Replace manual !klist_empty()+knote() with KNOTE().

OK mpi@

diff --git a/sys/net/bpf.c b/sys/net/bpf.c
index 14b6319..ce0fa81 100644 (file)
--- a/sys/net/bpf.c
+++ b/sys/net/bpf.c
@@ -1,4 +1,4 @@
-/*     $OpenBSD: bpf.c,v 1.211 2022/02/05 04:08:27 dlg Exp $   */
+/*     $OpenBSD: bpf.c,v 1.212 2022/02/11 07:28:29 visa Exp $  */
 /*     $NetBSD: bpf.c,v 1.33 1997/02/21 23:59:35 thorpej Exp $ */
 
 /*
@@ -580,16 +580,12 @@ out:
 void
 bpf_wakeup(struct bpf_d *d)
 {
-       struct klist *klist;
-
        MUTEX_ASSERT_LOCKED(&d->bd_mtx);
 
        if (d->bd_nreaders)
                wakeup(d);
 
-       klist = &d->bd_sel.si_note;
-       if (!klist_empty(klist))
-               knote(klist, 0);
+       KNOTE(&d->bd_sel.si_note, 0);
 
        /*
         * As long as pgsigio() and selwakeup() need to be protected