From: user5857@newsgrouper.org.invalid   
      
   Robert Finch  posted:   
      
   > Started working on yet another CPU – Qupls4. Fixed 40-bit instructions,   
   > 64 GPRs. GPRs may be used in pairs for 128-bit ops. Registers are named   
   > as if there were 32 GPRs, A0 (arg 0 register is r1) and A0H (arg 0 high   
   > is r33). Sameo for other registers. GPRs may contain either integer or   
   > floating-point values.   
   >   
   > Going with a bit result vector in any GPR for compares, then a branch on   
   > bit-set/clear for conditional branches. Might also include branch true /   
   > false.   
      
   I have both the bit-vector compare and branch, but also a compare to zero   
   and branch as a single instruction. I suggest you should too, if for no   
   other reason than:   
      
             if( p && p->next )   
      
   > Using operand routing for immediate constants and an operation size for   
   > the instruction. Constants and operation size may be specified   
   > independently. With 40-bit instruction words, constants may be 10,50,90   
   > or 130 bits.   
      
   My 66000 allows for occasional use of 128-bit values but is designed mainly   
   for 64-bit and smaller.   
      
   With 32-bit instructions, I provide, {5, 16, 32, and 64}-bit constants.   
      
   Just last week we discovered a case where HW can do a better job than SW.   
   Previously, the compiler would emit:   
      
             CVTfd    Rt,Rf   
             FMUL     Rt,Rt,#1.425D0   
             CVTdf    Rd,Rt   
      
   Which is subject to double rounding once at the FMUL and again at the   
   down conversion. I though about the problem and it seems fairly easy   
   to gate the 24-bit fraction into the multiplier tree along with the   
   53-bit fraction of the constant, and then normalize and round the   
   result dropping out of the tree--avoiding the double rounding case.   
      
   Now, the compiler emits:   
      
             FMULf   Rd,Rf,#1.425D0   
      
   saving 2 instructions alongwith the higher precision.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)