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 miibus" 41.Cd "device rl" 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, including the following: 48.Pp 49.Bl -bullet -compact -offset indent 50.It 51Accton ``Cheetah'' EN1207D (MPX 5030/5038; RealTek 8139 clone) 52.It 53Allied Telesyn AT2550 54.It 55Allied Telesyn AT2500TX 56.It 57Farallon NetLINE 10/100 PCI 58.It 59Genius GF100TXR, 60.It 61NDC Communications NE100TX-E 62.It 63OvisLink LEF-8129TX 64.It 65OvisLink LEF-8139TX, 66.It 67Netronix Inc. EA-1210 NetEther 10/100 68.It 69KTX-9130TX 10/100 Fast Ethernet, 70.It 71Encore ENL832-TX 10/100 M PCI, 72.It 73Longshine LCS-8038TX-R 74.It 75SMC EZ Card 10/100 PCI 1211-TX 76.It 77Compaq HNE-300 78.It 79LevelOne FPC-0106TX 80.It 81Edimax EP-4103DL CardBus 82.It 83Nortel Networks 10/100BaseTX 84.It 85D-Link DFE-690TXD 86.It 87D-Link DFE-530TX+ 88.It 89D-Link DFE-538TX (same as 530+?) 90.It 91Corega FEther CB-TXD 92.It 93Corega FEtherII CB-TXD 94.It 95Peppercon AG ROL-F 96.It 97Planex FNW-3800-TX 98.It 99Accton EN1207D 100.It 101GigaFast Ethernet EE100-AXP 102.It 103CompUSA no-name 10/100 PCI ethernet NIC 104.It 105Belkin F5D5000 106.El 107.Pp 108The RealTek 8129/8139 series controllers use bus master DMA but do not use a 109descriptor-based data transfer mechanism. 110The receiver uses a 111single fixed size ring buffer from which packets must be copied 112into mbufs. 113For transmission, there are only four outbound packet 114address registers which require all outgoing packets to be stored 115as contiguous buffers. 116Furthermore, outbound packet buffers must 117be longword aligned or else transmission will fail. 118.Pp 119The 8129 differs from the 8139 in that the 8139 has an internal 120PHY which is controlled through special direct access registers 121whereas the 8129 uses an external PHY via an MII bus. 122The 8139 123supports both 10 and 100Mbps speeds in either full or half duplex. 124The 8129 can support the same speeds and modes given an appropriate 125PHY chip. 126.Pp 127Note: support for the 8139C+ chip is provided by the 128.Xr re 4 129driver. 130.Pp 131The 132.Nm 133driver supports the following media types: 134.Pp 135.Bl -tag -width xxxxxxxxxxxxxxxxxxxx 136.It autoselect 137Enable autoselection of the media type and options. 138This is only 139supported if the PHY chip attached to the RealTek controller 140supports NWAY autonegotiation. 141The user can manually override 142the autoselected mode by adding media options to the 143.Pa /etc/rc.conf 144file. 145.It 10baseT/UTP 146Set 10Mbps operation. 147The 148.Ar mediaopt 149option can also be used to select either 150.Ar full-duplex 151or 152.Ar half-duplex 153modes. 154.It 100baseTX 155Set 100Mbps (fast ethernet) operation. 156The 157.Ar mediaopt 158option can also be used to select either 159.Ar full-duplex 160or 161.Ar half-duplex 162modes. 163.El 164.Pp 165The 166.Nm 167driver supports the following media options: 168.Pp 169.Bl -tag -width xxxxxxxxxxxxxxxxxxxx 170.It full-duplex 171Force full duplex operation 172.It half-duplex 173Force half duplex operation. 174.El 175.Pp 176Note that the 100baseTX media type is only available if supported 177by the adapter. 178For more information on configuring this device, see 179.Xr ifconfig 8 . 180.Sh DIAGNOSTICS 181.Bl -diag 182.It "rl%d: couldn't map memory" 183A fatal initialization error has occurred. 184.It "rl%d: couldn't map interrupt" 185A fatal initialization error has occurred. 186.It "rl%d: watchdog timeout" 187The device has stopped responding to the network, or there is a problem with 188the network connection (cable). 189.It "rl%d: no memory for rx list" 190The driver failed to allocate an mbuf for the receiver ring. 191.It "rl%d: no memory for tx list" 192The driver failed to allocate an mbuf for the transmitter ring when 193allocating a pad buffer or collapsing an mbuf chain into a cluster. 194.It "rl%d: chip is in D3 power state -- setting to D0" 195This message applies only to adapters which support power 196management. 197Some operating systems place the controller in low power 198mode when shutting down, and some PCI BIOSes fail to bring the chip 199out of this state before configuring it. 200The controller loses all of 201its PCI configuration in the D3 state, so if the BIOS does not set 202it back to full power mode in time, it won't be able to configure it 203correctly. 204The driver tries to detect this condition and bring 205the adapter back to the D0 (full power) state, but this may not be 206enough to return the driver to a fully operational condition. 207If 208you see this message at boot time and the driver fails to attach 209the device as a network interface, you will have to perform second 210warm boot to have the device properly configured. 211.Pp 212Note that this condition only occurs when warm booting from another 213operating system. 214If you power down your system prior to booting 215.Fx , 216the card should be configured correctly. 217.El 218.Sh SEE ALSO 219.Xr arp 4 , 220.Xr miibus 4 , 221.Xr netintro 4 , 222.Xr ng_ether 4 , 223.Xr ifconfig 8 224.Rs 225.%B The RealTek 8129, 8139 and 8139C+ datasheets 226.%O http://www.realtek.com.tw 227.Re 228.Sh HISTORY 229The 230.Nm 231device driver first appeared in 232.Fx 3.0 . 233.Sh AUTHORS 234The 235.Nm 236driver was written by 237.An Bill Paul Aq wpaul@ctr.columbia.edu . 238.Sh BUGS 239Since outbound packets must be longword aligned, the transmit 240routine has to copy an unaligned packet into an mbuf cluster buffer 241before transmission. 242The driver abuses the fact that the cluster buffer 243pool is allocated at system startup time in a contiguous region starting 244at a page boundary. 245Since cluster buffers are 2048 bytes, they are 246longword aligned by definition. 247The driver probably should not be 248depending on this characteristic. 249.Pp 250The RealTek data sheets are of especially poor quality, 251and there is a lot of information missing 252particularly concerning the receiver operation. 253One particularly 254important fact that the data sheets fail to mention relates to the 255way in which the chip fills in the receive buffer. 256When an interrupt 257is posted to signal that a frame has been received, it is possible that 258another frame might be in the process of being copied into the receive 259buffer while the driver is busy handling the first one. 260If the driver 261manages to finish processing the first frame before the chip is done 262DMAing the rest of the next frame, the driver may attempt to process 263the next frame in the buffer before the chip has had a chance to finish 264DMAing all of it. 265.Pp 266The driver can check for an incomplete frame by inspecting the frame 267length in the header preceding the actual packet data: an incomplete 268frame will have the magic length of 0xFFF0. 269When the driver encounters 270this value, it knows that it has finished processing all currently 271available packets. 272Neither this magic value nor its significance are 273documented anywhere in the RealTek data sheets. 274