Dear All, I am not able to understand why Execute-in-execute is IMP in O3CPU? any example to illustrate this
Execute-in-execute model ( http://www.m5sim.org/O3CPU#Execute-in-execute_model) For the O3CPU, we've made efforts to make it highly timing accurate. In order to do this, we use a model that actually executes instructions at the execute stage of the pipeline. Most simulator models will execute instructions either at the beginning or end of the pipeline; SimpleScalar and our old detailed CPU model both execute instructions at the beginning of the pipeline and then pass it to a timing backend. This presents two potential problems: first, there is the potential for error in the timing backend that would not show up in program results. Second, by executing at the beginning of the pipeline, the instructions are all executed in order and out-of-order load interaction is lost. Our model is able to avoid these deficiencies and provide an accurate timing model -- Thanks Newton MTech, IIT Bombay
_______________________________________________ gem5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
