xref: /freebsd/share/man/man4/rl.4 (revision 4cf49a43559ed9fdad601bdcccd2c55963008675)
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 i386
35.Os FreeBSD
36.Sh NAME
37.Nm rl
38.Nd
39RealTek 8129/8139 fast ethernet device driver
40.Sh SYNOPSIS
41.Cd "device rl0"
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. This 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. The receiver uses a
57single fixed size ring buffer from which packets must be copied
58into mbufs. For transmission, there are only four outbound packet
59address registers which require all outgoing packets to be stored
60as contiguous buffers. Furthermore, outbound packet buffers must
61be longword aligned or else transmission will fail.
62.Pp
63The 8129 differs from the 8139 in that the 8139 has an internal
64PHY which is controlled through special direct access registers
65whereas the 8129 uses an external PHY via an MII bus. The 8139
66supports both 10 and 100Mbps speeds in either full or half duplex.
67The 8129 can support the same speeds and modes given an appropriate
68PHY chip.
69.Pp
70The
71.Nm
72driver supports the following media types:
73.Pp
74.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
75.It autoselect
76Enable autoselection of the media type and options. This is only
77supported if the PHY chip attached to the RealTek controller
78supports NWAY autonegotiation. The user can manually override
79the autoselected mode by adding media options to the
80.Pa /etc/rc.conf
81file.
82.It 10baseT/UTP
83Set 10Mbps operation. The
84.Ar mediaopt
85option can also be used to select either
86.Ar full-duplex
87or
88.Ar half-duplex modes.
89.It 100baseTX
90Set 100Mbps (fast ethernet) operation. The
91.Ar mediaopt
92option can also be used to select either
93.Ar full-duplex
94or
95.Ar half-duplex
96modes.
97.El
98.Pp
99The
100.Nm
101driver supports the following media options:
102.Pp
103.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
104.It full-duplex
105Force full duplex operation
106.It half-duplex
107Force half duplex operation.
108.El
109.Pp
110Note that the 100baseTX media type is only available if supported
111by the adapter.
112For more information on configuring this device, see
113.Xr ifconfig 8 .
114.Sh DIAGNOSTICS
115.Bl -diag
116.It "rl%d: couldn't map memory"
117A fatal initialization error has occurred.
118.It "rl%d: couldn't map interrupt"
119A fatal initialization error has occurred.
120.It "rl%d: watchdog timeout"
121The device has stopped responding to the network, or there is a problem with
122the network connection (cable).
123.It "rl%d: no memory for rx list"
124The driver failed to allocate an mbuf for the receiver ring.
125.It "rl%d: no memory for tx list"
126The driver failed to allocate an mbuf for the transmitter ring when
127allocating a pad buffer or collapsing an mbuf chain into a cluster.
128.It "rl%d: chip is in D3 power state -- setting to D0"
129This message applies only to adapters which support power
130management. Some operating systems place the controller in low power
131mode when shutting down, and some PCI BIOSes fail to bring the chip
132out of this state before configuring it. The controller loses all of
133its PCI configuration in the D3 state, so if the BIOS does not set
134it back to full power mode in time, it won't be able to configure it
135correctly. The driver tries to detect this condition and bring
136the adapter back to the D0 (full power) state, but this may not be
137enough to return the driver to a fully operational condition. If
138you see this message at boot time and the driver fails to attach
139the device as a network interface, you will have to perform second
140warm boot to have the device properly configured.
141.Pp
142Note that this condition only occurs when warm booting from another
143operating system. If you power down your system prior to booting
144.Fx ,
145the card should be configured correctly.
146.El
147.Sh SEE ALSO
148.Xr arp 4 ,
149.Xr netintro 4 ,
150.Xr ifconfig 8
151.Rs
152.%B The RealTek 8129 and 8139 datasheets
153.%O ftp.realtek.com.tw:/lancard/data sheet
154.Re
155.Sh HISTORY
156The
157.Nm
158device driver first appeared in
159.Fx 3.0 .
160.Sh AUTHORS
161The
162.Nm
163driver was written by
164.An Bill Paul Aq wpaul@ctr.columbia.edu .
165.Sh BUGS
166Since outbound packets must be longword aligned, the transmit
167routine has to copy an unaligned packet into an mbuf cluster buffer
168before transmission. The driver abuses the fact that the cluster buffer
169pool is allocated at system startup time in a contiguous region starting
170at a page boundary. Since cluster buffers are 2048 bytes, they are
171longword aligned by definition. The driver probably should not be
172depending on this characteristic.
173.Pp
174The RealTek data sheets are of especially poor quality: the grammar
175and spelling are awful and there is a lot of information missing,
176particularly concerning the receiver operation. One particularly
177important fact that the data sheets fail to mention relates to the
178way in which the chip fills in the receive buffer. When an interrupt
179is posted to signal that a frame has been received, it is possible that
180another frame might be in the process of being copied into the receive
181buffer while the driver is busy handling the first one. If the driver
182manages to finish processing the first frame before the chip is done
183DMAing the rest of the next frame, the driver may attempt to process
184the next frame in the buffer before the chip has had a chance to finish
185DMAing all of it.
186.Pp
187The driver can check for an incomplete frame by inspecting the frame
188length in the header preceeding the actual packet data: an incomplete
189frame will have the magic length of 0xFFF0. When the driver encounters
190this value, it knows that it has finished processing all currently
191available packets. Neither this magic value nor its significance are
192documented anywhere in the RealTek data sheets.
193