Soumyaroop, Please check out a recent patch I posted to m5-dev update the instruction counts. All the regression tests need to be run, verified, etc. before pushing that patch into the tree.
There is also a conversation on m5-dev titled "Stats Changes" where we discussed the consistency in stats naming across CPU models. The most recent relevant post to you was Aug. 28th where I remarked that the O3 is not counting "nops" and "prefetches" which is throwing off the instruction counts in comparison stats. For the O3, the higher instruction count counts nops and prefetches while the lower count doesnt. The patch I'm providing aligns toward the higher instruction counts. On Tue, Sep 8, 2009 at 12:08 AM, Steve Reinhardt <[email protected]> wrote: > Korey, weren't you looking at this? I think the lower number may not > be counting no-ops or something like that. > > Steve > > On Fri, Sep 4, 2009 at 1:21 PM, soumyaroop roy<[email protected]> wrote: > > Hello all, > > > > I made an observation that has me confused and I need some > > clarification. On inspecting the o3-timing stats for 00.hello, I found > > the following: > > sim_insts = 6386 > > system.cpu.commit.commitCommittedInsts = 6403 > > system.cpu.committedInsts = 6386 > > > > How is it that total number of instructions simulated is lower than > > the total number of committed instructions? Also, shouldn't there be > > some relation like, (commit_count + squashed_count == sim_insts)? > > > > regards, > > Soumyaroop. > > > > -- > > Soumyaroop Roy > > Ph.D. Candidate > > Department of Computer Science and Engineering > > University of South Florida, Tampa > > http://www.csee.usf.edu/~sroy <http://www.csee.usf.edu/%7Esroy> > > _______________________________________________ > > m5-users mailing list > > [email protected] > > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > > > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > -- - Korey
_______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
