* [kernel][mem] Multiple instances of small memory allocation algorithm

* [kernel][mem] Change small memory management algorithm memory header flag

* [kernel][mem] Fix assertion problem

* [kernel][slab] Multiple instances of slab memory management algorithm

* [kernel][memheap] Remove rt_malloc/rt_free/rt_realloc and other related memory interfaces

* [kernel][mem] Clean up memory space of small memory management objects

* [kernel][kservice] Add memory application interface and thread protection interface

* [kernel][kservice] Fix function return value problem

* [kernel][memheap] Optimize memheaptrace print

* [kernel][memheap] Support best mode

* [kernel][memory] Remove semaphore lock

* [kernel][memheap] Add locked flag

* [kernel][memory] Support malloc memory in interrupt

* [kernel][memheap] Add 'memheapcheck' cmd

* [kernel][mem] Fix failure to request full memory

* [kernel][memheap] Fix compilation warning

* [kernel][mem] Fix mem realloc ASSERT

* [examples][testcases] Add small mem testcase

* [examples][mem_tc] Modify test memory size

* [examples][testcases] Add slab memory management algorithm test case

* [examples][testcases] fix small memory management algorithm test case

* [kernel][memory] Adjusting memory allocation algorithm object definition and interface

* [kernel][memory] Fix compilation warning

* [examples][utest] Fix mem test case

* [examples][utest] fix slab test case

* [utest][testcases] Shorten test time

* [kernel][memory] Formatting code

* [examples][utest] Adjust test run time

* [examples][utest] Formatting code

* [bsp] update all rtconfig.h

Signed-off-by: NWillianChan <chentingwei@rt-thread.com>

Signed-off-by: NWillianChan <chentingwei@rt-thread.com>

Signed-off-by: NWillianChan <chentingwei@rt-thread.com>

Update license for RT-Thread kernel.

- missing free old memory after malloc new
- free memory when new size is zero

This will ease the debugging of double-free bug.

When realloc a bigger space and the next node is free and big enough, we
should directly relocate the next node instead of doing alloc/memcpy.
The new method not only faster in this cases, it would avoid memory
fragment as well.

convert Tabs to spaces

git-svn-id: https://rt-thread.googlecode.com/svn/trunk@2569 bbd45198-f89e-11dd-88c7-29a3b14d5316