almost all pages in the area have been accessed
   in the desired order. */
-  const bool descending= page_id == low;
+  const bool descending= page_id != low;
 
   if (!descending && page_id != high_1)
     /* This is not a border page of the area */
@@ -555,7 +552,7 @@ fail:
                                uint32_t{buf_pool.read_ahead_area});
   page_id_t new_low= low, new_high_1= high_1;
   unsigned prev_accessed= 0;
-  for (page_id_t i= low; i != high_1; ++i)
+  for (page_id_t i= low; i <= high_1; ++i)
   {
     buf_pool_t::hash_chain &chain= buf_pool.page_hash.cell_get(i.fold());
     transactional_shared_lock_guard<page_hash_latch> g
@@ -583,12 +580,21 @@ failed:
       if (prev == FIL_NULL || next == FIL_NULL)
         goto fail;
       page_id_t id= page_id;
-      if (descending && next - 1 == page_id.page_no())
+      if (descending)
-        id.set_page_no(prev);
+      {
-      else if (!descending && prev + 1 == page_id.page_no())
+        if (id == high_1)
-        id.set_page_no(next);
+          ++id;
+        else if (next - 1 != page_id.page_no())
+          goto fail;
         else
-        goto fail; /* Successor or predecessor not in the right order */
+          id.set_page_no(prev);
+      }
+      else
+      {
+        if (prev + 1 != page_id.page_no())
+          goto fail;
+        id.set_page_no(next);
+      }
 
       new_low= id - (id.page_no() % buf_read_ahead_area);
       new_high_1= new_low + (buf_read_ahead_area - 1);
@@ -620,7 +626,7 @@ failed:
   /* If we got this far, read-ahead can be sensible: do it */
   count= 0;
   for (ulint ibuf_mode= ibuf ? BUF_READ_IBUF_PAGES_ONLY : BUF_READ_ANY_PAGE;
-       new_low != new_high_1; ++new_low)
+       new_low <= new_high_1; ++new_low)
   {
     if (ibuf_bitmap_page(new_low, zip_size))
       continue;
@@ -649,60 +655,35 @@ failed:
   return count;
 }
 
-/** @return whether a page has been freed */
+/** Schedule a page for recovery.
-inline bool fil_space_t::is_freed(uint32_t page)
+@param space    tablespace
+@param page_id  page identifier
+@param recs     log records
+@param init     page initialization, or nullptr if the page needs to be read */
+void buf_read_recover(fil_space_t *space, const page_id_t page_id,
+                      page_recv_t &recs, recv_init *init)
 {
-  std::lock_guard<std::mutex> freed_lock(freed_range_mutex);
+  ut_ad(space->id == page_id.space());
-  return freed_ranges.contains(page);
+  space->reacquire();
-}
-
-/** Issues read requests for pages which recovery wants to read in.
-@param space_id	tablespace identifier
-@param page_nos	page numbers to read, in ascending order */
-void buf_read_recv_pages(uint32_t space_id, st_::span<uint32_t> page_nos)
-{
-	fil_space_t* space = fil_space_t::get(space_id);
-
-	if (!space) {
-		/* The tablespace is missing or unreadable: do nothing */
-		return;
-	}
-
   const ulint zip_size= space->zip_size();
 
-	for (ulint i = 0; i < page_nos.size(); i++) {
+  if (init)
-
+  {
-		/* Ignore if the page already present in freed ranges. */
+    if (buf_page_t *bpage= buf_page_init_for_read(BUF_READ_ANY_PAGE, page_id,
-		if (space->is_freed(page_nos[i])) {
+                                                  zip_size, true))
-			continue;
+    {
+      ut_ad(bpage->in_file());
+      os_fake_read(IORequest{bpage, (buf_tmp_buffer_t*) &recs,
+                             UT_LIST_GET_FIRST(space->chain),
+                             IORequest::READ_ASYNC}, ptrdiff_t(init));
     }
-
-		const page_id_t	cur_page_id(space_id, page_nos[i]);
-
-		ulint limit = 0;
-		for (ulint j = 0; j < buf_pool.n_chunks; j++) {
-			limit += buf_pool.chunks[j].size / 2;
   }
-
+  else if (dberr_t err= buf_read_page_low(space, false, BUF_READ_ANY_PAGE,
-		if (os_aio_pending_reads() >= limit) {
+                                          page_id, zip_size, true))
-			os_aio_wait_until_no_pending_reads(false);
+  {
-		}
+    if (err != DB_SUCCESS_LOCKED_REC)
-
-		space->reacquire();
-		switch (buf_read_page_low(space, false, BUF_READ_ANY_PAGE,
-					  cur_page_id, zip_size, true)) {
-		case DB_SUCCESS: case DB_SUCCESS_LOCKED_REC:
-			break;
-		default:
       sql_print_error("InnoDB: Recovery failed to read page "
                       UINT32PF " from %s",
-					cur_page_id.page_no(),
+                      page_id.page_no(), space->chain.start->name);
-					space->chain.start->name);
   }
 }
-
-
-        DBUG_PRINT("ib_buf", ("recovery read (%zu pages) for %s",
-			      page_nos.size(), space->chain.start->name));
-	space->release();
-}

storage/innobase/dict/dict0stats.cc

@@ -204,7 +204,17 @@ static const dict_table_schema_t table_stats_schema =
   {
     {"database_name", DATA_VARMYSQL, DATA_NOT_NULL, 192},
     {"table_name", DATA_VARMYSQL, DATA_NOT_NULL, 597},
-    {"last_update", DATA_INT, DATA_NOT_NULL | DATA_UNSIGNED, 4},
+    /*
+      Don't check the DATA_UNSIGNED flag in last_update.
+      It presents if the server is running in a pure MariaDB installation,
+      because MariaDB's Field_timestampf::flags has UNSIGNED_FLAG.
+      But DATA_UNSIGNED misses when the server starts on a MySQL-5.7 directory
+      (during a migration), because MySQL's Field_timestampf::flags does not
+      have UNSIGNED_FLAG.
+      This is fine not to check DATA_UNSIGNED, because Field_timestampf
+      in both MariaDB and MySQL support only non-negative time_t values.
+    */
+    {"last_update", DATA_INT, DATA_NOT_NULL, 4},
     {"n_rows", DATA_INT, DATA_NOT_NULL | DATA_UNSIGNED, 8},
     {"clustered_index_size", DATA_INT, DATA_NOT_NULL | DATA_UNSIGNED, 8},
     {"sum_of_other_index_sizes", DATA_INT, DATA_NOT_NULL | DATA_UNSIGNED, 8},
@@ -218,7 +228,11 @@ static const dict_table_schema_t index_stats_schema =
     {"database_name", DATA_VARMYSQL, DATA_NOT_NULL, 192},
     {"table_name", DATA_VARMYSQL, DATA_NOT_NULL, 597},
     {"index_name", DATA_VARMYSQL, DATA_NOT_NULL, 192},
-    {"last_update", DATA_INT, DATA_NOT_NULL | DATA_UNSIGNED, 4},
+    /*
+      Don't check the DATA_UNSIGNED flag in last_update.
+      See comments about last_update in table_stats_schema above.
+    */
+    {"last_update", DATA_INT, DATA_NOT_NULL, 4},
     {"stat_name", DATA_VARMYSQL, DATA_NOT_NULL, 64*3},
     {"stat_value", DATA_INT, DATA_NOT_NULL | DATA_UNSIGNED, 8},
     {"sample_size", DATA_INT, DATA_UNSIGNED, 8},

storage/innobase/fil/fil0fil.cc

@@ -447,7 +447,9 @@ static bool fil_node_open_file(fil_node_t *node)
     }
   }
 
-  return fil_node_open_file_low(node);
+  /* The node can be opened beween releasing and acquiring fil_system.mutex
+  in the above code */
+  return node->is_open() || fil_node_open_file_low(node);
 }
 
 /** Close the file handle. */
@@ -1953,8 +1955,8 @@ err_exit:
 						     FIL_TYPE_TABLESPACE,
 						     crypt_data, mode, true)) {
 		fil_node_t* node = space->add(path, file, size, false, true);
-		mysql_mutex_unlock(&fil_system.mutex);
 		IF_WIN(node->find_metadata(), node->find_metadata(file, true));
+		mysql_mutex_unlock(&fil_system.mutex);
 		mtr.start();
 		mtr.set_named_space(space);
 		ut_a(fsp_header_init(space, size, &mtr) == DB_SUCCESS);
@@ -2775,38 +2777,41 @@ func_exit:
 
 #include <tpool.h>
 
-/** Callback for AIO completion */
+void IORequest::write_complete() const
-void fil_aio_callback(const IORequest &request)
 {
   ut_ad(fil_validate_skip());
-  ut_ad(request.node);
+  ut_ad(node);
+  ut_ad(is_write());
 
-  if (!request.bpage)
+  if (!bpage)
   {
     ut_ad(!srv_read_only_mode);
-    if (request.type == IORequest::DBLWR_BATCH)
+    if (type == IORequest::DBLWR_BATCH)
-      buf_dblwr.flush_buffered_writes_completed(request);
+      buf_dblwr.flush_buffered_writes_completed(*this);
     else
-      ut_ad(request.type == IORequest::WRITE_ASYNC);
+      ut_ad(type == IORequest::WRITE_ASYNC);
-write_completed:
-    request.node->complete_write();
-  }
-  else if (request.is_write())
-  {
-    buf_page_write_complete(request);
-    goto write_completed;
   }
   else
+    buf_page_write_complete(*this);
+
+  node->complete_write();
+  node->space->release();
+}
+
+void IORequest::read_complete() const
 {
-    ut_ad(request.is_read());
+  ut_ad(fil_validate_skip());
+  ut_ad(node);
+  ut_ad(is_read());
+  ut_ad(bpage);
 
   /* IMPORTANT: since i/o handling for reads will read also the insert
   buffer in fil_system.sys_space, we have to be very careful not to
-    introduce deadlocks. We never close fil_system.sys_space data
+  introduce deadlocks. We never close fil_system.sys_space data files
-    files and never issue asynchronous reads of change buffer pages. */
+  and never issue asynchronous reads of change buffer pages. */
-    const page_id_t id(request.bpage->id());
+  const page_id_t id(bpage->id());
 
-    if (dberr_t err= request.bpage->read_complete(*request.node))
+  if (dberr_t err= bpage->read_complete(*node))
   {
     if (recv_recovery_is_on() && !srv_force_recovery)
     {
@@ -2817,11 +2822,10 @@ write_completed:
 
     if (err != DB_FAIL)
       ib::error() << "Failed to read page " << id.page_no()
-                    << " from file '" << request.node->name << "': " << err;
+                  << " from file '" << node->name << "': " << err;
-    }
   }
 
-  request.node->space->release();
+  node->space->release();
 }
 
 /** Flush to disk the writes in file spaces of the given type

storage/innobase/fsp/fsp0fsp.cc

@@ -1284,24 +1284,21 @@ static dberr_t fsp_free_page(fil_space_t *space, page_no_t offset, mtr_t *mtr)
 			      + header->page.frame, frag_n_used - 1);
 	}
 
+	mtr->free(*space, static_cast<uint32_t>(offset));
+	xdes_set_free<true>(*xdes, descr, offset % FSP_EXTENT_SIZE, mtr);
+	ut_ad(err == DB_SUCCESS);
+
 	if (!xdes_get_n_used(descr)) {
 		/* The extent has become free: move it to another list */
 		err = flst_remove(header, FSP_HEADER_OFFSET + FSP_FREE_FRAG,
 				  xdes, xoffset, mtr);
-		if (UNIV_UNLIKELY(err != DB_SUCCESS)) {
+		if (err == DB_SUCCESS) {
-			return err;
-		}
 			err = fsp_free_extent(space, offset, mtr);
-		if (UNIV_UNLIKELY(err != DB_SUCCESS)) {
+		}
+	}
+
 	return err;
 }
-	}
-
-	mtr->free(*space, static_cast<uint32_t>(offset));
-	xdes_set_free<true>(*xdes, descr, offset % FSP_EXTENT_SIZE, mtr);
-
-	return DB_SUCCESS;
-}
 
 /** @return Number of segment inodes which fit on a single page */
 inline ulint FSP_SEG_INODES_PER_PAGE(ulint physical_size)

storage/innobase/handler/ha_innodb.cc

@@ -945,7 +945,8 @@ static SHOW_VAR innodb_status_variables[]= {
   {"buffer_pool_read_ahead", &buf_pool.stat.n_ra_pages_read, SHOW_SIZE_T},
   {"buffer_pool_read_ahead_evicted",
    &buf_pool.stat.n_ra_pages_evicted, SHOW_SIZE_T},
-  {"buffer_pool_read_requests", &buf_pool.stat.n_page_gets, SHOW_SIZE_T},
+  {"buffer_pool_read_requests",
+   &export_vars.innodb_buffer_pool_read_requests, SHOW_SIZE_T},
   {"buffer_pool_reads", &buf_pool.stat.n_pages_read, SHOW_SIZE_T},
   {"buffer_pool_wait_free", &buf_pool.stat.LRU_waits, SHOW_SIZE_T},
   {"buffer_pool_write_requests", &buf_pool.flush_list_requests, SHOW_SIZE_T},
@@ -1917,7 +1918,8 @@ static void innodb_disable_internal_writes(bool disable)
     sst_enable_innodb_writes();
 }
 
-static void wsrep_abort_transaction(handlerton*, THD *, THD *, my_bool);
+static void wsrep_abort_transaction(handlerton *, THD *, THD *, my_bool)
+    __attribute__((nonnull));
 static int innobase_wsrep_set_checkpoint(handlerton *hton, const XID *xid);
 static int innobase_wsrep_get_checkpoint(handlerton* hton, XID* xid);
 #endif /* WITH_WSREP */
@@ -18622,36 +18624,45 @@ void lock_wait_wsrep_kill(trx_t *bf_trx, ulong thd_id, trx_id_t trx_id)
                       wsrep_thd_client_mode_str(vthd),
                       wsrep_thd_transaction_state_str(vthd),
                       wsrep_thd_query(vthd));
-          /* Mark transaction as a victim for Galera abort */
-          vtrx->lock.set_wsrep_victim();
-          if (!wsrep_thd_set_wsrep_aborter(bf_thd, vthd))
           aborting= true;
-          else
-            WSREP_DEBUG("kill transaction skipped due to wsrep_aborter set");
         }
       }
       mysql_mutex_unlock(&lock_sys.wait_mutex);
       vtrx->mutex_unlock();
     }
-    wsrep_thd_UNLOCK(vthd);
+
-    if (aborting)
+    DEBUG_SYNC(bf_thd, "before_wsrep_thd_abort");
+    if (aborting && wsrep_thd_bf_abort(bf_thd, vthd, true))
     {
+      /* Need to grab mutexes again to ensure that the trx is still in
+         right state. */
+      lock_sys.wr_lock(SRW_LOCK_CALL);
+      mysql_mutex_lock(&lock_sys.wait_mutex);
+      vtrx->mutex_lock();
+
       /* if victim is waiting for some other lock, we have to cancel
          that waiting
       */
-      lock_sys.cancel_lock_wait_for_trx(vtrx);
+      if (vtrx->id == trx_id)
-
+      {
-      DEBUG_SYNC(bf_thd, "before_wsrep_thd_abort");
+        switch (vtrx->state) {
-      if (!wsrep_thd_bf_abort(bf_thd, vthd, true))
+        default:
+          break;
+        case TRX_STATE_ACTIVE:
+        case TRX_STATE_PREPARED:
+          lock_sys.cancel_lock_wait_for_wsrep_bf_abort(vtrx);
+        }
+      }
+      lock_sys.wr_unlock();
+      mysql_mutex_unlock(&lock_sys.wait_mutex);
+      vtrx->mutex_unlock();
+    }
+    else
     {
-        wsrep_thd_LOCK(vthd);
-        wsrep_thd_set_wsrep_aborter(NULL, vthd);
-        wsrep_thd_UNLOCK(vthd);
-
       WSREP_DEBUG("wsrep_thd_bf_abort has failed, victim %lu will survive",
                   thd_get_thread_id(vthd));
     }
-    }
+    wsrep_thd_UNLOCK(vthd);
     wsrep_thd_kill_UNLOCK(vthd);
   }
 }
@@ -18659,67 +18670,49 @@ void lock_wait_wsrep_kill(trx_t *bf_trx, ulong thd_id, trx_id_t trx_id)
 /** This function forces the victim transaction to abort. Aborting the
   transaction does NOT end it, it still has to be rolled back.
 
+  The caller must lock LOCK_thd_kill and LOCK_thd_data.
+
   @param bf_thd       brute force THD asking for the abort
   @param victim_thd   victim THD to be aborted
-
-  @return 0 victim was aborted
-  @return -1 victim thread was aborted (no transaction)
 */
-static
+static void wsrep_abort_transaction(handlerton *, THD *bf_thd, THD *victim_thd,
-void
-wsrep_abort_transaction(
-	handlerton*,
-	THD *bf_thd,
-	THD *victim_thd,
                                     my_bool signal)
 {
   DBUG_ENTER("wsrep_abort_transaction");
   ut_ad(bf_thd);
   ut_ad(victim_thd);
 
-	wsrep_thd_kill_LOCK(victim_thd);
-	wsrep_thd_LOCK(victim_thd);
   trx_t *victim_trx= thd_to_trx(victim_thd);
-	wsrep_thd_UNLOCK(victim_thd);
 
   WSREP_DEBUG("abort transaction: BF: %s victim: %s victim conf: %s",
-			wsrep_thd_query(bf_thd),
+              wsrep_thd_query(bf_thd), wsrep_thd_query(victim_thd),
-			wsrep_thd_query(victim_thd),
               wsrep_thd_transaction_state_str(victim_thd));
 
-	if (victim_trx) {
+  if (!victim_trx)
-		victim_trx->lock.set_wsrep_victim();
-
-		wsrep_thd_LOCK(victim_thd);
-		bool aborting= !wsrep_thd_set_wsrep_aborter(bf_thd, victim_thd);
-		wsrep_thd_UNLOCK(victim_thd);
-		if (aborting) {
-			DEBUG_SYNC(bf_thd, "before_wsrep_thd_abort");
-			DBUG_EXECUTE_IF("sync.before_wsrep_thd_abort",
   {
-					   const char act[]=
+    WSREP_DEBUG("abort transaction: victim did not exist");
-					     "now "
+    DBUG_VOID_RETURN;
-					     "SIGNAL sync.before_wsrep_thd_abort_reached "
-					     "WAIT_FOR signal.before_wsrep_thd_abort";
-					   DBUG_ASSERT(!debug_sync_set_action(bf_thd,
-									      STRING_WITH_LEN(act)));
-					 };);
-			wsrep_thd_bf_abort(bf_thd, victim_thd, signal);
-		}
-	} else {
-		DBUG_EXECUTE_IF("sync.before_wsrep_thd_abort",
-				 {
-				   const char act[]=
-				     "now "
-				     "SIGNAL sync.before_wsrep_thd_abort_reached "
-				     "WAIT_FOR signal.before_wsrep_thd_abort";
-				   DBUG_ASSERT(!debug_sync_set_action(bf_thd,
-								      STRING_WITH_LEN(act)));
-				 };);
-		wsrep_thd_bf_abort(bf_thd, victim_thd, signal);
   }
 
-	wsrep_thd_kill_UNLOCK(victim_thd);
+  lock_sys.wr_lock(SRW_LOCK_CALL);
+  mysql_mutex_lock(&lock_sys.wait_mutex);
+  victim_trx->mutex_lock();
+
+  switch (victim_trx->state) {
+  default:
+    break;
+  case TRX_STATE_ACTIVE:
+  case TRX_STATE_PREPARED:
+    /* Cancel lock wait if the victim is waiting for a lock in InnoDB.
+       The transaction which is blocked somewhere else (e.g. waiting
+       for next command or MDL) has been interrupted by THD::awake_no_mutex()
+       on server level before calling this function. */
+    lock_sys.cancel_lock_wait_for_wsrep_bf_abort(victim_trx);
+  }
+  lock_sys.wr_unlock();
+  mysql_mutex_unlock(&lock_sys.wait_mutex);
+  victim_trx->mutex_unlock();
+
   DBUG_VOID_RETURN;
 }
 

storage/innobase/ibuf/ibuf0ibuf.cc

@@ -2365,6 +2365,7 @@ tablespace_deleted:
 		}
 
 		const ulint zip_size = s->zip_size(), size = s->size;
+		s->x_lock();
 		s->release();
 		mtr_t mtr;
 
@@ -2382,13 +2383,17 @@ tablespace_deleted:
 				|| !page_is_leaf(block->page.frame);
 			mtr.commit();
 			if (err == DB_TABLESPACE_DELETED) {
+				s->x_unlock();
 				goto tablespace_deleted;
 			}
 			if (!remove) {
+				s->x_unlock();
 				continue;
 			}
 		}
 
+		s->x_unlock();
+
 		if (srv_shutdown_state == SRV_SHUTDOWN_NONE
 		    || srv_fast_shutdown) {
 			continue;
@@ -2417,7 +2422,7 @@ tablespace_deleted:
 		/* Prevent an infinite loop, by removing entries from
 		the change buffer in the case the bitmap bits were
 		wrongly clear even though buffered changes exist. */
-		ibuf_delete_recs(page_id_t(space_ids[i], page_nos[i]));
+		ibuf_delete_recs(page_id_t(space_id, page_nos[i]));
 	}
 }
 
@@ -4195,25 +4200,26 @@ dberr_t ibuf_merge_or_delete_for_page(buf_block_t *block,
 
 		ibuf_mtr_commit(&mtr);
 
-		if (bitmap_bits
+		if (!bitmap_bits) {
-		    && DB_SUCCESS
+		done:
+			/* No changes are buffered for this page. */
+			space->release();
+			return DB_SUCCESS;
+		}
+
+		if (!block
+		    || DB_SUCCESS
 		    == fseg_page_is_allocated(space, page_id.page_no())) {
 			ibuf_mtr_start(&mtr);
 			mtr.set_named_space(space);
 			ibuf_reset_bitmap(block, page_id, zip_size, &mtr);
 			ibuf_mtr_commit(&mtr);
-			bitmap_bits = 0;
 			if (!block
 			    || btr_page_get_index_id(block->page.frame)
 			    != DICT_IBUF_ID_MIN + IBUF_SPACE_ID) {
 				ibuf_delete_recs(page_id);
 			}
-		}
+			goto done;
-
-		if (!bitmap_bits) {
-			/* No changes are buffered for this page. */
-			space->release();
-			return DB_SUCCESS;
 		}
 	}
 

storage/innobase/include/buf0buf.h

@@ -75,8 +75,7 @@ struct buf_pool_info_t
 	ulint	flush_list_len;		/*!< Length of buf_pool.flush_list */
 	ulint	n_pend_unzip;		/*!< buf_pool.n_pend_unzip, pages
 					pending decompress */
-	ulint	n_pend_reads;		/*!< buf_pool.n_pend_reads, pages
+	ulint	n_pend_reads;		/*!< os_aio_pending_reads() */
-					pending read */
 	ulint	n_pending_flush_lru;	/*!< Pages pending flush in LRU */
 	ulint	n_pending_flush_list;	/*!< Pages pending flush in FLUSH
 					LIST */

storage/innobase/include/buf0rea.h

@@ -102,10 +102,13 @@ which could result in a deadlock if the OS does not support asynchronous io.
 ulint
 buf_read_ahead_linear(const page_id_t page_id, ulint zip_size, bool ibuf);
 
-/** Issue read requests for pages that need to be recovered.
+/** Schedule a page for recovery.
-@param space_id	tablespace identifier
+@param space    tablespace
-@param page_nos	page numbers to read, in ascending order */
+@param page_id  page identifier
-void buf_read_recv_pages(uint32_t space_id, st_::span<uint32_t> page_nos);
+@param recs     log records
+@param init     page initialization, or nullptr if the page needs to be read */
+void buf_read_recover(fil_space_t *space, const page_id_t page_id,
+                      page_recv_t &recs, recv_init *init);
 
 /** @name Modes used in read-ahead @{ */
 /** read only pages belonging to the insert buffer tree */

storage/innobase/include/fil0fil.h

@@ -638,8 +638,6 @@ public:
   /** Close all tablespace files at shutdown */
   static void close_all();
 
-  /** @return last_freed_lsn */
-  lsn_t get_last_freed_lsn() { return last_freed_lsn; }
   /** Update last_freed_lsn */
   void update_last_freed_lsn(lsn_t lsn) { last_freed_lsn= lsn; }
 

storage/innobase/include/lock0lock.h

@@ -955,6 +955,10 @@ public:
 
   /** Cancel possible lock waiting for a transaction */
   static void cancel_lock_wait_for_trx(trx_t *trx);
+#ifdef WITH_WSREP
+  /** Cancel lock waiting for a wsrep BF abort. */
+  static void cancel_lock_wait_for_wsrep_bf_abort(trx_t *trx);
+#endif /* WITH_WSREP */
 };
 
 /** The lock system */

storage/innobase/include/log0recv.h

@@ -38,9 +38,9 @@ Created 9/20/1997 Heikki Tuuri
 #define recv_recovery_is_on() UNIV_UNLIKELY(recv_sys.recovery_on)
 
 ATTRIBUTE_COLD MY_ATTRIBUTE((nonnull, warn_unused_result))
-/** Apply any buffered redo log to a page that was just read from a data file.
+/** Apply any buffered redo log to a page.
-@param[in,out]	space	tablespace
+@param space     tablespace
-@param[in,out]	bpage	buffer pool page
+@param bpage     buffer pool page
 @return whether the page was recovered correctly */
 bool recv_recover_page(fil_space_t* space, buf_page_t* bpage);
 
@@ -49,17 +49,6 @@ of first system tablespace page
 @return error code or DB_SUCCESS */
 dberr_t recv_recovery_from_checkpoint_start();
 
-/** Whether to store redo log records in recv_sys.pages */
-enum store_t {
-	/** Do not store redo log records. */
-	STORE_NO,
-	/** Store redo log records. */
-	STORE_YES,
-	/** Store redo log records if the tablespace exists. */
-	STORE_IF_EXISTS
-};
-
-
 /** Report an operation to create, delete, or rename a file during backup.
 @param[in]	space_id	tablespace identifier
 @param[in]	type		file operation redo log type
@@ -125,21 +114,15 @@ struct recv_dblwr_t
   list pages;
 };
 
-/** the recovery state and buffered records for a page */
+/** recv_sys.pages entry; protected by recv_sys.mutex */
 struct page_recv_t
 {
-  /** Recovery state; protected by recv_sys.mutex */
+  /** Recovery status: 0=not in progress, 1=log is being applied,
-  enum
+  -1=log has been applied and the entry may be erased.
-  {
+  Transitions from 1 to -1 are NOT protected by recv_sys.mutex. */
-    /** not yet processed */
+  Atomic_relaxed<int8_t> being_processed{0};
-    RECV_NOT_PROCESSED,
+  /** Whether reading the page will be skipped */
-    /** not processed; the page will be reinitialized */
+  bool skip_read= false;
-    RECV_WILL_NOT_READ,
-    /** page is being read */
-    RECV_BEING_READ,
-    /** log records are being applied on the page */
-    RECV_BEING_PROCESSED
-  } state= RECV_NOT_PROCESSED;
   /** Latest written byte offset when applying the log records.
   @see mtr_t::m_last_offset */
   uint16_t last_offset= 1;
@@ -162,6 +145,9 @@ struct page_recv_t
         head= recs;
       tail= recs;
     }
+    /** Remove the last records for the page
+    @param start_lsn   start of the removed log */
+    ATTRIBUTE_COLD void rewind(lsn_t start_lsn);
 
     /** @return the last log snippet */
     const log_rec_t* last() const { return tail; }
@@ -180,8 +166,8 @@ struct page_recv_t
     iterator begin() { return head; }
     iterator end() { return NULL; }
     bool empty() const { ut_ad(!head == !tail); return !head; }
-    /** Clear and free the records; @see recv_sys_t::alloc() */
+    /** Clear and free the records; @see recv_sys_t::add() */
-    inline void clear();
+    void clear();
   } log;
 
   /** Trim old log records for a page.
@@ -190,21 +176,27 @@ struct page_recv_t
   inline bool trim(lsn_t start_lsn);
   /** Ignore any earlier redo log records for this page. */
   inline void will_not_read();
-  /** @return whether the log records for the page are being processed */
+};
-  bool is_being_processed() const { return state == RECV_BEING_PROCESSED; }
+
+/** A page initialization operation that was parsed from the redo log */
+struct recv_init
+{
+  /** log sequence number of the page initialization */
+  lsn_t lsn;
+  /** Whether btr_page_create() avoided a read of the page.
+  At the end of the last recovery batch, mark_ibuf_exist()
+  will mark pages for which this flag is set. */
+  bool created;
 };
 
 /** Recovery system data structure */
 struct recv_sys_t
 {
-  /** mutex protecting apply_log_recs and page_recv_t::state */
+  using init= recv_init;
-  mysql_mutex_t mutex;
+
+  /** mutex protecting this as well as some of page_recv_t */
+  alignas(CPU_LEVEL1_DCACHE_LINESIZE) mysql_mutex_t mutex;
 private:
-  /** condition variable for
-  !apply_batch_on || pages.empty() || found_corrupt_log || found_corrupt_fs */
-  pthread_cond_t cond;
-  /** whether recv_apply_hashed_log_recs() is running */
-  bool apply_batch_on;
   /** set when finding a corrupt log block or record, or there is a
   log parsing buffer overflow */
   bool found_corrupt_log;
@@ -226,6 +218,8 @@ public:
   size_t offset;
   /** log sequence number of the first non-parsed record */
   lsn_t lsn;
+  /** log sequence number of the last parsed mini-transaction */
+  lsn_t scanned_lsn;
   /** log sequence number at the end of the FILE_CHECKPOINT record, or 0 */
   lsn_t file_checkpoint;
   /** the time when progress was last reported */
@@ -238,6 +232,9 @@ public:
   map pages;
 
 private:
+  /** iterator to pages, used by parse() */
+  map::iterator pages_it;
+
   /** Process a record that indicates that a tablespace size is being shrunk.
   @param page_id first page that is not in the file
   @param lsn     log sequence number of the shrink operation */
@@ -257,30 +254,42 @@ public:
   /** The contents of the doublewrite buffer */
   recv_dblwr_t dblwr;
 
-  /** Last added LSN to pages, before switching to STORE_NO */
-  lsn_t last_stored_lsn= 0;
-
   inline void read(os_offset_t offset, span<byte> buf);
   inline size_t files_size();
   void close_files() { files.clear(); files.shrink_to_fit(); }
 
+  /** Advance pages_it if it matches the iterator */
+  void pages_it_invalidate(const map::iterator &p)
+  {
+    mysql_mutex_assert_owner(&mutex);
+    if (pages_it == p)
+      pages_it++;
+  }
+  /** Invalidate pages_it if it points to the given tablespace */
+  void pages_it_invalidate(uint32_t space_id)
+  {
+    mysql_mutex_assert_owner(&mutex);
+    if (pages_it != pages.end() && pages_it->first.space() == space_id)
+      pages_it= pages.end();
+  }
+
 private:
   /** Attempt to initialize a page based on redo log records.
-  @param page_id  page identifier
+  @param p        iterator
-  @param p        iterator pointing to page_id
   @param mtr      mini-transaction
   @param b        pre-allocated buffer pool block
+  @param init     page initialization
   @return the recovered block
   @retval nullptr if the page cannot be initialized based on log records
   @retval -1      if the page cannot be recovered due to corruption */
-  inline buf_block_t *recover_low(const page_id_t page_id, map::iterator &p,
+  inline buf_block_t *recover_low(const map::iterator &p, mtr_t &mtr,
-                                  mtr_t &mtr, buf_block_t *b);
+                                  buf_block_t *b, init &init);
   /** Attempt to initialize a page based on redo log records.
   @param page_id  page identifier
   @return the recovered block
   @retval nullptr if the page cannot be initialized based on log records
   @retval -1      if the page cannot be recovered due to corruption */
-  buf_block_t *recover_low(const page_id_t page_id);
+  ATTRIBUTE_COLD buf_block_t *recover_low(const page_id_t page_id);
 
   /** All found log files (multiple ones are possible if we are upgrading
   from before MariaDB Server 10.5.1) */
@@ -289,10 +298,27 @@ private:
   /** Base node of the redo block list.
   List elements are linked via buf_block_t::unzip_LRU. */
   UT_LIST_BASE_NODE_T(buf_block_t) blocks;
+
+  /** Allocate a block from the buffer pool for recv_sys.pages */
+  ATTRIBUTE_COLD buf_block_t *add_block();
+
+  /** Wait for buffer pool to become available.
+  @param pages number of buffer pool pages needed */
+  ATTRIBUTE_COLD void wait_for_pool(size_t pages);
+
+  /** Free log for processed pages. */
+  void garbage_collect();
+
+  /** Apply a recovery batch.
+  @param space_id       current tablespace identifier
+  @param space          current tablespace
+  @param free_block     spare buffer block
+  @param last_batch     whether it is possible to write more redo log
+  @return whether the caller must provide a new free_block */
+  bool apply_batch(uint32_t space_id, fil_space_t *&space,
+                   buf_block_t *&free_block, bool last_batch);
+
 public:
-  /** Check whether the number of read redo log blocks exceeds the maximum.
-  @return whether the memory is exhausted */
-  inline bool is_memory_exhausted();
   /** Apply buffered log to persistent data pages.
   @param last_batch     whether it is possible to write more redo log */
   void apply(bool last_batch);
@@ -310,7 +336,7 @@ public:
   /** Clean up after create() */
   void close();
 
-  bool is_initialised() const { return last_stored_lsn != 0; }
+  bool is_initialised() const { return scanned_lsn != 0; }
 
   /** Find the latest checkpoint.
   @return error code or DB_SUCCESS */
@@ -321,60 +347,76 @@ public:
   @param start_lsn start LSN of the mini-transaction
   @param lsn      @see mtr_t::commit_lsn()
   @param l        redo log snippet
-  @param len      length of l, in bytes */
+  @param len      length of l, in bytes
-  inline void add(map::iterator it, lsn_t start_lsn, lsn_t lsn,
+  @return whether we ran out of memory */
+  bool add(map::iterator it, lsn_t start_lsn, lsn_t lsn,
            const byte *l, size_t len);
 
-  enum parse_mtr_result { OK, PREMATURE_EOF, GOT_EOF };
+  /** Parsing result */
+  enum parse_mtr_result {
+    /** a record was successfully parsed */
+    OK,
+    /** the log ended prematurely (need to read more) */
+    PREMATURE_EOF,
+    /** the end of the log was reached */
+    GOT_EOF,
+    /** parse<true>(l, false) ran out of memory */
+    GOT_OOM
+  };
 
 private:
   /** Parse and register one log_t::FORMAT_10_8 mini-transaction.
-  @param store   whether to store the records
+  @tparam store     whether to store the records
-  @param l       log data source */
+  @param  l         log data source
+  @param  if_exists if store: whether to check if the tablespace exists */
+  template<typename source,bool store>
+  inline parse_mtr_result parse(source &l, bool if_exists) noexcept;
+
+  /** Rewind a mini-transaction when parse() runs out of memory.
+  @param  l         log data source
+  @param  begin     start of the mini-transaction */
   template<typename source>
-  inline parse_mtr_result parse(store_t store, source& l) noexcept;
+  ATTRIBUTE_COLD void rewind(source &l, source &begin) noexcept;
+
+  /** Report progress in terms of LSN or pages remaining */
+  ATTRIBUTE_COLD void report_progress() const;
 public:
   /** Parse and register one log_t::FORMAT_10_8 mini-transaction,
   handling log_sys.is_pmem() buffer wrap-around.
-  @param store           whether to store the records */
+  @tparam store     whether to store the records
-  static parse_mtr_result parse_mtr(store_t store) noexcept;
+  @param  if_exists if store: whether to check if the tablespace exists */
+  template<bool store>
+  static parse_mtr_result parse_mtr(bool if_exists) noexcept;
 
   /** Parse and register one log_t::FORMAT_10_8 mini-transaction,
   handling log_sys.is_pmem() buffer wrap-around.
-  @param store   whether to store the records */
+  @tparam store     whether to store the records
-  static parse_mtr_result parse_pmem(store_t store) noexcept
+  @param  if_exists if store: whether to check if the tablespace exists */
+  template<bool store>
+  static parse_mtr_result parse_pmem(bool if_exists) noexcept
 #ifdef HAVE_PMEM
     ;
 #else
-  { return parse_mtr(store); }
+  { return parse_mtr<store>(if_exists); }
 #endif
 
+  /** Erase log records for a page. */
+  void erase(map::iterator p);
+
   /** Clear a fully processed set of stored redo log records. */
-  inline void clear();
+  void clear();
 
   /** Determine whether redo log recovery progress should be reported.
   @param time  the current time
   @return whether progress should be reported
   (the last report was at least 15 seconds ago) */
-  bool report(time_t time)
+  bool report(time_t time);
-  {
-    if (time - progress_time < 15)
-      return false;
-
-    progress_time= time;
-    return true;
-  }
 
   /** The alloc() memory alignment, in bytes */
   static constexpr size_t ALIGNMENT= sizeof(size_t);
 
-  /** Allocate memory for log_rec_t
-  @param len  allocation size, in bytes
-  @return pointer to len bytes of memory (never NULL) */
-  inline void *alloc(size_t len);
-
   /** Free a redo log snippet.
-  @param data buffer returned by alloc() */
+  @param data buffer allocated in add() */
   inline void free(const void *data);
 
   /** Remove records for a corrupted page.
@@ -386,8 +428,6 @@ public:
   ATTRIBUTE_COLD void set_corrupt_fs();
   /** Flag log file corruption during recovery. */
   ATTRIBUTE_COLD void set_corrupt_log();
-  /** Possibly finish a recovery batch. */
-  inline void maybe_finish_batch();
 
   /** @return whether data file corruption was found */
   bool is_corrupt_fs() const { return UNIV_UNLIKELY(found_corrupt_fs); }
@@ -405,12 +445,13 @@ public:
   }
 
   /** Try to recover a tablespace that was not readable earlier
-  @param p          iterator, initially pointing to page_id_t{space_id,0};
+  @param p          iterator
-                    the records will be freed and the iterator advanced
   @param name       tablespace file name
   @param free_block spare buffer block
-  @return whether recovery failed */
+  @return recovered tablespace
-  bool recover_deferred(map::iterator &p, const std::string &name,
+  @retval nullptr if recovery failed */
+  fil_space_t *recover_deferred(const map::iterator &p,
+                                const std::string &name,
                                 buf_block_t *&free_block);
 };
 

storage/innobase/include/os0file.h

@@ -212,6 +212,10 @@ public:
   bool is_LRU() const { return (type & (WRITE_LRU ^ WRITE_ASYNC)) != 0; }
   bool is_async() const { return (type & (READ_SYNC ^ READ_ASYNC)) != 0; }
 
+  void write_complete() const;
+  void read_complete() const;
+  void fake_read_complete(os_offset_t offset) const;
+
   /** If requested, free storage space associated with a section of the file.
   @param off   byte offset from the start (SEEK_SET)
   @param len   size of the hole in bytes
@@ -1040,6 +1044,11 @@ int os_aio_init();
 Frees the asynchronous io system. */
 void os_aio_free();
 
+/** Submit a fake read request during crash recovery.
+@param type   fake read request
+@param offset additional context */
+void os_fake_read(const IORequest &type, os_offset_t offset);
+
 /** Request a read or write.
 @param type		I/O request
 @param buf		buffer

storage/innobase/include/srv0srv.h

@@ -617,6 +617,8 @@ struct export_var_t{
 #ifdef UNIV_DEBUG
 	ulint innodb_buffer_pool_pages_latched;	/*!< Latched pages */
 #endif /* UNIV_DEBUG */
+	/** buf_pool.stat.n_page_gets (a sharded counter) */
+	ulint innodb_buffer_pool_read_requests;
 	ulint innodb_checkpoint_age;
 	ulint innodb_checkpoint_max_age;
 	ulint innodb_data_pending_reads;	/*!< Pending reads */

storage/innobase/include/trx0sys.h

@@ -1055,7 +1055,7 @@ public:
   void close();
 
   /** @return total number of active (non-prepared) transactions */
-  ulint any_active_transactions();
+  size_t any_active_transactions(size_t *prepared= nullptr);
 
 
   /**

storage/innobase/include/trx0undo.h

@@ -216,14 +216,6 @@ buf_block_t*
 trx_undo_assign_low(trx_t *trx, trx_rseg_t *rseg, trx_undo_t **undo,
                     mtr_t *mtr, dberr_t *err)
 	MY_ATTRIBUTE((nonnull, warn_unused_result));
-/******************************************************************//**
-Sets the state of the undo log segment at a transaction finish.
-@return undo log segment header page, x-latched */
-buf_block_t*
-trx_undo_set_state_at_finish(
-/*=========================*/
-	trx_undo_t*	undo,	/*!< in: undo log memory copy */
-	mtr_t*		mtr);	/*!< in: mtr */
 
 /** Set the state of the undo log segment at a XA PREPARE or XA ROLLBACK.
 @param[in,out]	trx		transaction

storage/innobase/lock/lock0lock.cc

@@ -5732,12 +5732,13 @@ static void lock_release_autoinc_locks(trx_t *trx)
 }
 
 /** Cancel a waiting lock request and release possibly waiting transactions */
-template <bool from_deadlock= false>
+template <bool from_deadlock= false, bool inner_trx_lock= true>
 void lock_cancel_waiting_and_release(lock_t *lock)
 {
   lock_sys.assert_locked(*lock);
   mysql_mutex_assert_owner(&lock_sys.wait_mutex);
   trx_t *trx= lock->trx;
+  if (inner_trx_lock)
     trx->mutex_lock();
   ut_d(const auto trx_state= trx->state);
   ut_ad(trx_state == TRX_STATE_COMMITTED_IN_MEMORY ||
@@ -5762,6 +5763,7 @@ void lock_cancel_waiting_and_release(lock_t *lock)
 
   lock_wait_end<from_deadlock>(trx);
 
+  if (inner_trx_lock)
     trx->mutex_unlock();
 }
 
@@ -5779,6 +5781,19 @@ void lock_sys_t::cancel_lock_wait_for_trx(trx_t *trx)
   mysql_mutex_unlock(&lock_sys.wait_mutex);
 }
 
+#ifdef WITH_WSREP
+void lock_sys_t::cancel_lock_wait_for_wsrep_bf_abort(trx_t *trx)
+{
+  lock_sys.assert_locked();
+  mysql_mutex_assert_owner(&lock_sys.wait_mutex);
+  ut_ad(trx->mutex_is_owner());
+  ut_ad(trx->state == TRX_STATE_ACTIVE || trx->state == TRX_STATE_PREPARED);
+  trx->lock.set_wsrep_victim();
+  if (lock_t *lock= trx->lock.wait_lock)
+    lock_cancel_waiting_and_release<false, false>(lock);
+}
+#endif /* WITH_WSREP */
+
 /** Cancel a waiting lock request.
 @tparam check_victim  whether to check for DB_DEADLOCK
 @param trx            active transaction

storage/innobase/os/os0file.cc

@@ -3427,15 +3427,12 @@ os_file_get_status(
 	return(ret);
 }
 
-
+static void io_callback_errorcheck(const tpool::aiocb *cb)
-extern void fil_aio_callback(const IORequest &request);
+{
-
+  if (cb->m_err != DB_SUCCESS)
-static void io_callback(tpool::aiocb *cb)
   {
     const IORequest &request= *static_cast<const IORequest*>
       (static_cast<const void*>(cb->m_userdata));
-  if (cb->m_err != DB_SUCCESS)
-  {
     ib::fatal() << "IO Error: " << cb->m_err << " during " <<
       (request.is_async() ? "async " : "sync ") <<
       (request.is_LRU() ? "lru " : "") <<
@@ -3443,19 +3440,36 @@ static void io_callback(tpool::aiocb *cb)
       " of " << cb->m_len << " bytes, for file " << cb->m_fh << ", returned " <<
       cb->m_ret_len;
   }
-  /* Return cb back to cache*/
+}
-  if (cb->m_opcode == tpool::aio_opcode::AIO_PREAD)
+
+static void fake_io_callback(void *c)
 {
+  tpool::aiocb *cb= static_cast<tpool::aiocb*>(c);
   ut_ad(read_slots->contains(cb));
-    fil_aio_callback(request);
+  static_cast<const IORequest*>(static_cast<const void*>(cb->m_userdata))->
+    fake_read_complete(cb->m_offset);
   read_slots->release(cb);
 }
-  else
+
+static void read_io_callback(void *c)
 {
-    ut_ad(write_slots->contains(cb));
+  tpool::aiocb *cb= static_cast<tpool::aiocb*>(c);
-    fil_aio_callback(request);
+  ut_ad(cb->m_opcode == tpool::aio_opcode::AIO_PREAD);
-    write_slots->release(cb);
+  io_callback_errorcheck(cb);
+  ut_ad(read_slots->contains(cb));
+  static_cast<const IORequest*>
+    (static_cast<const void*>(cb->m_userdata))->read_complete();
+  read_slots->release(cb);
 }
+
+static void write_io_callback(void *c)
+{
+  tpool::aiocb *cb= static_cast<tpool::aiocb*>(c);
+  ut_ad(cb->m_opcode == tpool::aio_opcode::AIO_PWRITE);
+  ut_ad(write_slots->contains(cb));
+  static_cast<const IORequest*>
+    (static_cast<const void*>(cb->m_userdata))->write_complete();
+  write_slots->release(cb);
 }
 
 #ifdef LINUX_NATIVE_AIO
@@ -3758,6 +3772,28 @@ void os_aio_wait_until_no_pending_reads(bool declare)
     tpool::tpool_wait_end();
 }
 
+/** Submit a fake read request during crash recovery.
+@param type  fake read request
+@param offset additional context */
+void os_fake_read(const IORequest &type, os_offset_t offset)
+{
+  tpool::aiocb *cb= read_slots->acquire();
+
+  cb->m_group= read_slots->get_task_group();
+  cb->m_fh= type.node->handle.m_file;
+  cb->m_buffer= nullptr;
+  cb->m_len= 0;
+  cb->m_offset= offset;
+  cb->m_opcode= tpool::aio_opcode::AIO_PREAD;
+  new (cb->m_userdata) IORequest{type};
+  cb->m_internal_task.m_func= fake_io_callback;
+  cb->m_internal_task.m_arg= cb;
+  cb->m_internal_task.m_group= cb->m_group;
+
+  srv_thread_pool->submit_task(&cb->m_internal_task);
+}
+
+
 /** Request a read or write.
 @param type		I/O request
 @param buf		buffer
@@ -3803,23 +3839,32 @@ func_exit:
 		return err;
 	}
 
+	io_slots* slots;
+	tpool::callback_func callback;
+	tpool::aio_opcode opcode;
+
 	if (type.is_read()) {
 		++os_n_file_reads;
+		slots = read_slots;
+		callback = read_io_callback;
+		opcode = tpool::aio_opcode::AIO_PREAD;
 	} else {
 		++os_n_file_writes;
+		slots = write_slots;
+		callback = write_io_callback;
+		opcode = tpool::aio_opcode::AIO_PWRITE;
 	}
 
 	compile_time_assert(sizeof(IORequest) <= tpool::MAX_AIO_USERDATA_LEN);
-	io_slots* slots= type.is_read() ? read_slots : write_slots;
 	tpool::aiocb* cb = slots->acquire();
 
 	cb->m_buffer = buf;
-	cb->m_callback = (tpool::callback_func)io_callback;
+	cb->m_callback = callback;
 	cb->m_group = slots->get_task_group();
 	cb->m_fh = type.node->handle.m_file;
 	cb->m_len = (int)n;
 	cb->m_offset = offset;
-	cb->m_opcode = type.is_read() ? tpool::aio_opcode::AIO_PREAD : tpool::aio_opcode::AIO_PWRITE;
+	cb->m_opcode = opcode;
 	new (cb->m_userdata) IORequest{type};
 
 	if (srv_thread_pool->submit_io(cb)) {
@@ -3827,6 +3872,7 @@ func_exit:
 		os_file_handle_error(type.node->name, type.is_read()
 				     ? "aio read" : "aio write");
 		err = DB_IO_ERROR;
+		type.node->space->release();
 	}
 
 	goto func_exit;

storage/innobase/row/row0ins.cc

@@ -214,14 +214,14 @@ row_ins_sec_index_entry_by_modify(
 		made to the clustered index, and completed the
 		secondary index creation before we got here. In this
 		case, the change would already be there. The CREATE
-		INDEX should be waiting for a MySQL meta-data lock
+		INDEX should be in wait_while_table_is_used() at least
-		upgrade at least until this INSERT or UPDATE
+		until this INSERT or UPDATE returns. After that point,
-		returns. After that point, set_committed(true)
+		set_committed(true) would be invoked in
-		would be invoked in commit_inplace_alter_table(). */
+		commit_inplace_alter_table(). */
 		ut_a(update->n_fields == 0);
-		ut_a(!cursor->index()->is_committed());
 		ut_ad(!dict_index_is_online_ddl(cursor->index()));
-		return(DB_SUCCESS);
+		return cursor->index()->is_committed()
+			? DB_CORRUPTION : DB_SUCCESS;
 	}
 
 	if (mode == BTR_MODIFY_LEAF) {

storage/innobase/row/row0purge.cc

@@ -615,6 +615,8 @@ row_purge_del_mark(
       const auto type= node->index->type;
       if (type & (DICT_FTS | DICT_CORRUPT))
         continue;
+      if (node->index->online_status > ONLINE_INDEX_CREATION)
+        continue;
       if (UNIV_UNLIKELY(DICT_VIRTUAL & type) && !node->index->is_committed() &&
           node->index->has_new_v_col())
         continue;
@@ -767,6 +769,11 @@ row_purge_upd_exist_or_extern_func(
 			continue;
 		}
 
+		if (node->index->online_status
+		    > ONLINE_INDEX_CREATION) {
+			continue;
+		}
+
 		if (row_upd_changes_ord_field_binary(node->index, node->update,
 						     thr, NULL, NULL)) {
 			/* Build the older version of the index entry */

storage/innobase/srv/srv0mon.cc

@@ -674,7 +674,7 @@ static monitor_info_t	innodb_counter_info[] =
 	{"trx_rseg_history_len", "transaction",
 	 "Length of the TRX_RSEG_HISTORY list",
 	 static_cast<monitor_type_t>(
-	 MONITOR_EXISTING | MONITOR_DISPLAY_CURRENT),
+	 MONITOR_EXISTING | MONITOR_DISPLAY_CURRENT | MONITOR_DEFAULT_ON),
 	 MONITOR_DEFAULT_START, MONITOR_RSEG_HISTORY_LEN},
 
 	{"trx_undo_slots_used", "transaction", "Number of undo slots used",