Re: [Gems-users] IPC information

["Zhang Yu" <thuzhangyu@xxxxxxxxx>]

Thanks for the response. This makes sense.

Thanks,
Yu

On Fri, Mar 28, 2008 at 9:58 PM, Dan Gibson <degibson@xxxxxxxx> wrote:

Why are the instructions_executed so unexpected? Its quite common in multithreaded code to have work imbalance, that eventually manifests itself as N-1 out of N threads spinning (quickly -- committing a LOT of instructions) on a lock or CV...

Regards,
Dan

Zhang Yu wrote:

Hi,

When I run a 16-thread application on a 16-core simulated machine, how should I collect the IPC information? I checked the statistics dumped by ruby, it seems that there is a big variation between different cores in terms of "instructions_executed". For example, some core only execute 10% instructions of some other core. According to the application I run, this shouldn't be the case.  Can somebody tell how to collect the IPC information of the simulated CMP machine correctly? Thanks a lot.

Yu

_______________________________________________
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.

  

-- 
http://www.cs.wisc.edu/~gibson [esc]:wq!

_______________________________________________
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.