On Tue, Apr 22, 2014 at 9:00 AM, <[email protected]>wrote:

> Send Marss86-Devel mailing list submissions to
>         [email protected]
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
> or, via email, send a message with subject or body 'help' to
>         [email protected]
>
> You can reach the person managing the list at
>         [email protected]
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Marss86-Devel digest..."
>
>
> Today's Topics:
>
>    1. multi-core sim (Pete Stevenson)
>    2. Measurement of cache line bouncing cycles in Marss        result?
> (ZH)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Mon, 21 Apr 2014 12:16:57 -0700
> From: Pete Stevenson <[email protected]>
> To: [email protected]
> Subject: [marss86-devel] multi-core sim
> Message-ID: <[email protected]>
> Content-Type: text/plain; charset=windows-1252
>
> Here are a Xeon config and a cache config that works for me. --Mu-Tien

xeon4820.conf:
----------
# vim: filetype=yaml
#
# All core and cache configuration will be in a separate file and that will
be
# included in the final config file using 'import: [filename, ...]'

# Import files that define various core/caches
import:
  - ooo_core.conf
  - cache_3_level.conf

memory:
  dram_cont:
    base: simple_dram_cont

machine:
  # Use run-time option '-machine [MACHINE_NAME]' to select
  xeon4820:
    description: OOO with shared L3
    min_contexts: 1
    cores: # The order in which core is defined is used to assign
           # the cores in a machine
      - type: ooo
        name_prefix: ooo_
    caches:
      - type: l1_32K_mesi
        name_prefix: L1_I_
        insts: $NUMCORES # Per core L1-I cache
        option:
            private: true
            last_private: false
      - type: l1_32K_mesi
        name_prefix: L1_D_
        insts: $NUMCORES # Per core L1-D cache
        option:
            private: true
            last_private: false
      - type: l2_256K_mesi
        name_prefix: L2_
        insts: $NUMCORES # Per core L1-D cache
        option:
            private: true
            last_private: true
      - type: l3_16M_wb
        name_prefix: L3_
        insts: 1 # Shared L3 config
    memory:
      - type: dram_cont
        name_prefix: MEM_
        insts: 1 # Single DRAM controller
        option:
            latency: 50 # In nano seconds
    interconnects:
      - type: p2p
        connections:
          - core_$: I
            L1_I_$: UPPER
          - core_$: D
            L1_D_$: UPPER
          - L1_I_$: LOWER
            L2_$: UPPER
          - L1_D_$: LOWER
            L2_$: UPPER2
          - L3_0: LOWER
            MEM_0: UPPER
      - type: split_bus
        connections:
            - L2_*: LOWER
              L3_0: UPPER
----------

And cache_3_level.conf looks like this:
----------
# vim: filetype=yaml

cache:
  l1_32K_mesi:
    base: mesi_cache
    params:
      SIZE: 32K
      LINE_SIZE: 64 # bytes
      ASSOC: 8
      LATENCY: 1
      READ_PORTS: 2
      WRITE_PORTS: 2
  l2_256K_mesi:
    base: mesi_cache
    params:
      SIZE: 256K
      LINE_SIZE: 64 # bytes
      ASSOC: 8
      LATENCY: 2
      READ_PORTS: 1
      WRITE_PORTS: 1
  l3_16M_wb:
    base: wb_cache
    params:
      SIZE: 16M
      LINE_SIZE: 64 # bytes
      ASSOC: 16
      LATENCY: 8
      READ_PORTS: 1
      WRITE_PORTS: 1

Hi All -
>
> I tried to mung the provided configurations to start simulating a
> multi-core xeon like architecture.  Specifically, I would like to simulate
> the following:
> 8 cores
> 32 kB L1I & L1D
> 256 kB L2
> 16 MB (or 20 MB) shared L3
>
> Unfortunately the simulator errors out with a message about the CPU
> pipeline possibly being deadlocked. Could someone reply with a working
> configuration file?  The configs included in the git repo don?t have L3
> caches (or they are single core).
>
> Regards,
> Pete Stevenson
>
>
>
>
> ------------------------------
>
> Message: 2
> Date: Mon, 21 Apr 2014 17:09:09 -0600
> From: "ZH" <[email protected]>
> To: <[email protected]>
> Subject: [marss86-devel] Measurement of cache line bouncing cycles in
>         Marss   result?
> Message-ID: <[email protected]>
> Content-Type: text/plain; charset="us-ascii"
>
> Hi all,
>
> I am trying to measure the cache line bouncing cycles in Linux on Marss.
> But
> in Marss result, there are so many measurement, does anyone know how to
> find
> out or calculate out the cache line bouncing cycles please?
>
>
>
> Regards,
>
> H
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <
> https://www.cs.binghamton.edu/mailman/private/marss86-devel/attachments/20140421/7c842076/attachment-0001.html
> >
>
> ------------------------------
>
> _______________________________________________
> http://www.marss86.org
> Marss86-Devel mailing list
> [email protected]
> https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel
>
>
> End of Marss86-Devel Digest, Vol 50, Issue 6
> ********************************************
>
_______________________________________________
http://www.marss86.org
Marss86-Devel mailing list
[email protected]
https://www.cs.binghamton.edu/mailman/listinfo/marss86-devel

Reply via email to