fix: fd operator segment fault because not being cached

net_netfd.go

@@ -137,6 +137,11 @@ func (c *netFD) connect(ctx context.Context, la, ra syscall.Sockaddr) (rsa sysca
 	}
 
 	c.pd = newPollDesc(c.fd)
+	defer func() {
+		// free operator to avoid leak
+		c.pd.operator.Free()
+		c.pd = nil
+	}()
 	for {
 		// Performing multiple connect system calls on a
 		// non-blocking socket under Unix variants does not

net_polldesc.go

@@ -24,12 +24,11 @@ import (
 func newPollDesc(fd int) *pollDesc {
 	pd := &pollDesc{}
 	poll := pollmanager.Pick()
-	pd.operator = &FDOperator{
-		poll:    poll,
-		FD:      fd,
-		OnWrite: pd.onwrite,
-		OnHup:   pd.onhup,
-	}
+	pd.operator = poll.Alloc()
+	pd.operator.poll = poll
+	pd.operator.FD = fd
+	pd.operator.OnWrite = pd.onwrite
+	pd.operator.OnHup = pd.onhup
 	pd.writeTrigger = make(chan struct{})
 	pd.closeTrigger = make(chan struct{})
 	return pd
@@ -60,10 +59,7 @@ func (pd *pollDesc) WaitWrite(ctx context.Context) (err error) {
 		err = nil
 	case <-ctx.Done(): // triggered by ctx
 		// deregister from poller, upper caller function will close fd
-		// detach first but there's a very small possibility that operator is doing in poller,
-		// so need call unused() to wait operator done
 		pd.detach()
-		pd.operator.unused()
 		err = mapErr(ctx.Err())
 	}
 	// double check close trigger