Hi,
    this is biswa from india,working in prefetching algorithms,i read in
your mail,you are getting wrong statistics for L3 cache,but M5 has only L2
cache right.If you have augmented it to L3 cache,can you tell me how to do
it,i am using splash2 benchmarks,i need to change the run.py file i guess,if
you donot mind you can share with me.

regards
biswa

2010/12/24 zhanglunkai <[email protected]>

>  Hi, I am adding a new prefetcher in m5’s last level cache system.
>
> But I have encountered some bugs when using the simulator:
>
> The simulation with new prefecther sometimes works out with absolutely
> wrong statistics (for example, in SPEC2006cpu.omnetpp, the L3 access counts
> are far smaller than simulation without the prefetcher).
>
>
>
> I find that in the wrong simulation, a ReadReq packet receive a hit in MSHR
> (this mshr is a prefetching request).
>
>
>
> I examine the cache_impl.hh code, and find there is a “todo” remark  in
> timingAcess() function:
>
> //@todo remove hw_pf here
>
> This remarks is in the section where access receive a miss in tags and a
> hit mshr, and the hit in mshr is a prefetch request (exact thing happened in
> my wrong simulation).
>
>
>
> I wonder whether such a situation causes the bug I have encountered. Does
> anybody can help figure out where is the problem?
>
>
>
>
>
> *发件人:* [email protected] [mailto:[email protected]] *代表
> *Steve Reinhardt
> *发送时间:* 2010年12月22日星期三 7:48
> *收件人:* M5 users mailing list
> *主题:* Re: [m5-users] Can't resolve Frequency error for FS Script File
>
>
>
> I think it's actually complaining about this line from src/sim/System.py:
>
>
>
>         boot_cpu_frequency = Param.Frequency(Self.cpu[0].clock.frequency,
>
>                                              "boot processor frequency")
>
>
>
> It's hard to say what the problem is without knowing more about the
> specific changes you've made to your config file.
>
>
>
> Steve
>
>
>
> On Mon, Dec 20, 2010 at 10:32 AM, Malek Musleh <[email protected]>
> wrote:
>
>  Hello,
>
> I am working on creating a script file to run my configured system
> (not using the standard Bus topology provided by the dev repo) and am
> receiving the following error when I try to run:
>
> File "/home/musleh/m5_modified_step/configs/common/Simulation.py",
> line 195, in run
>    m5.instantiate(root)
>  File "/home/musleh/m5_modified_step/src/python/m5/simulate.py", line
> 52, in instantiate
>    root.unproxy_all()
>  File "/home/musleh/m5_modified_step/src/python/m5/SimObject.py",
> line 680, in unproxy_all
>    self._children[child].unproxy_all()
>  File "/home/musleh/m5_modified_step/src/python/m5/SimObject.py",
> line 660, in unproxy_all
>    value = value.unproxy(self)
>  File "/home/musleh/m5_modified_step/src/python/m5/proxy.py", line
> 94, in unproxy
>    (self.path(), self._pdesc.ptype_str, base.path())
> AttributeError: Can't resolve proxy 'cpu[0].clock.frequency' of type
> 'Frequency' from 'system'
>
>
> In searching the mailing list, it seems to me that this type of "Can't
> resolve proxy error" is because I am missing defining the frequency
> for the CPU Model in the script file.
> However, as in the fs.py file, I have included the following:
>
>
> DriveCPUClass = AtomicSimpleCPU
> drive_mem_mode = 'atomic'
>
>
> (TestCPUClass, test_mem_mode, FutureClass) =
> Simulation.setCPUClass(options)
>
> TestCPUClass.clock = '2GHz'
> DriveCPUClass.clock = '2GHz'
>
>
> My system config script file is not yet complete, but I don't see
> where else a frequency is defined in fs.py. One difference between
> mine and fs.py is that I am not using the "MakeLinuXAlphaSystem" call
> in FSConfig.py, but I don't see where in that def would resolve that
> error.
>
> Any suggestions, or do I need to provide additional information?
>
> Malek
> _______________________________________________
> m5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
>
>
>
> _______________________________________________
> m5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
>



-- 

*thanks&regards
*
*BISWABANDAN PANDA*
*M.S.(RESEARCH SCHOLAR)*
*RISE LAB*
*IIT MADRAS*

http://www.cse.iitm.ac.in/~biswa/ <http://www.cse.iitm.ac.in/%7Ebiswa/>

Do not dwell in the past,
 do not dream of the future,
 concentrate the mind on the present moment.
_______________________________________________
m5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users

Reply via email to