Re: [Gems-users] Error when simulating 64 cores in Ruby

[Mishali Naik <mishali@xxxxxxxxxxx>]

Bernard,

Yes, I was thinking along the same lines that maybe the non-contiguous 
core id mapping is a problem. Does Ruby have access to the physical core 
ids though?
thanks a lot,
Mishali

BYONG WU CHONG wrote:
> Mishali,
>
> Let's work this out together. I am the one who posted one of the two posts you listed. I would also like to solve this issue so that we can have 32 thread simulation on 64 core setup of GEMS TM.
>
> My guess is that this problem could be related with the non-contiguous core id mapping on 64 core setting.
> https://www.simics.net/mwf/topic_show.pl?pid=110801 
>
> As you can see, the physical core id for 25th logical core id is 32. Not so intuitive, but if you read the above link you will see that it makes sense for 64 cores. Yet, GEMS use this type of function call.
>
> // $GEMS/ruby/simics/interface.C
> void SIMICS_unstall_proc(int cpuNumber)
> {
>   conf_object_t* proc_ptr = (conf_object_t *) SIM_proc_no_2_ptr(cpuNumber);
>   SIM_stall_cycle(proc_ptr, 0);
> }
>
> The problem seems that there is no remapping of the logical core id to physical core id.
>
> So maybe changing SIM_proc_no_2_ptr() to SIMICS_proc_no_2_ptr() wrapper function and making SIMICS_proc_no_2_ptr() wrapper function to call the SIM_proc_no_2_ptr() function with the correctly remapped physical core id maybe be the solution.
>
> - Bernard
>
>
> -----Original Message-----
> From: gems-users-bounces@xxxxxxxxxxx [mailto:gems-users-bounces@xxxxxxxxxxx] On Behalf Of Mishali Naik
> Sent: Monday, July 20, 2009 2:26 PM
> To: Gems Users
> Subject: [Gems-users] Error when simulating 64 cores in Ruby
>
> Can someone please help with this?
>
> thanks
>  - Mishali
>
> Mishali Naik wrote:
>   
> > Hi all,
> >
> > I am seeing the exact same error as the one mentioned in these two posts 
> > when I try to simulate 64 processor in Ruby.
> > https://lists.cs.wisc.edu/archive/gems-users/2008-May/msg00143.shtml
> >
> > https://lists.cs.wisc.edu/archive/gems-users/2009-April/msg00055.shtml
> >
> > Can someone please help with this issue? Any kind of help will be 
> > appreciated.
> > thanks a lot,
> > Mishali
> > _______________________________________________
> > Gems-users mailing list
> > Gems-users@xxxxxxxxxxx
> > https://lists.cs.wisc.edu/mailman/listinfo/gems-users
> > Use Google to search the GEMS Users mailing list by adding "site:https://lists.cs.wisc.edu/archive/gems-users/"; to your search.
> >   
> >     
> _______________________________________________
> Gems-users mailing list
> Gems-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/gems-users
> Use Google to search the GEMS Users mailing list by adding "site:https://lists.cs.wisc.edu/archive/gems-users/"; to your search.
>
> _______________________________________________
> Gems-users mailing list
> Gems-users@xxxxxxxxxxx
> https://lists.cs.wisc.edu/mailman/listinfo/gems-users
> Use Google to search the GEMS Users mailing list by adding "site:https://lists.cs.wisc.edu/archive/gems-users/"; to your search.
>