fix runc events error in cgroup v2

Signed-off-by: lifubang <[email protected]>

Author: lifubang

go.mod

@@ -9,6 +9,7 @@ require (
 	github.com/coreos/go-systemd/v22 v22.0.0
 	github.com/cyphar/filepath-securejoin v0.2.2
 	github.com/docker/go-units v0.4.0
+	github.com/fsnotify/fsnotify v1.4.9
 	github.com/godbus/dbus/v5 v5.0.3
 	github.com/golang/protobuf v1.3.5
 	github.com/moby/sys/mountinfo v0.1.3

go.sum

@@ -15,6 +15,8 @@ github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/docker/go-units v0.4.0 h1:3uh0PgVws3nIA0Q+MwDC8yjEPf9zjRfZZWXZYDct3Tw=
 github.com/docker/go-units v0.4.0/go.mod h1:fgPhTUdO+D/Jk86RDLlptpiXQzgHJF7gydDDbaIK4Dk=
+github.com/fsnotify/fsnotify v1.4.9 h1:hsms1Qyu0jgnwNXIxa+/V/PDsU6CfLf6CNO8H7IWoS4=
+github.com/fsnotify/fsnotify v1.4.9/go.mod h1:znqG4EE+3YCdAaPaxE2ZRY/06pZUdp0tY4IgpuI1SZQ=
 github.com/godbus/dbus/v5 v5.0.3 h1:ZqHaoEF7TBzh4jzPmqVhE/5A1z9of6orkAe5uHoAeME=
 github.com/godbus/dbus/v5 v5.0.3/go.mod h1:xhWf0FNVPg57R7Z0UbKHbJfkEywrmjJnf7w5xrFpKfA=
 github.com/golang/protobuf v1.3.5 h1:F768QJ1E9tib+q5Sc8MkdJi1RxLTbRcTf8LJV56aRls=
@@ -54,6 +56,7 @@ github.com/vishvananda/netns v0.0.0-20191106174202-0a2b9b5464df h1:OviZH7qLw/7Zo
 github.com/vishvananda/netns v0.0.0-20191106174202-0a2b9b5464df/go.mod h1:JP3t17pCcGlemwknint6hfoeCVQrEMVwxRLRjXpq+BU=
 golang.org/x/sys v0.0.0-20190422165155-953cdadca894/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190606203320-7fc4e5ec1444/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20191005200804-aed5e4c7ecf9/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20191115151921-52ab43148777/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20191120155948-bd437916bb0e/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200124204421-9fbb57f87de9/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=

libcontainer/container_linux.go

@@ -648,6 +648,9 @@ func (c *linuxContainer) NotifyOOM() (<-chan struct{}, error) {
 	if c.config.RootlessCgroups {
 		logrus.Warn("getting OOM notifications may fail if you don't have the full access to cgroups")
 	}
+	if cgroups.IsCgroup2UnifiedMode() {
+		return notifyOnOOMV2(c.cgroupManager.GetPaths())
+	}
 	return notifyOnOOM(c.cgroupManager.GetPaths())
 }
 

libcontainer/notify_linux.go

@@ -87,4 +87,4 @@ func notifyMemoryPressure(paths map[string]string, level PressureLevel) (<-chan
 
 	levelStr := []string{"low", "medium", "critical"}[level]
 	return registerMemoryEvent(dir, "memory.pressure_level", levelStr)
-}
+}

libcontainer/notify_linux_v2.go

@@ -0,0 +1,92 @@
+// +build linux
+
+package libcontainer
+
+import (
+	"bufio"
+	"fmt"
+	"io"
+	"os"
+	"path/filepath"
+	"strings"
+
+	"github.com/fsnotify/fsnotify"
+	"github.com/sirupsen/logrus"
+)
+
+func checkOOM(path string) (bool, error) {
+	file, err := os.Open(path)
+	defer file.Close()
+	if err != nil {
+		return false, err
+	}
+	buff := bufio.NewReader(file)
+	for {
+		line, err := buff.ReadString('\n')
+		if err == io.EOF {
+			break
+		}
+		arr := strings.Fields(line)
+		if len(arr) == 2 && (arr[0] == "oom" || arr[0] == "oom_kill") && arr[1] == "1" {
+			return true, nil
+		}
+	}
+	return false, nil
+}
+
+func registerMemoryEventV2(cgDir string, evName string, arg string) (<-chan struct{}, error) {
+	eventControlPath := filepath.Join(cgDir, evName)
+	watcher, err := fsnotify.NewWatcher()
+	if err != nil {
+		return nil, err
+	}
+	if err := watcher.Add(eventControlPath); err != nil {
+		return nil, err
+	}
+	ch := make(chan struct{})
+	go func() {
+		defer func() {
+			watcher.Close()
+			close(ch)
+		}()
+		for {
+			select {
+			case event := <-watcher.Events:
+				// When a cgroup is destroyed, an event is sent to eventfd.
+				// So if the control path is gone, return instead of notifying.
+				if _, err := os.Lstat(eventControlPath); os.IsNotExist(err) {
+					return
+				}
+				if event.Op&fsnotify.Write > 0 {
+					isOOM, err := checkOOM(eventControlPath)
+					if err != nil {
+						return
+					}
+					if isOOM {
+						ch <- struct{}{}
+						return
+					}
+				} else if event.Op&fsnotify.Remove > 0 {
+					return
+				}
+			case err := <-watcher.Errors:
+				if err != nil {
+					logrus.Warn(err)
+					return
+				}
+			}
+		}
+	}()
+	return ch, nil
+}
+
+// notifyOnOOMV2 returns channel on which you can expect event about OOM,
+// if process died without OOM this channel will be closed.
+func notifyOnOOMV2(paths map[string]string) (<-chan struct{}, error) {
+	dir := paths[oomCgroupName]
+	if dir == "" {
+		return nil, fmt.Errorf("path %q missing", oomCgroupName)
+	}
+
+	return registerMemoryEventV2(dir, "memory.events", "")
+}