OCPBUGS-66973: Ntpfailover backport 418

Dockerfile

@@ -8,6 +8,8 @@ FROM registry.ci.openshift.org/ocp/4.18:base-rhel9
 RUN yum -y update && yum -y update glibc && yum --setopt=skip_missing_names_on_install=False -y install linuxptp ethtool hwdata synce4l && yum clean all
 
 
+RUN yum install -y procps-ng
+RUN yum install -y chrony
 RUN yum install -y gpsd-minimal
 RUN yum install -y gpsd-minimal-clients
 

addons/generic/ntpfailover.go

@@ -0,0 +1,231 @@
+package generic
+
+import (
+	"encoding/json"
+	"regexp"
+	"sync"
+	"time"
+
+	"github.com/golang/glog"
+
+	"github.com/openshift/linuxptp-daemon/pkg/plugin"
+	ptpv1 "github.com/openshift/ptp-operator/api/v1"
+)
+
+type ntpFailoverPluginData struct {
+	gnssFailover    bool
+	cmdSetEnabled   map[string]func(bool)
+	pcfsmState      int
+	pcfsmMutex      sync.Mutex
+	pcfsmLocked     bool
+	ts2phcTolerance time.Duration
+	startupDelay    time.Duration
+	expiryTime      time.Time
+}
+
+type ntpFailoverOpts struct {
+	StartupDelay    string `json:"startupDelay"`
+	Ts2phcTolerance string `json:"ts2phcTolerance"` //nolint:stylecheck
+	GnssFailover    bool   `json:"gnssFailover"`
+}
+
+const ( // phc2sys/chronyd Finite State Machine States
+	pcsmsStartupDefault int = iota // Just started, both are unknown
+	pcsmsStartupPhc2sys            // phc2sys setting time
+	pcsmsStartupChronyd            // phc2sys setting time
+	pcsmsStartupBoth               // phc2sys setting time
+	pcsmsActive                    // phc2sys setting time
+	pcsmsOutOfSpec                 // switch to chronyd setting time
+	pcsmsFailover                  // chronyd setting time
+)
+
+const (
+	ts2phcPname  = "ts2phc"
+	chronydPname = "chronyd"
+	phc2sysPname = "phc2sys"
+)
+
+var (
+	ts2phcOffsetRegex  = regexp.MustCompile("offset .*s[23] freq")
+	chronydOnlineRegex = regexp.MustCompile("chronyd .* starting")
+)
+
+func onPTPConfigChangeNtpFailover(data *interface{}, nodeProfile *ptpv1.PtpProfile) error {
+	var _ntpFailoverOpts ntpFailoverOpts
+	_ntpFailoverOpts.StartupDelay = "90s"
+	_ntpFailoverOpts.Ts2phcTolerance = "5s"
+	_ntpFailoverOpts.GnssFailover = false
+	var err error
+	if data != nil {
+		_data := *data
+		var pluginData = _data.(*ntpFailoverPluginData)
+		if pluginData.cmdSetEnabled == nil {
+			pluginData.cmdSetEnabled = make(map[string]func(bool))
+		}
+		for name, opts := range (*nodeProfile).Plugins {
+			if name == "ntpfailover" {
+				optsByteArray, _ := json.Marshal(opts)
+				err = json.Unmarshal(optsByteArray, &_ntpFailoverOpts)
+				if err != nil {
+					glog.Error("ntpfailover failed to unmarshal opts: " + err.Error())
+				}
+			}
+		}
+
+		pluginData.gnssFailover = _ntpFailoverOpts.GnssFailover
+
+		pluginData.startupDelay, err = time.ParseDuration(_ntpFailoverOpts.StartupDelay)
+		if err != nil {
+			glog.Infof("Failed parsing startupDelay %s: %d.  Defaulting to 90 seconds.", _ntpFailoverOpts.StartupDelay, err)
+			pluginData.startupDelay, _ = time.ParseDuration("90s")
+		}
+		pluginData.expiryTime = time.Now().Add(pluginData.startupDelay)
+
+		pluginData.ts2phcTolerance, err = time.ParseDuration(_ntpFailoverOpts.Ts2phcTolerance)
+		if err != nil {
+			glog.Infof("Failed parsing ts2phcTolerance %s: %d.  Defaulting to 5 seconds.", _ntpFailoverOpts.Ts2phcTolerance, err)
+			pluginData.ts2phcTolerance, _ = time.ParseDuration("5s")
+		}
+	}
+	return nil
+}
+
+func registerProcessNtpFailover(data *interface{}, pname string, cmdSetEnabled func(bool)) {
+	if data != nil {
+		_data := *data
+
+		var pluginData = _data.(*ntpFailoverPluginData)
+		if pluginData.gnssFailover {
+			if pluginData.cmdSetEnabled == nil {
+				pluginData.cmdSetEnabled = make(map[string]func(bool))
+			}
+			pluginData.cmdSetEnabled[pname] = cmdSetEnabled
+		}
+	}
+}
+
+func processLogNtpFailover(data *interface{}, pname string, log string) string {
+	ret := log
+	if data != nil {
+		_data := *data
+
+		var pluginData = _data.(*ntpFailoverPluginData)
+		if pluginData.gnssFailover {
+			currentTime := time.Now()
+
+			if pname == ts2phcPname && ts2phcOffsetRegex.MatchString(log) {
+				pluginData.expiryTime = currentTime.Add(pluginData.ts2phcTolerance)
+			}
+
+			pluginData.pcfsmMutex.Lock()
+			ownLock := !pluginData.pcfsmLocked //If locked, then skip, otherwise take lock
+			pluginData.pcfsmMutex.Unlock()
+			if ownLock {
+			done:
+				for {
+					switch pluginData.pcfsmState {
+					case pcsmsStartupDefault:
+						_, foundChronyd := pluginData.cmdSetEnabled[chronydPname]
+						_, foundPhc2Sys := pluginData.cmdSetEnabled[phc2sysPname]
+						if foundChronyd && foundPhc2Sys {
+							pluginData.pcfsmState = pcsmsStartupBoth
+						} else if foundChronyd {
+							pluginData.pcfsmState = pcsmsStartupChronyd
+						} else if foundPhc2Sys {
+							pluginData.pcfsmState = pcsmsStartupPhc2sys
+						} else {
+							break done
+						}
+					case pcsmsStartupPhc2sys:
+						_, foundChronyd := pluginData.cmdSetEnabled[chronydPname]
+						if foundChronyd {
+							pluginData.pcfsmState = pcsmsStartupBoth
+						} else {
+							break done
+						}
+					case pcsmsStartupChronyd:
+						_, foundPhc2Sys := pluginData.cmdSetEnabled[phc2sysPname]
+						if foundPhc2Sys {
+							pluginData.pcfsmState = pcsmsStartupBoth
+						} else {
+							break done
+						}
+					case pcsmsStartupBoth:
+						chronydSetEnabled, ok := pluginData.cmdSetEnabled[chronydPname]
+						if ok {
+							chronydSetEnabled(false)
+						}
+						phc2sysSetEnabled, ok := pluginData.cmdSetEnabled[phc2sysPname]
+						if ok {
+							phc2sysSetEnabled(true)
+						}
+						pluginData.pcfsmState = pcsmsActive
+						continue
+					case pcsmsActive:
+						if pname == ts2phcPname {
+							if currentTime.After(pluginData.expiryTime) {
+								pluginData.pcfsmState = pcsmsOutOfSpec
+								continue
+							}
+						}
+						if pname == chronydPname && chronydOnlineRegex.MatchString(log) {
+							chronydSetEnabled, ok := pluginData.cmdSetEnabled[chronydPname]
+							if ok {
+								chronydSetEnabled(false)
+							}
+						}
+						break done
+					case pcsmsOutOfSpec:
+						if pname == ts2phcPname {
+							if currentTime.After(pluginData.expiryTime) {
+								pluginData.pcfsmState = pcsmsFailover
+								chronydSetEnabled, ok := pluginData.cmdSetEnabled[chronydPname]
+								if ok {
+									chronydSetEnabled(true)
+								}
+								phc2sysSetEnabled, ok := pluginData.cmdSetEnabled[phc2sysPname]
+								if ok {
+									phc2sysSetEnabled(false)
+								}
+								continue
+							}
+						}
+						break done
+					case pcsmsFailover:
+						if pname == ts2phcPname {
+							if currentTime.Before(pluginData.expiryTime) {
+								pluginData.pcfsmState = pcsmsStartupDefault
+								continue
+							}
+						}
+						break done
+					}
+				}
+				pluginData.pcfsmMutex.Lock()
+				pluginData.pcfsmLocked = false //If took lock, then return it
+				pluginData.pcfsmMutex.Unlock()
+			}
+		}
+	}
+
+	return ret
+}
+
+// NtpFailover initializes NtpFailover plugin
+func NtpFailover(name string) (*plugin.Plugin, *interface{}) {
+	if name != "ntpfailover" {
+		glog.Errorf("Plugin must be initialized as 'ntpfailover'")
+		return nil, nil
+	}
+	glog.Infof("registering ntpfailover plugin")
+	_plugin := plugin.Plugin{Name: "ntpfailover",
+		OnPTPConfigChange:      onPTPConfigChangeNtpFailover,
+		RegisterEnableCallback: registerProcessNtpFailover,
+		ProcessLog:             processLogNtpFailover,
+	}
+	pluginData := ntpFailoverPluginData{pcfsmState: pcsmsStartupDefault,
+		pcfsmMutex: sync.Mutex{}}
+	pluginData.cmdSetEnabled = make(map[string]func(bool))
+	var iface interface{} = &pluginData
+	return &_plugin, &iface
+}

addons/mapping.go

@@ -7,6 +7,7 @@ import (
 )
 
 var PluginMapping = map[string]plugin.New{
-	"reference": generic.Reference,
-	"e810":      intel.E810,
+	"reference":   generic.Reference,
+	"ntpfailover": generic.NtpFailover,
+	"e810":        intel.E810,
 }

go.mod

@@ -4,43 +4,44 @@ go 1.22.0
 
 require (
 	github.com/bigkevmcd/go-configparser v0.0.0-20240624060122-ccd05f93a9d2
-	github.com/facebook/time v0.0.0-20230529151911-512b3b30ab23
+	github.com/facebook/time v0.0.0-20241030181404-3e1b98825c29
 	github.com/golang/glog v1.2.4
 	github.com/google/goexpect v0.0.0-20210430020637-ab937bf7fd6f
 	github.com/jaypipes/ghw v0.12.0
 	github.com/mdlayher/genetlink v1.3.2
 	github.com/mdlayher/netlink v1.7.2
-	github.com/openshift/ptp-operator v0.0.0-20240623153039-2f8235d2dad8
+	github.com/openshift/ptp-operator v0.0.0-20251126211433-0dee6a03855c
 	github.com/prometheus/client_golang v1.16.0
-	github.com/sirupsen/logrus v1.9.0
+	github.com/sirupsen/logrus v1.9.3
 	github.com/stratoberry/go-gpsd v1.1.0
-	github.com/stretchr/testify v1.8.2
+	github.com/stretchr/testify v1.8.4
 	golang.org/x/sync v0.10.0
 	k8s.io/api v0.28.3
 	k8s.io/apimachinery v0.28.3
 	k8s.io/client-go v0.28.3
-	k8s.io/utils v0.0.0-20230726121419-3b25d923346b
-	sigs.k8s.io/yaml v1.3.0
+	k8s.io/utils v0.0.0-20240902221715-702e33fdd3c3
+	sigs.k8s.io/yaml v1.4.0
 )
 
 require (
+	github.com/Masterminds/semver/v3 v3.3.0 // indirect
 	github.com/StackExchange/wmi v1.2.1 // indirect
 	github.com/beorn7/perks v1.0.1 // indirect
 	github.com/cespare/xxhash/v2 v2.2.0 // indirect
-	github.com/davecgh/go-spew v1.1.1 // indirect
+	github.com/davecgh/go-spew v1.1.2-0.20180830191138-d8f796af33cc // indirect
 	github.com/emicklei/go-restful/v3 v3.11.0 // indirect
 	github.com/evanphx/json-patch v5.6.0+incompatible // indirect
 	github.com/evanphx/json-patch/v5 v5.6.0 // indirect
 	github.com/fsnotify/fsnotify v1.6.0 // indirect
 	github.com/ghodss/yaml v1.0.0 // indirect
-	github.com/go-logr/logr v1.2.4 // indirect
+	github.com/go-logr/logr v1.4.2 // indirect
 	github.com/go-ole/go-ole v1.2.6 // indirect
 	github.com/go-openapi/jsonpointer v0.19.6 // indirect
 	github.com/go-openapi/jsonreference v0.20.2 // indirect
-	github.com/go-openapi/swag v0.22.3 // indirect
+	github.com/go-openapi/swag v0.22.4 // indirect
 	github.com/gogo/protobuf v1.3.2 // indirect
 	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect
-	github.com/golang/protobuf v1.5.3 // indirect
+	github.com/golang/protobuf v1.5.4 // indirect
 	github.com/google/gnostic-models v0.6.8 // indirect
 	github.com/google/go-cmp v0.6.0 // indirect
 	github.com/google/gofuzz v1.2.0 // indirect
@@ -63,29 +64,31 @@ require (
 	github.com/prometheus/client_model v0.4.0 // indirect
 	github.com/prometheus/common v0.44.0 // indirect
 	github.com/prometheus/procfs v0.10.1 // indirect
+	github.com/rogpeppe/go-internal v1.11.0 // indirect
 	github.com/spf13/pflag v1.0.5 // indirect
 	golang.org/x/crypto v0.31.0 // indirect
-	golang.org/x/exp v0.0.0-20220722155223-a9213eeb770e // indirect
+	golang.org/x/exp v0.0.0-20220827204233-334a2380cb91 // indirect
 	golang.org/x/net v0.33.0 // indirect
-	golang.org/x/oauth2 v0.8.0 // indirect
+	golang.org/x/oauth2 v0.21.0 // indirect
 	golang.org/x/sys v0.28.0 // indirect
 	golang.org/x/term v0.27.0 // indirect
 	golang.org/x/text v0.21.0 // indirect
 	golang.org/x/time v0.3.0 // indirect
 	gomodules.xyz/jsonpatch/v2 v2.4.0 // indirect
-	google.golang.org/appengine v1.6.7 // indirect
 	google.golang.org/genproto/googleapis/rpc v0.0.0-20230525234030-28d5490b6b19 // indirect
-	google.golang.org/grpc v1.56.3 // indirect
-	google.golang.org/protobuf v1.33.0 // indirect
+	google.golang.org/grpc v1.54.0 // indirect
+	google.golang.org/protobuf v1.34.2 // indirect
 	gopkg.in/inf.v0 v0.9.1 // indirect
 	gopkg.in/yaml.v2 v2.4.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 	howett.net/plist v1.0.0 // indirect
 	k8s.io/apiextensions-apiserver v0.28.3 // indirect
 	k8s.io/component-base v0.28.3 // indirect
-	k8s.io/klog/v2 v2.100.1 // indirect
-	k8s.io/kube-openapi v0.0.0-20230717233707-2695361300d9 // indirect
+	k8s.io/klog/v2 v2.130.1 // indirect
+	k8s.io/kube-openapi v0.0.0-20240228011516-70dd3763d340 // indirect
 	sigs.k8s.io/controller-runtime v0.16.3 // indirect
 	sigs.k8s.io/json v0.0.0-20221116044647-bc3834ca7abd // indirect
-	sigs.k8s.io/structured-merge-diff/v4 v4.2.3 // indirect
+	sigs.k8s.io/structured-merge-diff/v4 v4.4.1 // indirect
 )
+
+replace github.com/openshift/ptp-operator => github.com/josephdrichard/ptp-operator v0.0.0-20251205063307-b04f1e9f5fc5