Re: [Gems-users] segmentation fault in MESI_CMP_filter_directory

[huitzu shih <huitzu.shih@xxxxxxxxx>]

I have the same problem in MESI_CMP_filter_directory.
I run Btree with 32 cores and TIMESTAMP.
But It seems normal when I use 4 cores.
Does anyone solve this problem?

Thank for help.

I  set
mfacet.run_sim_command('ruby0.setparam_str XACT_EAGER_CD true')               
mfacet.run_sim_command('ruby0.setparam_str XACT_LAZY_VM true')                 
mfacet.run_sim_command('ruby0.setparam_str XACT_CONFLICT_RES TIMESTAMP')       
mfacet.run_sim_command('ruby0.setparam_str XACT_NO_BACKOFF true')              

Error message:
Arithmetic exception (SIGFPE) in main thread
The simulation state has been corrupted. Simulation cannot continue.
Please restart Simics.

2008/4/30 <cge@xxxxxxxx>

Hi,
I am running btree microbenchmark in MESI_CMP_filter_directory in GEMS2.1.

When I set the conflict resolution policy to be TIMESTAMP and using 16
processors,
it will have arithmetic exception or segmentation fault.
However, it sometimes does not cause any exceptions and faults if I
use BASE resolution policy or set the XACT_LOG_BUFFER_SIZE to 512.

Does anyone have this kind of experiences?

Thanks,
Cen Kwang
_______________________________________________
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.