Re: [Gems-users] does simics processor number need to match ruby processor number?


Date: Thu, 29 Sep 2005 00:26:28 -0500
From: "ayu" <ayu@xxxxxxxxxxxxx>
Subject: Re: [Gems-users] does simics processor number need to match ruby processor number?
Well, what I'd like to do is simulate both shared and distributed memory
that uses many number of processors and allow it to share the data. Of
course that entails me writing/modifying the software to suit my needs.
However, if I run on a uniprocessor target machine can I use ruby to
simulate any number of processors that I need? 

Thanks,
Andy

----- Original Message Follows -----
> What exactly do you want to simulate? I see several options
> but only the last one made little sense to me. 
> 
> 1. Execute all processors in lock-step, but it is hardly real.
> 
> 2. Let all processors execute freely, but they
>    will interfere with each other (and crash) if they all
>    write to the same memory space.
> 
> 3. Just simulate the timing and assume each processor writes
>    to a different memory space? In this case, there would be
>    no sharing between processors.
> 
> 4. Just simulate the timing and assume all processors write the
>    same memory space. In this case, you have sharing, but
>    the sharing never changes program execution path.
> 
> Perhaps what you want is none of the above?
> 
> On Wed, 28 Sep 2005 ayu wrote :
> > 
> > Hello,
> > 
> >    I have a general question. If I have a craff file that contains a
> > single processor running on Simics, can I tell Ruby to "generate"
> > say, 64 processors and simulate as if I had a 64 processor system.
> > If that is indeed possible, how does Simics and Ruby talk to each
> > other to make that possible? Thanks for any info.
> > 
> > Andy
> > _______________________________________________
> > Gems-users mailing list
> > Gems-users@xxxxxxxxxxx
> > https://lists.cs.wisc.edu/mailman/listinfo/gems-users
> _______________________________________________
> Gems-users mailing list
> Gems-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/gems-users
[← Prev in Thread] Current Thread [Next in Thread→]