Adapt vfs kqfilter to the shared vnode lock used by zfs write vop. Use

vnode interlock to protect the knote fields [1]. The locking assumes
that shared vnode lock is held, thus we get exclusive access to knote
either by exclusive vnode lock protection, or by shared vnode lock +
vnode interlock.

Do not use kl_locked() method to assert either lock ownership or the
fact that curthread does not own the lock. For shared locks, ownership
is not recorded, e.g. VOP_ISLOCKED can return LK_SHARED for the shared
lock not owned by curthread, causing false positives in kqueue subsystem
assertions about knlist lock.

Remove kl_locked method from knlist lock vector, and add two separate
assertion methods kl_assert_locked and kl_assert_unlocked, that are
supposed to use proper asserts. Change knlist_init accordingly.

Add convenience function knlist_init_mtx to reduce number of arguments
for typical knlist initialization.

Submitted by:	jhb [1]
Noted by:	jhb [2]
Reviewed by:	jhb
Tested by:	rnoland

4 files changed, 5 insertions, 5 deletions

diff --git a/sys/net/bpf.c b/sys/net/bpf.c
index c74d09f..6c95fc6 100644
--- a/sys/net/bpf.c
+++ b/sys/net/bpf.c
@@ -648,7 +648,7 @@ bpfopen(struct cdev *dev, int flags, int fmt, struct thread *td)
 #endif
 	mtx_init(&d->bd_mtx, devtoname(dev), "bpf cdev lock", MTX_DEF);
 	callout_init(&d->bd_callout, CALLOUT_MPSAFE);
-	knlist_init(&d->bd_sel.si_note, &d->bd_mtx, NULL, NULL, NULL);
+	knlist_init_mtx(&d->bd_sel.si_note, &d->bd_mtx);
 
 	return (0);
 }
diff --git a/sys/net/if.c b/sys/net/if.c
index ba4cb1a..c95de07 100644
--- a/sys/net/if.c
+++ b/sys/net/if.c
@@ -445,7 +445,7 @@ vnet_net_iattach(const void *unused __unused)
 
 	TAILQ_INIT(&V_ifnet);
 	TAILQ_INIT(&V_ifg_head);
-	knlist_init(&V_ifklist, NULL, NULL, NULL, NULL);
+	knlist_init_mtx(&V_ifklist, NULL);
 	if_grow();				/* create initial table */
 
 	return (0);
@@ -548,7 +548,7 @@ if_alloc(u_char type)
 	TAILQ_INIT(&ifp->if_prefixhead);
 	TAILQ_INIT(&ifp->if_multiaddrs);
 	TAILQ_INIT(&ifp->if_groups);
-	knlist_init(&ifp->if_klist, NULL, NULL, NULL, NULL);
+	knlist_init_mtx(&ifp->if_klist, NULL);
 #ifdef MAC
 	mac_ifnet_init(ifp);
 #endif
diff --git a/sys/net/if_tap.c b/sys/net/if_tap.c
index 9256301..6668d94 100644
--- a/sys/net/if_tap.c
+++ b/sys/net/if_tap.c
@@ -462,7 +462,7 @@ tapcreate(struct cdev *dev)
 	tp->tap_flags |= TAP_INITED;
 	mtx_unlock(&tp->tap_mtx);
 
-	knlist_init(&tp->tap_rsel.si_note, NULL, NULL, NULL, NULL);
+	knlist_init_mtx(&tp->tap_rsel.si_note, NULL);
 
 	TAPDEBUG("interface %s is created. minor = %#x\n", 
 		ifp->if_xname, dev2unit(dev));
diff --git a/sys/net/if_tun.c b/sys/net/if_tun.c
index 6b352e8..03dd4ab 100644
--- a/sys/net/if_tun.c
+++ b/sys/net/if_tun.c
@@ -391,7 +391,7 @@ tuncreate(const char *name, struct cdev *dev)
 	IFQ_SET_MAXLEN(&ifp->if_snd, ifqmaxlen);
 	ifp->if_snd.ifq_drv_maxlen = 0;
 	IFQ_SET_READY(&ifp->if_snd);
-	knlist_init(&sc->tun_rsel.si_note, NULL, NULL, NULL, NULL);
+	knlist_init_mtx(&sc->tun_rsel.si_note, NULL);
 
 	if_attach(ifp);
 	bpfattach(ifp, DLT_NULL, sizeof(u_int32_t));