Re: [Gems-users] Problem with Simics segmentation fault

[Liz Mooney <mooneye@xxxxxxxxxxxx>]

Thanks for you help.  We figured this one out.  It does seem that certain configurations as you eluded to below dont quite work as expected.

Liz

---- Original message ----
>Date: Fri, 04 May 2007 23:00:53 -0500
>From: Mike Marty <mikem@xxxxxxxxxxx>  
>Subject: Re: [Gems-users] Problem with Simics segmentation fault  
>To: Gems Users <gems-users@xxxxxxxxxxx>
>
>The USER_MODE_DATA_ONLY appears to be a flag for LogTM only. 
>
>I would try to identify what exact parameter causes the segmentation 
>fault.  I wouldn't be surprised if it is the size of the data block 
>(since we haven't done much testing in using sizes other than 64). 
>
>--Mike
>
>
>Liz Mooney wrote:
>> Sorry..here is the line from rubyconfig.defaults that I changed to get ruby to run in user mode:
>>
>> USER_MODE_DATA_ONLY: false //I changed this to true
>>
>> L2_CACHE_NUM_SETS_BITS: 10 //changed this too
>> g_DATA_BLOCK_BYTES: 256    //and this but these are probably unrelated
>>
>> We also changed the SLICC interface file MOSI_SMP_directory_1level-cache.sm to add certain functionality. But this should not make a difference because it has been running fine in previous simulations with and without the user mode set.  
>>
>> Thanks,
>> Liz
>>
>> ---- Original message ----
>>   
>>> Date: Fri, 04 May 2007 20:47:16 -0500
>>> From: Dan Gibson <degibson@xxxxxxxx>  
>>> Subject: Re: [Gems-users] Problem with Simics segmentation fault  
>>> To: Gems Users <gems-users@xxxxxxxxxxx>
>>>
>>> I don't understand what you mean by "change(ing) the user mode". What 
>>> specific parameters did you change in rubyconfig.defaults?
>>>
>>> Also, have you made other changes to the Ruby codebase?
>>>
>>> Regards,
>>> Dan
>>>
>>> Liz Mooney wrote:
>>>
>>>     
>>>> Hello,
>>>>
>>>> I have been compiling and running Ruby on several different benchmarks just fine until this morning.  I wanted to change the user mode in /ruby/config in the rubyconfig.defaults and when I did this, recompiled Ruby, and tried loading it into Simics I got the following error after I tried typing something (like ls or pushing enter) in the Simics console:
>>>>
>>>> simics> c
>>>> Segmentation fault (SIGSEGV) in main thread
>>>> The simulation state has been corrupted. Simulation cannot continue.
>>>> Please restart Simics.
>>>>
>>>> This seems strange to me because I have been changing cache size, block size, and user mode in the configuration file in the past and it worked just fine. Nothing has changed since last night when I last ran a simulation. 
>>>>
>>>> I am loading Ruby to Simics correctly since I have checked this. Any help would be appreciated. 
>>>>
>>>> Thanks,
>>>> Liz
>>>> _______________________________________________
>>>> 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.
>>   
>_______________________________________________
>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.
>