Re: [Gems-users] in the stats file

[Dan Gibson <degibson@xxxxxxxx>]

Simics can clear the may_stall bit in its instruction fetches for whatever reason Simics desires. It does so very infrequently, but occasionally Ruby will observe an instruction fetch or data access that has may_stall set to 0 -- under those circumstances, since ruby /cannot/ stall the access, it doesn't profile the access. That is probably where the discrepancy between executed instructions and instruction requests arises.  Considering it appears to be a sixth-significant-digit effect for you, random noise should dominate.

As for the "two numbers", I see four in the [ ]'s, which are usually breakdowns on a per-processor basis. However, if that is the case, procs 1-3 appear to be silent, which seems a bit fishy. Can you be more specific where your "two numbers" are coming from?

Regards,
Dan

Joaquin Sanchez Sanchez wrote:

Looking in the .stats file that provides gems simultaion...i have doubts.

What is the difference between instructions executed and insn request?

For example...

instructions executed: 1640867276

insn request: 1640863213

 

what causes the difference: 1640867276-1640863213=4063?

 

and...when instructions executed...there is two numbers:

1640867276   XXXXXXXXX [1640867276    0    0    0]

 

What is the second number refers to?

---

Sé un Mejor Amante del Cine
¿Quieres saber cómo? ¡Deja que otras personas te ayuden! .

_______________________________________________
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.