Re: New assembly language instructions to support OO languages?

nmm1@cam.ac.uk wrote:
(snip, I wrote)


>>This is very much against the RISC idea that all instructions
>>take about the same amount of time.


> Yes. That particular dogma lasted for quite a long time after it was
> shown to be a total fantasy but - hey! - rationality and sanity are
> just words :-)


I used a Sun 4/110 for many years, which I believe used one of the
original SPARC processors. Instead of a multiply instruction, it has
multiply step such that you execute some number of such instructions
(maybe one per product bit) to do fixed point multiply. Since it was
too much work for compilers to generate multiply step, it was instead
emulated by the system, possibly after an interrupt for a
non-implemented opcode. SunOS, like most unix systems, keeps track of
user and system time, and multiply counted as system time.


Maybe that was the example that showed that it wasn't a good idea.


-- glen

Return to the comp.compilers page.
Search the comp.compilers archives again.