If you are looking at the M5 Wiki, then you've found the documentation available.
A lot of stuff you list below you'll just have to try and see if it works for you since it's kind of hard to anticipate what problems you may or may not have in the future. On Tue, Apr 28, 2009 at 9:14 AM, Javier Jose <[email protected]> wrote: > Hi Everyone, > > I know this question has been asked frequently now, but after looking > through tons of documentation and the mail-archive threads I am still not > clear at all how to get SPEC CPU2000 working on M5. Is it possible to find a > complete guideline/how_to that explains from source_code -> running (at > least in SE)? > > I have the CD and downloaded everything to my Ubuntu VM (where I have M5). > Since I've always used pre-compiled spec binaries I have never tried to > cross-compile the source code and it seems that a bunch of tools are needed > for it (for which the installation fails on my system) even if I just want > to cross-compile the source code using the alpha-gcc cross compiler given > the M5 wikipage. > > Thefore, in the case that i can't be redirected to a guidelin/how_to of some > sort, does anyone know how to succesfully do the following?: > > 1) Cross-compile SPEC CPU2000 benchmarks for ALPHA or at least get the > binaries. > 2) Run SPEC CPU 2000 in SE. Since I don't have the binaries I haven't been > able to try out the example given at the page > > $ build/ALPHA_SE/m5.debug configs/example/se.py --cmd=eon00 > --options="chair.control.cook > > chair.camera chair.surfaces chair.cook.ppm ppm pixels_out.cook" > > Does it work? I really don't wish to use cpu2000.py since > I want to run SMT and multicore SMT configurations in SE mode. > > 3) For SMT and multicore SMT configurtions would the above example > (considering that it works for 1 CPU 1 Thread) work by just replicating the > workload as processes and assiging them to each thread per cpu? Just as the > hello_world example? > > 4) For the configuration in 3) , does SIMPOINT/FastFwd work? > > I am trying to run this benchmarks as my last resort for my course project > due this week (and also for further research) since other benchmarks such as > ALPBench and PARSEC run fine in singlethreaded mode but, when trying to do > multicore, SMT or both they abort and fail and I haven't been able to fix it > yet. > > I am running out of time and I'd appreciate any help. > > Thanks > > Javier > Texas A&M University > Computer Science and Engineering Department > > _______________________________________________ > m5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/m5-users > -- =========== Korey L Sewell Computer Science & Engineering University of Michigan _______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
