simulator: Improve simulator resource model (#142)

This pull request improves the existing simulator resource model
that is at the core of all simulation models in OpenDC. 
Most importantly, we have changed the way of how metrics are reported by this layer.

* Add `SimResourceInterpreter` which is responsible for efficiently scheduling 
  communication between resources in OpenDC. The performance gain is in the 2x-5x range.
* Add uniform interface for exposing resource metrics (using `SimResourceCounters`).
* Split the CPUFreq subsystem in the compute simulator as it mixed responsibilities of 
  different layers.

**Breaking API Changes**

* Resource providers now accept a `SimResourceInterpreter` which is
   responsible for coordinating the communication between resources.
* `ScalingGovernor` is not part of the machine layer anymore. Instead, it should 
   move in the OS/Hypervisor layer.
* Workloads should now start CPU consumers using `cpu.startConsumer`.

1 files changed, 11 insertions, 17 deletions

diff --git a/opendc-simulator/opendc-simulator-compute/src/jmh/kotlin/org/opendc/simulator/compute/SimMachineBenchmarks.kt b/opendc-simulator/opendc-simulator-compute/src/jmh/kotlin/org/opendc/simulator/compute/SimMachineBenchmarks.kt
index 15714aca..fb753de2 100644
--- a/opendc-simulator/opendc-simulator-compute/src/jmh/kotlin/org/opendc/simulator/compute/SimMachineBenchmarks.kt
+++ b/opendc-simulator/opendc-simulator-compute/src/jmh/kotlin/org/opendc/simulator/compute/SimMachineBenchmarks.kt
@@ -25,17 +25,15 @@ package org.opendc.simulator.compute
 import kotlinx.coroutines.ExperimentalCoroutinesApi
 import kotlinx.coroutines.coroutineScope
 import kotlinx.coroutines.launch
-import org.opendc.simulator.compute.cpufreq.PerformanceScalingGovernor
-import org.opendc.simulator.compute.cpufreq.SimpleScalingDriver
 import org.opendc.simulator.compute.model.MemoryUnit
 import org.opendc.simulator.compute.model.ProcessingNode
 import org.opendc.simulator.compute.model.ProcessingUnit
 import org.opendc.simulator.compute.power.ConstantPowerModel
+import org.opendc.simulator.compute.power.SimplePowerDriver
 import org.opendc.simulator.compute.workload.SimTraceWorkload
 import org.opendc.simulator.core.SimulationCoroutineScope
 import org.opendc.simulator.core.runBlockingSimulation
-import org.opendc.simulator.resources.SimResourceScheduler
-import org.opendc.simulator.resources.SimResourceSchedulerTrampoline
+import org.opendc.simulator.resources.SimResourceInterpreter
 import org.openjdk.jmh.annotations.*
 import java.util.concurrent.TimeUnit
 
@@ -46,13 +44,13 @@ import java.util.concurrent.TimeUnit
 @OptIn(ExperimentalCoroutinesApi::class)
 class SimMachineBenchmarks {
     private lateinit var scope: SimulationCoroutineScope
-    private lateinit var scheduler: SimResourceScheduler
+    private lateinit var interpreter: SimResourceInterpreter
     private lateinit var machineModel: SimMachineModel
 
     @Setup
     fun setUp() {
         scope = SimulationCoroutineScope()
-        scheduler = SimResourceSchedulerTrampoline(scope.coroutineContext, scope.clock)
+        interpreter = SimResourceInterpreter(scope.coroutineContext, scope.clock)
 
         val cpuNode = ProcessingNode("Intel", "Xeon", "amd64", 2)
 
@@ -85,8 +83,7 @@ class SimMachineBenchmarks {
     fun benchmarkBareMetal(state: Workload) {
         return scope.runBlockingSimulation {
             val machine = SimBareMetalMachine(
-                coroutineContext, clock, machineModel, PerformanceScalingGovernor(),
-                SimpleScalingDriver(ConstantPowerModel(0.0))
+                interpreter, machineModel, SimplePowerDriver(ConstantPowerModel(0.0))
             )
             return@runBlockingSimulation machine.run(SimTraceWorkload(state.trace))
         }
@@ -96,10 +93,9 @@ class SimMachineBenchmarks {
     fun benchmarkSpaceSharedHypervisor(state: Workload) {
         return scope.runBlockingSimulation {
             val machine = SimBareMetalMachine(
-                coroutineContext, clock, machineModel, PerformanceScalingGovernor(),
-                SimpleScalingDriver(ConstantPowerModel(0.0))
+                interpreter, machineModel, SimplePowerDriver(ConstantPowerModel(0.0))
             )
-            val hypervisor = SimSpaceSharedHypervisor()
+            val hypervisor = SimSpaceSharedHypervisor(interpreter)
 
             launch { machine.run(hypervisor) }
 
@@ -118,10 +114,9 @@ class SimMachineBenchmarks {
     fun benchmarkFairShareHypervisorSingle(state: Workload) {
         return scope.runBlockingSimulation {
             val machine = SimBareMetalMachine(
-                coroutineContext, clock, machineModel, PerformanceScalingGovernor(),
-                SimpleScalingDriver(ConstantPowerModel(0.0))
+                interpreter, machineModel, SimplePowerDriver(ConstantPowerModel(0.0))
             )
-            val hypervisor = SimFairShareHypervisor(scheduler)
+            val hypervisor = SimFairShareHypervisor(interpreter)
 
             launch { machine.run(hypervisor) }
 
@@ -140,10 +135,9 @@ class SimMachineBenchmarks {
     fun benchmarkFairShareHypervisorDouble(state: Workload) {
         return scope.runBlockingSimulation {
             val machine = SimBareMetalMachine(
-                coroutineContext, clock, machineModel, PerformanceScalingGovernor(),
-                SimpleScalingDriver(ConstantPowerModel(0.0))
+                interpreter, machineModel, SimplePowerDriver(ConstantPowerModel(0.0))
             )
-            val hypervisor = SimFairShareHypervisor(scheduler)
+            val hypervisor = SimFairShareHypervisor(interpreter)
 
             launch { machine.run(hypervisor) }