Hmm... Maybe delete libc* from the image and then copy it over incase  
an old version is being used. You could also try statically linking  
the m5 binary with your complier. Did you compile the m5 binary with  
the same gcc/libc version?

Ali

On Mar 5, 2009, at 1:48 PM, Bob Nagel wrote:

> Hi Ali thanks for your help,
>
> I am getting this error:
>
> m5: setaffinity: Function not implemented
> Done :D
>
> I believe is because there is a problem with libc. This is what I did
> I went to the directory:
>
> /opt/crosstool/gcc-3.4.3-glibc-2.3.5/alpha-unknown-linux-gnu/alpha- 
> unknown-linux-gnu/lib
>
> And I copy all the libc* and libstdc++* files from there to the /lib
> directory in the disk image. Then I unmount I ran the simulation and I
> get the error I said before.
>
> Am I missing any file or something in the process to make it work?.
> Thanks a lot for your help.
>
>
> 2009/3/5 Ali Saidi <[email protected]>:
>> You need to overwrite the files you have with the ones from the cross
>> compiler. The problem is the version of libc on the disk doesn't
>> support the sched_setaffinity() call.
>>
>> Ali
>>
>> On Mar 5, 2009, at 1:33 AM, Bob Nagel wrote:
>>
>>> Hi Ali thanks a lot for your answer.
>>>
>>> I already have a bigger image so the option of copy the libc* and
>>> libstdc++* from one of the cross compilers works perfect for me.  
>>> In my
>>> image I already have some libc files in the lib directory, so I just
>>> copied the files that I did not have libstdc++.so.6.0.3,
>>> libstdc++.so.6, libstdc++.so. But it did not work. Can you provide
>>> more details with what files I need to copy and where specifically  
>>> in
>>> the disk image. Sorry for this simple question and thanks a lot  
>>> again.
>>>
>>> 2009/3/4 Ali Saidi <[email protected]>:
>>>> Hi Bob,
>>>>
>>>> The m5 binary on the disk image that we distribute doesn't support
>>>> the
>>>> pin command. You'll need to compile it yourself. Additionally, I
>>>> don't
>>>> believe libc on the disk image supports sched_setaffinity (needed  
>>>> by
>>>> pin). This is something we need to fix, but it unfortunately  
>>>> involves
>>>> distributing a new disk image to everyone and updating a bunch of
>>>> regressions. You can easily make a new disk image by starting  
>>>> with a
>>>> gentoo image or you can copy the libc* and libstdc++* from one of  
>>>> the
>>>> cross compilers on the m5sim.org website onto the disk image (you
>>>> might need to make the actual image bigger to do this).
>>>>
>>>> Ali
>>>>
>>>>
>>>>
>>>> On Mar 4, 2009, at 11:54 AM, Bob Nagel wrote:
>>>>
>>>>> Hello all,
>>>>>
>>>>> This message is related with this one,
>>>>> http://thread.gmane.org/gmane.comp.emulators.m5.users/3939/focus=3967
>>>>> .
>>>>>
>>>>> In that message Ali gives an explanation about how to use pin. I
>>>>> have
>>>>> been trying to use it but it seems that  when I am running the
>>>>> script
>>>>> m5 does not recognize the command, following you can see my script
>>>>> and
>>>>> the output.
>>>>>
>>>>> #!/bin/sh
>>>>> cd /parsec/install/bin
>>>>> /sbin/m5 pin 0 ./ferret /parsec/install/inputs/ferret/coreld lsh
>>>>> /parsec/install/inputs/ferret/queriesd 5 5 8
>>>>> /parsec/install/inputs/ferret/output.txt
>>>>> echo "Done :D"
>>>>> /sbin/m5 exit
>>>>>
>>>>> loading script...
>>>>> usage: m5 ivlb <interval>
>>>>>       m5 ivle <interval>
>>>>>       m5 initparam
>>>>>       m5 sw99param
>>>>>       m5 exit [delay]
>>>>>       m5 resetstats [delay [period]]
>>>>>       m5 dumpstats [delay [period]]
>>>>>       m5 dumpresetstats [delay [period]]
>>>>>       m5 checkpoint [delay [period]]
>>>>>       m5 readfile
>>>>>       m5 debugbreak
>>>>>       m5 switchcpu
>>>>>
>>>>> All times in nanoseconds!
>>>>> Done :D
>>>>>
>>>>> I looked to the code in /util/m5/m5.c and I see that the pin  
>>>>> command
>>>>> is only executed if linux is define. I do not understand why the  
>>>>> pin
>>>>> command is inside those defines. Also I tried to delete the  
>>>>> defines
>>>>> but that did not work. Any help about how to use it?.
>>>>>
>>>>> Thanks a lot in advance.
>>>>> _______________________________________________
>>>>> 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
>>>>
>>> _______________________________________________
>>> 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
>>
> _______________________________________________
> 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

Reply via email to