When we open merge children open error is normally handled early during
open phase. But there are two exceptions when error is handled later during
attach phase: CHECK/REPAIR TABLE and tables added by the pre-locking code.

The latter case wasn't considered by assertion in the merge code. This
assertion is corrected. Note that in MySQL-5.6 this assertion is removed.

Merging the following MySQL-5.6 changes:
- WL#5624: Collation customization improvements
  http://dev.mysql.com/worklog/task/?id=5624

- WL#4013: Unicode german2 collation
  http://dev.mysql.com/worklog/task/?id=4013

- Bug#62429 XML: ExtractValue, UpdateXML max arg length 127 chars
  http://bugs.mysql.com/bug.php?id=62429
  (required by WL#5624)

Analysis: There is debug assertion ut_ad(trx->start_file != 0); and ut_ad(trx->start_line != 0); on trx_start_low funcition at trx0trx.cc. These fields are initialized on include/trx0trx.h at function trx_start_if_not_started_xa. Thus at trx_prepare_for_mysql function should call trx_start_if_not_started_xa(trx); not trx_start_if_not_started_xa_low(trx) directly;

a special handling.

(pre fts)

(pre fts).

sql/create_options.cc:
  an utility function to compare two filled-in engine_option structures
sql/sql_table.cc:
  * two keys are different if their option_struct's differ
    (for ALTER TABLE DROP key, ADD key)
  * engines doing inplace alter must see the new frm image

Fix test results accordingly.

that is not run for embedded server

now, if the dynamic plugin puts itself in a distinct component, then
* ${plugin}.cnf file is generated and installed automatically
* a separate RPM is generated automatically
* necessary FILELIST/PROVIDES/OBSOLETES/REQUIRES are filled in automatically

Other fix of maybe_null problem and revert of revno: 3608 "MDEV-3873 & MDEV-3876 & MDEV-3912 : Wrong result (extra rows) with ALL subquery from a MERGE view."

Bundled jemalloc can only be on  Linux and OSX without problems.
  
On BSDs, build fails because make does not understand GNU extensions (also BSDs do not need  jemalloc, it is already system malloc).
On Solaris, build fails with compile error.

- Better error message when using huge pages
- Fixed link error
- Test suite should run even on system with huge pages

storage/tokudb/ft-index/cmake_modules/TokuThirdParty.cmake:
  Fixed that linking works on systems that uses lib64
storage/tokudb/ft-index/portability/huge_page_detection.cc:
  Better error message
storage/tokudb/mysql-test/rpl/suite.pm:
  Test suite should run even on system with huge pages
storage/tokudb/mysql-test/tokudb/suite.pm:
  Test suite should run even on system with huge pages

we don't build innodb plugin in 10.0 yet

SIZEOF_INT can never be 8. Removing the redundant #ifdef code.

Updated information in SHOW AUTHORS and SHOW CONTRIBUTORS
SHOW AUTHORS and SHOW CONTRIBUTORS are not depricated anymore.

CMakeLists.txt:
  Install TODO and CREDITS files
CREDITS:
  Information about the MariaDB foundation
README:
  Updated to include the MariaDB foundation
mysql-test/r/contributors.result:
  Updated contributors
mysql-test/r/show_check.result:
  SHOW AUTHORS and SHOW CONTRIBUTORS are not depricated anymore.
scripts/make_binary_distribution.sh:
  Added TODO and CREDITS
sql/authors.h:
  Updated list of active authors
sql/contributors.h:
  Updated list of active contributors and members of the MariaDB foundation
sql/sql_show.cc:
  Extend comment lengths
sql/sql_yacc.yy:
  SHOW AUTHORS and SHOW CONTRIBUTORS are not depricated anymore

MDEV-4735 - Assertion `! is_set()' fails in Diagnostics_area::set_ok_status on
            attempt to create a temporary SPIDER table connecting to
            non-existing source
MDEV-4737 - Server crashes in spider_mysql_handler::append_match_against on
            SELECT .. MATCH .. AGAINST .. BOOLEAN MODE
MDEV-4738 - INSERT DELAYED on a SPIDER table doesn't produce
            ER_DELAYED_NOT_SUPPORTED, doesn't work like delayed, and doesn't
            honor lock_wait_timeout like normal INSERT

- merged test cases for MySQL bug#27480
- fixed that LOCK TABLES was unable to open temporary table
  (covered by grant2 test, merged appropriate code from 5.6)
- commented lines that cause server crash in merge test, reported
  MDEV-5042 (not relevant to bug#27480)