the 'drop z' method reduced the impact of noise on omegaI, but it also
made us more sensitive to errors in accelerometer calibration and
scaling, as demonstated by the logs from Gabor here:

  http://diydrones.com/xn/detail/705844:Comment:834373

Simulation testing shows that the other noise suppression methods
applied in the DCM code, in particular the slope limiting on omegaI
the removal of the weighting and the upcoming use of a _omega_I_sum
buffer have reduced the impact of noise enough that we can now safely
include z in the acceleration calculation.

DataFlash_APM2 - moved CS_inactive call (which disables the dataflash) from the beginning to the end of all methods.  This means the dataflash does not monopolize the SPI bus.
    Also formatting changes to use tab instead of space. Sorry, should have done this as a separate check-in to the above changes.

this allows users to change the yaw gain in DCM

this adds local and remote radio noise levels

AP_MotorsMatrix - fixed stability patch issue in which it would not limit a motor's output unless an opposite motor had been defined.  This would only have affected Y6 frames.

AP_Motors - fixed last "shadowing" compiler warning caused by badly named parameter in auto_armed method.

AP_MotorsMatrix - fixed another compiler error regarding shadowing a variable.  Fixed by removing those  the angle and direction arrays which weren't even used

AP_MotorsHeli - fixed compiler warning re move_swash function declared with int parameters in .h file but actually using int16_t in .cpp file.  They're the same in arduino but perhaps the build checker is a little more strict that Arduino.

AP_Motors - fix compiler warning re shadowing a variable caused by badly named parameter in armed function.
Another compiler warning fixed in AP_MotorsMatrix.cpp caused by declaring "i" twice.

     1. it's step towards rearchitecting the current code base ahead of a move to RTOS.
     2. internally it uses a MatrixTable for the Quad, Octa, OctaQuad, Y6 frames.
     3. it implements the missing stability patch for Octa and OctaQuads (still missing for Y6)

Later check-ins will incorporate into the main ArduCopter code.

this allows us to put a serial port into non-blocking mode, so that
writes that don't fit in the transmit buffer are dropped. This will be
used in flight to prevent stray printf() calls from causing large time
delays in the code

when this is 1 (which is the default), we will get the declination
automatically via the AP_Declination library

when it is 0 we will use the value configured by the user

there was a comment saying this was too slow, but it actually costs
about 30 usec extra, which is trivial given the full auto declination
call costs 680 usec and its only called once per boot

AP_Declination: Update LUT based on the changes that I made in the LUT generation code. This LUT is the final version that has no truncation.

AP_Declination: Removed trailing zero deltas as they have no impact on the final value and only take up extra bytes.

AP_Declination: Updated compressed lookup value table. Located 9 more failure points due to truncation on the packed LUT.

AP_Declination: Found 7 of the 22 failure points and fixed them. The issue with those 7 was the way I originally packed the value array. The last delta values on some rows were getting truncated. The new number of failures is down to 15 - WIP.

AP_Declination: Updated test to run in 5 degree increments. Changed to only print failures and also print total pass vs total fail.