From: muta...@gmail.com   
      
   On Wednesday, October 12, 2022 at 9:01:03 PM UTC+8, muta...@gmail.com wrote:   
      
   > > Concerning huge mode: having function call for array indexing   
   > > is performance killer, people who care about performance   
   > > issues would not accept this.   
      
   > Huge memory model is a performance killer already. That's   
   > why it is rarely used. If someone wants to request a   
   > binary with a hardcoded segment manipulation (out of the   
   > many possible) of a shift of 4 bits rather than making   
   > it flexible, and rather than simply upgrading to a faster   
   > processor for their alleged speed requirements - so be it,   
   > hardcode the 4.   
      
   By the way, today I got huge memory model working in   
   PDPCLIB. This was my first experience with actually   
   using huge.   
      
   Watcom generates function calls already.   
      
   And I wrote routines that hardcode a shift of 4.   
      
   But what I realized while doing that is that I simply needed   
   to check a global variable. If it is non-zero, I branch to   
   whatever routine that is and it doesn't matter where it comes   
   from (e.g. the OS).   
      
   Otherwise, the hardcoded 4-bit segment shift version of the   
   code is executed.   
      
   Performance overhead on an 8086 is near-zero.   
      
   And people with modern CPUs don't care about such   
   trivialities.   
      
   The code I wrote can be found here:   
      
   https://sourceforge.net/p/pdos/gitcode/ci/master/tree/pdpclib/dossupa.asm#l552   
      
   It's designed to be called with register parameters and   
   ideally written with a little bit of assembler.   
      
   I converted it into a C call because I have enough trouble   
   even getting it to work in C (take a look at the history!).   
      
   BFN. Paul.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)