>From Steve's response, it looks like I'm jumping in the conversation on the wrong page.
To be clear, Nilay were you optimizing the lookup() calls or trying to reduce the number of times lookup gets called? My MRU comments and keeping things in the the SimpleCPU was directed toward the latter. On Thu, Jan 27, 2011 at 8:29 AM, Korey Sewell <[email protected]> wrote: > > I tried caching the index for the MRU block, so that the hash table need >> not be looked up. It is hard to point if there is a speed up or not. When >> I run m5.prof, profile results show that time taken by >> CacheMemory::lookup() drops from ~ 5.5% to ~ 4%. But when I run m5.fast, >> the time of execution increases by 2%. >> >> > Hi Nilay, > when you did the MRU block caching did it reduce the # of sendFetch calls? > I would think that that # should go down from say 140 million to something > drastically smaller (...say 50 million)... > > My next question why did you choose caching the MRU index rather than the > MRU block. Ideally, what could happen is that in the TimingSimpleCPU you > could check to to see if that block is cached inside the CPU model and if so > you would not to have send a fetch to the memory system at all (thus > reducing lookup() calls). Is this what is happening here? > > -- > - Korey > -- - Korey
_______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
