In general, yes.

I say 'yes' in the sense that it's all C++ object models so whatever you
would like to do you are free to go into the source code and add yourself
if it's not there out of the box. However, it's likely that to do something
interesting you'll have to make some minor edits to some code since all the
things "out the box" are probably architectures that someone has looked at
before.

For your specific question, it is too generic to answer. The better
question is how will the CPUs access this shared structure? Through the
memory system? Through some special instruction? Does this structure
automatically extract data at some time interval? In any event, since
caches can be shared amongst different CPUs then I'm going to say yes you
can define whatever structure you like. But again, you'll need to figure
out a number of specific things about your structure plus become somewhat
familiar with the code you want to change so you can do whatever you want
to do correctly.

On Mon, Feb 20, 2012 at 10:39 AM, Heba Saadeldeen <[email protected]> wrote:

> Hi,
>
> I am trying to run different workloads on multiple cpus, I need to define
> a structure that is shared among the different cpus, is there a way to do
> that?
>
> Thanks,
> --
> Heba
>
> _______________________________________________
> gem5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>



-- 
- Korey
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to