home bbs files messages ]

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

   comp.compilers      Compiler construction, theory, etc. (Mod      2,753 messages   

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

   Message 1,314 of 2,753   
   Sid Touati to All   
   Re: Progress in instruction scheduling?   
   22 Apr 08 11:18:26   
   
   From: SidTouati@inria.fr   
      
   > There is some recent work on optimal scheduling which compares with   
   > the list scheduling heuristic.  The problem is that the list scheduler   
   > is very close to optimal and much faster than the optimal one.   
      
   Sorry to contradict you, list schedulers are  far from the optimal,   
   twice the optimal in extreme case...   
   For very small basic blocks, list schedulers are good, as any scheduling   
   technique.   
      
   The problem comes when considering large basic blocks, which is the most   
   common codes: for large basic blocks, you cannot compute optimal   
   instruction schedules easily (because it is an NP-complete problem), so   
   you cannot compare list schedulers to optimal schedulers.   
      
   S.T.   
      
   --- 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