MPORTANT: This new option requires lua-md5 library

	sudo apt-get install lua-md5

We prefer to use size and shift than to get the size by reading the mask
because the bits in mask might not be contiguous.

This commit lets the user define the fields in the .wb that are going to
forwarded to the EPICS database.

if we have in the *.wb

~~~~~~~~~~~~{.wb}
      field {
         name = "Amplitude";
         ...
         epicsdb_egu  = "V";
         epicsdb_hopr = "1.5";
      };
~~~~~~~~~~~~

We will obtain in the *.db

~~~~~~~~~~~~{.wb}
record(ao, "$(user):AdcAmpSet")
{
	...
	field(DESC ,"Amplitude")
	field(EGU  ,"V")
	field(HOPR ,"1.5")
}
~~~~~~~~~~~~

The precision is automatically given according to the number of fixed
point available for this field.

When using:

* sign=0 (We are using unsigned field)
* sign=1 (We are using the most significant bit to give the signess)
* sign=2 (We are using two complements signess)

For example if we have 8bits we will have:

* sign=0 => val=[0,255]
* sign=1 => val=[-127,127]
* sign=2 => val=[-128,127] //because we don't have 0 and -0

You only need to add sdb_vendor in your *.wb to generate the file

{...
hdl_entity = "wb_clk_counters";
sdb_vendor  = 0x7501;
sdb_version = 1;
...
}

Then you can generate with the new example/Makefile by calling:

	make pts

or

	wbgen2 --vo=wb_clk_counters.vhdl --vpo=wb_clk_counters_pkg.vhdl
--hstyle=record wb_clk_counters.wb

IMPORTANT: This new option requires lua-md5 library

	sudo apt-get install lua-md5

This field is usefull when the register works with fixed point value

We can get a warning when trying to shift a 32bits '1' 32 times, so we use a 64bit '1'.

Generate extended headers with name, prefix, description, access type.

cgen_vhdl: fixed f_x_to_zero() function in generated VHDL package causing synthesis error on Mentor Precision

initial support for 'optional' parameter (conditional instantiation of RAMs and FIFOs depending on a user-defined generic value)

fixed readback value of write-only (monostable) bits to 0. See Wesley's message on white-rabbit-dev, dated 12/02/2013 for explanation