From iragwar@erols.com Tue Feb 1 20:28:24 2000 Date: Tue Feb 1 20:28:24 2000 From: Got Osimbo iragwar@erols.com Subject: Building driver module with the Kernel This is a multi-part message in MIME format. ------=_NextPart_000_0014_01BF6CF2.B267DD60 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hello, I have NIC which uses rtl8139 driver. I am in a process of building this = driver with kernel and I am not good at this domain. The instructions for compiling the rtl8139 driver is extracted from Donald Backer's = rtl8139 driver's source code and it is shown below: I wonder which compile = command should I use and is the command include double and single quotes and = also brackets. ____________________________________________________________________ Local variables: * compile-command: "gcc -DMODULE -D__KERNEL__ -Wall = -Wstrict-prototypes -O6 -c rtl8139.c `[ -f /usr/include/linux/modversions.h ] && echo = -DMODVERSIONS`" * SMP-compile-command: "gcc -D__SMP__ -DMODULE -D__KERNEL__ -Wall -Wstrict-prototypes -O6 -c rtl8139.c `[ -f = /usr/include/linux/modversions.h ] && echo -DMODVERSIONS`" * cardbus-compile-command: "gcc -DCARDBUS -DMODULE -D__KERNEL__ -Wall -Wstrict-prototypes -O6 -c rtl8139.c -o realtek_cb.o -I/usr/src/pcmcia-cs-3.0.9/include/" * c-indent-level: 4 * c-basic-offset: 4 * tab-width: 4 * End: */ _____________________________________________________________________ Thanks Got ------=_NextPart_000_0014_01BF6CF2.B267DD60 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hello,
 
I have NIC which uses rtl8139 driver. I = am in a=20 process of building this driver
with kernel and I am not good at this = =20 domain. The  instructions for
compiling the rtl8139 driver  = is=20 extracted from  Donald Backer's  rtl8139
driver's source = code and=20 it is shown below:  I wonder  which compile = command
should I=20 use and is the command include double and single quotes and also=20 brackets.
 
________________________________________________________________= ____
Local variables:
 *  = compile-command:=20 "gcc -DMODULE -D__KERNEL__ -Wall -Wstrict-prototypes -O6 = -c
 rtl8139.c=20 `[ -f /usr/include/linux/modversions.h ] && echo=20 -DMODVERSIONS`"
 *  SMP-compile-command: "gcc -D__SMP__ = -DMODULE=20 -D__KERNEL__ -Wall
-Wstrict-prototypes -O6 -c rtl8139.c `[ -f=20 /usr/include/linux/modversions.h ] &&
 echo=20 -DMODVERSIONS`"
 *  cardbus-compile-command: "gcc -DCARDBUS = -DMODULE -D__KERNEL__ -Wall
-Wstrict-prototypes -O6 -c rtl8139.c -o=20 realtek_cb.o
-I/usr/src/pcmcia-cs-3.0.9/include/"
 * =20 c-indent-level: 4 *  c-basic-offset: 4 *  tab-width: 4 * End:=20 */

_______________________________________________________________= ______
 
Thanks
 
Got



------=_NextPart_000_0014_01BF6CF2.B267DD60-- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From osmondng@writeme.com Tue Feb 1 23:39:41 2000 Date: Tue Feb 1 23:39:41 2000 From: Osmond Ng osmondng@writeme.com Subject: echo subscribe | /bin/mail linux-realtek-request@beowulf.gsfc.nasa.gov | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Wed Feb 2 04:55:52 2000 Date: Wed Feb 2 04:55:52 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: Building driver module with the Kernel > I have NIC which uses rtl8139 driver. I am in a process of building this driver > with kernel and I am not good at this domain. The instructions for > compiling the rtl8139 driver is extracted from Donald Backer's rtl8139 > driver's source code and it is shown below: I wonder which compile command > should I use and is the command include double and single quotes and also brackets. > > ____________________________________________________________________ > Local variables: > * compile-command: "gcc -DMODULE -D__KERNEL__ -Wall -Wstrict-prototypes -O6 -c > rtl8139.c `[ -f /usr/include/linux/modversions.h ] && echo -DMODVERSIONS`" > * SMP-compile-command: "gcc -D__SMP__ -DMODULE -D__KERNEL__ -Wall > -Wstrict-prototypes -O6 -c rtl8139.c `[ -f /usr/include/linux/modversions.h ] && > echo -DMODVERSIONS`" > * cardbus-compile-command: "gcc -DCARDBUS -DMODULE -D__KERNEL__ -Wall > -Wstrict-prototypes -O6 -c rtl8139.c -o realtek_cb.o > -I/usr/src/pcmcia-cs-3.0.9/include/" > * c-indent-level: 4 * c-basic-offset: 4 * tab-width: 4 * End: */ If you don't know which command is correct for your system, then the first one is probably the correct one. If your system is SMP, you would probably know that. Same for cardbus. If it's a standard PCI card being put in a computer with only one CPU, the first one is what you want. Do not type the double quotes when you use the command. Do type the single quotes and brackets. Make sure that your kernel sources are up to date. If your distribution does things properly, then it is probably better to substitute /usr/src/linux/include/linux in place of /usr/include/linux. Of course, the word "properly" is open to interpretation. In many cases, it is not actually necessary to do any of this, though. For most people it is sufficient to use the driver which is distributed with the kernel sources, or precompiled by your distributor. Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From rd.it@gabel.net Wed Feb 2 08:27:58 2000 Date: Wed Feb 2 08:27:58 2000 From: Paolo Panizza rd.it@gabel.net Subject: getting two directly connected rtl81 Hi. 1) diag report: when you use -m, diag_1.01 inverts BMCR and BMSR values; the (relevant line) src code should be : inw(ioaddr + MII_BMCR), inw(ioaddr + MII_BMSR) instead of: inw(ioaddr + MII_BMSR), inw(ioaddr + MII_BMCR) Also, since you compiled with -DLIBMII, -mm makes diag go through show_mii_details() function with -mm: --------- #ifdef LIBMII if (show_mii > 1) { show_mii_details(ioaddr, -1); --------- but the -1par. (ie PHY #-1) doesn't work neither for me; I changed it to 32 (ie PHY #32) as used in mii-diag.c and got the correct numbers from diag (ie: same answers/results as from RTK's own setup prg/diag) 2) to tell your NIC how to behave, you should use the -w option of diag; but I did not use it and can't tell much more; however, it may be useful for you my card values as ref. (and perhaps get the RTL8139 datasheet from www.realtek.com.tw): =========== The RTL8139 does not use a MII transceiver. It does have internal MII-compatible registers: Basic mode control register 0x1000. Basic mode status register 0x782d. Autonegotiation Advertisement 0x05e1. Link Partner Ability register 0x40a1. Autonegotiation expansion 0x0001. Disconnects 0x0000. False carrier sense counter 0x0000. NWay test register 0x0004. Receive frame error count 0x0000. MII PHY #32 transceiver registers: 1000 782d 0000 0000 05e1 40a1 0001 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000. Basic mode control register 0x1000: Auto-negotiation enabled. Basic mode status register 0x782d ... 782d. Link status: established. Capable of 100baseTx-FD 100baseTx 10baseT-FD 10baseT. Able to perform Auto-negotiation, negotiation complete. This transceiver has no vendor identification. I'm advertising 05e1: Flow-control 100baseTx-FD 100baseTx 10baseT-FD 10baseT Advertising no additional info pages. IEEE 802.3 CSMA/CD protocol. Link partner capability is 40a1: 100baseTx 10baseT. ======= Which is correct [hub is an AT-FH716SW] BTW: ping -f -s1500 gets about 5.5MBps btween 2 x86-200MHz PCs; file xfer rates at ca. 4.2MBps. Bye. Paolo ---------------------------------------------------- From: David -dof- Goodwin To: linux-realtek@beowulf.gsfc.nasa.gov; Centralino (Arianna) Subject: getting two directly connected rtl8139 cards to goto 100mb/s Date: Wed Feb 2 08:27:58 2000 <> <> Hello... I currently have two computers running slackware linux connected together with a cross over cable. (they also have another ne2k-pci type card each, which they use to talk to another LAN - sm.net (private ip setup thing)) The nics connected to this cable are both rtl8139 based, but neither wants to run at 100mb/s (they are stuck at 10mb/s) Is there any way that i can tell them to run at 100mb/s ? (eg a command to put in the /etc/conf.modules or something similar, or a boottime command to go in /etc/lilo.conf ?) I've read most of the realtrek related webpages by Donald Becker, but they don't seem to mention anything about this. I've run the rtl8139 diagnostic program which was on his webpage, and the results are attached below. As you can see neither card seems to want to autonegotiate. Thanks in advance for any help. David Goodwin | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From rd.it@gabel.net Wed Feb 2 08:28:00 2000 Date: Wed Feb 2 08:28:00 2000 From: Paolo Panizza rd.it@gabel.net Subject: i have this problem: redhat linux 52 Hi. Excluding troubles on the other machine... seems to me your m.board doesn't have pci-master capable slots (they say this NIC must be installed on such a slot) or is faulty. You should try the latest driver, however: I'm running RH 5.2 on an Asus' MB (with UWSCSI on-board) and K6, the lancard is a basic 8139 from Level-1, with latest driver from Becker's site and am having no troubles (so far...). bye. Paolo ---------- From: Dan Uscatu To: linux-realtek@beowulf.gsfc.nasa.gov; Centralino (Arianna) Date: Wed Feb 2 08:28:00 2000 hi i have this problem: redhat linux 52 machine, kernel 2.0.36, realtek 8139 fast ethernet adaptor connected via switch to the lan. when i open a terminal and 'cat' a long file, it scrolls very fast and the ethernet card freezes. syslog message is: Jan 31 13:08:43 machine kernel: eth0: Transmit timeout, status 0d 0000. Jan 31 13:09:09 machine last message repeated 5 times i do a net restart and all goes fine, if i do it fast i dont even lose my terminal session, and i am able to see the rest of the file. i have changed the port on the switch, the switch, the cable, the ethernet card. still the same. can anyone help ? i am using this version of driver (provided by the kernel sources): rtl8139.c:v0.99B 4/7/98 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html Dan Uscatu Network Administrator DigiCom Systems | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From furtado@predialnet.com.br Sat Feb 5 10:47:02 2000 Date: Sat Feb 5 10:47:02 2000 From: Miguel B. Furtado Jr. furtado@predialnet.com.br Subject: problems wiht rtl8139 This is a multi-part message in MIME format. ------=_NextPart_000_0009_01BF6FE0.328E3080 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi. I have an EA-1210 Series 10/100M Fast Ethernet PCI Ethernet Adapter = (made by Netronix with the RTL8139b chip, www.netronixinc.com) and I'm = trying to install it on Linux Red Hat 6.0. I have followed the = instructions on http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html = but when I type the command "insmod rtl8139.o" I receive the message = "device is busy". If I use netconf to configure my card, I can't find the rtl8139 module = on the kernel modules avaible there. Anyone could help me to configure my ethernet card? Thanks. Miguel B. Furtado Jr. e-mail: furtado@predialnet.com.br ------=_NextPart_000_0009_01BF6FE0.328E3080 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi. I have an EA-1210 Series 10/100M Fast Ethernet = PCI=20 Ethernet Adapter (made by Netronix with the RTL8139b chip, www.netronixinc.com) and = I'm trying=20 to install it on Linux Red Hat 6.0. I have followed the instructions on = http://ce= sdis.gsfc.nasa.gov/linux/drivers/rtl8139.html but=20 when I type the command "insmod rtl8139.o" I receive the message "device = is=20 busy".
If I use netconf to configure my card, I can't find = the=20 rtl8139 module on the kernel modules avaible there.
Anyone could help me to configure my ethernet=20 card?
Thanks.
 Miguel B. Furtado Jr.
 e-mail: furtado@predialnet.com.br
------=_NextPart_000_0009_01BF6FE0.328E3080-- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Sun Feb 6 05:56:39 2000 Date: Sun Feb 6 05:56:39 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: problems wiht rtl8139 > Hi. I have an EA-1210 Series 10/100M Fast Ethernet PCI Ethernet Adapter = > (made by Netronix with the RTL8139b chip, www.netronixinc.com) and I'm = > trying to install it on Linux Red Hat 6.0. I have followed the = > instructions on http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html = > but when I type the command "insmod rtl8139.o" I receive the message = > "device is busy". Is the device listed when you cat /proc/pci? There should be an IRQ listed there. Is it being used by another device? Check /proc/interrupts. Do you have any other device drivers running for this card? > If I use netconf to configure my card, I can't find the rtl8139 module = > on the kernel modules avaible there. I do not have any experience with netconf, so I can't help you there. Plain text email is preferred. At least by me. Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.com Mon Feb 7 07:17:08 2000 Date: Mon Feb 7 07:17:08 2000 From: Jeff Garzik jgarzik@mandrakesoft.com Subject: ANN: Public beta test of new RTL-8139 Fast Ethernet driver This is a multi-part message in MIME format. --------------CD86E0A5D2F26D665C837C79 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit The first public beta test release of "8139too" is now available for download, testing, and non-production use from http://gtf.org/garzik/drivers/8139too/ Kernel 2.3.40 or later is required for use, though a 2.2.x-compatible version is currently being developed. "8139too" driver features improved performance, bug fixes, and portability over the existing experimental "rtl8139" driver, from which 8139too is derived. Testing feedback and patches are greatly appreciated. The README is attached below for more details. Jeff Garzik MandrakeSoft, Inc. -- --------------CD86E0A5D2F26D665C837C79 Content-Type: text/plain; charset=us-ascii; name="README.txt" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="README.txt" "8139too" Fast Ethernet driver for Linux 2.4.0 Improved support for RTL-8139 Fast Ethernet adapters Copyright 2000 Jeff Garzik Version 0.7.0 - a.k.a. the first public beta release February 7, 2000 Architectures supported (all PCI platforms): x86, Alpha AXP, PowerPC, Sparc64 Disclaimer ---------- THIS DRIVER IS A DEVELOPMENT RELEASE FOR A DEVELOPMENT KERNEL. DO NOT USE IN A PRODUCTION ENVIRONMENT. DO NOT CONTACT DONALD BECKER FOR SUPPORT OF THIS DRIVER, his driver is completely different and maintained independently of the 8139too code base. Requirements ------------ Kernel 2.3.40 or later. RTL-8139-based Fast Ethernet adapters. Introduction ------------ The "8139too" Fast Ethernet driver for Linux 2.4.0 is a substantial modification of the experimental rtl8139 driver from Donald Becker, some versions of which appeared in 2.2.x and 2.3.x kernels. The RTL-8139 is a very low-cost Fast Ethernet chip, which makes it very popular. The step from 2.2.x to 2.4.x kernels brings many new features to Linux device drivers. Features for MMIO resources, a standard hot-plug API, and other interfaces are now becoming requirements, as drivers move off the x86 platform. With that in mind, I have begun updating the RTL-8139 driver to current 2.3.x (2.4) kernel standards and APIs, and fixing the problems that users have been encountering. Features of 8139too ------------------- [note - this list intended for people familiar with kernel drivers] ** 100% MMIO, for full speed operation. All users (so far) have reported performance increases over their existing RTL drivers. ** Multi-platform support: x86, Alpha, PPC, ... ** Use proper SMP spinlocking, fixing SMP interrupt bugs, making the driver portable to non-x86 SMP platforms in the process. ** Use new PCI driver API for seamless, low-maintenance hot-plug support ** Several bugs fixes from original rtl8139 1.08r (October 5, 1999), including the very common "transmit timeout" problem. * Use new resource allocation API, required for hot-plug support * Use new register read/write macros * initcall support (module_init/exit) * vastly improved debug tracing support * code formatting in many places for readability * use new init_etherdev() facilities ...and probably some other less important changes which I forgot. Installation ------------ OPTION 1: Build inside kernel tree (into kernel image, or as module) 1) cp 8139too.c $my_source_tree/drivers/net/rtl8139.c 2) cp kcompat24.[ch] $my_source_tree/drivers/net 3) Apply the patch "patch-2.3.x". OPTION 2: Build outside kernel tree Use the included Makefile. Tested Adapters --------------- AOpen ALN-325C KTI KF-230TX KTI KF-230TX/2 (please add your adapter model to this list) Status of Platform Support -------------------------- (see errata below for details) x86: tested, stable Alpha AXP: tested, stable PowerPC: tested, unstable Sparc64: not tested Special Thanks -------------- The following people contributed invaluable testing time, feedback and/or patches during the development of this driver. Thanks to all of them. Donald Becker, Alan Cox, Richard Stallman, Linus Torvalds - inspiration Alan Cox, Gerard Roudier - insight on posted MMIO writes Martin Mares - code review Tigran Aivazian - testing, code review, and a bug fix Chmouel Boudjnah, Alexander Dietrich, Oleg Drokin, James Fidell, Taso Hatzi, Peter K - intrepid test team And thanks to every supporter free software. Known Bugs / Errata / To-Do --------------------------- The following issues are known, and are actively being pursued. Patches to resolve these issues is welcome. If a problem occurs which is not in the list, please report it. That's why we do beta releases, after all... 1) Work with Donald to merge fixes and updates into his driver. 2) 2.2.x COMPATIBILITY SUPPORT IS BROKEN. DO NOT USE IT. It is included only for enterprising hackers willing to help fix it. 3) PPC platform has stability problems. 4) Sparc64 platform not tested at all. 5) RTL-8129 support. 8129 support was temporarily removed for lack of any test adapters. Owners of 8129 chips help test this driver! 6) CardBus support. Hot-plug support is mostly complete, but not tested at all. Do not rely on hot-plug support. 7) N-Way auto-negotiation is known to fail in some cases. This problem also occurs in the rtl8139 driver in kernels 2.2.x/2.3.x. 8) Much improved command line / module parameter setup. (patches and suggestions welcome) 9) Better documentation. (patches welcome) 10) User-mode (or maybe optional /proc) diagnostics program. [EOF] --------------CD86E0A5D2F26D665C837C79-- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Mon Feb 7 07:35:56 2000 Date: Mon Feb 7 07:35:56 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: ANN: Public beta test of new RTL-8139 Fast Ethernet driver > The first public beta test release of "8139too" is now available for > download, testing, and non-production use from > http://gtf.org/garzik/drivers/8139too Interesting. Why was it necessary to fork this driver code? Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From bhawkins@oz.net Mon Feb 7 08:00:27 2000 Date: Mon Feb 7 08:00:27 2000 From: Byron Hawkins bhawkins@oz.net Subject: via-rhine and multiple cards Can someone tell me if it is possible to support two D-Link DFE-530TX cards with the via-rhine driver? The via-rhine works fine for one of my cards, but it doesn't see the other one, and there appears to be no conf.modules option for specifying the location of the second card. Thanks in advance. Byron -- bhawkins@oz.net www.oz.net/~bhawkins | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From alexander.dietrich@iname.com Mon Feb 7 09:33:42 2000 Date: Mon Feb 7 09:33:42 2000 From: Alexander Dietrich alexander.dietrich@iname.com Subject: RTL8139 still "jams" Hi, for some reasons I didn't receive my posting from the list, so I'll try again: -------------------------------------------------------------- I still have a problem with an RTL8139A-based card: I have two computers with the same network card. Computer A is a K6-2 300 with a VIA MVP4 chipset and Computer B is a K6-3 400 with an ALi Aladdin V chipset. Both run on RedHat 6.0 with kernel 2.2.14, both are connected to a 100 Mbit hub. The problem: when transfering large files from B to A, A stops responding after a couple of megabytes. It is not even possible to ping between the computers anymore, local traffic is not affected. I don't receive any error messages of transmission failures or things like that in /var/log/messages or elsewhere. I can get the card working again by taking the interface down and up on A with the /etc/rc.d/init.d/network script. I tried driver versions 1.07, 1.08 and 1.08a and also changed the PCI slot the card is in and exchanged the cards between A and B, nothing helped. When copying the same files from A to B I don't have any problems. Another twist: the same card worked in an older third computer with a P90 and some old Intel chipset. So what is happening here ? Is it the card itself or the driver failing ? And why is the card failing only on this new system, is the data coming in too fast ? But computer B is even faster, so why doesn't it have any problems ? I'm going to borrow a 3Com card from a friend and see if it works better. This is neither a cheap nor nice "solution" to the problem, but this card/driver is starting to waste my time and I need the system working. I had always thought of the Realtek cards as cheap but usable, too bad. Does anybody have any ideas ? Regards, Alexander Dietrich -- | Alexander Dietrich | Norderstedt, Germany | | e-mail: alexander.dietrich@iname.com | | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Mon Feb 7 09:50:56 2000 Date: Mon Feb 7 09:50:56 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: RTL8139 still "jams" > The problem: when transfering large files from B to A, A stops > responding after a couple of megabytes. It is not even possible > to ping between the computers anymore, local traffic is not affected. That seems to be a common problem. > So what is happening here ? Is it the card itself or the driver failing ? It seems to me that it must be the driver. If you (and every one of us who has reported this problem in the past) can get the card to start talking again by only restarting the driver, then the driver must be at fault. > And why is the card failing only on this new system, is the data coming in > too fast ? But computer B is even faster, so why doesn't it have any > problems ? The only thing I can say is, "join the club". I have four machines and four 8139-based cards and three different types of motherboards. Two machines work ok, one occasionally shows this problem and one consistently shows this problem. The two machines that show this have the same motherboard. Swapping the ethernet cards around changes nothing. > I'm going to borrow a 3Com card from a friend and see if it works better. It undoubtedly will. I've never had a problem with any of my machines which have 3Com cards. > Does anybody have any ideas ? I suppose you could try the new driver which was just mentioned in the previous post: The first public beta test release of "8139too" is now available for download, testing, and non-production use from http://gtf.org/garzik/drivers/8139too/ Kernel 2.3.40 or later is required for use, though a 2.2.x-compatible version is currently being developed. "8139too" driver features improved performance, bug fixes, and portability over the existing experimental "rtl8139" driver, from which 8139too is derived. Testing feedback and patches are greatly appreciated. The README is attached below for more details. Jeff Garzik MandrakeSoft, Inc. Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From bhawkins@oz.net Mon Feb 7 15:56:45 2000 Date: Mon Feb 7 15:56:45 2000 From: Byron Hawkins bhawkins@oz.net Subject: Never mind: via-rhine and multiple cards Never mind this question. Turns out my second card isn't a D-Link at all! -----Original Message----- From: Byron Hawkins To: linux-realtek@beowulf.gsfc.nasa.gov Date: Mon Feb 7 15:56:45 2000 Subject: via-rhine and multiple cards > >Can someone tell me if it is possible to support two D-Link DFE-530TX >cards with the via-rhine driver? The via-rhine works fine for one of my >cards, but it doesn't see the other one, and there appears to be no >conf.modules option for specifying the location of the second card. >Thanks in advance. > >Byron > >-- > >bhawkins@oz.net >www.oz.net/~bhawkins > > > > | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the > | body of the mail, include only the text: > | unsubscribe this-list-name youraddress@wherever.org > | You will be unsubscribed as speedily as possible. > | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From akuehn@HTWM.De Tue Feb 8 08:44:50 2000 Date: Tue Feb 8 08:44:50 2000 From: Nagilum akuehn@HTWM.De Subject: Wakeup frame Hi, I know this is a bit off topic, but only a bit. I recently tried to wake up my box using the program below but either this program doesn't work or my hardware setup is wrong because it did not work. I would like to know if the following program works with a rtl8139 chip and if not if there are any other programs for this purpose.. regards, Alexander. PS: The program comes from http://www.berens-partner.de/wann/computer/wol.html but i applied the netbsd patches to it, now it works on linux too ;) . <----------------------snip-----------------------> /* WAKE ON LAN - Remote Power On hpwake.c: Sendet eine Broadcast Message an alle Clients des Subnetzes, die eine Abfolge von Bytes enthaelt um einen Rechner zu booten. Unterstuetzte OS: HP-UX 10.20 und 9.05 Kompilieren: cc -o hpwake hpwake.c Aufruf: hpwake xx:xx:xx:xx:xx:xx xx:xx:xx:xx:xx:xx ist die MAC Adresse der Ethernetkarte in Hexadezimalwerten, man erhaelt sie fuer einen entsprechenden Rechner durch Eingabe von: Unter HP-UX: lanscan Unter Windoof NT: ipconfig -all Unter Windoof 95: winipcfg email: wann@server7.ik.mb.uni-siegen.de */ #include #include #include #include #include #include #include #include u_char magicpacket[500]; int get_magicpacket(unsigned char *,char *); int main(int,char *[]); int get_magicpacket(packet,arg) unsigned char *packet; char *arg; { unsigned char ether_addr[6]; int dstaddr[6]; int i; int j; int k; char *a; for (a = arg; *a; a++) if (*a != ':' && !isxdigit(*a)) { fprintf(stderr, "MAC address must be given in hex! [00:11:22:33:44:55]\n"); exit(2); } k = sscanf(arg, "%2x:%2x:%2x:%2x:%2x:%2x", &dstaddr[0], &dstaddr[1], &dstaddr[2], &dstaddr[3], &dstaddr[4], &dstaddr[5]); if (k != 6) { fprintf(stderr, "Invalid MAC address! [00:11:22:33:44:55]\n"); exit(2); } for (i = 0; i < 6; i++) ether_addr[i] = dstaddr[i]; memcpy(packet, ether_addr, 6); packet[6] = 0x00; packet[7] = 0x74; j = 8; memset(packet+j, 0xff, 6); j += 6; for (i = 0; i < 16; i++) { memcpy(packet+j, ether_addr, 6); j += 6; } packet[111] = 0x40; packet[112] = 0x00; packet[113] = 0x90; packet[114] = 0x90; packet[115] = 0x40; packet[116] = 0x00; j += 6; return j; } int main(argc,argv) int argc; char *argv[]; { struct sockaddr_in client, server; int one = 1; int packetsize; int s; int i; if (argc < 2) { fprintf(stderr, "Usage: %s xx:xx:xx:xx:xx:xx\n", argv[0]); exit(1); } packetsize = get_magicpacket(magicpacket, argv[1]); if ((s = socket(AF_INET, SOCK_DGRAM, 0)) < 0) perror("socket"); if (setsockopt(s, SOL_SOCKET, SO_BROADCAST, (char *)&one, sizeof(one)) < 0) perror("setsockopt"); server.sin_family = AF_INET; server.sin_port = 0; server.sin_addr.s_addr = INADDR_ANY; client.sin_family = AF_INET; client.sin_port = 32768+666; /* invalid port */ client.sin_addr.s_addr = INADDR_BROADCAST; if (bind(s, (struct sockaddr *)&server, sizeof(server)) < 0) perror("bind"); if ((i = sendto(s, magicpacket, packetsize, 0, (struct sockaddr *)&client, sizeof(client))) < 0) perror("sendto"); close(s); return 0; } <----------------------snip-----------------------> -- ======================================================================== # _ __ _ __http://home.htwm.de/akuehn/MPeople/MPeople.html# # / |/ /__ ____ _(_) /_ ____ _ dedicated to one of the best.. # # / / _ `/ _ `/ / / // / ' \ nagilum@chillout.org or +01776461165 # # /_/|_/\_,_/\_, /_/_/\_,_/_/_/_/ Amiga OS/NetBSD(68k)/Linux(68k/PPC) # # /___/ x86: Linux / OpenBSD / FreeBSD / Win 98 # ======================================================================== | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Tue Feb 8 08:59:24 2000 Date: Tue Feb 8 08:59:24 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: Wakeup frame > I know this is a bit off topic, but only a bit. I recently tried to wake > up my box using the program below but either this program doesn't work > or my hardware setup is wrong because it did not work. Can you elaborate on what your hardware setup is? AFAIK, a generic PCI LAN card cannot wake up a PC. It need to be specially designed for this purpose and have a cable which goes directly from the card to the motherboard. And it only works in ATX systems, of course. > I would like to know if the following program works with a rtl8139 chip > and if not if there are any other programs for this purpose.. I don't have any capable hardware, so I can't really say if this program works or not. Wake on LAN is only useful if you ever turn your machines off, which I don't. 'cept my laptop, of course. Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From akuehn@HTWM.De Tue Feb 8 09:34:49 2000 Date: Tue Feb 8 09:34:49 2000 From: Nagilum akuehn@HTWM.De Subject: Wakeup frame sharkey@superk.physics.sunysb.edu wrote: > > I know this is a bit off topic, but only a bit. I recently tried to wake > > up my box using the program below but either this program doesn't work > > or my hardware setup is wrong because it did not work. > > Can you elaborate on what your hardware setup is? AFAIK, a generic PCI > LAN card cannot wake up a PC. It need to be specially designed for this > purpose and have a cable which goes directly from the card to the > motherboard. And it only works in ATX systems, of course. Of course i have ATX and a network card which is supposed to be able to wake up the computer (and i have the little power cable connecting the board and the nw-card plugged in) The board is a MS6167 the network adapter is a LevelONE FNC-0107TX. Alexander. -- ======================================================================== # _ __ _ __http://home.htwm.de/akuehn/MPeople/MPeople.html# # / |/ /__ ____ _(_) /_ ____ _ dedicated to one of the best.. # # / / _ `/ _ `/ / / // / ' \ nagilum@chillout.org or +01776461165 # # /_/|_/\_,_/\_, /_/_/\_,_/_/_/_/ Amiga OS/NetBSD(68k)/Linux(68k/PPC) # # /___/ x86: Linux / OpenBSD / FreeBSD / Win 98 # ======================================================================== | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From manty@i.am Sun Feb 13 14:42:23 2000 Date: Sun Feb 13 14:42:23 2000 From: Santiago Garcia Mantinan manty@i.am Subject: rtl8139 problem This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. Send mail to mime@docserver.cac.washington.edu for more info. ---1463811839-1771354263-949763780=:435 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Content-ID: Hi! It is long time since I found this problem, wich I believe I can now replicate. I have written about this to Donald but I think he must be too busy to take a look at it, so I'd like to see if anyone on the list can confirmate this. The problem was found on a Linux bridge with rtl8139 based cards, where it causes that windows clients conected to the bridge loose packages. Now I have found what I think is the problem, and it is not related to any bridge or anything, so it is easy to see on any network. The thing is that you do, for example a ping on a network and try to see its length with for example a tcpdump -e, well, the hosts with rtl8129 cards will see a legth 4 bytes greater than the hosts with other cards, on the ping example, the correct length is 98, while the tcpdump on the rtl8129 shows 102. This happens with all the packages, and wether the rtl8139 card is on promiscuous mode or not, and wether the package was for it, or from it or not. This is what is causing the problems on the bridge when packages are near the mtu limit, but it can cause many other problems, so let's see if any of you can confirmate this so it can be fixed. Thanks in advance. Regards... Manty/BestiaTester -> http://www.mantianito.ddns.org ---1463811839-1771354263-949763780=:435-- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From patrik@ein.cz Mon Feb 14 04:24:26 2000 Date: Mon Feb 14 04:24:26 2000 From: Patrik Rak patrik@ein.cz Subject: RTL8139 TX timeouts Hi! FYI, from time to time I get error message like this: eth0: Transmit timeout, status 0d 2000 media 08. eth0: Tx queue start entry 4026122 dirty entry 4026118. eth0: Tx descriptor 0 is 100805ea. eth0: Tx descriptor 1 is 400845ea. eth0: Tx descriptor 2 is 000805ea. (queue head) eth0: Tx descriptor 3 is 000805ea. eth0: MII #32 registers are: 1000 782d 0000 0000 05e1 0000 0000 0000. The kernel in question is 2.2.14. Contact me in case you want more related hw/sw info. Patrik | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From mrenzmann@compulan.de Mon Feb 14 09:18:06 2000 Date: Mon Feb 14 09:18:06 2000 From: Michael Renzmann mrenzmann@compulan.de Subject: RTL8139 TX timeouts Hi. > FYI, from time to time I get error message like this: [...] I encountered a similar problem a few minutes ago. Kernel is 2.2.14, I use two RTL8139-Cards. Trying to ping IPs within our network failed, then the following message came up: eth0: Transmit timeout, status 01 0000 media 80 eth0: Tx queue start entry 4 dirty entry 0. After this message pinging worked again (from linux and from other computers within the same network). Any hints what to do against this behaviour? I plan to use the rtl8139 within some of our products, so those "i donīt want to ping"-feature wouldnīt be good :) Bye, Mike --------------------------------------------- Development - Wireless Solutions CompuLAN Europe GmbH, Germany Tel: ++49-6003-815299 Fax: ++49-6003-815222 --------------------------------------------- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.com Mon Feb 14 09:46:03 2000 Date: Mon Feb 14 09:46:03 2000 From: Jeff Garzik jgarzik@mandrakesoft.com Subject: RTL8139 TX timeouts Michael Renzmann wrote: > > Hi. > > > FYI, from time to time I get error message like this: > [...] > > I encountered a similar problem a few minutes ago. Kernel is 2.2.14, > I use two RTL8139-Cards. Trying to ping IPs within our network > failed, then the following message came up: > > eth0: Transmit timeout, status 01 0000 media 80 > eth0: Tx queue start entry 4 dirty entry 0. > > After this message pinging worked again (from linux and from other > computers within the same network). Any hints what to do against this > behaviour? I plan to use the rtl8139 within some of our products, so > those "i donīt want to ping"-feature wouldnīt be good :) The 8139too driver in Linux kernel 2.3.45 and later fix this problem. Jeff -- Jeff Garzik | Writing software is Building 1024 | more fun than working. MandrakeSoft, Inc. | -Anon | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From mrenzmann@compulan.de Mon Feb 14 10:48:37 2000 Date: Mon Feb 14 10:48:37 2000 From: Michael Renzmann mrenzmann@compulan.de Subject: RTL8139 TX timeouts Hi Jeff. > > I encountered a similar problem a few minutes ago. Kernel is 2.2.14, > > I use two RTL8139-Cards. Trying to ping IPs within our network > > failed, then the following message came up: [...] > The 8139too driver in Linux kernel 2.3.45 and later fix this problem. Is it possible to simply replace the version from kernel 2.2.14 with the source from 2.3.45? Bye, Mike ------------------------------------ Development - Wireless Solutions CompuLAN Europe GmbH, Germany Tel: ++49-6003-815299 Fax: ++49-6003-815222 ------------------------------------ | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.mandrakesoft.com Mon Feb 14 10:52:54 2000 Date: Mon Feb 14 10:52:54 2000 From: Jeff Garzik jgarzik@mandrakesoft.mandrakesoft.com Subject: RTL8139 TX timeouts On Mon, 14 Feb 2000, Michael Renzmann wrote: > Hi Jeff. > > > > I encountered a similar problem a few minutes ago. Kernel is 2.2.14, > > > I use two RTL8139-Cards. Trying to ping IPs within our network > > > failed, then the following message came up: [...] > > The 8139too driver in Linux kernel 2.3.45 and later fix this problem. > > Is it possible to simply replace the version from kernel 2.2.14 with > the source from 2.3.45? I will answer a different question, to avoid confusion... :) You can boot a 2.2.x kernel and a 2.3.x kernel without having to change anything but LILO. If you are using the kernel in a production environment, you should not upgrade to 2.3.x unless 2.2.x simply fails to work. Regards, Jeff | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.mandrakesoft.com Mon Feb 14 11:03:06 2000 Date: Mon Feb 14 11:03:06 2000 From: Jeff Garzik jgarzik@mandrakesoft.mandrakesoft.com Subject: RTL8139 TX timeouts On Mon, 14 Feb 2000, Michael Renzmann wrote: > Is it possible to simply replace the version from kernel 2.2.14 with > the source from 2.3.45? I'm sorry, I didn't understand your question at first. No, it is not possible to use the 8139too driver under kernel 2.2.x. Jeff | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From remonb@gelrevision.nl Mon Feb 14 16:50:11 2000 Date: Mon Feb 14 16:50:11 2000 From: Remon Boonstra remonb@gelrevision.nl Subject: Ain't getting it to work at 100Mbps This is a multi-part message in MIME format. ------=_NextPart_000_0009_01BF773D.C586A7F0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi, I just don't get my 2 RTL cards to work they are connected with an = twisted cable,=20 and one side with winNT(RTL1) the other side Linux Slackware (RTL2 and = NE2000Clone) The NE2000 clone works fine, and the RTL's too, but the RTL's work at = 10Mbps, not 100!??!? Why not? it doesn't help setting it in linux with mii-diag -F 100baseT Thanks in Advance... ------=_NextPart_000_0009_01BF773D.C586A7F0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi,
 
I just don't get my 2 RTL cards to work they are connected with an = twisted=20 cable,
and one side with winNT(RTL1) the other side Linux Slackware (RTL2 = and=20 NE2000Clone)
The NE2000 clone works fine, and the RTL's too, but the RTL's work = at=20 10Mbps, not 100!??!?
 
Why not?
 
it doesn't help setting it in linux with mii-diag -F 100baseT
 
Thanks in Advance...
 ------=_NextPart_000_0009_01BF773D.C586A7F0-- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.com Tue Feb 15 11:41:16 2000 Date: Tue Feb 15 11:41:16 2000 From: Jeff Garzik jgarzik@mandrakesoft.com Subject: 8139too update available for testing This is a multi-part message in MIME format. --------------2988B51D83FAD47457F2DCAB Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello 8139 users! Here is an update against Linux kernel 2.3.45 which imports some features and fixes from the FreeBSD driver. If anyone wanted to give this a stress test, that would be great. -- Jeff Garzik | Writing software is Building 1024 | more fun than working. MandrakeSoft, Inc. | -Anon --------------2988B51D83FAD47457F2DCAB Content-Type: text/plain; charset=us-ascii; name="8139too.patch" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="8139too.patch" Index: drivers/net/8139too.c =================================================================== RCS file: /g/cvslan/linux_2_3/drivers/net/Attic/8139too.c,v retrieving revision 1.1.1.1 retrieving revision 1.1.1.1.8.2 diff -u -r1.1.1.1 -r1.1.1.1.8.2 --- drivers/net/8139too.c 2000/02/13 19:14:32 1.1.1.1 +++ drivers/net/8139too.c 2000/02/15 16:04:21 1.1.1.1.8.2 @@ -19,6 +19,8 @@ David S. Miller - PCI DMA and softnet updates + Ernst Gill - fixes ported from BSD driver + ----------------------------------------------------------------------------- Theory of Operation @@ -82,7 +84,7 @@ #include -#define RTL8139_VERSION "0.9.0" +#define RTL8139_VERSION "0.9.1" #define RTL8139_MODULE_NAME "8139too" #define RTL8139_DRIVER_NAME RTL8139_MODULE_NAME " Fast Ethernet driver " RTL8139_VERSION #define PFX RTL8139_MODULE_NAME ": " @@ -154,15 +156,17 @@ #define RTL8139_CAPS HAS_CHIP_XCVR|HAS_LNK_CHNG typedef enum { - UNKNOWN = 0, RTL8139, RTL8139_CB, SMC1211TX, - MPX5030, + /*MPX5030,*/ SIS900, SIS7016, + DELTA8139, + ADDTRON8139, } chip_t; + static struct { chip_t chip; const char *name; @@ -170,19 +174,24 @@ { RTL8139, "RealTek RTL8139 Fast Ethernet"}, { RTL8139_CB, "RealTek RTL8139B PCI/CardBus"}, { SMC1211TX, "SMC1211TX EZCard 10/100 (RealTek RTL8139)"}, - { MPX5030, "Accton MPX5030 (RealTek RTL8139)"}, +/* { MPX5030, "Accton MPX5030 (RealTek RTL8139)"},*/ { SIS900, "SiS 900 (RealTek RTL8139) Fast Ethernet"}, { SIS7016, "SiS 7016 (RealTek RTL8139) Fast Ethernet"}, + { DELTA8139, "Delta Electronics 8139 10/100BaseTX"}, + { ADDTRON8139, "Addtron Technolgy 8139 10/100BaseTX"}, {0,}, }; + static struct pci_device_id rtl8139_pci_tbl[] __devinitdata = { {0x10ec, 0x8139, PCI_ANY_ID, PCI_ANY_ID, 0, 0, RTL8139 }, {0x10ec, 0x8138, PCI_ANY_ID, PCI_ANY_ID, 0, 0, RTL8139_CB }, {0x1113, 0x1211, PCI_ANY_ID, PCI_ANY_ID, 0, 0, SMC1211TX }, - {0x1113, 0x1211, PCI_ANY_ID, PCI_ANY_ID, 0, 0, MPX5030 }, +/* {0x1113, 0x1211, PCI_ANY_ID, PCI_ANY_ID, 0, 0, MPX5030 },*/ {0x1039, 0x0900, PCI_ANY_ID, PCI_ANY_ID, 0, 0, SIS900 }, {0x1039, 0x7016, PCI_ANY_ID, PCI_ANY_ID, 0, 0, SIS7016 }, + {0x1500, 0x1360, PCI_ANY_ID, PCI_ANY_ID, 0, 0, DELTA8139 }, + {0x4033, 0x1360, PCI_ANY_ID, PCI_ANY_ID, 0, 0, ADDTRON8139 }, {0,}, }; MODULE_DEVICE_TABLE (pci, rtl8139_pci_tbl); @@ -379,6 +388,12 @@ #define RTL_R32(reg) readl (ioaddr + reg) +static const u16 rtl8139_intr_mask = + PCIErr | PCSTimeout | RxUnderrun | RxOverflow | RxFIFOOver | + TxErr | TxOK | RxErr | RxOK; + + + static const char * __devinit rtl8139_name_from_chip (chip_t chip) { int i; @@ -897,9 +912,7 @@ RTL_W8 (ChipCmd, CmdRxEnb | CmdTxEnb); /* Enable all known interrupts by setting the interrupt mask. */ - RTL_W16 (IntrMask, - PCIErr | PCSTimeout | RxUnderrun | RxOverflow | RxFIFOOver - | TxErr | TxOK | RxErr | RxOK); + RTL_W16 (IntrMask, rtl8139_intr_mask); DPRINTK ("%s: rtl8139_open() ioaddr %#lx IRQ %d" " GP Pins %2.2x %s-duplex.\n", @@ -974,9 +987,7 @@ rtl8139_set_rx_mode (dev); RTL_W8 (ChipCmd, CmdRxEnb | CmdTxEnb); /* Enable all known interrupts by setting the interrupt mask. */ - RTL_W16 (IntrMask, - PCIErr | PCSTimeout | RxUnderrun | RxOverflow | RxFIFOOver - | TxErr | TxOK | RxErr | RxOK); + RTL_W16 (IntrMask, rtl8139_intr_mask); netif_start_queue (dev); @@ -1359,6 +1370,18 @@ printk (".\n"); #endif + /* E. Gill */ + /* Note from BSD driver: + * Here's a totally undocumented fact for you. When the + * RealTek chip is in the process of copying a packet into + * RAM for you, the length will be 0xfff0. If you spot a + * packet header with this value, you need to stop. The + * datasheet makes absolutely no mention of this and + * RealTek should be shot for this. + */ + if (rx_size == 0xfff0) + break; + if (rx_status & (RxBadSymbol | RxRunt | RxTooLong | RxCRCErr | RxBadAlign)) { @@ -1524,14 +1547,37 @@ spin_lock_irq (&tp->lock); + /* disable interrupt generation while handling this interrupt */ + RTL_W16 (IntrMask, 0x0000); + do { int status = readw (ioaddr + IntrStatus); /* Acknowledge all of the current interrupt sources ASAP, but an first get an additional status bit from CSCR. */ if (status & RxUnderrun) link_changed = readw (ioaddr + CSCR) & CSCR_LinkChangeBit; - RTL_W16 (IntrStatus, status); + /* E. Gill */ + /* In case of an RxFIFOOver we must also clear the RxOverflow + bit to avoid dropping frames for ever. Believe me, I got a + lot of troubles copying huge data (approximately 2 RxFIFOOver + errors per 1GB data transfer). + The following is written in the 'p-guide.pdf' file (RTL8139(A/B) + Programming guide V0.1, from 1999/1/15) on page 9 from REALTEC. + ----------------------------------------------------------- + 2. RxFIFOOvw handling: + When RxFIFOOvw occurs, all incoming packets are discarded. + Clear ISR(RxFIFOOvw) doesn't dismiss RxFIFOOvw event. To + dismiss RxFIFOOvw event, the ISR(RxBufOvw) must be written + with a '1'. + ----------------------------------------------------------- + Unfortunately I was not able to find any reason for the + RxFIFOOver error (I got the feeling this depends on the + CPU speed, lower CPU speed --> more errors). + After clearing the RxOverflow bit the transfer of the + packet was repeated and all data are error free transfered */ + RTL_W16 (IntrStatus, (status & RxFIFOOver) ? (status | RxOverflow) : status); + DPRINTK ("%s: interrupt status=%#4.4x new intstat=%#4.4x.\n", dev->name, status, readw (ioaddr + IntrStatus)); @@ -1565,11 +1611,14 @@ } } while (1); - DPRINTK ("%s: exiting interrupt, intr_status=%#4.4x.\n", - dev->name, readw (ioaddr + IntrStatus)); + /* Enable all known interrupts by setting the interrupt mask. */ + RTL_W16 (IntrMask, rtl8139_intr_mask); spin_unlock_irq (&tp->lock); - return; + + DPRINTK ("%s: exiting interrupt, intr_status=%#4.4x.\n", + dev->name, RTL_R16 (IntrStatus)); + } static int rtl8139_close (struct net_device *dev) --------------2988B51D83FAD47457F2DCAB-- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Fri Feb 18 10:39:55 2000 Date: Fri Feb 18 10:39:55 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: problems wiht rtl8139 Sorry it's taken me a few days to reply. It's better if you direct mail to the list rather than to just me. When you send mail to me alone, you're relying on me finding the time to respond, which isn't always easy. If you write to the list, whoever has free time can answer you. You'll get faster, and usually better answers that way. > Here is an attachment from cat proc/pci and about the other questions (if I > compiled the kernel myself) the answer is no. > My Linux distribution is Red Hat 6.0. During installation, the program asked > me about installing drivers for my adapter and I tryed the rtl8129 and > rtl8139 compatible but the system answered "Can't find such device". That's interesting. This is the relevant part of your /proc/pci: Bus 0, device 11, function 0: Ethernet controller: Unknown vendor Unknown device (rev 16). Vendor id=1. Device id=8139. Medium devsel. Fast back-to-back capable. IRQ 10. Master Capable. Latency=32. Min Gnt=32.Max Lat=64. I/O at 0xd000 [0xd001]. Non-prefetchable 32 bit memory at 0xe0800000 [0xe0800000]. Here's mine: Bus 0, device 10, function 0: Ethernet controller: Realtek 8139 (rev 16). Medium devsel. Fast back-to-back capable. IRQ 12. Master Capable. Latency=64. Min Gnt=32.Max Lat=64. I/O at 0xec00 [0xec01]. Non-prefetchable 32 bit memory at 0xea002000 [0xea002000]. For some reason your card is not being recognized as a Realtek device. The vendor id for your card is set to 1, which I don't think is correct. I don't know what Realtek's vendor id actually is off the top of my head, but, 1 seems unlikely. Unfortunately, I have no idea what to do about that. Maybe Donald can offer a suggestion. Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.com Fri Feb 18 11:42:29 2000 Date: Fri Feb 18 11:42:29 2000 From: Jeff Garzik jgarzik@mandrakesoft.com Subject: problems wiht rtl8139 sharkey@ale.physics.sunysb.edu wrote: > > Here is an attachment from cat proc/pci and about the other questions (if I > > compiled the kernel myself) the answer is no. > > My Linux distribution is Red Hat 6.0. During installation, the program asked > > me about installing drivers for my adapter and I tryed the rtl8129 and > > rtl8139 compatible but the system answered "Can't find such device". > > That's interesting. This is the relevant part of your /proc/pci: > > Bus 0, device 11, function 0: > Ethernet controller: Unknown vendor Unknown device (rev 16). > Vendor id=1. Device id=8139. > Medium devsel. Fast back-to-back capable. IRQ 10. Master Capable. Latency=32. Min Gnt=32.Max Lat=64. > I/O at 0xd000 [0xd001]. > Non-prefetchable 32 bit memory at 0xe0800000 [0xe0800000]. Your vendor id is 1, your I/O region size is 1, and your MMIO region size is zero. consider that the driver needs 0x80 bytes worth of registers, at least, it is not surprising that things do not work. Miguel, can you * download latest pciutils (2.1.5, listed on http://freshmeat.net/) * boot into a Linux 2.3.x kernel * e-mail the linux-realtek list with the output of "lspci -vvvxxx" If this command LOCKS YOUR COMPUTER, which it might, reboot and run "lspci -vvvx" Regards, Jeff -- Jeff Garzik | Writing software is Building 1024 | more fun than working. MandrakeSoft, Inc. | -Anon | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From andreas.kaltenbach@balcab.ch Fri Feb 18 13:58:45 2000 Date: Fri Feb 18 13:58:45 2000 From: akaltenbach andreas.kaltenbach@balcab.ch Subject: network is unreachable Halo, I have got your address through the following. boot log switching from suse 6.2 to 6.3 brought up this problem. Using / lib/modules/2.2.13/net/rtl8139.o rtl8139.c:v1.08 6/25/99 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html etho: RealTek RTL8139 Fast Ethernet at 0*9400, IRQ 11, 00:00:64:90:3d:79 PCI latency timer (CFLT) is unreasonably low at 0. Setting to 64 clocks. .... .... Setting up network device eth0 Setting up routing using /etc/route.conf Error Error while excuting /sbin/route add default gw ............ SIOCADDRT: Network is unreachable Setting up routing failed Starting syslog services can you tell me why I can't get connected to my network? All my settings are correct and I can't find the problem. IRQ 11 is shared with the graphic card thanks in advance Andreas Basel Switzerland | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Sat Feb 19 01:46:05 2000 Date: Sat Feb 19 01:46:05 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: network is unreachable > Using / lib/modules/2.2.13/net/rtl8139.o > rtl8139.c:v1.08 6/25/99 Donald Becker > http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html > http://cesdis.gsfc.nasa.gov/linux/drivers/rtl8139.html > etho: RealTek RTL8139 Fast Ethernet at 0*9400, IRQ 11, > 00:00:64:90:3d:79 > PCI latency timer (CFLT) is unreasonably low at 0. Setting to 64 clocks. > > .... > .... > Setting up network device eth0 It looks like the network card driver is set up and working properly. > Setting up routing > using /etc/route.conf > Error Error while excuting > /sbin/route add default gw ............ > SIOCADDRT: Network is unreachable That's a routing problem, not a network driver problem. That can happen if the gateway you specified is not reachable *within one hop*. Either you've specified the wrong gateway, the wrong network, or the wrong subnet mask. Without your routing table, it's hard to guess which. Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From daniel.kobras@student.uni-tuebingen.de Sun Feb 20 12:02:24 2000 Date: Sun Feb 20 12:02:24 2000 From: Daniel Kobras daniel.kobras@student.uni-tuebingen.de Subject: 8139too: wrong I/O size Hi Jeff! I found out why 8139too (as in 2.3.46) refused to load on my machine: You define RTL_IO_SIZE to 256 while my card only registers a window of 128 byte. As the larger size apparently works for you, I suspect there are different configs floating around. Defining RTL_IO_SIZE to 128 is one way to fix it, the patch below maybe is a bit cleaner. Regards, Daniel. --[snip]-- --- 8139too.c.orig Sun Feb 20 16:22:26 2000 +++ 8139too.c Sun Feb 20 16:52:14 2000 @@ -154,8 +154,6 @@ HAS_LNK_CHNG = 0x040000, }; -#define RTL_IO_SIZE 256 - #define RTL8139_CAPS HAS_CHIP_XCVR|HAS_LNK_CHNG typedef enum { @@ -243,6 +241,8 @@ PARA7c = 0x7c, /* Magic transceiver parameter register. */ }; +#define RTL_IO_WIDTH 0x7f + enum ChipCmdBits { CmdReset = 0x10, CmdRxEnb = 0x08, @@ -437,8 +437,8 @@ void *ioaddr = NULL; u8 tmp8; int rc; - u32 pio_start, pio_end, pio_flags; - u32 mmio_start, mmio_end, mmio_flags; + u32 pio_start, pio_end, pio_size, pio_flags; + u32 mmio_start, mmio_end, mmio_size, mmio_flags; DPRINTK ("ENTER\n"); @@ -449,12 +449,20 @@ pio_start = pci_resource_start (pdev, 0); pio_end = pci_resource_end (pdev, 0); + pio_size = pio_end - pio_start + 1; pio_flags = pci_resource_flags (pdev, 0); mmio_start = pci_resource_start (pdev, 1); mmio_end = pci_resource_end (pdev, 1); + mmio_size = mmio_end - mmio_start + 1; mmio_flags = pci_resource_flags (pdev, 1); + /* make sure IO window is large enough */ + if (pio_size <= RTL_IO_WIDTH || mmio_size <= RTL_IO_WIDTH) { + printk (KERN_ERR PFX "I/O window too small, aborting\n"); + return -ENODEV; + } + /* make sure PCI base addr 0 is PIO */ if (pio_start == 0 || pio_end <= pio_start || (!(pio_flags & IORESOURCE_IO))) { @@ -470,14 +478,14 @@ } /* make sure our PIO region in PCI space is available */ - if (!request_region (pio_start, RTL_IO_SIZE, RTL8139_MODULE_NAME)) { + if (!request_region (pio_start, pio_size, RTL8139_MODULE_NAME)) { printk (KERN_ERR PFX "no I/O resource available, aborting\n"); return -EBUSY; } /* make sure our MMIO region in PCI space is available */ - if (!request_mem_region (mmio_start, RTL_IO_SIZE, RTL8139_MODULE_NAME)) { - release_region (pio_start, RTL_IO_SIZE); + if (!request_mem_region (mmio_start, mmio_size, RTL8139_MODULE_NAME)) { + release_region (pio_start, pio_size); printk (KERN_ERR PFX "no mem resource available, aborting\n"); return -EBUSY; } @@ -487,7 +495,7 @@ pci_set_master (pdev); /* ioremap MMIO region */ - ioaddr = ioremap (mmio_start, RTL_IO_SIZE); + ioaddr = ioremap (mmio_start, mmio_size); if (ioaddr == NULL) { printk (KERN_ERR PFX "cannot remap MMIO, aborting\n"); rc = -EIO; @@ -523,8 +531,8 @@ err_out: if (ioaddr) iounmap (ioaddr); - release_region (pio_start, RTL_IO_SIZE); - release_mem_region (mmio_start, RTL_IO_SIZE); + release_region (pio_start, pio_size); + release_mem_region (mmio_start, mmio_size); DPRINTK ("EXIT, returning %d\n", rc); return rc; } @@ -641,8 +649,12 @@ unregister_netdev (dev); iounmap (np->mmio_addr); - release_region (pci_resource_start (pdev, 0), RTL_IO_SIZE); - release_mem_region (pci_resource_start (pdev, 1), RTL_IO_SIZE); + release_region (pci_resource_start (pdev, 0), + pci_resource_end (pdev, 0) - + pci_resource_start (pdev, 0) + 1); + release_mem_region (pci_resource_start (pdev, 1), + pci_resource_end (pdev, 1) - + pci_resource_start (pdev, 1) + 1); #ifndef RTL8139_NDEBUG /* poison memory before freeing */ @@ -1865,6 +1877,8 @@ printk (KERN_INFO RTL8139_DRIVER_NAME " loaded (%d device%s registered)\n", rc, rc > 1 ? "s" : ""); + } else { + pci_unregister_driver (&rtl8139_pci_driver); } DPRINTK ("EXIT\n"); | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.com Sun Feb 20 13:05:19 2000 Date: Sun Feb 20 13:05:19 2000 From: Jeff Garzik jgarzik@mandrakesoft.com Subject: 8139too: wrong I/O size > I found out why 8139too (as in 2.3.46) refused to load on my machine: > You define RTL_IO_SIZE to 256 while my card only registers a window of 128 > byte. As the larger size apparently works for you, I suspect there are > different configs floating around. Defining RTL_IO_SIZE to 128 is one way > to fix it, the patch below maybe is a bit cleaner. Thanks for the patch. I'm going to apply the fix but not the patch, as there are little picky things I want to clean up, like using pci_resource_len() to get the PCI region length in a portable manner. Your basic method for determining I/O size is correct though... If you look through the source, you'll notice that I have begun adding parts of your patch which will flush MMIO writes only when necessary. A patch to remove MMIO_FLUSH_AUDIT_COMPLETE, or works towards that, is definitely welcome. :) -- Jeff Garzik | I never vote for anyone. Building 1024 | I always vote against. MandrakeSoft, Inc. | -- W.C. Fields | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From jgarzik@mandrakesoft.com Tue Feb 22 19:52:08 2000 Date: Tue Feb 22 19:52:08 2000 From: Jeff Garzik jgarzik@mandrakesoft.com Subject: ANN: 8139too second public beta release Another release of 8139too is unleashed upon the world... If you missed the first announcement, 8139too is an improved (hacked) rtl8139 driver, with many improvements for stability and speed. Some people have reported a _10%_ performance improvement over the existing available drivers. Available at: http://gtf.org/garzik/drivers/8139too/ Changes, from README.txt: * Begin integration of Daniel Kobras' MMIO flush patch (disabled for now) * Softnet logic updates to fix bugs and improve performance * Dynamic sizing of I/O resources (0x80 for older chips, 0xFF for newer ones) * Remove bogus SiS entries from PCI probe table * Add support for cards "Delta Electronics 8139 10/100BaseTX" "Addtron Technolgy 8139 10/100BaseTX" * Fix major bug with rx ring buffer size (also present in rtl8139.c 1.08r) * PCI DMA mapping by Dave Miller * Complete rewrite of SMP locking logic * Hotplug support * Call rtl8139_hw_start from rtl8139_open, and remove duplicated code from rtl8139_open * Reset NWay registers to sane defaults on rtl8139_open/hw_start * Miscellaneous code cleanup -- Jeff Garzik | Building 1024 | Viva la open source! MandrakeSoft, Inc. | | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From golfking@purdue.edu Thu Feb 24 23:36:47 2000 Date: Thu Feb 24 23:36:47 2000 From: Kristopher King golfking@purdue.edu Subject: realtek-linux problems This is a multi-part message in MIME format. ------=_NextPart_000_0036_01BF7F20.2E664AC0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi, I am just new to linux. Previously I have used the Realtek 8139 in = win98 and it has worked just fine. But I just installed linux on a = second HD. I didn't install LILO. So I have to put in a LILO disk to = start up linux. But when I do my ethernet card doesn't work. I have a = cable modem which is connected to a hub and 3 computers come out of that = hub and grab 3 ips. My is one of them. I am pretty sure I have it = configured correctly. When using win98, I used a DCHP config to for it = to work-I didn't specify any ips. But in linux I gave all the = information it needed and it won't get online. =20 I have verified that the kernel module is set to rtl8139. This = automatically was set for me. but when I check the lights on the NIC = card they don't light up like they do in windows. I don't know what to = do?? The other problem is... I can't use my floppy in linux. It keeps giving = me errors when I try to access it. I logged in as superuser and gave = myself access but it still will not function. It can only read my LILO = disk. Need help??? thanks=20 kris ------=_NextPart_000_0036_01BF7F20.2E664AC0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi, I am just new to linux.  = Previously I have=20 used the Realtek 8139 in win98 and it has worked just fine.  But I = just=20 installed linux on a second HD.  I didn't install LILO.  So I = have to=20 put in a LILO disk to start up linux.  But when I do my = ethernet card=20 doesn't work.  I have a cable modem which is connected to a hub and = 3=20 computers come out of that hub and grab 3 ips.  My is one of = them.  I=20 am pretty sure I have it configured correctly.  When using win98, I = used a=20 DCHP config to for it to work-I didn't specify any ips.  But in = linux I=20 gave all the information it needed and it won't get online.  =
 
I have verified that the kernel module = is set to=20 rtl8139.  This automatically was set for me.  but when I check = the=20 lights on the NIC card they don't light up like they do in = windows.  I=20 don't know what to do??
 
The other problem is... I can't use my = floppy in=20 linux.  It keeps giving me errors when I try to access it.  I = logged=20 in as superuser and gave myself access but it still will not = function.  It=20 can only read my LILO disk.
 
Need help???
 
 
thanks
 
kris
------=_NextPart_000_0036_01BF7F20.2E664AC0-- | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Fri Feb 25 04:49:42 2000 Date: Fri Feb 25 04:49:42 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: realtek-linux problems Step 1) Don't send annoying HTML-mail to mailing lists. If you only want to send text, then just send text. Please do not send your messages both as text and HTML. > Hi, I am just new to linux. Previously I have used the Realtek 8139 in = > win98 and it has worked just fine. But I just installed linux on a = > second HD. I didn't install LILO. So I have to put in a LILO disk to = > start up linux. That's fine. > But when I do my ethernet card doesn't work. Can you be more specific? > I have a = > cable modem Lucky you. > which is connected to a hub and 3 computers come out of that = > hub and grab 3 ips. Your cable service provider allows multiple IP's from a single modem? > My is one of them. I am pretty sure I have it = > configured correctly. Can you be more specific? > When using win98, I used a DCHP config to for it = > to work-I didn't specify any ips. But in linux I gave all the = > information it needed and it won't get online. =20 What happens if you use DHCP under Linux? > I have verified that the kernel module is set to rtl8139. This = > automatically was set for me. but when I check the lights on the NIC = > card they don't light up like they do in windows. Not even the link light? > The other problem is... I can't use my floppy in linux. It keeps giving = > me errors when I try to access it. Can you be more specific? Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible. From sharkey@superk.physics.sunysb.edu Fri Feb 25 20:51:14 2000 Date: Fri Feb 25 20:51:14 2000 From: sharkey@superk.physics.sunysb.edu sharkey@superk.physics.sunysb.edu Subject: realtek-linux problems Please direct your responses to the mailing list, not to me personally. If I can't find time to reply, someone else might. > step 1) plain text is checked in the format menu buddy. relax. get over > it. Who's helping who here? > After I log into linux, I get "could not look up internett address for > CTXXXXX-A.lafaytX.in.home.com. This will prevent GNOME from operating > correctly. This can be fixed by /etc/hosts" So I looked in /etc/hosts, > there is no directory hosts. /etc/hosts is a file not a directory. http://linuxdoc.org/HOWTO/Net-HOWTO-5.html#ss5.5 And that error is not necessarily an indication of a hardware error. It may just be a DNS problem. Can you ping the other hosts on your LAN using raw IP addresses? > I went in to linuxconf and punch in all the info I have from the cable modem > company. I tried switching it to DCHP in linux and ethernet card FAILS on > boot up. *How* does it fail? Be specific. Quote error messages. > the link light on the hub is on. On boot up the light on the card is on > through the LILO process and starts blinking when it checks hardware(the > blinks are regular paced and bright full blinks, time spaced). I'm not really sure what that means. > The floppy just looks at the boot disk and no other disks. won't reconize a > disk. If that's all you can say, then, I'm afraid I can't help you. You need to be much more precise. Quote the exact text you see on the screen that tells you that things aren't working. Eric | To unsubscribe, send mail to Majordomo@cesdis.gsfc.nasa.gov, and within the | body of the mail, include only the text: | unsubscribe this-list-name youraddress@wherever.org | You will be unsubscribed as speedily as possible.