From: jameskuyper@alumni.caltech.edu   
      
   On 2026-01-03 21:19, Lawrence D’Oliveiro wrote:   
   > On Sat, 3 Jan 2026 17:24:54 -0800, Andrey Tarasevich wrote:   
   >   
   >> The compiler on that embedded system is, of course, aware of the   
   >> fact that address 0x00000000 is perfectly valid and should be left   
   >> accessible. So, for that reason, the compiler is supposed to choose   
   >> some other physical representation for null pointers ...   
   >   
   > What if the entire machine address space is valid? Are C pointer types   
   > supposed to add an extra “invalid” value on top of that?   
      
   Either that, or set aside one piece of addressable memory that is not   
   available to user code. Note, in particular, that it might be a piece of   
   memory used by the implementation of C, or by the operating system. In   
   which case, the undefined behavior that can occur as a result of   
   dereferencing a null point would take the form of messing up the C   
   runtime or the operating system.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)