(Mostly) markup fixes.

Approved by:	re (blanket)

16 files changed, 235 insertions, 211 deletions

diff --git a/share/man/man3/pthread_atfork.3 b/share/man/man3/pthread_atfork.3
index e2f240a..125bb43 100644
--- a/share/man/man3/pthread_atfork.3
+++ b/share/man/man3/pthread_atfork.3
@@ -59,9 +59,9 @@ in the context of the thread that called
 .Xr fork 2 .
 .Pp
 The handlers registered with
-.Nm
+.Fn pthread_atfork
 are called at the moments in time described below:
-.Bl -tag -width "prepare"
+.Bl -tag -width ".Fa prepare"
 .It Fa prepare
 Before
 .Xr fork 2
@@ -86,7 +86,7 @@ If more than one
 .Fa child
 handler is registered they will be called in the same order
 they were registered.
-.El.
+.El
 .Pp
 If no handling is desired at one or more of these three points,
 a null pointer may be passed as the corresponding fork handler.
@@ -102,6 +102,7 @@ function will fail if:
 .Bl -tag -width Er
 .It Bq Er ENOMEM
 Insufficient table space exists to record the fork handler addresses.
+.El
 .Sh SEE ALSO
 .Xr fork 2 ,
 .Xr pthread 3
diff --git a/share/man/man4/carp.4 b/share/man/man4/carp.4
index 749e777..5535c3b 100644
--- a/share/man/man4/carp.4
+++ b/share/man/man4/carp.4
@@ -73,12 +73,12 @@ advertisements.
 The
 .Cm advbase
 parameter stands for
-.Qq advertisement base .
+.Dq "advertisement base" .
 It is measured in seconds and specifies the base of the adverisement interval.
 The
 .Cm advskew
 parameter stands for
-.Qq advertisement skew .
+.Dq "advertisement skew" .
 It is measured in 1/256 of seconds.
 It is added to the base advertisement interval to make one host advertise
 a bit slower that the other does.
@@ -86,7 +86,7 @@ Both
 .Cm advbase
 and
 .Cm advskew
-are put inside CARP advertisments.
+are put inside CARP advertisements.
 These configurations can be done using
 .Xr ifconfig 8 ,
 or through the
@@ -225,9 +225,9 @@ It cannot balance traffic that crosses a router, because the router
 itself will always be balanced to the same virtual host.
 .Sh SEE ALSO
 .Xr inet 4 ,
-.Xr ifconfig 8 ,
 .Xr pfsync 4 ,
 .Xr rc.conf 5 ,
+.Xr ifconfig 8 ,
 .Xr sysctl 8
 .Sh HISTORY
 The
diff --git a/share/man/man4/cdce.4 b/share/man/man4/cdce.4
index 50ea08d..b4516b4 100644
--- a/share/man/man4/cdce.4
+++ b/share/man/man4/cdce.4
@@ -119,6 +119,7 @@ device driver first appeared in
 and
 .Fx 6.0 .
 .Sh AUTHORS
+.An -nosplit
 The
 .Nm
 driver was written by
diff --git a/share/man/man4/if_bridge.4 b/share/man/man4/if_bridge.4
index 1164aeb..c271116 100644
--- a/share/man/man4/if_bridge.4
+++ b/share/man/man4/if_bridge.4
@@ -34,20 +34,21 @@
 .\" POSSIBILITY OF SUCH DAMAGE.
 .\"
 .\" $FreeBSD$
-.Dd January 4, 2004
+.\"
+.Dd June 10, 2005
 .Dt IF_BRIDGE 4
 .Os
 .Sh NAME
 .Nm if_bridge
 .Nd network bridge device
 .Sh SYNOPSIS
-.Cd device if_bridge
+.Cd "device if_bridge"
 .Sh DESCRIPTION
 The
 .Nm
 driver creates a logical link between two or more IEEE 802 networks
 that use the same (or
-.Dq similar enough )
+.Dq "similar enough" )
 framing format.
 For example, it is possible to bridge Ethernet and 802.11 networks together,
 but it is not possible to bridge Ethernet and Token Ring together.
@@ -82,57 +83,65 @@ Spanning Tree is used to detect and remove loops in a network topology.
 When filtering is enabled, bridged packets will pass through the filter
 inbound on the originating interface, on the bridge interface and outbound on
 the appropriate interfaces.
-Either stage can be disabled, this behaviour can be controlled using 
+Either stage can be disabled, this behaviour can be controlled using
 .Xr sysctl 8 :
 .Bl -tag -width ".Va net.link.bridge.pfil_member"
 .It Va net.link.bridge.pfil_member
 Set to
 .Li 1
-to enable enable filtering on the incoming and outgoing member interfaces, set
-to 
+to enable filtering on the incoming and outgoing member interfaces, set
+to
 .Li 0
 to disable it.
 .It Va net.link.bridge.pfil_bridge
 Set to
 .Li 1
-to enable enable filtering on the bridge interface, set
-to 
+to enable filtering on the bridge interface, set
+to
 .Li 0
 to disable it.
 .It Va net.link.bridge.ipfw
 Set to
 .Li 1
-to enable enable layer2 filtering with
-.Xr ipfirewall 4
-, set to 
+to enable layer2 filtering with
+.Xr ipfirewall 4 ,
+set to
 .Li 0
 to disable it.
 This needs to be enabled for
 .Xr dummynet 4
 support.
-When ipfw is enabled pfil_bridge and pfil_member will be disabled so that IPFW
-is not run twice, these can be re-enabled if desired.
+When
+.Va ipfw
+is enabled,
+.Va pfil_bridge
+and
+.Va pfil_member
+will be disabled so that IPFW
+is not run twice; these can be re-enabled if desired.
 .El
 .Pp
 ARP and REVARP packets are forwarded without being filtered and others
-that are not IP nor IPv6 packets are not forwarded when 
+that are not IP nor IPv6 packets are not forwarded when
 .Xr pfil 9
 filtering is enabled.
-IPFW can filter Ethernet types using 'mac-type' so all packets are passed to
+IPFW can filter Ethernet types using
+.Cm mac-type
+so all packets are passed to
 the filter for processing.
 .Pp
 Note that packets to and from the bridging host will be seen by the
 filter on the interface with the appropriate address configured as well
 as on the interface on which the packet arrives or departs.
 .Sh EXAMPLES
-The following then placed in the file
+The following when placed in the file
 .Pa /etc/rc.conf
-will cause the a bridge called
-.Sq bridge0
+will cause a bridge called
+.Dq Li bridge0
 to be created, and will add the interfaces
-.Sq wi0
+.Dq Li wi0
 and
-.Sq fxp0
+.Dq Li fxp0
 to the bridge, and then enable packet forwarding.
 Such a configuration could be used to implement a simple
 802.11-to-Ethernet bridge (assuming the 802.11 interface is
@@ -182,5 +191,5 @@ driver has been heavily modified from the original version by
 .Sh BUGS
 The
 .Nm
-driver currently supports only Ethernet and Ethernet-like (e.g. 802.11)
+driver currently supports only Ethernet and Ethernet-like (e.g., 802.11)
 network devices, with exactly the same interface MTU size as the bridge device.
diff --git a/share/man/man4/man4.i386/acpi_fujitsu.4 b/share/man/man4/man4.i386/acpi_fujitsu.4
index 565f1be..e634f0d 100644
--- a/share/man/man4/man4.i386/acpi_fujitsu.4
+++ b/share/man/man4/man4.i386/acpi_fujitsu.4
@@ -62,11 +62,10 @@ Defaults for these sysctls can be set in
 .Xr sysctl.conf 5 .
 .Sh EXAMPLES
 The following can be added to
-.Xr devd.conf 5 
+.Xr devd.conf 5
 in order to pass button events to a
 .Pa /usr/local/sbin/acpi_oem_exec.sh
 script:
-.Pp
 .Bd -literal -offset indent
 notify 10 {
         match "system"		"ACPI";
diff --git a/share/man/man4/ng_ccatm.4 b/share/man/man4/ng_ccatm.4
index 3fecdea..598b5ed 100644
--- a/share/man/man4/ng_ccatm.4
+++ b/share/man/man4/ng_ccatm.4
@@ -32,49 +32,51 @@
 .\" $FreeBSD$
 .\"
 .Dd March 10, 2005
-.Dt ng_ccatm 4
-.Os FreeBSD
+.Dt NG_CCATM 4
+.Os
 .Sh NAME
 .Nm ng_ccatm
-.Nd netgraph ATM Call Control node type
+.Nd "ATM Call Control netgraph node type"
 .Sh SYNOPSIS
-.Fd #include <netnatm/unimsg.h>
-.Fd #include <netnatm/msg/unistruct.h>
-.Fd #include <netnatm/sig/unidef.h>
-.Fd #include <netnatm/api/unisap.h>
-.Fd #include <netnatm/api/atmapi.h>
-.Fd #include <netnatm/api/ccatm.h>
-.Fd #include <netgraph.h>
-.Fd #include <netgraph/ng_uni.h>
-.Fd #include <netgraph/ng_ccatm.h>
+.In netnatm/unimsg.h
+.In netnatm/msg/unistruct.h
+.In netnatm/sig/unidef.h
+.In netnatm/api/unisap.h
+.In netnatm/api/atmapi.h
+.In netnatm/api/ccatm.h
+.In netgraph.h
+.In netgraph/ng_uni.h
+.In netgraph/ng_ccatm.h
 .Sh DESCIPTION
 The
 .Nm
 node implements the API specified by the ATM Forum for access to ATM services
-(see ATM-Forum document af-saa-0108).
+(see ATM-Forum document
+.Pa af-saa-0108 ) .
 This document specifies the semantics
 of the API, not the exact language binding.
-For this reason it is expected, that
+For this reason, it is expected that
 this implementation is neither compile-time nor binary compatible with
 other implementations of this API.
 It should, however, be fairly straightforward
 to convert between different API implementations.
 .Pp
 This node is usually stacked on top of one or more UNI nodes (see
-.Xr ng_uni 4 ).
+.Xr ng_uni 4 ) .
 Each of these hooks appears as a
-.Ql port
+.Dq port
 to the user of the node.
-It has also one hook connected to the ILMI daemon for management purposes.
+It also has one hook connected to the ILMI daemon for management purposes.
 .Pp
 The node is removed when it receives a
 .Dv NGM_SHUTDOWN
-messages or all hooks are disconnected.
+messages or when all hooks are disconnected.
 .Sh HOOKS
 The node understands a number of hooks with predefined names and an
-unlimited number of hooks for user connections. The predefined names are:
-.Bl -tag -width orphans
-.It Dv uniNNN
+unlimited number of hooks for user connections.
+The predefined names are:
+.Bl -tag -width ".Va orphans"
+.It Va uni Ns Ar NNN
 These hooks stack the
 .Nm
 node on top of a UNI stack.
@@ -84,27 +86,27 @@ to conform to the upper interface specified in
 These hooks are forced into queuing mode, so that there are no circular
 calls from call control to UNI and UNI back to call control.
 The
-.Dv NNN
+.Ar NNN
 in the hook name is the decimal port number and should not be zero.
 The port number is a 32-bit unsigned integer.
-.It Dv manage
+.It Va manage
 This hook should be connected to the ILMI daemon.
 No data is ever sent on this hook and all received data is discarded.
 The hook is used to send control messages along.
-.It Dv dump
+.It Va dump
 On receipt of a
-.Dv
-NGM_DUMP command a textual description of the current state of the node is sent
+.Dv NGM_CCATM_DUMP
+command a textual description of the current state of the node is sent
 out of this hook.
 This text is sent as one large message consisting of more
 than one
-.Fa mbuf .
+.Vt mbuf .
 .El
 .Pp
 All other hook names are taken to be user hooks and correspond to an
-ATM endpoint as specified in the ATM Forum document. The interface on these
-hooks is defined in
-.Pa atmapi.h
+ATM endpoint as specified in the ATM Forum document.
+The interface on these hooks is defined in
+.In atmapi.h
 and uses a structure
 .Bd -literal
 struct ccatm_op {
@@ -144,14 +146,14 @@ enum atmop {
 .Pp
 These codes correspond directly to the operations specified in the ATM
 Forum document with the following exceptions:
-.Bl -tag -width xxx
+.Bl -tag -width indent
 .It Dv ATMOP_RESP
 As discussed in
-.Xr ng_uni 4
+.Xr ng_uni 4 ,
 this is used to
-.Ql synchronify
+.Dq synchronify
 the interface.
-The argument is a struct
+The argument is a
 .Bd -literal
 struct atm_resp {
 	int32_t		resp;
@@ -160,14 +162,14 @@ struct atm_resp {
 .Ed
 .Pp
 If the response code
-.Fa resp
+.Va resp
 is zero, the node has accepted the user request.
 If something goes wrong,
-.Fa resp
+.Va resp
 contains an error code.
 For requests that return data,
-.Fa data
-contains a code describing the type of data and the data itseld
+.Va data
+contains a code describing the type of data and the data itself
 starts immediately after the structure.
 .It Dv ATMOP_QUERY_CONNECTION_ATTRIBUTES_X
 This is the same as
@@ -183,19 +185,19 @@ The list of attributes is followed directly by the attributes in the same
 order as they appear in the list.
 .El
 .Pp
-If a user hook is disconnected an active connection on that hook is released.
+If a user hook is disconnected, an active connection on that hook is released.
 Incoming connections waiting to be accepted are reoffered to other
-listeing hooks or rejected.
+listening hooks or rejected.
 .Sh CONTROL MESSAGES
 Besides the generic messages the node understands the following special
 messages:
-.Bl -tag -width xxx
+.Bl -tag -width indent
 .It Dv NGM_CCATM_DUMP
 This causes the internal state of the node to be dumped in ASCII to the
-.Dv dump
+.Va dump
 hook.
 .It Dv NGM_CCATM_STOP
-This message causes all connections on that port to be aborted (not released!)
+This message causes all connections on that port to be aborted (not released!\&)
 and all ATM endpoints which are bound to that port to be closed.
 It stops processing of all messages from the UNI stack on that port UNI stack.
 The argument is a
@@ -208,19 +210,19 @@ struct ngm_ccatm_port {
 .It Dv NGM_CCATM_START
 Start processing on the port.
 The argument is a
-.Fa ngm_ccatm_port
+.Vt ngm_ccatm_port
 structure.
 .It Dv NGM_CCATM_CLEAR
 This message takes a
-.Fa ngm_ccatm_port
+.Vt ngm_ccatm_port
 structure and clears all prefixes and addresses on that port.
 If the port number is zero, all ports are cleared.
 .It Dv NGM_CCATM_GET_ADDRESSES
 Get the list of all registered addresses on the given port.
 The argument is a
-.Fa ngm_ccatm_port
+.Vt ngm_ccatm_port
 structure and the result is a
-.Fa ngm_ccatm_get_addresses
+.Vt ngm_ccatm_get_addresses
 structure:
 .Bd -literal
 struct ngm_ccatm_get_addresses {
@@ -234,27 +236,28 @@ struct ngm_ccatm_address_req {
 .Ed
 .Pp
 If the
-.Fa port
+.Va port
 field is zero in the request, all addresses on all ports
-are returned, if it is not zero only the addresses on that port are reported.
+are returned.
+If it is not zero, only the addresses on that port are reported.
 The number of addresses is returned in the
-.Fa count
+.Va count
 field.
 .It Dv NGM_CCATM_ADDRESS_REGISTERED
 This message is used by ILMI to inform the
 .Nm
-node, that a previous address registration request was successful.
+node that a previous address registration request was successful.
 This causes the node to activate that address.
 The argument to the message is a
-.Fa ngm_ccatm_address_req
+.Vt ngm_ccatm_address_req
 structure.
 .It Dv NGM_CCATM_ADDRESS_UNREGISTERED
 This message is used by ILMI to inform the
 .Nm
-node, that an address has been unregistered.
+node that an address has been unregistered.
 The node clears that address from its tables.
 The argument is a
-.Fa ngm_ccatm_address_req
+.Vt ngm_ccatm_address_req
 structure.
 .It Dv NGM_CCATM_SET_PORT_PARAM
 This request sets the parameters on the given port.
@@ -276,14 +279,14 @@ struct ngm_ccatm_atm_port {
 This should be used only by ILMI and when that port is stopped and the
 address and prefix tables of that port are empty.
 The
-.Fa num_addr
+.Va num_addr
 field is ignored.
 .It Dv NGM_CCATM_GET_PORT_PARAM
 Retrieve the parameters of the given port.
 The argument is a
-.Fa ngm_ccatm_port
+.Vt ngm_ccatm_port
 and the result a
-.Fa ngm_ccatm_atm_port .
+.Vt ngm_ccatm_atm_port .
 .It Dv NGM_CCATM_GET_PORTLIST
 Get a list of all available ports on that node.
 This is returned as a
@@ -296,16 +299,16 @@ struct ngm_ccatm_portlist {
 .It Dv NGM_CCATM_GETSTATE
 Return the state of a port.
 The argument is a
-.Fa struct ngm_ccatm_port
+.Vt "struct ngm_ccatm_port"
 and the return values as a
-.Ft uint32_t .
+.Vt uint32_t .
 .It Dv NGM_CCATM_SETLOG
 This requests sets a new logging level and returns the previous one.
 The argument is either a
-.Ft uint32_t
+.Vt uint32_t
 in which case it specifies the new logging level, or may be empty
 in which case just the old level is returned as a
-.Ft uint32_t .
+.Vt uint32_t .
 .It Dv NGM_CCATM_RESET
 Reset the node.
 This is allowed only if the number of user hooks and connected UNI stacks is
@@ -317,5 +320,5 @@ Return extended status information from the node.
 .Xr netgraph 4 ,
 .Xr ng_uni 4 ,
 .Xr ngctl 8
-.Sh AUTHOR
+.Sh AUTHORS
 .An Harti Brandt Aq harti@FreeBSD.org
diff --git a/share/man/man4/ng_nat.4 b/share/man/man4/ng_nat.4
index 73b7827..3d6d04f 100644
--- a/share/man/man4/ng_nat.4
+++ b/share/man/man4/ng_nat.4
@@ -29,7 +29,7 @@
 .Os
 .Sh NAME
 .Nm ng_nat
-.Nd NAT netgraph node type
+.Nd "NAT netgraph node type"
 .Sh SYNOPSIS
 .In netgraph/ng_nat.h
 .Sh DESCRIPTION
@@ -37,39 +37,38 @@ An
 .Nm
 node performs network address translation (NAT) of packets
 passing through it.
-An
-.Nm
+A
+.Nm nat
 node uses
 .Xr libalias 3
 engine for packet aliasing.
-At this moment it supports the basic functionality of the library.
+At this moment it supports only the basic functionality of the library.
 .Sh HOOKS
 This node type has two hooks:
-.Pp
-.Bl -tag -width foobar
-.It Dv out
-Packets received on this hook are considered outgoing and would be
-masqueraded to configured address.
-.It Dv in
-Packets coming on this hook are considered incoming and would be
+.Bl -tag -width indent
+.It Va out
+Packets received on this hook are considered outgoing and will be
+masqueraded to a configured address.
+.It Va in
+Packets coming on this hook are considered incoming and will be
 dealiased.
 .El
 .Sh CONTROL MESSAGES
 This node type supports the generic control messages, plus the following:
-.Bl -tag -width foo
+.Bl -tag -width indent
 .It Dv NGM_NAT_SET_IPADDR Pq Li setaliasaddr
-Configure aliasing address for node.
-After both hooks are connected and aliasing address configured,
-node is ready for aliasing operation.
+Configure aliasing address for a node.
+After both hooks have been connected and aliasing address was configured,
+a node is ready for aliasing operation.
 .El
 .Sh SHUTDOWN
 This node shuts down upon receipt of a
 .Dv NGM_SHUTDOWN
 control message, or when both hooks are disconnected.
 .Sh EXAMPLES
-In the following example packets are injected into
-.Nm
-node with help of
+In the following example, the packets are injected into a
+.Nm nat
+node using the
 .Xr ng_ipfw 4
 node.
 .Bd -literal -offset indent
@@ -87,11 +86,12 @@ ipfw add 400 netgraph 60 all from any to any out via fxp0
 sysctl net.inet.ip.fw.one_pass=0
 .Ed
 .Pp
+The
 .Nm
-node can simply be inserted right after
-.Xr ng_iface
-node.
-In the following example we perform masquerading on a
+node can be inserted right after the
+.Xr ng_iface 4
+node in the graph.
+In the following example, we perform masquerading on a
 serial line with HDLC encapsulation.
 .Bd -literal -offset indent
 /usr/sbin/ngctl -f- <<-SEQ
@@ -106,8 +106,8 @@ ifconfig ng0 x.y.8.35 x.y.8.1
 .Ed
 .Sh SEE ALSO
 .Xr libalias 3 ,
-.Xr natd 8 ,
 .Xr ng_ipfw 4 ,
+.Xr natd 8 ,
 .Xr ngctl 8
 .Sh HISTORY
 The
diff --git a/share/man/man4/ng_tcpmss.4 b/share/man/man4/ng_tcpmss.4
index a5ea22a..fc610b6 100644
--- a/share/man/man4/ng_tcpmss.4
+++ b/share/man/man4/ng_tcpmss.4
@@ -29,8 +29,9 @@
 .Os
 .Sh NAME
 .Nm ng_tcpmss
-.Nd netgraph node to adjust TCP MSS option
+.Nd "netgraph node to adjust TCP MSS option"
 .Sh SYNOPSIS
+.In netgraph.h
 .In netgraph/ng_tcpmss.h
 .Sh DESCRIPTION
 The
@@ -39,18 +40,18 @@ node type is designed to alter the Maximum Segment Size option
 of TCP packets.
 This node accepts an arbitrary number of hooks.
 Initially a new hook is considered unconfigured.
-To configure a hook, user should send
+The
 .Dv NG_TCPMSS_CONFIG
-control message to node.
+control message is used to configure a hook.
 .Sh CONTROL MESSAGES
 This node type supports the generic control messages, plus the following.
-.Bl -tag -width foo
-.It Dv NGM_TCPMSS_CONFIG Pq Ic config
+.Bl -tag -width indent
+.It Dv NGM_TCPMSS_CONFIG Pq Li config
 This control message configures node to do given MSS adjusting on
-particular hook.
+a particular hook.
 It requires the
 .Vt "struct ng_tcpmss_config"
-to be supplied as argument:
+to be supplied as an argument:
 .Bd -literal
 struct ng_tcpmss_config {
 	char		inHook[NG_HOOKSIZ];
@@ -60,16 +61,16 @@ struct ng_tcpmss_config {
 .Ed
 .Pp
 This means: packets received on
-.Qq inHook
+.Va inHook
 would be checked for TCP MSS option and the latter would be
 reduced down to
-.Qq maxMSS ,
+.Va maxMSS
 if it exceeds
-.Qq maxMSS .
-After that packets would be sent to hook
-.Qq outHook .
-.It Dv NGM_TCPMSS_GET_STATS Pq Ic getstats
-This control message obtains statistics for the given hook.
+.Va maxMSS .
+After that, packets would be sent to hook
+.Va outHook .
+.It Dv NGM_TCPMSS_GET_STATS Pq Li getstats
+This control message obtains statistics for a given hook.
 The statistics are returned in
 .Vt "struct ng_tcpmss_hookstat" :
 .Bd -literal
@@ -81,23 +82,22 @@ struct ng_tcpmss_hookstat {
 	uint64_t	FixedPkts;	/* changed packets */
 };
 .Ed
-.Pp
-.It Dv NGM_TCPMSS_CLR_STATS Pq Ic clrstats
-This control message clears statistics for the given hook.
-.It Dv NGM_TCPMSS_GETCLR_STATS Pq Ic getclrstats
-This control message obtains and clears statistics for the given hook.
+.It Dv NGM_TCPMSS_CLR_STATS Pq Li clrstats
+This control message clears statistics for a given hook.
+.It Dv NGM_TCPMSS_GETCLR_STATS Pq Li getclrstats
+This control message obtains and clears statistics for a given hook.
 .El
 .Sh EXAMPLES
-In the following example packets are injected into
-.Nm
-node with help of
+In the following example, packets are injected into the
+.Nm tcpmss
+node using the
 .Xr ng_ipfw 4
 node.
 .Bd -literal -offset indent
 # Create tcpmss node and connect it to ng_ipfw node
 ngctl mkpeer ipfw: tcpmss 100 qqq
 
-# Adjust MSS to 1452 
+# Adjust MSS to 1452
 ngctl msg ipfw:100 config '{ inHook="qqq" outHook="qqq" maxMSS=1452 }'
 
 # Divert traffic into tcpmss node
@@ -120,7 +120,6 @@ node type was implemented in
 .Fx 6.0 .
 .Sh AUTHORS
 .An Alexey Popov Aq lollypop@flexuser.ru
-and
 .An Gleb Smirnoff Aq glebius@FreeBSD.org
 .Sh BUGS
-When running on SMP system statistics may be broken.
+When running on SMP, system statistics may be broken.
diff --git a/share/man/man4/nve.4 b/share/man/man4/nve.4
index 35fd018..a3101d0 100644
--- a/share/man/man4/nve.4
+++ b/share/man/man4/nve.4
@@ -27,7 +27,7 @@
 .\" $FreeBSD$
 .\"
 .Dd May 7, 2005
-.Dt NVE 4 i386
+.Dt NVE 4
 .Os
 .Sh NAME
 .Nm nve
@@ -43,7 +43,8 @@ networking adapter that is embedded in the southbridge of most
 nForce and nForce2 motherboards.
 .Pp
 This driver is a reimplementation of the NVIDIA supported Linux
-nvnet driver and uses the same closed source API library to access
+.Nm nvnet
+driver and uses the same closed source API library to access
 the underlying hardware.
 There is currently no programming documentation available for this
 device, and therefore little is known about the internal architecture
@@ -52,22 +53,20 @@ of the MAC engine itself.
 The
 .Nm
 driver supports the following media types:
-.Pp
-.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
-.It autoselect
+.Bl -tag -width ".Cm 10baseT/UTP"
+.It Cm autoselect
 Enable autoselection of the media type and options.
-.It 10baseT/UTP
+.It Cm 10baseT/UTP
 Set 10Mbps operation.
-.It 100baseTX
+.It Cm 100baseTX
 Set 100Mbps (Fast Ethernet) operation.
 .El
 .Pp
 The
 .Nm
 driver supports the following media options:
-.Pp
-.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
-.It full-duplex
+.Bl -tag -width ".Cm 10baseT/UTP"
+.It Cm full-duplex
 Set full duplex operation.
 .El
 .Pp
@@ -91,24 +90,25 @@ nForce2
 A fatal initialization error has occurred.
 .It "nve%d: couldn't map interrupt"
 A fatal initialization error has occurred.
-.It "nve%d: failed to allocate memory" 
+.It "nve%d: failed to allocate memory"
 There are not enough mbufs available for allocation.
 .It "nve%d: device timeout"
 The device has stopped responding to the network, or there is a problem with
 the network connection (cable).
 .El
-.Sh HISTORY
-The
-.Nm
-driver first appeared in
-.Fx 5.5 .
 .Sh SEE ALSO
 .Xr arp 4 ,
 .Xr miibus 4 ,
 .Xr netintro 4 ,
 .Xr ng_ether 4 ,
 .Xr ifconfig 8
+.Sh HISTORY
+The
+.Nm
+driver first appeared in
+.Fx 5.5 .
 .Sh AUTHORS
+.An -nosplit
 The
 .Nm
 driver was written by
diff --git a/share/man/man4/sched_4bsd.4 b/share/man/man4/sched_4bsd.4
index 44b2303..5ffaa0c 100644
--- a/share/man/man4/sched_4bsd.4
+++ b/share/man/man4/sched_4bsd.4
@@ -29,28 +29,30 @@
 .Dt SCHED_4BSD 4
 .Sh NAME
 .Nm sched_4bsd
-.Nd 4.4BSD Scheduler
+.Nd "4.4BSD scheduler"
 .Sh SYNOPSIS
 .Cd "options SCHED_4BSD"
 .Sh DESCRIPTION
+The
 .Nm
+scheduler
 is the default system scheduler, providing both high throughput and solid
 interactive response in the presence of load.
 .Pp
 The following sysctls are relevant to the operation of
 .Nm :
 .Bl -tag -width indent
-.It kern.sched.name
+.It Va kern.sched.name
 This read-only sysctl reports the name of the active scheduler.
-.It kern.sched.quantum
+.It Va kern.sched.quantum
 This read-write sysctl reports or sets the length of the quantum (in
 micro-seconds) granted to a thread.
-.It kern.sched.ipiwakeup.enabled
+.It Va kern.sched.ipiwakeup.enabled
 This read-write sysctl sets whether or not the scheduler will generate an
 inter-processor interrupt (IPI) to an idle CPU when a thread is woken up.
 Otherwise, idle CPUs will wait until the next clock tick before looking for
 new work.
-.It kern.sched.preemption
+.It Va kern.sched.preemption
 This read-only sysctl reports whether or not the kernel is configured to
 support preemption, which reduces the latency to run lower priority threads
 on wakeup.
diff --git a/share/man/man4/sched_ule.4 b/share/man/man4/sched_ule.4
index 237535e..b73e956 100644
--- a/share/man/man4/sched_ule.4
+++ b/share/man/man4/sched_ule.4
@@ -33,7 +33,9 @@
 .Sh SYNOPSIS
 .Cd "options SCHED_ULE"
 .Sh DESCRIPTION
+The
 .Nm
+scheduler
 is an experimental scheduler that provides a number of advanced scheduler
 features not present in
 .Xr sched_4bsd 4 ,
@@ -55,13 +57,17 @@ them preferentially under high load.
 The following sysctls are relevant to the operation of
 .Nm :
 .Bl -tag -width indent
-.It kern.sched.name
+.It Va kern.sched.name
 This read-only sysctl reports the name of the active scheduler.
 .El
 .Pp
+The
 .Nm
+scheduler
 is disabled by default as it is experimental.
-See the bugs section for more information.
+See the
+.Sx BUGS
+section for more information.
 .Sh SEE ALSO
 .Xr sched_4bsd 4 ,
 .Xr sysctl 8
diff --git a/share/man/man4/snd_via8233.4 b/share/man/man4/snd_via8233.4
index 45b6677..09b76bc 100644
--- a/share/man/man4/snd_via8233.4
+++ b/share/man/man4/snd_via8233.4
@@ -66,7 +66,9 @@ The
 device driver first appeared in
 .Fx 4.7 .
 .Sh BUGS
+The
 .Nm
+driver
 does not support S/PDIF.
 There is partial support in the code, so implementing it should be fairly
 easy if the right hardware is available.
diff --git a/share/man/man5/devfs.conf.5 b/share/man/man5/devfs.conf.5
index 4b1acdc..ecb2b7b 100644
--- a/share/man/man5/devfs.conf.5
+++ b/share/man/man5/devfs.conf.5
@@ -51,45 +51,50 @@ The lines that specify
 .Nm
 rules consist of three parameters separated by whitespace:
 .Bl -tag -width indent
-.It Li action
+.It Ar action
 The action to take for the device.
 The action names are only significant to the first unique character.
-.It Li devname
+.It Ar devname
 The name of the device created by
-.Xr devfs 5
-.It Li arg
+.Xr devfs 5 .
+.It Ar arg
 The argument of the
-.Dq Li action
+.Ar action .
 .El
 .Pp
 The actions currently supported are:
 .Bl -tag -width indent
-.It Li link
+.It Ic link
 This action creates a symbolic link named
-.Dq Li arg
+.Ar arg
 that points to
-Dq Li devname ,
+.Ar devname ,
 the name of the device created by
 .Xr devfs 5 .
-.It Li own
+.It Ic own
 This action changes the ownership of
-.Dq Li devname .
+.Ar devname .
 The
-.Dq Li arg
+.Ar arg
 parameter must be in the form of an
-.Li "owner:group"
+.Ar owner Ns : Ns Ar group
 pair, in the same format used by
 .Xr chown 8 .
-.It Li perm
+.It Ic perm
 This action changes the permissions of
-.Dq Li devname .
+.Ar devname .
 The
-.Dq Li arg
+.Ar arg
 paramater must be a
-.Pa mode
+.Ar mode
 as explained in
 .Xr chmod 1 .
 .El
+.Sh FILES
+.Bl -tag -compact
+.It Pa /etc/devfs.conf
+.It Pa /usr/share/examples/etc/devfs.conf
+.El
 .Sh EXAMPLES
 To create a
 .Pa /dev/cdrom
@@ -112,23 +117,18 @@ link	acd0	cdrom
 .Ed
 .Pp
 To set the owner of a device, the
-.Dq Li own
+.Ic own
 action may be specified:
 .Bd -literal -offset indent
 own	cd0	root:cdrom
 .Ed
 .Pp
 To set the permissions of a device, a
-.Dq Li perm
+.Ic perm
 action should be used:
 .Bd -literal -offset indent
 perm	cd0	0660
 .Ed
-.Sh FILES
-.Bl -tag -compact
-.It Pa /etc/devfs.conf
-.It Pa /usr/share/examples/etc/devfs.conf
-.El
 .Sh SEE ALSO
 .Xr chmod 1 ,
 .Xr devfs 5 ,
@@ -136,4 +136,5 @@ perm	cd0	0660
 .Xr usbd.conf 5 ,
 .Xr chown 8
 .Sh AUTHORS
-.An "Roland Smith" Aq rsmith@xs4all.nl
+This manual page was written by
+.An "Roland Smith" Aq rsmith@xs4all.nl .
diff --git a/share/man/man5/devfs.rules.5 b/share/man/man5/devfs.rules.5
index 7d07a58..075e71b 100644
--- a/share/man/man5/devfs.rules.5
+++ b/share/man/man5/devfs.rules.5
@@ -51,7 +51,7 @@ separated by an equal sign.
 Other lines are rule specifications as documented in
 .Xr devfs 8 ,
 in the section
-.Dq "Rule Specification" .
+.Sx "Rule Specification" .
 These lines are prepended with
 .Dq Li rule
 and are passed to
@@ -60,33 +60,32 @@ by the startup scripts of the system.
 It is important to put path elements that contain
 .Xr glob 3
 special characters between quotes.
+.Sh FILES
+.Bl -tag -compact
+.It Pa /etc/devfs.rules
+.It Pa /etc/defaults/devfs.rules
+.El
 .Sh EXAMPLES
 To make all the partitions of
 .Xr da 4
 devices readable and writable by their owner and the
 .Dq Li usb
 group, the following rule may be used:
-.Bd -literal -offset indent
-add path 'da*s*' mode 0660 group usb
-.Ed
+.Pp
+.Dl "add path 'da*s*' mode 0660 group usb"
 .Pp
 To make all the
 .Xr uscanner 4
 devices accessible to their owner and the
 .Dq Li usb
 group, a similar rule may be used:
-.Bd -literal -offset indent
-add path 'uscanner*' mode 0660 group usb
-.Ed
-.Sh FILES
-.Bl -tag -compact
-.It Pa /etc/devfs.rules
-.It Pa /etc/defaults/devfs.rules
-.El
+.Pp
+.Dl "add path 'uscanner*' mode 0660 group usb"
 .Sh SEE ALSO
 .Xr glob 3 ,
 .Xr devfs 5 ,
 .Xr devfs.conf 5 ,
 .Xr devfs 8
 .Sh AUTHORS
-.An "Roland Smith" Aq rsmith@xs4all.nl
+This manual page was written by
+.An "Roland Smith" Aq rsmith@xs4all.nl .
diff --git a/share/man/man9/alloc_unr.9 b/share/man/man9/alloc_unr.9
index 90b145d..9d81453 100644
--- a/share/man/man9/alloc_unr.9
+++ b/share/man/man9/alloc_unr.9
@@ -48,37 +48,39 @@
 .Sh DESCRIPTION
 The kernel unit number allocator is a generic facility, which allows to allocate
 unit numbers within a specified range.
-.Bl -ohang
+.Bl -tag -width indent
 .It Fn new_unrhdr low high mutex
 Initialize a new unit number allocator entity.
+The
 .Fa low
 and
 .Fa high
+arguments
 specify minimum and maximum number of unit numbers.
 There is no cost associated with the range of unit numbers, so unless the resource
 really is finite,
-.Va INT_MAX
+.Dv INT_MAX
 can be used.
 If
 .Fa mutex
 is not
-.Va NULL ,
+.Dv NULL ,
 it is used for locking when allocating and freeing units.
-Otherwise internal mutex is used.
+Otherwise, internal mutex is used.
 .It Fn delete_unrhdr uh
 Destroy specified unit number allocator entity.
 .It Fn alloc_unr uh
 Return a new unit number.
 The lowest free number is always allocated.
 This function does not allocate memory and never sleeps, however it may
-block on mutex.
-If no free unit numbers are left
-.Va -1
+block on a mutex.
+If no free unit numbers are left,
+.Li \-1
 is returned.
 .It Fn alloc_unrl uh
 Same as
 .Fn alloc_unr
-except that mutex is assumed to be locked already and thus is not used.
+except that mutex is assumed to be already locked and thus is not used.
 .It Fn free_unr uh
 Free a previously allocated unit number.
 This function may require allocating memory, and thus it can sleep.
diff --git a/share/man/man9/devclass_get_drivers.9 b/share/man/man9/devclass_get_drivers.9
index ee6629b..b5eaf1c 100644
--- a/share/man/man9/devclass_get_drivers.9
+++ b/share/man/man9/devclass_get_drivers.9
@@ -33,7 +33,7 @@
 .Os
 .Sh NAME
 .Nm devclass_get_drivers
-.Nd get a list of drivers in a devclass
+.Nd "get a list of drivers in a devclass"
 .Sh SYNOPSIS
 .In sys/param.h
 .In sys/bus.h
@@ -46,8 +46,8 @@ devclass and return the list in
 and the number of drivers in the list in
 .Fa *countp .
 The memory allocated for the list should be freed using
-.Fn free "*listp" "M_TEMP" .
-even if    
+.Fn free *listp M_TEMP ,
+even if
 .Fa *countp
 is 0.
 .Sh RETURN VALUES