xref: /openbsd-src/share/man/man4/rl.4 (revision b2ea75c1b17e1a9a339660e7ed45cd24946b230e)
1.\"	$OpenBSD: rl.4,v 1.11 2001/03/08 06:18:34 aaron Exp $
2.\"
3.\" Copyright (c) 1997, 1998
4.\"	Bill Paul <wpaul@ctr.columbia.edu>. All rights reserved.
5.\"
6.\" Redistribution and use in source and binary forms, with or without
7.\" modification, are permitted provided that the following conditions
8.\" are met:
9.\" 1. Redistributions of source code must retain the above copyright
10.\"    notice, this list of conditions and the following disclaimer.
11.\" 2. Redistributions in binary form must reproduce the above copyright
12.\"    notice, this list of conditions and the following disclaimer in the
13.\"    documentation and/or other materials provided with the distribution.
14.\" 3. All advertising materials mentioning features or use of this software
15.\"    must display the following acknowledgement:
16.\"	This product includes software developed by Bill Paul.
17.\" 4. Neither the name of the author nor the names of any co-contributors
18.\"    may be used to endorse or promote products derived from this software
19.\"   without specific prior written permission.
20.\"
21.\" THIS SOFTWARE IS PROVIDED BY Bill Paul AND CONTRIBUTORS ``AS IS'' AND
22.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24.\" ARE DISCLAIMED.  IN NO EVENT SHALL Bill Paul OR THE VOICES IN HIS HEAD
25.\" BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
26.\" CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
27.\" SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
28.\" INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
29.\" CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
30.\" ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
31.\" THE POSSIBILITY OF SUCH DAMAGE.
32.\"
33.\"	$FreeBSD: rl.4,v 1.7 1998/12/24 18:52:47 wpaul Exp $
34.\"
35.Dd November 4, 1998
36.Dt RL 4
37.Os
38.Sh NAME
39.Nm rl
40.Nd RealTek 8129/8139 Fast Ethernet driver
41.Sh SYNOPSIS
42.Cd "rl* at pci? dev ? function ?"
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. This includes the Allied Telesyn AT2550, Genius 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.
53.Pp
54The RealTek controllers use bus master DMA but do not use a
55descriptor-based data transfer mechanism. The receiver uses a
56single fixed size ring buffer from which packets must be copied
57into mbufs. For transmission, there are only four outbound packet
58address registers which require all outgoing packets to be stored
59as contiguous buffers. Furthermore, outbound packet buffers must
60be longword aligned or else transmission will fail.
61.Pp
62The 8129 differs from the 8139 in that the 8139 has an internal
63PHY which is controlled through special direct access registers
64whereas the 8129 uses an external PHY via an MII bus. The 8139
65supports both 10 and 100Mbps speeds in either full or half duplex.
66The 8129 can support the same speeds and modes given an appropriate
67PHY chip.
68.Pp
69The
70.Nm
71driver supports the following media types:
72.Pp
73.Bl -tag -width xxxxxxxxxxxxxxxxxxxx
74.It autoselect
75Enable autoselection of the media type and options. This is only
76supported if the PHY chip attached to the RealTek controller
77supports NWAY autonegotiation. The user can manually override
78the autoselected mode by adding media options to the appropriate
79.Pa /etc/hostname.rlX
80file.
81.It 10baseT/UTP
82Set 10Mbps operation. The
83.Ar mediaopt
84option can also be used to select either
85.Ar full-duplex
86or
87.Ar half-duplex
88modes.
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.Ox ,
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.Ox
161support first appeared in
162.Ox 2.5 .
163.Sh AUTHORS
164The
165.Nm
166driver was written by
167.An Bill Paul Aq wpaul@ctr.columbia.edu .
168.Sh BUGS
169Since outbound packets must be longword aligned, the transmit
170routine has to copy an unaligned packet into an mbuf cluster buffer
171before transmission. The driver abuses the fact that the cluster buffer
172pool is allocated at system startup time in a contiguous region starting
173at a page boundary. Since cluster buffers are 2048 bytes, they are
174longword aligned by definition. The driver probably should not be
175depending on this characteristic.
176.Pp
177The RealTek data sheets are of especially poor quality: the grammar
178and spelling are awful and there is a lot of information missing,
179particularly concerning the receiver operation. One particularly
180important fact that the data sheets fail to mention relates to the
181way in which the chip fills in the receive buffer. When an interrupt
182is posted to signal that a frame has been received, it is possible that
183another frame might be in the process of being copied into the receive
184buffer while the driver is busy handling the first one. If the driver
185manages to finish processing the first frame before the chip is done
186DMAing the rest of the next frame, the driver may attempt to process
187the next frame in the buffer before the chip has had a chance to finish
188DMAing all of it.
189.Pp
190The driver can check for an incomplete frame by inspecting the frame
191length in the header preceding the actual packet data: an incomplete
192frame will have the magic length of 0xFFF0. When the driver encounters
193this value, it knows that it has finished processing all currently
194available packets. Neither this magic value nor its significance are
195documented anywhere in the RealTek data sheets.
196