update CMSIS version to 3.20

fix msh_exec(char* cmd, rt_size_t length) bug

I have to mention that this bug is introduced by myself, in the last
commit <fd07f30f511335f506ae6557e300093d4dde421e> to dfs_nfs.c

fix dfs_unmount bug

more infomation:
http://www.rt-thread.org/phpBB3/topic2965.html

fix #211

empty is confused in dfs_filesystem_lookup.

clean code with pointer rather than index

decrease indentation to make cleaner code
fix mkfs bug when there is no mkfs implementation

1. rt_console_get_device should be call to set shell device only when it is null, or it will make finsh_set_device useless.
2. RT_USING_CONSOLE can be disabled in bsp/rtconfig.h

On Windows platform, please set `autocrlf = true` and on Linux, please
set `autocrlf = input`.

Signed-off-by: armink <armink.ztl@gmail.com>

Signed-off-by: armink <armink.ztl@gmail.com>

This reverts commit 07bd09fc0f.
The validation of the context is done with in `rt_set_errno`. So it's
safe to call it here.

The errno is associated with threads. It will crash the thread if we set
it in interrupt context.

Return early if the size is 0 as there is nothing to do.

Finsh ex

We split the history handling form the key handling. So we could handle
the direction key even if the history is disabled. As a "side effect", I
also remove the unnecessary "use_history" bit.

UART DMA tx_complete() with buffer information

_TIME_T_DEFINED was pre-defined in order to build jffs2 with VC.
While it may cause many side effects.

DFS: add dfs_file_flush prototype in dfs_file.h

The first parameter of memmove is the dst, not src. Besides, fix the
wrong move size.

components/utilities: add ymodem

Real-YModem implemented a flexible YModem support. It use callback-based
structure to let the user application to deal with the data. It contains
3 examples:

    1. echo.c: write the data recieved on YModem to an other device
    2. null.c: discard the YModem data
    3. tofile.c: write the data to the file system

Currently, it does not support batch file transmission.

found by haitao5198
clear mount parition device talbe when jffs2 mount falied

found by haitao52198
URL:http://www.rt-thread.org/phpBB3/viewtopic.php?f=3&t=3112&p=17153#p17153

*buffer) with buffer pointer

logtrace only use odd number of log levels. So the use could set some
custom level amount them.

Because the device could still remain opened when closed by finsh, the
old rx_indicate is useless for finsh. Some buggy driver will still
generate rx_indicate even after the device has been closed. So FinSh
should unregister the rx_indicate when releasing the old device.

With this patch, one could define two alias for the same function like:

    FINSH_FUNCTION_EXPORT_ALIAS(rt_hw_cpu_reset, reboot, reboot the cpu);
    FINSH_FUNCTION_EXPORT_ALIAS(rt_hw_cpu_reset, __cmd_reboot, reboot the cpu);

Suitable for being used in MSH/Finsh.

When forwarding rx/tx callback from the underlaying device(pipe), the
"dev" argument should be the portal. So the portal callback could think
itself as called.

It does not make sense to open portal without a flag. One should open a
portal with RT_DEVICE_OFLAG_RDWR in most cases.

Portal is a device that connect devices. Currently, you can only connect
pipes in portal. Pipes are unidirectional. But with portal, you can
construct a bidirectional device with two pipes.

Pipe have many differences with char device. The main difference is
altough pipe have both read and write, it is unidirectional in nature.

Some reader(FinSh again) would like to block on a semaphore which is
released in rx_indicate. So we should invoke rx_indicate in
_rt_pipe_resume_reader.

The previous implementation will always blocks the reader/writer.
However, at least FinSh would expect the device to be nonblocking ---
read should return 0 when there is no data in it.

Add the APIs that will discard the old data when rb is full.

RT_RINGBUFFER_SIZE could mean "the size of the whole buffer", "the size
of the empty space" or "the size of the data". Moreover, it's never a
micro anymore. Change it to rt_ringbuffer_data_len before it's too late.
Also, RT_RINGBUFFER_EMPTY is changed to rt_ringbuffer_space_len.