1. Always drop merged_for_insert flag on cleanup (there could be errors which prevent TABLE to be assigned)
2. Make more precise cleanup of select parts which was touched

st_select_lex::handle_derived() and mysql_handle_list_of_derived() had
exactly the same implementations.

- Adding a new method LEX::handle_list_of_derived() instead
- Removing public function mysql_handle_list_of_derived()
- Reusing LEX::handle_list_of_derived() in st_select_lex::handle_derived()

candle.exe's preprocessor flags (-dHaveUpgradeWizard=0 -DHaveInnodb=1)
were not passed correctly to EXECUTE_PROCESS

Fix is to make a list out of the EXTRA_WIX_PREPROCESSOR_FLAGS string,
and use the preprocessor flags list in EXECUTE_PROCESS.

For multi-table views with LOAD, updates are not allowed, so
we should just throw an error.

ha_innobase::commit_inplace_alter_table(): Do not crash if
innobase_update_foreign_cache() returns an error. It can return
an error on ALTER TABLE if an inconsistent FOREIGN KEY constraint
was created earlier when SET foreign_key_checks=0 was in effect.
Instead, report a warning to the client that constraints cannot
be loaded.

ha_innobase::prepare_inplace_alter_table(): Filter out duplicates
from ha_alter_info->alter_info->drop_list.elements.

innobase_rename_column_try(): Declare fk_evict as std::set
instead of std::list, in order to filter out duplicates.

For singe-table views, we need to find the bottom most base table in the embedded views
and then update that table

@@open_files_limit now behaves differenly and cannot
be used to skip the test anymore.

Reverting part of da34c7de that was already fixed by MDEV-17531 by Marko

Disable LOAD DATA LOCAL INFILE suport by default and
auto-enable it for the duration of one query, if the query
string starts with the word "load". In all other cases the application
should enable LOAD DATA LOCAL INFILE support explicitly.

If the rlimit.rlim_cur value returned by getrlimit is not the
RLIM_INFINITY magic constant, but a *very* large number, we can allocate
too many open files. Restrict set_max_open_files to only return at most
max_file_limit, as passed via its parameter.

commit 6a6a1f37798
Author: Olivier Bertrand <bertrandop@gmail.com>
Date:   Fri Jan 4 12:31:52 2019 +0100

    - Fix a few bug mainly concerning discovery and call from OEM
      (and prepare new table types)

      modified:   storage/connect/tabjson.cpp
      modified:   storage/connect/tabjson.h
      modified:   storage/connect/tabxml.cpp
      modified:   storage/connect/tabxml.h

    - Fix wrong line estimate
      modified:   storage/connect/mysql-test/connect/r/part_table.result
      modified:   storage/connect/mysql-test/connect/t/part_table.test

commit bd7d2e912d9
Author: Olivier Bertrand <bertrandop@gmail.com>
Date:   Tue Dec 4 23:35:09 2018 +0100

    Fix wrong version number

commit 4933680e7ab
Author: Olivier Bertrand <bertrandop@gmail.com>
Date:   Sun Dec 2 00:25:05 2018 +0100

    - Make PlugSubAlloc to be exportable
      Suppress unused parameter from PlugSubSet
      modified:   storage/connect/global.h
      modified:   storage/connect/plugutil.cpp
      modified:   storage/connect/jsonudf.cpp
      modified:   storage/connect/tabjson.cpp
      modified:   storage/connect/user_connect.cc

    - Fix a bug making column catalog XML tables fail
      modified:   storage/connect/tabxml.cpp

    - Comment out wrong message
      modified:   storage/connect/ha_connect.cc

    - Update error message when sorting an ODBC table fails
      modified:   storage/connect/tabodbc.cpp

    - Add error message when gettting an address
      from an OEM fails.
      modified:   storage/connect/reldef.cpp

    - Make some modifications useful for OEM module writting
      Export discovery functions for CSV, JDBC and XML
      Remove unuseful include from tabjson.h
      Move TDBXML::data_charset function from header file to source
      modified:   storage/connect/tabfmt.h
      modified:   storage/connect/tabjson.h
      modified:   storage/connect/tabxml.cpp
      modified:   storage/connect/tabxml.h

    - Update test result
      modified:   storage/connect/mysql-test/connect/r/jdbc_oracle.result

remove attempts to track "candidate keys", use what was already
decided in create_table_impl().

Calling st_select_lex::update_used_tables in JOIN::optimize_unflattened_subqueries
only when we are sure that the join have not been cleaned up.
This can happen for a case when we have a non-merged semi-join and an impossible
where which would lead to the cleanup of the join which has the non-merged semi-join

pcre needs symbols from pcreposix
(but this is only an issue when linking with system static libraries)

Problem:
========
Server fails to notify the engine by not setting the ADD_PK_INDEX and
DROP_PK_INDEX When there is a
 i) Change in candidate for primary key.
 ii) New candidate for primary key.

Fix:
====
Server sets the ADD_PK_INDEX and DROP_PK_INDEX while doing alter for the
above problematic case.

Close connection handler on connection failure. This fixes 14 failing tests in
main suite under clang+ASAN build.

ASAN report for main.connect looks like this:
=================================================================
==25495==ERROR: LeakSanitizer: detected memory leaks

Direct leak of 146280 byte(s) in 115 object(s) allocated from:
    #0 0x4fba47 in calloc /fun/cpp_projects/llvm_toolchain/llvm/projects/compiler-rt/lib/asan/asan_malloc_linux.cc:138
    #1 0x5a7a02 in mysql_init /work/mariadb/libmariadb/libmariadb/mariadb_lib.c:977:26
    #2 0x570a7a in do_connect(st_command*) /work/mariadb/client/mysqltest.cc:6096:26
    #3 0x584c39 in main /work/mariadb/client/mysqltest.cc:9321:9
    #4 0x7fd15514db96 in __libc_start_main /build/glibc-OTsEL5/glibc-2.27/csu/../csu/libc-start.c:310

Indirect leak of 7065600 byte(s) in 115 object(s) allocated from:
    #0 0x4fb80f in __interceptor_malloc /fun/cpp_projects/llvm_toolchain/llvm/projects/compiler-rt/lib/asan/asan_malloc_linux.cc:129
    #1 0x637a83 in my_context_init /work/mariadb/libmariadb/libmariadb/ma_context.c:367:23
    #2 0x59fd16 in mysql_optionsv /work/mariadb/libmariadb/libmariadb/mariadb_lib.c:2738:9
    #3 0x5bc1d4 in mysql_options /work/mariadb/libmariadb/libmariadb/mariadb_lib.c:3242:10
    #4 0x570b94 in do_connect(st_command*) /work/mariadb/client/mysqltest.cc:6103:7
    #5 0x584c39 in main /work/mariadb/client/mysqltest.cc:9321:9
    #6 0x7fd15514db96 in __libc_start_main /build/glibc-OTsEL5/glibc-2.27/csu/../csu/libc-start.c:310

Indirect leak of 940240 byte(s) in 115 object(s) allocated from:
    #0 0x4fb80f in __interceptor_malloc /fun/cpp_projects/llvm_toolchain/llvm/projects/compiler-rt/lib/asan/asan_malloc_linux.cc:129
    #1 0x64386e in ma_init_dynamic_array /work/mariadb/libmariadb/libmariadb/ma_array.c:49:31
    #2 0x649ead in _hash_init /work/mariadb/libmariadb/libmariadb/ma_hash.c:52:7
    #3 0x5a3080 in mysql_optionsv /work/mariadb/libmariadb/libmariadb/mariadb_lib.c:2938:13
    #4 0x5bc20c in mysql_options4 /work/mariadb/libmariadb/libmariadb/mariadb_lib.c:3248:10
    #5 0x56f63b in connect_n_handle_errors(st_command*, st_mysql*, char const*, char const*, char const*, char const*, int, char const*) /work/mariadb/client/mysqltest.cc:5874:3
    #6 0x57146b in do_connect(st_command*) /work/mariadb/client/mysqltest.cc:6193:7
    #7 0x584c39 in main /work/mariadb/client/mysqltest.cc:9321:9
    #8 0x7fd15514db96 in __libc_start_main /build/glibc-OTsEL5/glibc-2.27/csu/../csu/libc-start.c:310
...

Closes #809

           32 bit int

Row-based slave applier could not parse correctly the table id when
the value exceeded the max of 32 bit unsigned int.
The reason turns out in that the being parsed value placeholder
was sized as 4 bytes.

The type is fixed to ulonglong.

Additionally the patch works around Rows_log_event::m_table_id 4 bytes
size on 32 bits platforms. In case of last_table_id value overflows
the 4 byte max, there won't be the zero value for m_table_id generated
and the first wrapped-around value is one, this is thanks to excluding
UINT_MAX32 + 1 from TABLE_SHARE::table_map_id.

don't run `su - mysql` is $USER is already mysql

test case

Issue:
------
When a subquery contains UNION the count of the number of
subquery columns is calculated incorrectly. Only the first
query block in the subquery's UNION is considered and an
array indexing goes out-of-bounds, and this is caught by an
assert.

Solution:
---------
Sum up the columns from all query blocks of the query
expression.

Change specific to 5.6/5.5:
---------------------------
The "child" points to the last query block of the UNION
(as opposed to 5.7+ where it points to the first member of
UNION). So "child->master_unit()->first_select()" is used
to reach the first query block of UNION.