remeber a real backtrace for every allocation.
make safemalloc to tract C++ new/delete too.
collateral fixes to make the test suite pass.

- leaking mutex in lf_hash_destroy
- pthread_getspecific() before pthread_key_create() in my_thread_var_dbug()
(called by static C++ object constructors called in sys_vars)
- perfschema destroys mutexes that were  not created.

Default to no SAFEMALLOC on Windows, because C runtime malloc
has this functionslity already

Replace statements connected with bitwise OR with series of "if"s.
The later is guaranteed to execute in order, bitwise OR does not have
specific order for statement execution.

cmake/maintainer.cmake:
  don't do -Werror just yet
config.h.cmake:
  according to MSDN PSAPI_VERSION should be 1 in a portable application
mysys/my_thr_init.c:
  first, reset THR_KEY_mysys, and then free dbug data,
  because dbug data are automacially created on the next dbug call,
  unless THR_KEY_mysys is null.

- build executables we have in 5.3 (mysql_install_db.exe, mysq_upgrade_service.exe, upgrade wizard), and MSI

- add some missing headers to windows specific source files. 
This needs to be done since 5.5 is using WIN32_LEAN_AND_MEAN preprocessor constant thus windows.h 
no more includes whiole Windows

- do not deliver perl scripts  (mysql_install_db.pl & friends)  -they do not work, are not documented, and we 
have native executables for this functionality. do not pack echo.exe, replace.exe  into MSI, they 
are not needed. Do not build resolveip on Windows, it is not used.

- precache results of  of system checks in cmake/os/WindowsCache.cmake (like it is alreay done for  majority of tests
to speed up cmake run with VS)

- make feedback plugin DEFAULT on Windows (so MSI works if user enables plugin), 
fix null pointer access in PSI_register

A variable XTRADB_PREFER_STATIC_LIBAIO should be set to 1 (or TRUE
 or ON) for static linking.

Even if mysqld can avoid dependency on shared libaio, shared libraries 
libmysqld.so or ha_innodb.so cannot link without it.

Given that the patch primarily addresses building tar.gz package, and 
shared libraries mentioned above deemed less important than mysqld
executable, we accept shared lib dependency on libaio.so

Hardcode -MariaDB suffix for MYSQL_SERVER_VERSION in mysql_version.h

Make max_user_connections signed, with min allowed value being -1.

So removed.

the depth of subquery nestedness to less than 31
(sizeof(ulong)-1).

This bug in the function Loose_scan_opt::check_ref_access_part1 could lead
to choosing an invalid execution plan employing a loose scan access to a
semi-join table even in the cases when such access could not be used at all.
This could result in wrong answers for some queries with IN subqueries.

Analysis:
The optimizer distinguishes two kinds of 'constant' conditions:
expensive ones, and non-expensive ones. The non-expensive conditions
are evaluated inside make_join_select(), and if false, already the
optimizer detects empty query results.

In order to avoid arbitrarily expensive optimization, the evaluation of
expensive constant conditions is delayed until execution. These conditions
are attached to JOIN::exec_const_cond and evaluated in the beginning of
JOIN::exec. The relevant execution logic is:

JOIN::exec()
{
  if (! join->exec_const_cond->val_int())
  {
    produce an empty result;
    stop execution
  }
  continue execution
  execute the original WHERE clause (that contains exec_const_cond)
 ...
}

As a result, when an expensive constant condition is
TRUE, it is evaluated twice - once through
JOIN::exec_const_cond, and once through JOIN::cond.
When the expensive constant condition is a subquery,
predicate, the subquery is evaluated twice. If we have
many levels of subqueries, this logic results in a chain
of recursive subquery executions that walk a perfect
binary tree. The result is that for subquries with depth N,
JOIN::exec is executed O(2^N) times.

Solution:
Notice that the second execution of the constant conditions
happens inside do_select(), in the branch:
if (join->table_count == join->const_tables) { ... }
In this case exec_const_cond is equivalent to the whole WHERE
clause, therefore the WHERE clause has already been checked in
the beginnig of JOIN::exec, and has been found to be true.
The bug is addressed by not evaluating the WHERE clause if there
was exec_const_conds, and it was TRUE.

set to 'on' by default.