feat: switch logging framework (#10220)

This updates our logging framework from legacy freetext strings using the `log` package to structured log entries using `log/slog`. I have updated all INFO or higher level entries, but not yet DEBUG (😓)... So, at a high level: There is a slight change in log levels, effectively adding a new warning level: - DEBUG is still debug (ideally not for users but developers, though this is something we need to work on) - INFO is still info, though I've added more data here, effectively making Syncthing more verbose by default (more on this below) - WARNING is a new log level that is different from the _old_ WARNING (more below) - ERROR is what was WARNING before -- problems that must be dealt with, and also bubbled as a popup in the GUI. A new feature is that the logging level can be set per package to something other than just debug or info, and hence I feel that we can add a bit more things into INFO while moving some (in fact, most) current INFO level warnings into WARNING. For example, I think it's justified to get a log of synced files in INFO and sync failures in WARNING. These are things that have historically been tricky to debug properly, and having more information by default will be useful to many, while still making it possible get close to told level of inscrutability by setting the log level to WARNING. I'd like to get to a stage where DEBUG is never necessary to just figure out what's going on, as opposed to trying to narrow down a likely bug. Code wise: - Our logging object, generally known as `l` in each package, is now a new adapter object that provides the old API on top of the newer one. (This should go away once all old log entries are migrated.) This is only for `l.Debugln` and `l.Debugf`. - There is a new level tracker that keeps the log level for each package. - There is a nested setup of handlers, since the structure mandated by `log/slog` is slightly convoluted (imho). We do this because we need to do formatting at a "medium" level internally so we can buffer log lines in text format but with separate timestamp and log level for the API/GUI to consume. - The `debug` API call becomes a `loglevels` API call, which can set the log level to `DEBUG`, `INFO`, `WARNING` or `ERROR` per package. The GUI is updated to handle this. - Our custom `sync` package provided some debugging of mutexes quite strongly integrated into the old logging framework, only turned on when `STTRACE` was set to certain values at startup, etc. It's been a long time since this has been useful; I removed it. - The `STTRACE` env var remains and can be used the same way as before, while additionally permitting specific log levels to be specified, `STTRACE=model:WARN,scanner:DEBUG`. - There is a new command line option `--log-level=INFO` to set the default log level. - The command line options `--log-flags` and `--verbose` go away, but are currently retained as hidden & ignored options since we set them by default in some of our startup examples and Syncthing would otherwise fail to start. Sample format messages: ``` 2009-02-13 23:31:30 INF A basic info line (attr1="val with spaces" attr2=2 attr3="val\"quote" a=a log.pkg=slogutil) 2009-02-13 23:31:30 INF An info line with grouped values (attr1=val1 foo.attr2=2 foo.bar.attr3=3 a=a log.pkg=slogutil) 2009-02-13 23:31:30 INF An info line with grouped values via logger (foo.attr1=val1 foo.attr2=2 a=a log.pkg=slogutil) 2009-02-13 23:31:30 INF An info line with nested grouped values via logger (bar.foo.attr1=val1 bar.foo.attr2=2 a=a log.pkg=slogutil) 2009-02-13 23:31:30 WRN A warning entry (a=a log.pkg=slogutil) 2009-02-13 23:31:30 ERR An error (a=a log.pkg=slogutil) ``` --------- Co-authored-by: Ross Smith II <ross@smithii.com>
2025-08-07 11:19:36 +02:00
parent 49462448d0
commit 836045ee87
149 changed files with 1797 additions and 2641 deletions
@@ -19,17 +19,18 @@ import (
 	"errors"
 	"fmt"
 	"io"
+	"log/slog"
 	"math"
 	"net"
 	"net/url"
 	"slices"
 	"strings"
-	stdsync "sync"
+	"sync"
 	"time"

 	"github.com/thejerf/suture/v4"
-	"golang.org/x/time/rate"

+	"github.com/syncthing/syncthing/internal/slogutil"
 	"github.com/syncthing/syncthing/lib/build"
 	"github.com/syncthing/syncthing/lib/config"
 	"github.com/syncthing/syncthing/lib/connections/registry"
@@ -42,7 +43,6 @@ import (
 	"github.com/syncthing/syncthing/lib/sliceutil"
 	"github.com/syncthing/syncthing/lib/stringutil"
 	"github.com/syncthing/syncthing/lib/svcutil"
-	"github.com/syncthing/syncthing/lib/sync"

 	// Registers NAT service providers
 	_ "github.com/syncthing/syncthing/lib/pmp"
@@ -185,7 +185,7 @@ type service struct {
 }

 func NewService(cfg config.Wrapper, myID protocol.DeviceID, mdl Model, tlsCfg *tls.Config, discoverer discover.Finder, bepProtocolName string, tlsDefaultCommonName string, evLogger events.Logger, registry *registry.Registry, keyGen *protocol.KeyGenerator) Service {
-	spec := svcutil.SpecWithInfoLogger(l)
+	spec := svcutil.SpecWithInfoLogger()
 	service := &service{
 		Supervisor:              suture.New("connections.Service", spec),
 		connectionStatusHandler: newConnectionStatusHandler(),
@@ -206,11 +206,9 @@ func NewService(cfg config.Wrapper, myID protocol.DeviceID, mdl Model, tlsCfg *t
 		keyGen:               keyGen,
 		lanChecker:           &lanChecker{cfg},

-		dialNowDevicesMut: sync.NewMutex(),
-		dialNow:           make(chan struct{}, 1),
-		dialNowDevices:    make(map[protocol.DeviceID]struct{}),
+		dialNow:        make(chan struct{}, 1),
+		dialNowDevices: make(map[protocol.DeviceID]struct{}),

-		listenersMut:   sync.NewRWMutex(),
 		listeners:      make(map[string]genericListener),
 		listenerTokens: make(map[string]suture.ServiceToken),
 	}
@@ -257,7 +255,7 @@ func (s *service) handleConns(ctx context.Context) error {
 		// because there are implementations out there that don't support
 		// protocol negotiation (iOS for one...).
 		if cs.NegotiatedProtocol != s.bepProtocolName {
-			l.Infof("Peer at %s did not negotiate bep/1.0", c)
+			slog.WarnContext(ctx, "Peer at did not negotiate bep/1.0", slogutil.Address(c.RemoteAddr()))
 		}

 		// We should have received exactly one certificate from the other
@@ -265,7 +263,7 @@ func (s *service) handleConns(ctx context.Context) error {
 		// connection.
 		certs := cs.PeerCertificates
 		if cl := len(certs); cl != 1 {
-			l.Infof("Got peer certificate list of length %d != 1 from peer at %s; protocol error", cl, c)
+			slog.WarnContext(ctx, "Got peer certificate list of incorrect length", slog.Int("length", cl), slogutil.Address(c.RemoteAddr()))
 			c.Close()
 			continue
 		}
@@ -276,17 +274,13 @@ func (s *service) handleConns(ctx context.Context) error {
 		// though, especially in the presence of NAT hairpinning, multiple
 		// clients between the same NAT gateway, and global discovery.
 		if remoteID == s.myID {
-			l.Debugf("Connected to myself (%s) at %s", remoteID, c)
+			slog.DebugContext(ctx, "Connected to myself", "id", remoteID, "addr", c)
 			c.Close()
 			continue
 		}

 		if err := s.connectionCheckEarly(remoteID, c); err != nil {
-			if errors.Is(err, errDeviceAlreadyConnected) {
-				l.Debugf("Connection from %s at %s (%s) rejected: %v", remoteID, c.RemoteAddr(), c.Type(), err)
-			} else {
-				l.Infof("Connection from %s at %s (%s) rejected: %v", remoteID, c.RemoteAddr(), c.Type(), err)
-			}
+			slog.DebugContext(ctx, "Connection rejected", remoteID.LogAttr(), slogutil.Address(c.RemoteAddr()), slog.String("type", c.Type()), slogutil.Error(err))
 			c.Close()
 			continue
 		}
@@ -382,22 +376,10 @@ func (s *service) handleHellos(ctx context.Context) error {

 		if err != nil {
 			if protocol.IsVersionMismatch(err) {
-				// The error will be a relatively user friendly description
-				// of what's wrong with the version compatibility. By
-				// default identify the other side by device ID and IP.
-				remote := fmt.Sprintf("%v (%v)", remoteID, c.RemoteAddr())
-				if hello.DeviceName != "" {
-					// If the name was set in the hello return, use that to
-					// give the user more info about which device is the
-					// affected one. It probably says more than the remote
-					// IP.
-					remote = fmt.Sprintf("%q (%s %s, %v)", hello.DeviceName, hello.ClientName, hello.ClientVersion, remoteID)
-				}
-				msg := fmt.Sprintf("Connecting to %s: %s", remote, err)
-				warningFor(remoteID, msg)
+				slog.WarnContext(ctx, "Remote device is too old", remoteID.LogAttr(), slogutil.Address(c.RemoteAddr()), slogutil.Error(err))
 			} else {
 				// It's something else - connection reset or whatever
-				l.Infof("Failed to exchange Hello messages with %s at %s: %s", remoteID, c, err)
+				slog.WarnContext(ctx, "Failed to exchange Hello messages", remoteID.LogAttr(), slogutil.Address(c.RemoteAddr()), slogutil.Error(err))
 			}
 			c.Close()
 			continue
@@ -407,14 +389,14 @@ func (s *service) handleHellos(ctx context.Context) error {
 		// The Model will return an error for devices that we don't want to
 		// have a connection with for whatever reason, for example unknown devices.
 		if err := s.model.OnHello(remoteID, c.RemoteAddr(), hello); err != nil {
-			l.Infof("Connection from %s at %s (%s) rejected: %v", remoteID, c.RemoteAddr(), c.Type(), err)
+			slog.WarnContext(ctx, "Connection rejected", remoteID.LogAttr(), slogutil.Address(c.RemoteAddr()), slog.Any("type", c.Type()), slogutil.Error(err))
 			c.Close()
 			continue
 		}

 		deviceCfg, ok := s.cfg.Device(remoteID)
 		if !ok {
-			l.Infof("Device %s removed from config during connection attempt at %s", remoteID, c)
+			slog.WarnContext(ctx, "Device removed from config during connection attempt", remoteID.LogAttr(), slogutil.Address(c.RemoteAddr()))
 			c.Close()
 			continue
 		}
@@ -434,7 +416,7 @@ func (s *service) handleHellos(ctx context.Context) error {
 			// Incorrect certificate name is something the user most
 			// likely wants to know about, since it's an advanced
 			// config. Warn instead of Info.
-			l.Warnf("Bad certificate from %s at %s: %v", remoteID, c, err)
+			slog.ErrorContext(ctx, "Bad certificate from remote", remoteID.LogAttr(), slogutil.Address(c.RemoteAddr()), slogutil.Error(err))
 			c.Close()
 			continue
 		}
@@ -455,7 +437,7 @@ func (s *service) handleHellos(ctx context.Context) error {
 			s.dialNowDevicesMut.Unlock()
 		}()

-		l.Infof("Established secure connection to %s at %s", remoteID.Short(), c)
+		slog.InfoContext(ctx, "Established secure connection", remoteID.LogAttr(), slog.Any("connection", c))

 		s.model.AddConnection(protoConn, hello)
 		continue
@@ -477,9 +459,9 @@ func (s *service) connect(ctx context.Context) error {
 		bestDialerPriority := s.bestDialerPriority(cfg)
 		isInitialRampup := initialRampup < stdConnectionLoopSleep

-		l.Debugln("Connection loop")
+		slog.DebugContext(ctx, "Connection loop")
 		if isInitialRampup {
-			l.Debugln("Connection loop in initial rampup")
+			slog.DebugContext(ctx, "Connection loop in initial rampup")
 		}

 		// Used for consistency throughout this loop run, as time passes
@@ -616,9 +598,9 @@ func (s *service) dialDevices(ctx context.Context, now time.Time, cfg config.Con
 	// Perform dials according to the queue, stopping when we've reached the
 	// allowed additional number of connections (if limited).
 	numConns := 0
-	var numConnsMut stdsync.Mutex
+	var numConnsMut sync.Mutex
 	dialSemaphore := semaphore.New(dialMaxParallel)
-	dialWG := new(stdsync.WaitGroup)
+	dialWG := new(sync.WaitGroup)
 	dialCtx, dialCancel := context.WithCancel(ctx)
 	defer func() {
 		dialWG.Wait()
@@ -675,7 +657,7 @@ func (s *service) resolveDialTargets(ctx context.Context, now time.Time, cfg con
 		uri, err := url.Parse(addr)
 		if err != nil {
 			s.setConnectionStatus(addr, err)
-			l.Infof("Parsing dialer address %s: %v", addr, err)
+			slog.WarnContext(ctx, "Failed to parse dialer address", slogutil.Address(addr), slogutil.Error(err))
 			continue
 		}

@@ -695,7 +677,7 @@ func (s *service) resolveDialTargets(ctx context.Context, now time.Time, cfg con
 			l.Debugf("Dialer for %v: %v", uri, err)
 			continue
 		} else if err != nil {
-			l.Infof("Dialer for %v: %v", uri, err)
+			slog.WarnContext(ctx, "Failed to get dialer", slogutil.URI(uri), slogutil.Error(err))
 			continue
 		}

@@ -711,7 +693,7 @@ func (s *service) resolveDialTargets(ctx context.Context, now time.Time, cfg con
 			continue
 		}
 		if currentConns >= s.desiredConnectionsToDevice(deviceCfg.DeviceID) && priority == priorityCutoff {
-			l.Debugf("Not dialing %s at %s using %s as priority is equal and we already have %d/%d connections", deviceID.Short(), addr, dialerFactory, currentConns, deviceCfg.NumConnections)
+			l.Debugf("Not dialing %s at %s using %s as priority is equal and we already have %d/%d connections", deviceID.Short(), addr, dialerFactory, currentConns, deviceCfg.NumConnections())
 			continue
 		}

@@ -818,7 +800,7 @@ func (s *lanChecker) isLAN(addr net.Addr) bool {
 func (s *service) createListener(factory listenerFactory, uri *url.URL) bool {
 	// must be called with listenerMut held

-	l.Debugln("Starting listener", uri)
+	slog.Debug("Starting listener", "uri", uri)

 	listener := factory.New(uri, s.cfg, s.tlsCfg, s.conns, s.natService, s.registry, s.lanChecker)
 	listener.OnAddressesChanged(s.logListenAddressesChangedEvent)
@@ -826,7 +808,7 @@ func (s *service) createListener(factory listenerFactory, uri *url.URL) bool {
 	// Retrying a listener many times in rapid succession is unlikely to help,
 	// thus back off quickly. A listener may soon be functional again, e.g. due
 	// to a network interface coming back online - retry every minute.
-	spec := svcutil.SpecWithInfoLogger(l)
+	spec := svcutil.SpecWithInfoLogger()
 	spec.FailureThreshold = 2
 	spec.FailureBackoff = time.Minute
 	sup := suture.New(fmt.Sprintf("listenerSupervisor@%v", listener), spec)
@@ -854,9 +836,6 @@ func (s *service) CommitConfiguration(from, to config.Configuration) bool {

 	for _, dev := range from.Devices {
 		if !newDevices[dev.DeviceID] {
-			warningLimitersMut.Lock()
-			delete(warningLimiters, dev.DeviceID)
-			warningLimitersMut.Unlock()
 			metricDeviceActiveConnections.DeleteLabelValues(dev.DeviceID.String())
 		}
 	}
@@ -875,7 +854,7 @@ func (s *service) CommitConfiguration(from, to config.Configuration) bool {

 		uri, err := url.Parse(addr)
 		if err != nil {
-			l.Warnf("Skipping malformed listener URL %q: %v", addr, err)
+			slog.Error("Skipping malformed listener URL", slogutil.URI(addr), slogutil.Error(err))
 			continue
 		}

@@ -886,7 +865,7 @@ func (s *service) CommitConfiguration(from, to config.Configuration) bool {
 		// mean something entirely different to the computer (e.g.,
 		// tcp:/127.0.0.1:22000 in fact being equivalent to tcp://:22000).
 		if canonical := uri.String(); canonical != addr {
-			l.Warnf("Skipping malformed listener URL %q (not canonical)", addr)
+			slog.Error("Skipping malformed listener URL (not canonical)", slogutil.URI(addr))
 			continue
 		}

@@ -900,7 +879,7 @@ func (s *service) CommitConfiguration(from, to config.Configuration) bool {
 			l.Debugf("Listener for %v: %v", uri, err)
 			continue
 		} else if err != nil {
-			l.Infof("Listener for %v: %v", uri, err)
+			slog.Warn("Failed to get listener", slogutil.URI(uri), slogutil.Error(err))
 			continue
 		}

@@ -1007,8 +986,7 @@ type connectionStatusHandler struct {

 func newConnectionStatusHandler() connectionStatusHandler {
 	return connectionStatusHandler{
-		connectionStatusMut: sync.NewRWMutex(),
-		connectionStatus:    make(map[string]ConnectionStatusEntry),
+		connectionStatus: make(map[string]ConnectionStatusEntry),
 	}
 }

@@ -1082,24 +1060,6 @@ func urlsToStrings(urls []*url.URL) []string {
 	return strings
 }

-var (
-	warningLimiters    = make(map[protocol.DeviceID]*rate.Limiter)
-	warningLimitersMut = sync.NewMutex()
-)
-
-func warningFor(dev protocol.DeviceID, msg string) {
-	warningLimitersMut.Lock()
-	defer warningLimitersMut.Unlock()
-	lim, ok := warningLimiters[dev]
-	if !ok {
-		lim = rate.NewLimiter(rate.Every(perDeviceWarningIntv), 1)
-		warningLimiters[dev] = lim
-	}
-	if lim.Allow() {
-		l.Warnln(msg)
-	}
-}
-
 func tlsTimedHandshake(tc *tls.Conn) error {
 	tc.SetDeadline(time.Now().Add(tlsHandshakeTimeout))
 	defer tc.SetDeadline(time.Time{})
@@ -1156,7 +1116,7 @@ func (s *service) dialParallel(ctx context.Context, deviceID protocol.DeviceID,
 	for _, prio := range priorities {
 		tgts := dialTargetBuckets[prio]
 		res := make(chan internalConn, len(tgts))
-		wg := stdsync.WaitGroup{}
+		wg := sync.WaitGroup{}
 		for _, tgt := range tgts {
 			sema.Take(1)
 			wg.Add(1)
@@ -1215,7 +1175,7 @@ func (s *service) validateIdentity(c internalConn, expectedID protocol.DeviceID)
 	// connection.
 	certs := cs.PeerCertificates
 	if cl := len(certs); cl != 1 {
-		l.Infof("Got peer certificate list of length %d != 1 from peer at %s; protocol error", cl, c)
+		slog.Warn("Got peer certificate list of incorrect length", slog.Int("length", cl), slogutil.Address(c.RemoteAddr()))
 		c.Close()
 		return fmt.Errorf("expected 1 certificate, got %d", cl)
 	}
@@ -1364,7 +1324,7 @@ func (s *service) desiredConnectionsToDevice(deviceID protocol.DeviceID) int {
 // connected to and how many connections we have to each device. It also
 // tracks how many connections they are willing to use.
 type deviceConnectionTracker struct {
-	connectionsMut  stdsync.Mutex
+	connectionsMut  sync.Mutex
 	connections     map[protocol.DeviceID][]protocol.Connection // current connections
 	wantConnections map[protocol.DeviceID]int                   // number of connections they want
 }