Preprocessor macros USE_MYSYS_NEW and -DSAFEMALLOC are 
used to conditionally compile safemalloc or overwritten new/delete.

-Define dummy symbol in my_new.cc in case -DUSE_MYSYS_NEW is not set.
This avoids compiler/linker warnings about an essentially empty file
being compiled.

nes prefixed with .\ or ./
- Add my_basename() to mysys.
- Do not compile files that are not needed on Windows (my_addr_resolve, an
d safemalloc related stuff it it is not used)
Avoids linker warnings about compilation of essentially empty files.

The root cause is that after recent fixes around --debug variable (lp:909051)
the variable is now available in both release and debug builds, such that MTR
cannot tell a debug compiled server from optimized one.

To fix, assign a special default value 'disabled' for 'debug' variable in optimized build
and fix MTR to check for this special value to recognize optimized build.

find_set() in typelib.c expected a zero-terminated string

include/typelib.h:
  disable unused flags

client/CMakeLists.txt:
  1. -rdynamic is a linker flag.
  2. it should be used in all builds, not debug only
libmysql/get_password.c:
  prefer a standard function, when possible
  (otherwise a plugin will need to load it from the client)

sql/share/errmsg-utf8.txt:
  correct the error message, as query_cache_type variable is not read-ony anymore
sql/sql_cache.cc:
  the caller should verify that query cache resize
  is possible, before trying it
sql/sys_vars.cc:
  * test if qc resize is possible in the sysvar on_check() funntion,
    not in the on_update() function.
  * use the error message that better describes the problem

disable test cases for bugs not fixed in xtradb.

mysql-test/suite/innodb/t/innodb-lock.test:
  instead of disabling the test in disabled.def
  (hoping that somebody someday will enable it),
  skip it only for xtradb (not for innodb), and
  only for the current xtradb version. the test
  will enable itself automatically on the next
  xtradb merge
mysql-test/suite/innodb/t/innodb_cmp_drop_table.test:
  instead of disabling the test in disabled.def
  (hoping that somebody someday will enable it),
  skip it only for xtradb (not for innodb), and
  only for the current xtradb version. the test
  will enable itself automatically on the next
  xtradb merge
mysql-test/suite/innodb/t/innodb_corrupt_bit.test:
  fix the error number
storage/innobase/handler/ha_innodb.cc:
  fix the error number
storage/xtradb/handler/ha_innodb.cc:
  fix the error number

mysql-test/mysql-test-run.pl:
  strace-client does not take an argument

when looking for tables to delete from

e.g 
cmake . -DWITH_XXXX_STORAGE_ENGINE=1

fix misleading test name

Fix for LP BUG#908269 Wrong result with subquery in select list, EXISTS, constant MyISAM/Aria table.

Problem: When building the condition for JOIN::outer_ref_cond the optimizer forgot to take into account
that this condition could depend on constant tables as well.

Packagers may attempt to outsmart MariaDB/MySQL build system -DBUILD_SHARED_LIBS=1, we need to minimize the damage of such attempts.

mysql-test/r/information_schema_all_engines.result:
  Update result
mysql-test/t/information_schema_all_engines.test:
  Added --sorted-results as tables in information_schema are not sorted.

Split IO threads into ones that handle only read completion and ones that handle only write completion, as it was originally done, but got lost with "completion port" patch. The reason we need to have dedicated read and dedicated write threads is that read completion routine can block waiting for write io to complete, and in rare cases where all io threads are handling async reads, it can deadlock.