Gfex calib
This merge request implements the calibration as it is done in the Csim with hard code values for the noise cuts and slopes.
Integer arithmetic is used to insure that calculation is the same on all architectures.
Currently all slopes are set to their nominal value of 1024. The only noise cuts are set to -100 except for the column 0 and 11 of FPGA C where they are 176 = 0xB0>
The explicit noise cuts added in an earlier merge have been removed.
Changes have been made to gFEXFPGA.cxx where the calibration method has been introduced the same way as in the Csim.
Eventually we will add a connection to COOL and the values used to generate the lookup tables can be retrieved.
For each channel we store an offset, noise cut and slope.