> I don't think you really need to be able to "cancel" those events if every > distinct hardware-event was simulated. In case of a mispredicted instruction > the immediately following events don't necessarily need to commit any > lasting cpu state changes just like in a real out of order cpu. So I'm > guessing the primary reasons why you allow de- and reallocation are > efficiency and code simplicity.
Well, the effect of the event must be prevented from happening, so yes, it must be cancelled. Think of a watchdog timer, there's no other way to think of it. Must it be done in the event queue? No, but it is more convenient. Nate _______________________________________________ m5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/m5-dev
