runspec forks the "actual" SPEC process that runs the benchmarks. So when
you use $? to get the PID of runspec, you're really just getting a PID
that is responsible for a process that quickly forks another (the one you
really want to wait on). Other benchmarks (like stream) don't fork/execute
in this manner, so they won't suffer from these problems.
Tyler
> Hello guys,
>
> I have been using the following bash file in order to run some
> benchmarks of SPEC2006 simultaneously. I have ran it on my own host
> machine
> and also on qemu without any problem and corresponding results have been
> generated correctly inside the "result" folder of SPEC. However, when I
> want to run this bash file on the simulation environment with the
> "private_L2" machine configuration, after passing about 5 days, the only
> thing that I have got is a number of result files with the size of 0
> (nothing inside them).
>
> SPEC_Run.bash
> =============================================================
> ./parallel_commands.sh "runspec --config=mymachine.cfg --size=test
> --noreportable --tune=base --iterations=1 libquantum" "runspec
> --config=mymachine.cfg --size=test --noreportable --tune=base
> --iterations=1 hmmer" "runspec --config=mymachine.cfg --size=test
> --noreportable --tune=base --iterations=1 h264ref" "runspec
> --config=mymachine.cfg --size=test --noreportable --tune=base
> --iterations=1 omnetpp"
>
>
>
> parallel_commands.sh
> ===============================================================
>
> #!/bin/bash
>
> for cmd in "$@"; do {
> echo "Process \"$cmd\" started";
> $cmd & pid=$!
> PID_LIST+=" $pid";
> } done
>
> trap "kill $PID_LIST" SIGINT
>
> echo "Parallel processes have started";
>
> wait $PID_LIST
>
> echo
> echo "All processes have completed";
>
> ===============================================================
>
>
>
> I was wondering what can be the problem since I have tried the private_L2
> machine with other benchmarks such as stream and I believe it works fine.
>
>
> Thanks
>
> Mansour
> _______________________________________________
> http://www.marss86.org
> Marss86-Devel mailing list
> [email protected]
> https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
>
_______________________________________________
http://www.marss86.org
Marss86-Devel mailing list
[email protected]
https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
