xref: /freebsd/share/man/man4/rl.4 (revision 2a4a1db342263067035ce69a4017c645da63455d)
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, Farallon NetLINE 10/100 PCI,
48Genius GF100TXR,
49NDC Communications NE100TX-E, OvisLink LEF-8129TX, OvisLink LEF-8139TX,
50Netronix Inc. EA-1210 NetEther 10/100, KTX-9130TX 10/100 Fast Ethernet,
51Encore ENL832-TX 10/100 M PCI, Longshine LCS-8038TX-R, the
52SMC EZ Card 10/100 PCI 1211-TX, and various other cheap adapters.
53It also supports the Accton EN1207D which has a
54chip labeled MPX5030 (or MPX5038) which appears to be a RealTek workalike.
55.Pp
56The RealTek controllers use bus master DMA but do not use a
57descriptor-based data transfer mechanism.
58The receiver uses a
59single fixed size ring buffer from which packets must be copied
60into mbufs.
61For transmission, there are only four outbound packet
62address registers which require all outgoing packets to be stored
63as contiguous buffers.
64Furthermore, outbound packet buffers must
65be longword aligned or else transmission will fail.
66.Pp
67The 8129 differs from the 8139 in that the 8139 has an internal
68PHY which is controlled through special direct access registers
69whereas the 8129 uses an external PHY via an MII bus.
70The 8139
71supports both 10 and 100Mbps speeds in either full or half duplex.
72The 8129 can support the same speeds and modes given an appropriate
73PHY chip.
74.Pp
75The
76.Nm
77driver supports the following media types:
78.Pp
79.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
80.It autoselect
81Enable autoselection of the media type and options.
82This is only
83supported if the PHY chip attached to the RealTek controller
84supports NWAY autonegotiation.
85The user can manually override
86the autoselected mode by adding media options to the
87.Pa /etc/rc.conf
88file.
89.It 10baseT/UTP
90Set 10Mbps operation.
91The
92.Ar mediaopt
93option can also be used to select either
94.Ar full-duplex
95or
96.Ar half-duplex
97modes.
98.It 100baseTX
99Set 100Mbps (fast ethernet) operation.
100The
101.Ar mediaopt
102option can also be used to select either
103.Ar full-duplex
104or
105.Ar half-duplex
106modes.
107.El
108.Pp
109The
110.Nm
111driver supports the following media options:
112.Pp
113.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
114.It full-duplex
115Force full duplex operation
116.It half-duplex
117Force half duplex operation.
118.El
119.Pp
120Note that the 100baseTX media type is only available if supported
121by the adapter.
122For more information on configuring this device, see
123.Xr ifconfig 8 .
124.Sh DIAGNOSTICS
125.Bl -diag
126.It "rl%d: couldn't map memory"
127A fatal initialization error has occurred.
128.It "rl%d: couldn't map interrupt"
129A fatal initialization error has occurred.
130.It "rl%d: watchdog timeout"
131The device has stopped responding to the network, or there is a problem with
132the network connection (cable).
133.It "rl%d: no memory for rx list"
134The driver failed to allocate an mbuf for the receiver ring.
135.It "rl%d: no memory for tx list"
136The driver failed to allocate an mbuf for the transmitter ring when
137allocating a pad buffer or collapsing an mbuf chain into a cluster.
138.It "rl%d: chip is in D3 power state -- setting to D0"
139This message applies only to adapters which support power
140management.
141Some operating systems place the controller in low power
142mode when shutting down, and some PCI BIOSes fail to bring the chip
143out of this state before configuring it.
144The controller loses all of
145its PCI configuration in the D3 state, so if the BIOS does not set
146it back to full power mode in time, it won't be able to configure it
147correctly.
148The driver tries to detect this condition and bring
149the adapter back to the D0 (full power) state, but this may not be
150enough to return the driver to a fully operational condition.
151If
152you see this message at boot time and the driver fails to attach
153the device as a network interface, you will have to perform second
154warm boot to have the device properly configured.
155.Pp
156Note that this condition only occurs when warm booting from another
157operating system.
158If you power down your system prior to booting
159.Fx ,
160the card should be configured correctly.
161.El
162.Sh SEE ALSO
163.Xr arp 4 ,
164.Xr netintro 4 ,
165.Xr ng_ether 4 ,
166.Xr ifconfig 8
167.Rs
168.%B The RealTek 8129 and 8139 datasheets
169.%O ftp.realtek.com.tw:/lancard/data sheet
170.Re
171.Sh HISTORY
172The
173.Nm
174device driver first appeared in
175.Fx 3.0 .
176.Sh AUTHORS
177The
178.Nm
179driver was written by
180.An Bill Paul Aq wpaul@ctr.columbia.edu .
181.Sh BUGS
182Since outbound packets must be longword aligned, the transmit
183routine has to copy an unaligned packet into an mbuf cluster buffer
184before transmission.
185The driver abuses the fact that the cluster buffer
186pool is allocated at system startup time in a contiguous region starting
187at a page boundary.
188Since cluster buffers are 2048 bytes, they are
189longword aligned by definition.
190The driver probably should not be
191depending on this characteristic.
192.Pp
193The RealTek data sheets are of especially poor quality: the grammar
194and spelling are awful and there is a lot of information missing,
195particularly concerning the receiver operation.
196One particularly
197important fact that the data sheets fail to mention relates to the
198way in which the chip fills in the receive buffer.
199When an interrupt
200is posted to signal that a frame has been received, it is possible that
201another frame might be in the process of being copied into the receive
202buffer while the driver is busy handling the first one.
203If the driver
204manages to finish processing the first frame before the chip is done
205DMAing the rest of the next frame, the driver may attempt to process
206the next frame in the buffer before the chip has had a chance to finish
207DMAing all of it.
208.Pp
209The driver can check for an incomplete frame by inspecting the frame
210length in the header preceding the actual packet data: an incomplete
211frame will have the magic length of 0xFFF0.
212When the driver encounters
213this value, it knows that it has finished processing all currently
214available packets.
215Neither this magic value nor its significance are
216documented anywhere in the RealTek data sheets.
217