xref: /freebsd/share/man/man4/rl.4 (revision 6990ffd8a95caaba6858ad44ff1b3157d1efba8f)
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 rl"
41.Sh DESCRIPTION
42The
43.Nm
44driver provides support for PCI ethernet adapters and embedded
45controllers based on the RealTek 8129 and 8139 fast ethernet controller
46chips.
47This includes the Allied Telesyn AT2550, Genius GF100TXR,
48NDC Communications NE100TX-E, OvisLink LEF-8129TX, OvisLink LEF-8139TX,
49Netronix Inc. EA-1210 NetEther 10/100, KTX-9130TX 10/100 Fast Ethernet,
50Encore ENL832-TX 10/100 M PCI, Longshine LCS-8038TX-R, the
51SMC EZ Card 10/100 PCI 1211-TX, and various other cheap adapters.
52It also supports the Accton EN1207D which has a
53chip labeled MPX5030 (or MPX5038) which appears to be a RealTek workalike.
54.Pp
55The RealTek controllers use bus master DMA but do not use a
56descriptor-based data transfer mechanism.
57The receiver uses a
58single fixed size ring buffer from which packets must be copied
59into mbufs.
60For transmission, there are only four outbound packet
61address registers which require all outgoing packets to be stored
62as contiguous buffers.
63Furthermore, outbound packet buffers must
64be longword aligned or else transmission will fail.
65.Pp
66The 8129 differs from the 8139 in that the 8139 has an internal
67PHY which is controlled through special direct access registers
68whereas the 8129 uses an external PHY via an MII bus.
69The 8139
70supports both 10 and 100Mbps speeds in either full or half duplex.
71The 8129 can support the same speeds and modes given an appropriate
72PHY chip.
73.Pp
74The
75.Nm
76driver supports the following media types:
77.Pp
78.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
79.It autoselect
80Enable autoselection of the media type and options.
81This is only
82supported if the PHY chip attached to the RealTek controller
83supports NWAY autonegotiation.
84The user can manually override
85the autoselected mode by adding media options to the
86.Pa /etc/rc.conf
87file.
88.It 10baseT/UTP
89Set 10Mbps operation.
90The
91.Ar mediaopt
92option can also be used to select either
93.Ar full-duplex
94or
95.Ar half-duplex
96modes.
97.It 100baseTX
98Set 100Mbps (fast ethernet) operation.
99The
100.Ar mediaopt
101option can also be used to select either
102.Ar full-duplex
103or
104.Ar half-duplex
105modes.
106.El
107.Pp
108The
109.Nm
110driver supports the following media options:
111.Pp
112.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
113.It full-duplex
114Force full duplex operation
115.It half-duplex
116Force half duplex operation.
117.El
118.Pp
119Note that the 100baseTX media type is only available if supported
120by the adapter.
121For more information on configuring this device, see
122.Xr ifconfig 8 .
123.Sh DIAGNOSTICS
124.Bl -diag
125.It "rl%d: couldn't map memory"
126A fatal initialization error has occurred.
127.It "rl%d: couldn't map interrupt"
128A fatal initialization error has occurred.
129.It "rl%d: watchdog timeout"
130The device has stopped responding to the network, or there is a problem with
131the network connection (cable).
132.It "rl%d: no memory for rx list"
133The driver failed to allocate an mbuf for the receiver ring.
134.It "rl%d: no memory for tx list"
135The driver failed to allocate an mbuf for the transmitter ring when
136allocating a pad buffer or collapsing an mbuf chain into a cluster.
137.It "rl%d: chip is in D3 power state -- setting to D0"
138This message applies only to adapters which support power
139management.
140Some operating systems place the controller in low power
141mode when shutting down, and some PCI BIOSes fail to bring the chip
142out of this state before configuring it.
143The controller loses all of
144its PCI configuration in the D3 state, so if the BIOS does not set
145it back to full power mode in time, it won't be able to configure it
146correctly.
147The driver tries to detect this condition and bring
148the adapter back to the D0 (full power) state, but this may not be
149enough to return the driver to a fully operational condition.
150If
151you see this message at boot time and the driver fails to attach
152the device as a network interface, you will have to perform second
153warm boot to have the device properly configured.
154.Pp
155Note that this condition only occurs when warm booting from another
156operating system.
157If you power down your system prior to booting
158.Fx ,
159the card should be configured correctly.
160.El
161.Sh SEE ALSO
162.Xr arp 4 ,
163.Xr netintro 4 ,
164.Xr ng_ether 4 ,
165.Xr ifconfig 8
166.Rs
167.%B The RealTek 8129 and 8139 datasheets
168.%O ftp.realtek.com.tw:/lancard/data sheet
169.Re
170.Sh HISTORY
171The
172.Nm
173device driver first appeared in
174.Fx 3.0 .
175.Sh AUTHORS
176The
177.Nm
178driver was written by
179.An Bill Paul Aq wpaul@ctr.columbia.edu .
180.Sh BUGS
181Since outbound packets must be longword aligned, the transmit
182routine has to copy an unaligned packet into an mbuf cluster buffer
183before transmission.
184The driver abuses the fact that the cluster buffer
185pool is allocated at system startup time in a contiguous region starting
186at a page boundary.
187Since cluster buffers are 2048 bytes, they are
188longword aligned by definition.
189The driver probably should not be
190depending on this characteristic.
191.Pp
192The RealTek data sheets are of especially poor quality: the grammar
193and spelling are awful and there is a lot of information missing,
194particularly concerning the receiver operation.
195One particularly
196important fact that the data sheets fail to mention relates to the
197way in which the chip fills in the receive buffer.
198When an interrupt
199is posted to signal that a frame has been received, it is possible that
200another frame might be in the process of being copied into the receive
201buffer while the driver is busy handling the first one.
202If the driver
203manages to finish processing the first frame before the chip is done
204DMAing the rest of the next frame, the driver may attempt to process
205the next frame in the buffer before the chip has had a chance to finish
206DMAing all of it.
207.Pp
208The driver can check for an incomplete frame by inspecting the frame
209length in the header preceding the actual packet data: an incomplete
210frame will have the magic length of 0xFFF0.
211When the driver encounters
212this value, it knows that it has finished processing all currently
213available packets.
214Neither this magic value nor its significance are
215documented anywhere in the RealTek data sheets.
216