From: user5857@newsgrouper.org.invalid   
      
   EricP  posted:   
      
   > MitchAlsup wrote:   
   > > Does any one have {data on} or a pointer to {literature on} ST   
   > > to LD forwarding occurrences that isolate any of the following   
   > > patterns::   
   > >   
   > >      ST  Rs,[Rbase+smallDisp]   
   > >      ...   
   > >      LD  Rl,[Rbase+smallDisp]   
   > >   
   > > where Rbase is the 5-bit register specifier (not value), and:   
   > >       ST-Rbase is the same as LD-Rbase AND (of course)   
   > >       ST-smallDisp == LD-smallDISP,   
   > >   
   > > Where ST-Rbase is different than LD-Rbase but forwarding   
   > > happens none-the-less,   
   > >   
   > > Where ST-Rbase is the same as LD-Rbase, but Rbase has been   
   > > modified between ST and LD but forwarding happens none-the-   
   > > less,   
   > >   
   > >      ST  Rs,[Rbase+Rindex< >      ...   
   > >      LD  Rl,[Rbase+Rindex< >   
   > > Where ST-Rbase is the same as LD-Rbase but   
   > >       St-Rindex is different than LD-Rindex but   
   > > forwarding happens none-the-less,   
   > >   
   > > Where ST-Rbase is the same as LD-Rbase but Rbase has been   
   > > modified between ST and LD but forwarding happens none-the-   
   > > less,   
   > >   
   > > Where ST-Rindex is the same as LD-Rindex but Rindex has been   
   > > modified between ST and LD but forwarding happens none-the-   
   > > less,   
   > >   
   > > ???   
   > >   
   > > Thanks   
   > > Mitch   
   >   
   > There are tons of papers on various aspects of store to load forwarding.   
   >   
   > You appear to be interested in detecting not when the ST and LD address   
   > are the same but whether registers have changed but give the same address.   
      
   Yes, I want to do the ST-LD earlier than AGEN and/or use fewer   
   than 64-bits (the address) to do the compares.   
      
   I am also interested in the patterns presented to AGEN and which   
   ones tend to have ST-LD opportunities.   
      
   > Are you thinking of ST-LD elimination, where a push forwards to a pop   
   > and the ST and LD to the stack memory can be eliminated?   
      
   Not do much elimination, just getting a value to the LD faster   
   than waiting on ST data.   
   >   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)