From: user5857@newsgrouper.org.invalid   
      
   anton@mips.complang.tuwien.ac.at (Anton Ertl) posted:   
      
   > Thomas Koenig  writes:   
   > >Anton Ertl  schrieb:   
   > >> So does gfortran support assigned goto, too?   
   > >   
   > >Yes.   
   >   
   > Cool.   
   >   
   > >>What problems in   
   > >> interaction with other features do you see?   
   > >   
   > >In this case, it is more the problem of modern architeectures.   
   > >On 32-bit architectures, it might have been possible to stash   
   > >the address of a jump target in an actual INTEGER variable and   
   > >GO TO there.  On a 64-bit architecture, this is not possible, so   
   > >you need to have a shadow variable for the pointer   
   >   
   > Implementation options that come to my mind are:   
   >   
   > 1) Have the code in the bottom 4GB (or maybe 2GB), and a 32-bit   
   > variable is sufficient.  AFAIK on some 64-bit architectures the   
   > default memory model puts the code in the bottom 4GB or 2GB.   
   >   
   > 2) Put the offset from the start of the function or compilation unit   
   > (whatever scope the assigned goto can be used in) in the 32-bit   
   > variable.  32 bits should be enough for that.   
      
   After 4 years of looking, we are still waiting for a single function   
   that needs more than a scaled 16-bit displacement from current IP   
   {±17-bits} to reach all labels within the function.   
      
   >                                                Of course, if Fortran   
   > assigns labels between shared libraries and the main program, that   
   > approach probably does not work, but does anybody really do that?   
   >   
   > How does ifort deal with this problem?   
   >   
   > - anton   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)