Re: [Gems-users] regarding cmp configuration


Date: Tue, 28 Jun 2005 08:57:11 -0500 (CDT)
From: Bradford Beckmann <beckmann@xxxxxxxxxxx>
Subject: Re: [Gems-users] regarding cmp configuration
...just a quick addendum

Any of the automatically generated non-ordered networks should work with
the CMP directory protocols.  Both PT_TO_PT and TORUS_2D topologies will
create the associated inter-chip network with the intra-chip network
being a single switch connecting all on chip components.

Brad


On Tue, 28 Jun 2005, Mike Marty wrote:

> It looks like you are loading a checkpoint for 32-processors, but Ruby is
> configured for 16 processors??
>
> You can use the PT_TO_PT topology to create CMP networks for any number of
> processors.  Otherwise create a network file in
> ruby/network/simple/Network_Files.  You can base one off of the others by
> removing components and links where needed.
>
> --Mike
>
>
> > hi
> > I was trying to simulate a cmp configuration (from the tutorial and
> > following some of the messages from the list)
> > I am getting this error message
> > -----------------------------------------------
> > Turning I-STC off and flushing old data
> > Turning D-STC off and flushing old data
> > successful installation of the ruby timing model.
> > Ruby Timing Mode
> > Creating event queue...
> > Creating event queue done
> > Creating system...
> >   Processors: 16
> > Creating system done
> > Ruby initialization complete
> > simics> c
> > ***  Simics getting shaky, switching to 'safe' mode.
> > ***  Simics (main thread) received a segmentation fault. Will try to
> > recuperate.
> > <Simics is running in 'safe' mode>
> > -----------------------------------
> > I am using simics version 2.0.16 I have been able to run the SMP
> > protocols successfully, but all the CMP protocols (non-token) seem to
> > give me this error
> > here is my .simics file , i m trying to load ruby after a checkpoint
> > read-configuration ./checkpoints/basic_32p_splash
> > instruction-fetch-mode instruction-fetch-trace
> > istc-disable
> > dstc-disable
> > load-module ruby
> > ruby0.setparam g_NUM_PROCESSORS 16
> > ruby0.setparam g_MEMORY_SIZE_BYTES 536870912  ## this is to match the
> > initial memory size in sarek
> > ruby0.setparam g_PROCS_PER_CHIP 4
> > ruby0.setparam g_NUM_MEMORIES 16
> > ruby0.setparam g_NUM_L2_BANKS 4
> > ruby0.setparam NUMBER_OF_VIRTUAL_NETWORKS 5
> > ruby0.setparam g_endpoint_bandwidth 1000
> > ruby0.setparam_str g_NETWORK_TOPOLOGY FILE_SPECIFIED
> > ##ruby0.setparam_str  REMOVE_SINGLE_CYCLE_DCACHE_FAST_PATH false
> > ruby0.init
> > #run-python-file "./ruby_config.py" ## this file is for supporting magic
> > Could u tell me if I have set any parameter incorrectly. I tried to use
> > the debug facility ( i can send you a snapshot of the debug o/p file
> > too, I couldnt figure out what went wrong )
> >
> > Also another thing that I am interested in is regarding the CMP
> > configurations u offer, does the network topology have to be
> > FILE_SPECIFIED with a CMP configration, becoz I couldnt find any
> > 2-processor-per- chip CMP network file?  Is there a way to make one ?
> >
> > -fl
> >
> > _______________________________________________
> > 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
>

-----------------------------------------------------------------
 Department of Computer Science         Residence
 University of Wisconsin
 1210 W. Dayton St. #6366               918 E. Dayton St. #4
 Madison, WI 53706                      Madison, WI 53703
 (608)265-2702				(608)442-0187
-----------------------------------------------------------------
[← Prev in Thread] Current Thread [Next in Thread→]