(Continued from previous message)   
      
   achieve a multiply, or use shifts. Modern Intel processors are much like   
   a cross between the two.  They have CISC instruction sets, yet they have   
   multiple processors on one CPU to allow concurrent instructions like the   
   RISC processors.  While I'd like to play around with writing RISC based   
   assembly....  I wouldn't want to do it on a regular basis!  Much like   
   80486+ optimizations, you write your code and you constantly look at the   
   next and previous instruction to make sure that your not using the same   
   register in the next or previous instruction, nor accessing memory when   
   you need an instruction to run at the same time as another.... Any use   
   of the same instruction on the next instruction causes a stall, any use   
   of memory read or write causes a stall, etc...  If you want to bog down   
   a 80486, just do something like this:   
      
                   L0:  MOV EAX, EBX   
                        MOV WORD PTR [DI], EAX   
                        SUB EAX, EAX   
                        JMP FAR L1   
                   L1:  JMP FAR L2   
                   L2:  MOV EBX, EAX   
                        SUB EBX, EBX   
                        RET   
      
   Pretty much every instruction generates a stall and causes the CPU to   
   wait for the previous instruction to finish before starting another....   
   Basically doing away with most of the benifiets of the 80486 and making   
   it more like a 80386.... ;-)   
   ---   
    þ OLXWin 1.00b þ MSI - Connecting The World (tm)   
    * Origin: electronic chicken bbs - bbs.electronicchicken.com (1:229/200)