xref: /dragonfly/share/man/man4/rl.4 (revision f02303f9)
1.\" Copyright (c) 1997, 1998
2.\"	Bill Paul <wpaul@ctr.columbia.edu>. All rights reserved.
3.\"
4.\" Redistribution and use in source and binary forms, with or without
5.\" modification, are permitted provided that the following conditions
6.\" are met:
7.\" 1. Redistributions of source code must retain the above copyright
8.\"    notice, this list of conditions and the following disclaimer.
9.\" 2. Redistributions in binary form must reproduce the above copyright
10.\"    notice, this list of conditions and the following disclaimer in the
11.\"    documentation and/or other materials provided with the distribution.
12.\" 3. All advertising materials mentioning features or use of this software
13.\"    must display the following acknowledgement:
14.\"	This product includes software developed by Bill Paul.
15.\" 4. Neither the name of the author nor the names of any co-contributors
16.\"    may be used to endorse or promote products derived from this software
17.\"   without specific prior written permission.
18.\"
19.\" THIS SOFTWARE IS PROVIDED BY Bill Paul AND CONTRIBUTORS ``AS IS'' AND
20.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22.\" ARE DISCLAIMED.  IN NO EVENT SHALL Bill Paul OR THE VOICES IN HIS HEAD
23.\" BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24.\" CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25.\" SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26.\" INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27.\" CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28.\" ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
29.\" THE POSSIBILITY OF SUCH DAMAGE.
30.\"
31.\" $FreeBSD: src/share/man/man4/rl.4,v 1.15.2.8 2003/02/17 21:20:39 trhodes Exp $
32.\" $DragonFly: src/share/man/man4/rl.4,v 1.6 2007/03/11 10:22:52 swildner Exp $
33.\"
34.Dd November 4, 1998
35.Dt RL 4
36.Os
37.Sh NAME
38.Nm rl
39.Nd RealTek 8129/8139 fast ethernet device driver
40.Sh SYNOPSIS
41.Cd "device miibus"
42.Cd "device rl"
43.Sh DESCRIPTION
44The
45.Nm
46driver provides support for PCI ethernet adapters and embedded
47controllers based on the RealTek 8129 and 8139 fast ethernet controller
48chips.
49This includes the Allied Telesyn AT2550, Farallon NetLINE 10/100 PCI,
50Genius GF100TXR,
51NDC Communications NE100TX-E, OvisLink LEF-8129TX, OvisLink LEF-8139TX,
52Netronix Inc. EA-1210 NetEther 10/100, KTX-9130TX 10/100 Fast Ethernet,
53Encore ENL832-TX 10/100 M PCI, Longshine LCS-8038TX-R, the
54SMC EZ Card 10/100 PCI 1211-TX, and various other cheap adapters.
55It also supports the Accton EN1207D which has a
56chip labeled MPX5030 (or MPX5038) which appears to be a RealTek workalike.
57.Pp
58The RealTek controllers use bus master DMA but do not use a
59descriptor-based data transfer mechanism.
60The receiver uses a
61single fixed size ring buffer from which packets must be copied
62into mbufs.
63For transmission, there are only four outbound packet
64address registers which require all outgoing packets to be stored
65as contiguous buffers.
66Furthermore, outbound packet buffers must
67be longword aligned or else transmission will fail.
68.Pp
69The 8129 differs from the 8139 in that the 8139 has an internal
70PHY which is controlled through special direct access registers
71whereas the 8129 uses an external PHY via an MII bus.
72The 8139
73supports both 10 and 100Mbps speeds in either full or half duplex.
74The 8129 can support the same speeds and modes given an appropriate
75PHY chip.
76.Pp
77The
78.Nm
79driver supports the following media types:
80.Pp
81.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
82.It autoselect
83Enable autoselection of the media type and options.
84This is only
85supported if the PHY chip attached to the RealTek controller
86supports NWAY autonegotiation.
87The user can manually override
88the autoselected mode by adding media options to the
89.Pa /etc/rc.conf
90file.
91.It 10baseT/UTP
92Set 10Mbps operation.
93The
94.Ar mediaopt
95option can also be used to select either
96.Ar full-duplex
97or
98.Ar half-duplex
99modes.
100.It 100baseTX
101Set 100Mbps (fast ethernet) operation.
102The
103.Ar mediaopt
104option can also be used to select either
105.Ar full-duplex
106or
107.Ar half-duplex
108modes.
109.El
110.Pp
111The
112.Nm
113driver supports the following media options:
114.Pp
115.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
116.It full-duplex
117Force full duplex operation
118.It half-duplex
119Force half duplex operation.
120.El
121.Pp
122Note that the 100baseTX media type is only available if supported
123by the adapter.
124For more information on configuring this device, see
125.Xr ifconfig 8 .
126.Sh DIAGNOSTICS
127.Bl -diag
128.It "rl%d: couldn't map memory"
129A fatal initialization error has occurred.
130.It "rl%d: couldn't map interrupt"
131A fatal initialization error has occurred.
132.It "rl%d: watchdog timeout"
133The device has stopped responding to the network, or there is a problem with
134the network connection (cable).
135.It "rl%d: no memory for rx list"
136The driver failed to allocate an mbuf for the receiver ring.
137.It "rl%d: no memory for tx list"
138The driver failed to allocate an mbuf for the transmitter ring when
139allocating a pad buffer or collapsing an mbuf chain into a cluster.
140.It "rl%d: chip is in D3 power state -- setting to D0"
141This message applies only to adapters which support power
142management.
143Some operating systems place the controller in low power
144mode when shutting down, and some PCI BIOSes fail to bring the chip
145out of this state before configuring it.
146The controller loses all of
147its PCI configuration in the D3 state, so if the BIOS does not set
148it back to full power mode in time, it won't be able to configure it
149correctly.
150The driver tries to detect this condition and bring
151the adapter back to the D0 (full power) state, but this may not be
152enough to return the driver to a fully operational condition.
153If
154you see this message at boot time and the driver fails to attach
155the device as a network interface, you will have to perform second
156warm boot to have the device properly configured.
157.Pp
158Note that this condition only occurs when warm booting from another
159operating system.
160If you power down your system prior to booting
161.Dx ,
162the card should be configured correctly.
163.El
164.Sh SEE ALSO
165.Xr arp 4 ,
166.Xr ifmedia 4 ,
167.Xr miibus 4 ,
168.Xr netintro 4 ,
169.Xr ng_ether 4 ,
170.Xr polling 4 ,
171.Xr ifconfig 8
172.Rs
173.%B The RealTek 8129 and 8139 datasheets
174.%O ftp.realtek.com.tw:/lancard/data sheet
175.Re
176.Sh HISTORY
177The
178.Nm
179device driver first appeared in
180.Fx 3.0 .
181.Sh AUTHORS
182The
183.Nm
184driver was written by
185.An Bill Paul Aq wpaul@ctr.columbia.edu .
186.Sh BUGS
187Since outbound packets must be longword aligned, the transmit
188routine has to copy an unaligned packet into an mbuf cluster buffer
189before transmission.
190The driver abuses the fact that the cluster buffer
191pool is allocated at system startup time in a contiguous region starting
192at a page boundary.
193Since cluster buffers are 2048 bytes, they are
194longword aligned by definition.
195The driver probably should not be
196depending on this characteristic.
197.Pp
198The RealTek data sheets are of especially poor quality,
199and there is a lot of information missing
200particularly concerning the receiver operation.
201One particularly
202important fact that the data sheets fail to mention relates to the
203way in which the chip fills in the receive buffer.
204When an interrupt
205is posted to signal that a frame has been received, it is possible that
206another frame might be in the process of being copied into the receive
207buffer while the driver is busy handling the first one.
208If the driver
209manages to finish processing the first frame before the chip is done
210DMAing the rest of the next frame, the driver may attempt to process
211the next frame in the buffer before the chip has had a chance to finish
212DMAing all of it.
213.Pp
214The driver can check for an incomplete frame by inspecting the frame
215length in the header preceding the actual packet data: an incomplete
216frame will have the magic length of 0xFFF0.
217When the driver encounters
218this value, it knows that it has finished processing all currently
219available packets.
220Neither this magic value nor its significance are
221documented anywhere in the RealTek data sheets.
222