home bbs files messages ]

Forums before death by AOL, social media and spammers... "We can't have nice things"

   alt.os.development      Operating system development chatter      4,255 messages   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]

   Message 3,332 of 4,255   
   mutazilah@gmail.com to Joe Monk   
   Re: segmentation   
   22 Oct 22 07:10:56   
   
   From: muta...@gmail.com   
      
   On Friday, October 21, 2022 at 8:10:35 AM UTC+8, Joe Monk wrote:   
   > > A single far pointer (no manipulation of the segment register)   
   > > is restricted to accessing a single 64k block of memory.   
   > "In a segmented architecture computer, a far pointer is a pointer which   
   includes a segment selector, making it possible to point to addresses outside   
   of the default segment."   
   >   
   > https://en.wikipedia.org/wiki/Far_pointer   
      
   It is possible if you manipulate the segment.   
      
   C compilers, even in the large memory model,   
   will not do that for you, because it is inefficient.   
      
   They wrap back to 0 when you increment   
   past ffff   
      
   Only a huge pointer will get the compiler to   
   manipulate the segment when you attempt to   
   go past ffff (effectively).   
      
   As soon as a decision is made to manipulate   
   a segment, you have a problem.   
      
   That's the point at which software is   
   typically tied down to a 4 bit segment shift.   
      
   It doesn't need to be like that though.   
      
   A flexible segment manipulation method   
   could be implemented.   
      
   Yes, there's a cost, but huge pointers are   
   rare in the first place.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   

[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]


(c) 1994,  bbs@darkrealms.ca