Yes, in SE multicore mode you can either run multithreaded workloads (where all threads share the same memory) or multiprogrammed workloads (where each thread has its own memory). Right now there's no facility for sharing part of the workload but not all of it (e.g., just the binary). That could be added though, by checking for situations where you're loading an already-loaded binary and reusing the existing text segment. It would require some hacking in the loader and possibly other parts of the code.
Steve On Mon, Jan 27, 2014 at 1:48 AM, Fernando Endo <[email protected]>wrote: > Hello, > > I would say that this issue is specific to the SE mode in gem5. It seems > that they copy-paste the binaries in the memory for each core (even if the > binary is the same). > > Regards, > > -- > Fernando A. Endo, PhD student and researcher > > Université de Grenoble, UJF > France > > > > 2014-01-26 kunta Zhang <[email protected]> > >> I'm using gem5 to simulate a multi core system in which each core >> executes the same workload with a shared low level instruction cache. >> >> I found that gem5 will assign the workload of each core a different >> physical address in memory even if the workload is exactly the same! That >> behavior is unexpected for me because I'm studying the sharing mechanism of >> the shared instruction cache in the system. >> >> What I want is that the sharing workload will be place in a single, >> shared address in the shared instruction cache and the memory. >> >> For example, in this simple run: >> build/ALPHA/gem5.opt --debug-flags=TLB,Cache,Bus,Fetch >> configs/example/se.py --l2cache --cpu-type=timing -n 4 --caches >> >> --cmd="tests/test-progs/hello/bin/alpha/tru64/hello;tests/test-progs/hello/bin/alpha/tru64/hello;tests/test-progs/hello/bin/alpha/tru64/hello;tests/test-progs/hello/bin/alpha/tru64/hello" >> >> Gem5 will assign each "hello" binary to a different physical address. >> >> Can anyone help? Thanks ahead! >> >> _______________________________________________ >> gem5-users mailing list >> [email protected] >> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users >> > > > _______________________________________________ > gem5-users mailing list > [email protected] > http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users >
_______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
