dict0dict.cc

buf_LRU_drop_page_hash_for_tablespace(): Return whether any adaptive
hash index entries existed. If yes, the caller should keep retrying to
drop the adaptive hash index.

row_import_for_mysql(), row_truncate_table_for_mysql(),
row_drop_table_for_mysql(): Ensure that the adaptive hash index was
entirely dropped for the table.

Before this patch if no default database was set the server threw
an error for any table name reference that was not fully qualified by
database name. In particular it happened for table names referenced
CTE tables. This was incorrect.
The error message was thrown at the parser stage when the names referencing
different tables were not resolved yet.
Now if no default database is set and  a with clause is used in the
processed statement  any table reference is just supplied with a dummy
database name "*none*" at the parser stage. Later after a call
of check_dependencies_in_with_clauses() when the names for CTE tables
can be resolved error messages are thrown only for those names that
refer to non-CTE tables. This is done in open_and_process_table().

register changes of last_gtid

it only worked if mroonga plugin wasn't installed before (normal case),
but then it didn't need to delete anything.

if, by some glitch, mroonga was already installed, it would delete
mroonga from mysql.plugin, but INSTALL would fail (as mroonga was running),
and the script aborted, leaving mroonga not in mysql.plugin at all.

Solution for Debian/Ubuntu: install a trigger to restart mysqld
automatically whenever a package changes something in /etc/mysql
or in /etc/systemd/system/mariadb.service.d

remove manual libcurl3 dependency.
it'll be automatically added (libcurl3 or libcurl4, whatever was
actually used)

to avoid "using unique option prefix 'backup' is error-prone", there is
already --backup option there.

There is a tiny chance for race condition during MDL acquisition.

If table is renamed just prior to
SELECT 1 FROM <table_name> LIMIT 0

then this query would  fail, yet mariabackup --backup does not handle
it as fatal error and continues, only to fail later during file copy.

The fix is to die on error, of MDL lock query fails.

make it not to fail when `show engine innodb status` output
contains a double quote

don't tell users to set the root password if they're using unix_socket

there's an assert that catches it

Don't let SET PASSWORD to set the password, if auth_string is set.

Now SET PASSWORD always sets the plugin/auth_string fields and clears
the password field (on pre-plugin mysql.user table it works as before).

MDEV-14578: mysql_install_db install unix_socket plugin when --auth-root-authentication-method=socket

post-merge fixes

MDEV-14578: mysql_install_db install unix_socket plugin when --auth-root-authentication-method=socket

$ cmake -DPLUGIN_AUTH_SOCKET=STATIC ../mariadb-server-10.2/

$ scripts/mysql_install_db --datadir=/tmp/mysqldatadir-auth-static --auth-root-socket-user=dan --auth-root-authentication-method=socket --builddir=.  --srcdir=../mariadb-server-10.2
Installing MariaDB/MySQL system tables in '/tmp/mysqldatadir-auth-static' ...
OK

$ client/mysql -S /tmp/mysql.sock -e 'show create user dan@localhost'
+----------------------------------------------------------+
| CREATE USER for dan@localhost                            |
+----------------------------------------------------------+
| CREATE USER 'dan'@'localhost' IDENTIFIED VIA unix_socket |
+----------------------------------------------------------+

$ cmake -DPLUGIN_AUTH_SOCKET=NO ../mariadb-server-10.2/
$ scripts/mysql_install_db --datadir=/tmp/mysqldatadir-auth-none --auth-root-socket-user=dan --auth-root-authentication-method=socket --builddir=.  --srcdir=../mariadb-server-10.2
Installing MariaDB/MySQL system tables in '/tmp/mysqldatadir-auth-none' ...
ERROR: 1126  Can't open shared library '.../plugin/auth_socket/auth_socket.so' (errno: 2, cannot open shared object file: No such file or directory)
2018-05-01 11:38:56 0 [ERROR] Aborting

$ cmake -DPLUGIN_AUTH_SOCKET=DYNAMIC ../mariadb-server-10.2/
$ scripts/mysql_install_db --datadir=/tmp/mysqldatadir-auth-dyn --auth-root-socket-user=dan --auth-root-authentication-method=socket --builddir=.  --srcdir=../mariadb-server-10.2
Installing MariaDB/MySQL system tables in '/tmp/mysqldatadir-auth-dyn' ...
OK
$ ./sql/mysqld  --datadir=/tmp/mysqldatadir-auth-dyn  --lc-messages-dir=${PWD}/sql/share --plugin-dir=./plugin/auth_socket/
$ client/mysql -S /tmp/mysql.sock -e 'show create user dan@localhost'
+----------------------------------------------------------+
| CREATE USER for dan@localhost                            |
+----------------------------------------------------------+
| CREATE USER 'dan'@'localhost' IDENTIFIED VIA unix_socket |
+----------------------------------------------------------+

$ sudo make install
$ sudo chmod a+rwx /usr/local/mysql/data
$ cd /usr/local/mysql/
$ scripts/mysql_install_db  --auth-root-socket-user=dan --auth-root-authentication-method=socket
Installing MariaDB/MySQL system tables in './data' ...
OK
..
$ bin/mysqld_safe
$ client/mysql -S /tmp/mysql.sock -e 'show create user dan@localhost'
+----------------------------------------------------------+
| CREATE USER for dan@localhost                            |
+----------------------------------------------------------+
| CREATE USER 'dan'@'localhost' IDENTIFIED VIA unix_socket |
+----------------------------------------------------------+

Merges #767

In RPM/DEB packages - always ld-preload jemalloc, instead
of linking ha_tokudb.so with it.

Keep linking in bintars, because they don't install cnf files
in the correct locations.

This reverts commit d39629f0.

Because running mtr for many hours with no output whatsoever
is not really what we should do.

And in 5.5 `make test` just works anyway, nothing to fix here.

my_option strucures for command-line variables should be set properly.

settings.