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 1,457 of 2,753    |
|    Pertti Kellomaki to Tim Frink    |
|    Re: Software Pipelining    |
|    17 Sep 08 15:15:26    |
      From: pertti.kellomaki@tut.fi              Tim Frink wrote:       >>> And if profiling might be exploited here.       >> I didn't get you there i.e. profiling doesn't seem related to       >> pipelining as described above.       >       > This was just a question if it would make sense to somehow       > combine software pipelining with the most frequently executed       > path (determined by profiling). I can't imagine any useful       > combination of these two techniques, but maybe there are some?              Software pipelining does not provide much performance       increase for loops with low trip counts, but it increases       code size. So using it anywhere else but in the frequently       executed paths would be rather ill adviced. Profiling is a       good way to find those paths.       --       Pertti              --- 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