On 2/4/08, André Pouliot <[EMAIL PROTECTED]> wrote: > I found that the bug was in the result of the multiplication. It was a > rounding that's present in the software model and is not present in the > rtl of the multiplication. > > Now the question is : do we care about the rounding? Adding the rounding > to the fp multiplicator would add another stage with a 19 bits adder. I > did a quick test with the ogsim program there seem to have some small > variation on the edge of the triangle. I attach the resulting image for > the 2 run of the program with and without the rounding enabled in the > float25 class. > > Could someone tell me if we do without the rounding would the compound > error would be to important or we can neglect it? >
I can't see any difference. When it comes to pixel values, there are no loops in OGA to compound errors, and the numbers get chopped to 8-bit values going out to memory. However, it is coordinates I'm concerned about. I don't remember if we ever perform a multiply on a float coordinate. If we don't, then I suggest that we take careful and prominent note of the fact that we don't do rounding in the multiplies, and see if anyone notices a problem that could be fixed by having it. -- Timothy Normand Miller http://www.cse.ohio-state.edu/~millerti Open Graphics Project _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
