On Tue, Apr 1, 2008 at 9:27 PM, Darren New <[EMAIL PROTECTED]> wrote:
>
> I don't think so. I think when memory is as fast as CPU, having complex
> addressing modes works well. When CPUs need multiple levels of memory
> cache to run efficiently, it's easy to do several CPU operations to
> calculate an address.
>
> Look at the original FORTRAN compilers. You could write
> X(I), X(I+3), X(I+J), but not X(I+J+3)
> because that last one didn't correspond to an indexed address mode, so
> it would turn into multiple assembler instructions, making it slower.
My Fortran II ingrained memory says that the allowable form of a
subscript was C1 * V + C2.
Where C1 and C2 are integer constants, and V is an integer variable.
But I won't try to look up.
carl
--
carl lowenstein marine physical lab u.c. san diego
[EMAIL PROTECTED]
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-lpsg
