new make target - abi_update

libservices/HOWTO:
  remove references to Makefile.am
  small tweaks

mysql-test/suite/innodb/t/group_commit_crash.test:
  remove autoincrement to avoid rbr being used for insert ... select
mysql-test/suite/innodb/t/group_commit_crash_no_optimize_thread.test:
  remove autoincrement to avoid rbr being used for insert ... select
mysys/my_addr_resolve.c:
  a pointer to a buffer is returned to the caller -> the buffer cannot be on the stack
mysys/stacktrace.c:
  my_vsnprintf() is ok here, in 5.5

bug in semisync plugin. It didn't check thd->killed before waiting on mysys->current_cond,
and thus an attepmt to kill the thread (on shutdown) was lost

plugin/semisync/semisync_master.cc:
  mdev:201 lp:962540

include/mysql_com.h:
  remove "shutdown levels" that aren't shutdown levels from mysql_enum_shutdown_level
mysys/my_addr_resolve.c:
  my_snprintf in 5.5 (but not in 5.3) supports %p
sql/item_func.cc:
  use a method (that exists only in 5.5) instead of directly accessing a member
sql/item_subselect.cc:
  use a method (that exists only in 5.5) instead of directly accessing a member
sql/opt_subselect.cc:
  use a method (that exists only in 5.5) instead of directly accessing a member
sql/sql_select.cc:
  use a method (that exists only in 5.5) instead of directly accessing a member

- Use native memcmp() supplied with C runtime instead of hand-unrolled loop ptr_compare_N loop
Prior to fix  ptr_compare_0()  has  3.7% samples in OLTP-RO in-memory. 
Fix brings this down to 1.8% (all memcmp samples)

- Innodb : fix UT_RELAX_CPU to  be defined  as YieldProcessor, as  was also originally intended 
(but intention was lost in the #ifdef maze

This reduces number of ut_delay() samples in profile from 1.5% to 0.5%

Improve filesort performance for small sorts: Don't write pointers to records that we will never use.

        modified for MySQL 5.5. Logger service moved to the
        plugin/sql_errlog directory to be properly used later.

plugin/sql_errlog/sql_errlog.c:
  Fixes for bugs #956427 (SQL_ERROR_LOG plugin produces bogus warnings about sql-error-log-size-limit value) and #956463 (Server crashes if SQL_ERROR_LOG fails to initialize) they're also MDEV-184 and MDEV-183
    
  The sql_error_log_deinit() should be prepared for the logger_file to be NULL.
  The logger_file_size_limit upper limit wasn't properly set.

it makes "./mtr --debug" unusable

revert the mysql fix for bug#46165.
implement shared FILE's with reference counting

- Don't call update_virtual_fields() if table->vfield is not set
- Don't prealloc memory for in open_tables() as this is very seldom used.


sql/records.cc:
  Don't call update_virtual_fields() if table->vfield is not set
sql/sql_base.cc:
  Don't prealloc memory for in open_tables() as this is very seldom used.
  Don't call update_virtual_fields() if table->vfield is not set
sql/sql_delete.cc:
  Don't call update_virtual_fields() if table->vfield is not set
sql/sql_handler.cc:
  Don't call update_virtual_fields() if table->vfield is not set
sql/sql_join_cache.cc:
  Don't call update_virtual_fields() if table->vfield is not set
  Move some frequent values to local variables
sql/sql_table.cc:
  Don't call update_virtual_fields() if table->vfield is not set
sql/sql_update.cc:
  Don't call update_virtual_fields() if table->vfield is not set
sql/table.cc:
  Assert if update_virtual_fields is called with wrong parameters

- Remove all references of MAX_TABLES from JOIN struct and make these dynamic
- Updated Join_plan_state to allocate just as many elements as it's needed


sql/opt_subselect.cc:
  Optimized version of Join_plan_state
sql/sql_select.cc:
  Set join->positions and join->best_positions dynamicly
  Don't call update_virtual_fields() if table->vfield is not set.
sql/sql_select.h:
  Remove all references of MAX_TABLES from JOIN struct and Join_plan_state and make these dynamic

- Optimize away calls to hp_rec_hashnr() by cashing hash
- Try to get more rows / block (to minimize overhead of HP_PTRS) in HEAP tables.


storage/heap/_check.c:
  Optimize away calls to hp_rec_hashnr() by cashing hash.
  Print cleanups
storage/heap/heapdef.h:
  Added place to hold calculated hash value for row
storage/heap/hp_create.c:
  Try to get more rows / block (to minimize overhead of HP_PTRS)
storage/heap/hp_delete.c:
  Optimize away calls to hp_rec_hashnr() by cashing hash.
storage/heap/hp_hash.c:
  Optimize away calls to hp_rec_hashnr() by cashing hash.
  Remove some not needed DBUG_PRINT
storage/heap/hp_test2.c:
  Increased max table size as now heap tables takes a bit more space (a few %)
storage/heap/hp_write.c:
  Optimize away calls to hp_rec_hashnr() by cashing hash.
  Remove duplicated code
  More DBUG_PRINT
storage/maria/ma_create.c:
  More DBUG_PRINT

suppress these harmless but confusing warnings.
fix the program name (MY_INIT) in mysqldump

client/mysqldump.c:
  for backward compatibility, prefix mysqldump error messages with "mysqldump", not with the full path of the executable

If the first component of a ref key happened to be a constant appeared
after constant row substitution then no store_key element should be
created for such a component. Yet create_ref_for_key() erroneously could
create such an element that caused construction of invalid ref keys and
wrong results for some joins.

Do not use Valgrind client requests in a normal release build, they have a small but noticable performance impact.

lp:933959 Assertion `0' failed in net_end_statement(THD*) on concurrent SELECT FROM I_S.INNODB_SYS_INDEXES and ALTER TABLE

Workaround: report a generic error if an I_S plugin failed silently.

mark the corresponding I_S table as OPTIMIZE_I_S_TABLE, to let the I_S optimizer
figure out whether files need to be opened, and don't open the tables unless
I_S optimizer says so.

This bug in the constructor SEL_IMERGE::SEL_IMERGE could 
cause huge excessive memory requests.

Make the slave options --replicate-* dynamic variables so that these
options can be changed dynamically while the server is running,
which enables users to modify replication filtering rules without
having to stop and restart the server.

This is accomplished by just requiring that the slave threads are
stopped when these options are set dynamically. Since filtering
rules are only used by the SQL slave thread, setting them while the
thread is not running avoids the need for locking.

Fix race condition in rpl_stop_start_slave. after kill connection, wait until it is gone in processlist

 - Avoid needless load/stores in my_hash_sort_simple due to possible aliasing
 - Avoid expensive Join_plan_state constructor in choose_subquery_plan when no subquery
 - Avoid calling update_virtual_fields for every row when no virtual fields.

No commit message

Fixed lp:947474 "Assertion `table->file->stats.records > 0 || error' failed in join_read_const_table on concurrent SELECT and ALTER, constant Aria table"
Remove Aria state history for drop/rename

mysql-test/suite/maria/r/maria-recovery2.result:
  Updated old (wrong) test result
sql/handler.cc:
  Fixed wrong argument to implict_commit
storage/maria/ha_maria.cc:
  Ensure that we don't use file->trn if THD_TRN is 0. (This means that implict_commit() has been called and the trn object is not ours anymore)
storage/maria/ma_extra.c:
  Remove Aria state history for drop/rename
storage/maria/ma_rename.c:
  Remove Aria state history for rename
storage/maria/ma_state.c:
  More DBUG_PRINT

- Don't use SAFEMALLOC on valgrind builds (slows things down)
- Added back lost option from 5.3: debug-mutex-deadlock-detector
- Flush pages before taking lock mutex (speeds up closing of Aria tables).

BUILD/SETUP.sh:
  - Don't use SAFEMALLOC on valgrind builds (slows things down)
sql/lock.cc:
  Make default argument explicit (improves readability)
sql/mysqld.cc:
  Removed compiler warnings
  Sorted debug options alphabetically
  Added back lost option from 5.3: debug-mutex-deadlock-detector
storage/maria/ma_close.c:
  Flush pages before taking lock mutex (speeds up closing of Aria tables).
storage/maria/ma_open.c:
  More DBUG_PRINT
storage/maria/maria_def.h:
  Better DBUG_PRINT
storage/maria/trnman.c:
  Better DBUG_PRINT

- Take into account that there may exist Item_field objects with context==NULL.

BUG#952372: Server crashes on 2nd execution of PS in find_field_in_tables with semijoin+materialization
- The problem was that convert_subq_to_jtbm() attached the semi-join
  TABLE_LIST object into the wrong list: they used to attach it to the 
  end of parent_lex->leaf_tables.head()->next_local->...->next_local.
  This was apparently inccorect, as one can construct an example where 
  JTBM nest is attached to a table that is inside some mergeable VIEW, which
  breaks (causes crash) for name resolution on the subsequent statement
  re-execution.
- Solution: Attach to the "right" list. The "wording" was copied from
  st_select_lex::handle_derived.