SHARC "short" float

Started by Randy Yates July 4, 2008
The SHARC 2136x has a "short float" of 16 bits, but I cannot find a
corresponding type in the compiler documentation. Does their C/C++
support 16-bit floats? Pointers would be appreciated.
-- 
%  Randy Yates                  % "Though you ride on the wheels of tomorrow,
%% Fuquay-Varina, NC            %  you still wander the fields of your
%%% 919-577-9882                %  sorrow."
%%%% <yates@ieee.org>           % '21st Century Man', *Time*, ELO
http://www.digitalsignallabs.com
Randy Yates wrote:
> The SHARC 2136x has a "short float" of 16 bits, but I cannot find a > corresponding type in the compiler documentation. Does their C/C++ > support 16-bit floats? Pointers would be appreciated.
If the compiler doesn't support short floats, it probably doesn't support pointers to them either. (sorry, couldn't resist...). If they use the gnu toolchain, it would be much more difficult to insert a new (and to my mind somewhat dubious) new data type into the language. If you're programming in C++, why not make your own data type? Arithmetic types are particularly easy to implement -- all of the operator overloading is very well defined for you. If you do it right you'll incur the cost of a function call on each operation, but not much more. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" gives you just what it says. See details at http://www.wescottdesign.com/actfes/actfes.html
Tim Wescott wrote:
> If they use the gnu toolchain, it would be much more difficult to insert > a new (and to my mind somewhat dubious) new data type into the language.
It is not *that* dubious. 16 bit floating point values have become somewhat of a standard to store light intensities in graphics. Most graphic cards on PC even support such float-modes as a native representation and can do calculations in it. Nils