From: nowhere@nevernet.at   
      
   On 13/09/2022 14:21, muta...@gmail.com wrote:   
   [...]   
   >> Even if gate a20 was a logical solution,   
   >> would it have been reasonable for msdos to   
   >> have an explicit bios call to enable wraparound?   
      
   back then there were a lot (also non-M$) programs around   
   which relied on wraparound for a reason:   
   catch any overrun to indicate an error.   
      
   you can see overrun/wrap also on segments, it became   
   handy to restart code at offset zero with two bytes   
   at the end of a code segment [FFFE: EB 00  jmp 0000].   
      
   But even this looks similar to A20 warp it is another   
   story.   
   the top of the last 64K in the 1MB range belong to BIOS   
   which is/was ROM anyway.   
      
   >> Even if existing bioses don't implement that   
   >> call, it doesn't matter, because the 8086 is   
   >> going to wrap around anyway.   
      
   my early OS (~1985..1997) didn't use the BIOS for   
   A20 check and enable, it just talked the hardware.   
      
   ...   
   > Unless you are on a machine with more than 1 MB   
   > of memory, in which case the high memory   
   > area is populated with those 4 bytes.   
      
   what four bytes to you mean here?   
   "far" CALL/JMP are 5 byte within 16-bit.   
      
   > Actually, why is disabling a20 ever   
   > needed? Won't you always be in a position   
   > to populate high memory if it exists,   
   > and if it doesn't, there's no issue anyway.   
      
   I never used a disabled A20, mine were always ON   
   it's just a historical remain (aka tomb stone).   
      
   > This is why I like to discuss things.   
   > I can't derive anything in a vacuum.   
      
   if this is a question I haven't got the sense :)   
   __   
   wolfgang   
   the day when an M$-product wont suck is the day   
   when they start producing vacuum cleaners.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)