from mysql-next-mr-bugfixing into mysql-trunk-bugfixing.

NOTE: the "utf8_phone_ci" collation does not exist in mysql-trunk yet,
so another collation with 2-byte collation ID is used: "utf8_test_ci".

This patch will be null-merged to mysql-next-mr-bugfixing.

Original revision:
------------------------------------------------------------
revision-id: bar@mysql.com-20091207121153-hs3bqbmr0719ws21
committer: Alexander Barkov <bar@mysql.com>
branch nick: mysql-next-mr.b47756
timestamp: Mon 2009-12-07 16:11:53 +0400
message:
  Bug#47756 Setting 2byte collation ID with 'set names' crashes the server
  
  The problem is not actually related to 2byte collation IDs.
  The same crash happens if you change the collation ID in
  mysql-test/str_data/Index.xml to a value smaller than 256.
  
  Crash happened in SQL parser, because the "ident_map" and "state_map"
  arrays were not initialized in loadable utf8 collations.
  
  Fix: adding proper initialization of the "ident_map" and "state_map"
  members for loadable utf8 collations.
------------------------------------------------------------

Added back n_frees, use 'clear' instead of 'free' since memory is
not freed here.

Before this patch, semisync assumed transactions running in parallel
can not be larger than max_connections, but this is not true when
the event scheduler is executing events, and cause semisync run out
of preallocated transaction nodes.

Fix the problem by allocating transaction nodes dynamically.

This patch also fixed a possible deadlock when running UNINSTALL
PLUGIN rpl_semi_sync_master and updating in parallel. Fixed by
releasing the internal Delegate lock before unlock the plugins.

    the 'shared-memory-base-name' was added to the [client] section of the my.cnf file.
    That option isn't supported by the mysqltest_embedded.

per-file comments:
  mysql-test/lib/My/ConfigFactory.pm

  In fact this crashes in normal (not embedded) run also.
  The problem is in the memory mapping. Handling the ha_myisammrg::extra(MMAP)
  the MERGE engine tries to mmap all the tables it unites.
  Though some can be empty and then in the mi_dynmap_file()
  we call the my_mmap(0). Normally this call returns MAP_FAILED,
  but not on FreeBSD. There it returns like a 'normal' value,
  and after the consequitive munmap systems gets unstable and
  crashes on some system call later.

per-file comments:
  storage/myisam/mi_dynrec.c
Bug #47139      Test "merge" crashes in "embedded" run
    don't try to mmap zero-length area, just return at once.

   The additional patch. That 'loadxml.test' failure was actually about our testing system,
   not the code.
   Firstly we need a new mysqltest command, wich i called 'send_eval'. So the expression
   can be evaluated, then started in a parallel thread. We only have separane 'send' and
   'eval' commands at the moment.
   Then we need to add the waiting code after the 'KILL' to our test, so the thread will be killed
   before the test goes further. The present 'reap' command doesn't handle the killed threads
   well.
      
per-file comments:
  client/mysqltest.cc
Bug#42520      killing load .. infile Assertion failed: ! is_set(), file .\sql_error.cc, line 8
    The 'send_eval' command implemented.

  mysql-test/r/loadxml.result
Bug#42520      killing load .. infile Assertion failed: ! is_set(), file .\sql_error.cc, line 8
   test result updated.

  mysql-test/t/loadxml.test
Bug#42520      killing load .. infile Assertion failed: ! is_set(), file .\sql_error.cc, line 8
   test case added.

(revision-id: kristofer.pettersson@sun.com-20091019074333-g2ces0lo4c2ejar7),
because it broke the tree and was reverted later.