On Thu, 23 Aug 2012, Xi Chen wrote:
Hi all,
I have a question about calculating request latency and L1 data miss rate.
My configuration is ALPHA, MESI_directory, 2 level cache, garnet network.
When L1 has a miss, it should generate a request and send it into the
network, then wait for the response. I wonder is there any place I can find
to calculate the latency of the request.
In /mem/request.hh, does "contextId" represent which CPU this request is
from?
The request can be generated from CPU and sent to L1, also can be generated
from L1 and sent into the network. Is there any way I can specifically find
and calculate the latency of the second request which is generated by L1
and sent into the network?
Also is there any way I can find L1 data cache miss number and access/hit
number of each core?
Have you taken a look at file ruby.stats which is generated at the end of
the simulation?
--
Nilay
_______________________________________________
gem5-users mailing list
gem5-users@gem5.org
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
