First of all, thanks for your answer and your time. I really appreciate it.

If you focus on the first number of -o option, I ask to blackscholes for 2
threads. I suppose that the program is creating more threads than I
specify. To solve this quickly, I simply put more cores than threads.

In another moment, I going to research this topic a little bit more, If
someone have any idea will be appreciated as well.

Best,
Cano.



2017-05-12 17:25 GMT+02:00 Gutierrez, Anthony <[email protected]>:

> Yes, it is possible, but you need to specify the number of CPUs prior to
> starting the simulation. You have to do because thread contexts may not be
> created dynamically in the current code. It looks like you’re only running
> with 2 CPUs, so I your running blackscholes with more than 2 threads you
> need to adjust what you pass via -n.
>
>
>
> -Tony
>
>
>
> *From:* gem5-users [mailto:[email protected]] *On Behalf Of *Javier
> Cano Cano
> *Sent:* Friday, May 12, 2017 8:16 AM
> *To:* gem5 users mailing list <[email protected]>
> *Subject:* [gem5-users] Multi threads on syscall emulations mode
>
>
>
> Hi everyone,
>
> I'm trying to run a multi threaded benchmark on syscall emulation mode,
> however I'm getting this error:
>
> fatal: clone: no spare thread context in system
>
> So, Is possible to run multi threaded applications on SE mode? If the
> answer is yes, an extra work to build the benchmark is required?
>
> I'm using the following commands to build and run gem5:
>
> scons ./build/X86/gem5.opt PROTOCOL=MOESI_CMP_directory RUBY=True -j7
>
> ./build/X86/gem5.opt configs/example/se.py --ruby -n 2 --caches --l2cache
> --num-l2cache=2 --num-dir=2 --topology=Mesh_XY --network=garnet2.0
> --mesh-rows=1 -c 
> tests/parsec-3.0/pkgs/apps/blackscholes/inst/amd64-linux.gcc/bin/blackscholes
> -o '2 tests/parsec-3.0/pkgs/apps/blackscholes/inputs/in_4.txt out.txt'
> --force-vef-trace --link-width-bits=672 --routing-algorithm=1
>
> Thanks for your time guys.
>
> Best,
>
> Cano.
>
> _______________________________________________
> 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

Reply via email to