refactor(cpuidle): move cgroupMgr into cpuIdleTracing struct

Converge per-instance state from package-level variable into struct field, per rule 5 global variable scope contraction.

Author: hao022

core/autotracing/cpuidle.go

@@ -1,4 +1,4 @@
-// Copyright 2025 The HuaTuo Authors
+// Copyright 2025, 2026 The HuaTuo Authors
 //
 // Licensed under the Apache License, Version 2.0 (the "License");
 // you may not use this file except in compliance with the License.
@@ -40,26 +40,23 @@ func init() {
 	tracing.RegisterEventTracing("cpuidle", newCPUIdle)
 }
 
-var cgroupMgr cgroups.Cgroup
-
 func newCPUIdle() (*tracing.EventTracingAttr, error) {
+	// NewManager only returns nil alongside an error (default branch); v1/v2 never return nil on success
 	cgroup, err := cgroups.NewManager()
 	if err != nil {
 		return nil, err
 	}
-	if cgroup == nil {
-		return nil, fmt.Errorf("cgroup manager is nil")
-	}
-	cgroupMgr = cgroup
 
 	return &tracing.EventTracingAttr{
-		TracingData: &cpuIdleTracing{},
+		TracingData: &cpuIdleTracing{cgroupMgr: cgroup},
 		Interval:    20,
 		Flag:        tracing.FlagTracing,
 	}, nil
 }
 
-type cpuIdleTracing struct{}
+type cpuIdleTracing struct {
+	cgroupMgr cgroups.Cgroup
+}
 
 type cpuStats struct {
 	user  int64
@@ -122,14 +119,14 @@ func updateContainersCPUIdle(f *matcher.ContainerMatcher) error {
 	return nil
 }
 
-func detectCPUIdleContainer(threshold *cpuIdleThreshold) (*containerCPUInfo, error) {
+func (c *cpuIdleTracing) detectCPUIdleContainer(threshold *cpuIdleThreshold) (*containerCPUInfo, error) {
 	for id, container := range containersCPUIdle {
 		if !container.alive {
 			delete(containersCPUIdle, id)
 		} else {
 			container.alive = false
 
-			if err := updateContainerCpuUsage(container); err != nil {
+			if err := c.updateContainerCpuUsage(container); err != nil {
 				log.Debugf("cpuidle update container [%s]: %v", container.path, err)
 				continue
 			}
@@ -161,8 +158,8 @@ func containerCpuUsageDelta(cpu1, cpu2 *cpuStats) cpuStats {
 	}
 }
 
-func updateContainerCpuUsage(container *containerCPUInfo) error {
-	cpuQuotaPeriod, err := cgroupMgr.CpuQuotaAndPeriod(container.path)
+func (c *cpuIdleTracing) updateContainerCpuUsage(container *containerCPUInfo) error {
+	cpuQuotaPeriod, err := c.cgroupMgr.CpuQuotaAndPeriod(container.path)
 	if err != nil {
 		return err
 	}
@@ -178,7 +175,7 @@ func updateContainerCpuUsage(container *containerCPUInfo) error {
 		return fmt.Errorf("cpu too small")
 	}
 
-	usage, err := cgroupMgr.CpuUsage(container.path)
+	usage, err := c.cgroupMgr.CpuUsage(container.path)
 	if err != nil {
 		return err
 	}
@@ -330,7 +327,7 @@ func (c *cpuIdleTracing) Start(ctx context.Context) error {
 				return err
 			}
 
-			container, err := detectCPUIdleContainer(threshold)
+			container, err := c.detectCPUIdleContainer(threshold)
 			if err != nil {
 				continue
 			}