Hi, I am planning to work with M5 InOrder model for ALPHA in the SE mode. If I understand correctly, the way inorder is currently set up is that a CPU has a resource pool with a bunch of resources and each instruction has to create a schedule for itself(as to what it wants to do in each pipeline stage) at it starts. And then, each tick we go over all the pipeline stages and check which instruction needs to get what done in that stage.
But what I want to do is to implement a kind of a side-by-side parallel pipeline w.r.t the main instruction pipeline in the same cpu, which will do some stuff in each of its stages but the work would not be tied to any specific instructions. What I mean is that this parallel pipeline might be able to watch over the main instruction pipeline, but would do not instruction processing. I could potentially think of it in similar lines with the Pipeline stage class, but the PipelineStage class allows the "instruction" to define what tasks it wants to do in a particular stage, but this new non-instruction pipeline does some book keeping work and is not acted upon by instructions. Any suggestions how I can implement this in the in-order model? Thanks, Reena
_______________________________________________ m5-users mailing list [email protected] http://m5sim.org/cgi-bin/mailman/listinfo/m5-users
