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