xref: /freebsd/share/man/man4/rl.4 (revision 7660b554bc59a07be0431c17e0e33815818baa69)
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.\" $FreeBSD$
32.\"
33.Dd November 4, 1998
34.Dt RL 4
35.Os
36.Sh NAME
37.Nm rl
38.Nd RealTek 8129/8139 fast ethernet device driver
39.Sh SYNOPSIS
40.Cd "device miibus"
41.Cd "device rl"
42.Sh DESCRIPTION
43The
44.Nm
45driver provides support for PCI ethernet adapters and embedded
46controllers based on the RealTek 8129 and 8139 fast ethernet controller
47chips, including the following:
48.Pp
49.Bl -bullet -compact -offset indent
50.It
51Accton ``Cheetah'' EN1207D (MPX 5030/5038; RealTek 8139 clone)
52.It
53Allied Telesyn AT2550
54.It
55Allied Telesyn AT2500TX
56.It
57Farallon NetLINE 10/100 PCI
58.It
59Genius GF100TXR,
60.It
61NDC Communications NE100TX-E
62.It
63OvisLink LEF-8129TX
64.It
65OvisLink LEF-8139TX,
66.It
67Netronix Inc. EA-1210 NetEther 10/100
68.It
69KTX-9130TX 10/100 Fast Ethernet,
70.It
71Encore ENL832-TX 10/100 M PCI,
72.It
73Longshine LCS-8038TX-R
74.It
75SMC EZ Card 10/100 PCI 1211-TX
76.It
77Compaq HNE-300
78.It
79LevelOne FPC-0106TX
80.It
81Edimax EP-4103DL CardBus
82.It
83Nortel Networks 10/100BaseTX
84.It
85D-Link DFE-690TXD
86.It
87D-Link DFE-530TX+
88.It
89D-Link DFE-538TX (same as 530+?)
90.It
91Corega FEther CB-TXD
92.It
93Corega FEtherII CB-TXD
94.It
95Peppercon AG ROL-F
96.It
97Planex FNW-3800-TX
98.It
99Accton EN1207D
100.It
101GigaFast Ethernet EE100-AXP
102.It
103CompUSA no-name 10/100 PCI ethernet NIC
104.It
105Belkin F5D5000
106.El
107.Pp
108The RealTek 8129/8139 series controllers use bus master DMA but do not use a
109descriptor-based data transfer mechanism.
110The receiver uses a
111single fixed size ring buffer from which packets must be copied
112into mbufs.
113For transmission, there are only four outbound packet
114address registers which require all outgoing packets to be stored
115as contiguous buffers.
116Furthermore, outbound packet buffers must
117be longword aligned or else transmission will fail.
118.Pp
119The 8129 differs from the 8139 in that the 8139 has an internal
120PHY which is controlled through special direct access registers
121whereas the 8129 uses an external PHY via an MII bus.
122The 8139
123supports both 10 and 100Mbps speeds in either full or half duplex.
124The 8129 can support the same speeds and modes given an appropriate
125PHY chip.
126.Pp
127Note: support for the 8139C+ chip is provided by the
128.Xr re 4
129driver.
130.Pp
131The
132.Nm
133driver supports the following media types:
134.Pp
135.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
136.It autoselect
137Enable autoselection of the media type and options.
138This is only
139supported if the PHY chip attached to the RealTek controller
140supports NWAY autonegotiation.
141The user can manually override
142the autoselected mode by adding media options to the
143.Pa /etc/rc.conf
144file.
145.It 10baseT/UTP
146Set 10Mbps operation.
147The
148.Ar mediaopt
149option can also be used to select either
150.Ar full-duplex
151or
152.Ar half-duplex
153modes.
154.It 100baseTX
155Set 100Mbps (fast ethernet) operation.
156The
157.Ar mediaopt
158option can also be used to select either
159.Ar full-duplex
160or
161.Ar half-duplex
162modes.
163.El
164.Pp
165The
166.Nm
167driver supports the following media options:
168.Pp
169.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
170.It full-duplex
171Force full duplex operation
172.It half-duplex
173Force half duplex operation.
174.El
175.Pp
176Note that the 100baseTX media type is only available if supported
177by the adapter.
178For more information on configuring this device, see
179.Xr ifconfig 8 .
180.Sh DIAGNOSTICS
181.Bl -diag
182.It "rl%d: couldn't map memory"
183A fatal initialization error has occurred.
184.It "rl%d: couldn't map interrupt"
185A fatal initialization error has occurred.
186.It "rl%d: watchdog timeout"
187The device has stopped responding to the network, or there is a problem with
188the network connection (cable).
189.It "rl%d: no memory for rx list"
190The driver failed to allocate an mbuf for the receiver ring.
191.It "rl%d: no memory for tx list"
192The driver failed to allocate an mbuf for the transmitter ring when
193allocating a pad buffer or collapsing an mbuf chain into a cluster.
194.It "rl%d: chip is in D3 power state -- setting to D0"
195This message applies only to adapters which support power
196management.
197Some operating systems place the controller in low power
198mode when shutting down, and some PCI BIOSes fail to bring the chip
199out of this state before configuring it.
200The controller loses all of
201its PCI configuration in the D3 state, so if the BIOS does not set
202it back to full power mode in time, it won't be able to configure it
203correctly.
204The driver tries to detect this condition and bring
205the adapter back to the D0 (full power) state, but this may not be
206enough to return the driver to a fully operational condition.
207If
208you see this message at boot time and the driver fails to attach
209the device as a network interface, you will have to perform second
210warm boot to have the device properly configured.
211.Pp
212Note that this condition only occurs when warm booting from another
213operating system.
214If you power down your system prior to booting
215.Fx ,
216the card should be configured correctly.
217.El
218.Sh SEE ALSO
219.Xr arp 4 ,
220.Xr miibus 4 ,
221.Xr netintro 4 ,
222.Xr ng_ether 4 ,
223.Xr ifconfig 8
224.Rs
225.%B The RealTek 8129, 8139 and 8139C+ datasheets
226.%O http://www.realtek.com.tw
227.Re
228.Sh HISTORY
229The
230.Nm
231device driver first appeared in
232.Fx 3.0 .
233.Sh AUTHORS
234The
235.Nm
236driver was written by
237.An Bill Paul Aq wpaul@ctr.columbia.edu .
238.Sh BUGS
239Since outbound packets must be longword aligned, the transmit
240routine has to copy an unaligned packet into an mbuf cluster buffer
241before transmission.
242The driver abuses the fact that the cluster buffer
243pool is allocated at system startup time in a contiguous region starting
244at a page boundary.
245Since cluster buffers are 2048 bytes, they are
246longword aligned by definition.
247The driver probably should not be
248depending on this characteristic.
249.Pp
250The RealTek data sheets are of especially poor quality,
251and there is a lot of information missing
252particularly concerning the receiver operation.
253One particularly
254important fact that the data sheets fail to mention relates to the
255way in which the chip fills in the receive buffer.
256When an interrupt
257is posted to signal that a frame has been received, it is possible that
258another frame might be in the process of being copied into the receive
259buffer while the driver is busy handling the first one.
260If the driver
261manages to finish processing the first frame before the chip is done
262DMAing the rest of the next frame, the driver may attempt to process
263the next frame in the buffer before the chip has had a chance to finish
264DMAing all of it.
265.Pp
266The driver can check for an incomplete frame by inspecting the frame
267length in the header preceding the actual packet data: an incomplete
268frame will have the magic length of 0xFFF0.
269When the driver encounters
270this value, it knows that it has finished processing all currently
271available packets.
272Neither this magic value nor its significance are
273documented anywhere in the RealTek data sheets.
274