From: Digestifier To: Linux-Development@senator-bedfellow.mit.edu Reply-To: Linux-Development@senator-bedfellow.mit.edu Date: Fri, 30 Sep 94 20:13:10 EDT Subject: Linux-Development Digest #246 Linux-Development Digest #246, Volume #2 Fri, 30 Sep 94 20:13:10 EDT Contents: Re: Try this IPX bridging code ... (Alan Cox) Re: File /lib/ld.so not found! (Mitchum DSouza) PCnet32 ("LANCE") PCI ethernet card with Linux. (Donald Becker) Re: TMC-850 on IRQ 11 no workee... (Ed Carp) Running out of sockets... (Alan Lenton) Ungerman-Bass Ether Cards (Alan Hannan) Re: xircom pocket ethernet support ??? (Gary Paul Gortmaker) Re: [STATUS] Linus Floppy Driver Development (Eric J. Schwertfeger) Q: Using a AIC 6360 VLB (Gerald Beuchelt) Re: AX25 & KISS Amateur Radio Protocols in Linux?? (Ian Logan KC5DAS) Re: Could TCP/IP be implemented over SCSI? (jbarrett@onramp.net) Re: how to install SCSI tape drive (Rob Janssen) Re: Could TCP/IP be implemented over SCSI? (Daniel Garcia) Re: i486 Word length, anyone? (Dan Pop) EXTREMELY ALPHA ARCnet drivers ready for testing (Avery Pennarun) Re: LOOK FIRST-- FORGED SPAM (Viktor T. Toth) Re: [?] DIP with auto-redial? (Harald Milz) ---------------------------------------------------------------------------- From: iialan@iifeak.swan.ac.uk (Alan Cox) Subject: Re: Try this IPX bridging code ... Date: Fri, 30 Sep 1994 10:12:45 GMT In article <1994Sep25.223539.260@acad.ursinus.edu> STEVO@acad.ursinus.edu (Steve Kneizys) writes: >If somebody wanted to isolate an IPX net/server from the main net >in terms of packet density but did not want to change the net >numbers, well, bridging would be an option! I may decide to add >it to my above bridge, as bridging is faster than routing. No bridging is normally slower than routing as you process more packets at the software level. Alan -- ..-----------,,----------------------------,,----------------------------,, // Alan Cox // iialan@www.linux.org.uk // GW4PTS@GB7SWN.#45.GBR.EU // ``----------'`----------------------------'`----------------------------'' ------------------------------ From: Mitchum.DSouza@mrc-apu.cam.ac.uk (Mitchum DSouza) Subject: Re: File /lib/ld.so not found! Date: 28 Sep 1994 10:10:34 GMT In article <1994Sep27.145421.18521@news.research.ptt.nl>, r.g.heller@research.ptt.nl (Ron Heller) writes: |> If the file /lib/ld.so does not exist, how or where can I get it??? Read the GCC-FAQ Question 44 on sunsite.unc.edu:/pub/Linux/docs/faqs Mitch ------------------------------ From: becker@cesdis.gsfc.nasa.gov (Donald Becker) Subject: PCnet32 ("LANCE") PCI ethernet card with Linux. Date: 29 Sep 1994 00:56:22 -0400 A few quick notes on getting PCnet32/PCI boards to work with Linux: 0. Use the new driver from my machine. It can be anonymously NFS mounted. beowulf.gsfc.nasa.gov:/linux/drivers/net/lance.c or 'ftp'ed beowulf.gsfc.nasa.gov:/pub/linux/drivers/lance.c The driver in 1.1.51 should work as well, but the earlier ones probably won't. 1a. Find the address that your BIOS put the ethernet card at. The way I did it was the running the following command as 'root': od -A x -t x /dev/port | grep 57 The board will appear with the ethernet address in the first six bytes, and '57' in bytes. 1b. Add this address to the probe list at the top of lance.c. [[ These steps won't be necessary when I integrate Drew's PCI-BIOS32 code into the new PCnet32 driver, but that's not done yet. ]] 2. Build and install a new kernel with the "LANCE" driver selected. 3. Configure your PCI slot for *edge triggered* interrupts. On the ASUS SP3G this means moving the jumpers rather than just changing the BIOS. If you don't do this you machine *will* hang when autoIRQ is done. With level-triggered interrupts you must clear the interrupt line or disable the interrupt. The current autoIRQ code doesn't do this. 4. The board will be detected as PCnet32/VLB instead of the PCnet32/PCI. This is harmless. (The driver follows the datasheet, but the hardware doesn't :^< ) 5. Ping away! -- Donald Becker becker@cesdis.gsfc.nasa.gov USRA-CESDIS, Center of Excellence in Space Data and Information Sciences. Code 930.5, Goddard Space Flight Center, Greenbelt, MD. 20771 301-286-0882 http://cesdis.gsfc.nasa.gov/pub/people/becker/whoiam.html ------------------------------ Crossposted-To: comp.os.linux.help,comp.os.linux.misc,comp.os.linux.questions From: ecarp@netcom.com (Ed Carp) Subject: Re: TMC-850 on IRQ 11 no workee... Date: Thu, 29 Sep 1994 12:36:37 GMT "... and the answer is..." Although several people responded to my post, no one really had an answer I hadn't tried before. :( In searching throught the SCSI-HOWTO, however, I *did* come across a note that proved to be the answer to the problem. Around line 840 or so, there appears three lines of code like thus: DATA = ... CONTROL = .... | (stuff...); Swapping the order from DATA, CONTROL to CONTROL, DATA worked. Thanks to all who emailed with suggestions!! -- Ed Carp, N7EKG Ed.Carp@linux.org, ecarp@netcom.com Finger ecarp@netcom.com for PGP 2.5 public key an88744@anon.penet.fi If you want magic, let go of your armor. Magic is so much stronger than steel! -- Richard Bach, "The Bridge Across Forever" ------------------------------ From: yeti@feddev.demon.co.uk (Alan Lenton) Subject: Running out of sockets... Reply-To: Alan@feddev.demon.co.uk Date: Fri, 30 Sep 1994 15:09:25 +0000 I have a couple of applications running that I suspect are 'leaking' sockets. First: are there any tools that will let me look at what sockets are allocated? Netstat only seems to show sockets that are connected to something. Second: (This might be the cause of the problem...) I am using shutdown(fd,2), to close the (TCP/IP) sockets, because I don't want them hanging around. Do I have to then do a close(fd) to actually get rid of the socket? All the documentation I have says that the shutdown is equivalent to the close but doesn't wait even if the protocol gives guaranteed delivery. It doesn't say anything about having to close the socket afterwards... Info: System is Slackware 2, with the kernal recompiled to allow for 512 sockets. There are three applications on this box, all designed to run continuously, and all using the socket interface. We have no idea which one (if it is only one!) is misbehaving! Any help gratefully received. Alan -- Alan Lenton - yeti@feddev.demon.co.uk ------------------------------ From: alan@mid.net (Alan Hannan) Subject: Ungerman-Bass Ether Cards Date: 30 Sep 1994 16:03:14 GMT I am interested in using an Ungerman Bass Ethernet card, model UB NIU PC/EOTP. However, I do not seem to be able to locate information regarding linux and this card. My searches have yielded nothing. If anyone has used, or knows about compatibility of this card, I would greatly appreciate it. It is possible that the card emulates a 3com or something else more predominant, and I would love to hear from you if you know anything about these Ungerman-Bass cards as they relate to Linux. Thanks! -- + alan@mid.net Network Operations Center (402)/472-0242, Fax (402)/472-0240 + + + + + + + + + + + + + + + + + + + + ++ + + + + + + + + + + + + + + + + + + + + +============\\ "Whenever people agree with me I always feel I must be wrong." + +/|/| * |\ \\____________________________ - Oscar Wilde + +|| | | |/ net (A Global Internet Company)| + + + + + + + + + + + + + + + + + + ------------------------------ From: gpg109@huxley.anu.edu.au (Gary Paul Gortmaker) Subject: Re: xircom pocket ethernet support ??? Date: 29 Sep 1994 15:36:12 +1000 root@nimir.demon.co.uk () writes: >Is anyone working on a driver for the Xircom parallel >port ethernet adaptor ???? No. >What is the difference between this and the dlink ???? D-Link supports open environments and gives out the low-level programming information needed to write a driver. Xircom doesn't. Some people have been able to return the Xircom parts via some sort of "fitness for purpose clause" - YMMV of course. Paul. ------------------------------ From: eric@pandora.Las-Vegas.NV.US (Eric J. Schwertfeger) Subject: Re: [STATUS] Linus Floppy Driver Development Date: 29 Sep 1994 14:15:52 GMT Alain Knaff (knaff@ngulu) wrote: : Yes, but what we need to know is not when the floppy has been : changed, but rather when it is inserted. The DISK CHANGE LINE is set : when the floppy is _removed_ not when it is inserted. It is cleared : when a disk is in the drive _and_ a seek is done. Thus, in order to : detect a disk _insertion_ the drive has to be seeked endlessly. If the : period of the poll is too small, the noise gets annoying, if it is too : long the feature gets useless. A trick that Amiga users could tell you about (causes problems with older drives, but should work because the Amiga drives were slightly modified 720K floppy drives) is stepping from track 0 to track -1. the result was either a silent drive that detects insertion, or a very noisy drive that could damage itself actually trying to step (older drives). : feature could not be used "as is" in Linux, because only one drive at : a time may be monitored that way, and many machines have several : drives attached to them. Moreover, the drive motor would have to be : spinning while waiting. For a backup program, these limitations are : ok, because it knows where to expect the disk, and because usually the : waiting period is not that long. No, the motor doesn't need to spin (at least on the Amiga), and you can poll different drives. I can't say for certain that this would work, but I do know of people that used normal IBM 1.44M drives on their Amiga to provide for GCR disks with 1.40M of data. ------------------------------ From: a2823ca@cd1.lrz-muenchen.de (Gerald Beuchelt) Subject: Q: Using a AIC 6360 VLB Date: 27 Sep 1994 05:01:46 GMT Hi. Does anybody have some experience using a AIC 6360 VLB, BIOS 1.01L with Linux? I got some problems using it (Kernel 1.1.50, i486-50 DX) Please mail to: beuchelt@lrz-muenchen.de Thanks, Gerald ------------------------------ From: ian@myhost.subdomain.domain (Ian Logan KC5DAS) Subject: Re: AX25 & KISS Amateur Radio Protocols in Linux?? Date: 30 Sep 1994 15:58:23 GMT Alan Cox (iialan@iifeak.swan.ac.uk) wrote: : In article bart@dunedin.es.co.nz (Bart Kindt) writes: : >Hi! Is any development going on to support AX25 and KISS protocols, so that : >Linux can be used as a Internet <> Amateur Radio Gateway (Just like KA9Q's : >NOS/NET program)? Hello, I'm on of the sysops for the NMSU gateway. We are currently working on switching over from our dos based gateway (crashes constantly.) If we ever get the routing to work on the Linux box (just a matter of time) then we will be running our gateway from Linux. We're using JNOS for Linux and some other stuff, just so you know. Ian KC5DAS ------------------------------ From: jbarrett@onramp.net Subject: Re: Could TCP/IP be implemented over SCSI? Date: Wed, 28 Sep 94 14:22:09 PDT writes: > > It is possible, yes. I've heard of it being done before. It's not > terribly practical though, because the maximum length of a scsi bus > is on the other of 6 meters, which doesn't give you much to work > with... You've also gotta be careful not to have the two systems > both try to access one drive at the same time. > 18-20 feet sounds OK to me... that will easily reach any machine installed in my 10x13 office... and then some... and SCSI2 extends the distance.... Drive access conflicts are no problem... the drive will return a busy status until the first access completes and the Linux driver has auto retry on busy (Nice SCSI Driver Code) so the request will complete after the other machine finishes it's access to the device.... small performace hit.. but no huhu So back to SCSI/NET: Drew Eckhart just mailed me and suggested that we take this thread to E-Mail or the SCSI channel on the Mailing List.. and I agree... cause I think we could hack this idea out in 4-6 weeks max.. (plus a packet driver for ASPI compatible SCSI devices on PC's maybe???) Someboby tell me where the mail list server is, and let's meet there (or should I start up a server here just for SCSI/NET fanatics???) John Barrett ------------------------------ Crossposted-To: comp.os.linux.admin From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: how to install SCSI tape drive Reply-To: pe1chl@rabo.nl Date: Fri, 30 Sep 1994 07:42:26 GMT In woody@hermes.stmarys-ca.edu (Woody Weaver) writes: >gentlebeings of cola and cold: Please don't post questions like this to .development! Your question is not related to it... >I can talk to the card fine: $ mount -t iso9660 /dev/sr0 /mnt >mounts the CD ROM normally, and gives me access. However, I don't have any >st0 devices. I tried >$ mknod /dev/st0 c 46 0 ^^ >(and a couple of other minor numbers, 8 and 128) but each time I try >$ mt -f /dev/st0 fsf >or whatever, I get "/dev/st0: No such device". Am I being particularly >dense? What is the fix? 46?? Who told you to use 46?? The devices for SCSI tape are: crw-r--r-- 1 root root 9, 128 Mar 19 1994 nrmt0 crw-r--r-- 1 root root 9, 0 Mar 19 1994 rmt0 Rob -- ========================================================================= | Rob Janssen | AMPRnet: rob@pe1chl.ampr.org | | e-mail: pe1chl@rabo.nl | AX.25 BBS: PE1CHL@PI8UTR.#UTR.NLD.EU | ========================================================================= ------------------------------ From: kender@leviathan.ccnet.com (Daniel Garcia) Subject: Re: Could TCP/IP be implemented over SCSI? Date: 30 Sep 1994 15:53:33 GMT Reply-To: kender@esu.edu Slaving away in a dark room, ianm@qualcomm.com (Ian McCloghrie) produced: >jason@idiom.berkeley.ca.us (Jason Venner) writes: >>So: given a little change in technology, networking via scsi will make a lot >>of sense. >It's a cool idea, and perhaps useful in a Desk-Area-Network concept >(put an xterminal on your system via scsi and use that as your main >display, etc). But it's not well suited as a general LAN technology. Actually, something liek this is currently being done w/Fibre Channel. There are currently standards for doing both IP _and_ SCSI over fibre channel, and I believe they are also considering both IP and SCSI on the same fabric (though, there, you get into some security problems). Someone I'm working with is considering writing a driver to do both IP and SCSI simultaneously out of a fibre channel port. I'm still trying to convince them to 'loan' me a PCI-FC and an EISE-FC card so that I can develop a driver for linux & osf/1 (pentium & alpha machine at home - heh - home area network - 256Mbps, connection to the net, 28.8Kbps ;). D -- ===========.,======Coming=soon=to=a=PhD=Program=near=you=====.,==Carpe=Diem=== Ethernet is||Daniel Garcia - ATP Group - LLNL Gigabit Testbed||AntiClipperCens for Sissies|| Lawrence Livermore Nat'l Labs - Livermore, CA ||orshipFightingF Disclaimer.||dgarcia@cohl.llnl.gov<-==->kender@eri.erinet.com||reedomLovingCru Try Linux || This .sig file (c)1994 by Daniel Garcia ||saderForRights. ----hi-----'`--PGP-key-avail-finger-kender@client.erinet.com-'`---Coram-Deo--- http://http.esu.edu/~kender <<==-- WWW --==>> http://client.erinet.com/~kender Fibre Channel - Unix - TCP/IP - Music - MIDI - Biking - Networking - Reading GCS/MU d? -p+ c++(----) l++ u+ e+(*) m+@ s/+ !n h f+@ !g w++ t++ r y? ------------------------------ From: danpop@cernapo.cern.ch (Dan Pop) Subject: Re: i486 Word length, anyone? Date: Fri, 30 Sep 1994 12:45:21 GMT In <36fan0$92c@nz12.rz.uni-karlsruhe.de> ig25@fg70.rz.uni-karlsruhe.de (Thomas Koenig) writes: >The registers of a 486SX also have 32 bits, it't the external >bus that's 16 bits. Wow, when did they change the size of the 486SX external bus? Last time I checked it was 32 bits. Or is the '4' in 486 a typo? Dan -- Dan Pop CERN, CN Division Email: danpop@cernapo.cern.ch Mail: CERN - PPE, Bat. 31 R-004, CH-1211 Geneve 23, Switzerland ------------------------------ From: apenwarr@tourism.807-city.on.ca (Avery Pennarun) Subject: EXTREMELY ALPHA ARCnet drivers ready for testing Date: 30 Sep 1994 01:53:50 GMT Yes, here it is... for those of us who have been waiting such a long time! I have written ARCnet drivers for Linux. Present testing version is 0.10 ALPHA. Here's the info. NOTE: If you are afraid of ALPHA drivers, then you should be especially afraid of this one. It is in early stages, and while it works quite well for ME, no one else I know has made these work. (of course, no one else I know has ARCnet cards :) NOTE: Installation is difficult at the moment. This is for two reasons: a) because I'm far too lazy to make diffs; and b) because this way only people who are reasonably computer literate will try to install the drivers. If you aren't, then please wait for a beta or a real release (which should be part of the kernel) NOTE: These drivers aren't compatible with ARCether for DOS, yet. Close, but not quite. Anyone interested in tweaking, go ahead, but please send me the patches so we can stay organized. NOTE: There are various other bugs. Please read the top of the arcnet.c file for more information. NOTE: This driver was tested only on kernels 1.1.45 and higher, so it may be in your best interests to upgrade to that if nothing seems to work. Installation ============ It's kind of a pain to install, since you have to edit a few kernel source files. Nothing too challenging if you know the least bit about C. First, fix /usr/src/linux/config.in to include a line for CONFIG_ARCNET. This will make sense when you look at the other netcard entries in the file, I hope :) Then you have to change /usr/src/linux/drivers/net/Space.c to probe for my arcnet at some point if CONFIG_ARCNET is defined. This also is easy if you look at the other entries. The function you need is arcnet_probe. Hmm... also you will need to edit /usr/src/linux/drivers/net/Makefile to link in my driver. Then it should be a simple matter of: make config make dep make clean make And you'll be fine :) Where do I get the drivers? =========================== Right now, they are ONLY available via e-mail to me. Send mail to one of the addresses listed in my signature with the word "ARCNET" at the beginning of the subject line, in all capitals. If the demand gets overwhelming (how likely is THAT?) I may upload to an FTP site. Right now I want to control release as much as possible, since the code may be quite dangerous (although it hasn't locked MY machine lately) and I don't want flames from people who don't know what they're doing... :( Where do I discuss the drivers? =============================== My news server is weird, so the NET channel of linux-activists may be the best idea. Or, via direct email to one of the addresses below, or even to this newsgroup if you're particularly lazy (but I can't guarantee I'll see it that way) When is a release expected? =========================== When enough people have tried the alpha and verified the abscence of bugs. This isn't likely to happen soon, since even _I_ know of some bugs; however, it does work reasonably well (ie. it IS usable for most things) and I know there is a great demand for ARCnet drivers, so I'm releasing this now. This is the first truly public announcement of the driver, so NO ONE BUT ME has tried it yet; this means it is still PRE-ALPHA as far as I am concerned. This is your last warning... :) -- Avery Pennarun 807-CITY - Technical Services / Information Provider Training Committee Internet: apenwarr@tourism.807-city.on.ca [OR] avery.pennarun@norlink.com ------------------------------ From: vttoth@vttoth.com (Viktor T. Toth) Crossposted-To: comp.os.linux.help,comp.os.minix Subject: Re: LOOK FIRST-- FORGED SPAM Date: Thu, 29 Sep 1994 21:44:12 In article mmendis@splinter.coe.neu.edu (Michael E. Mendis) writes: > If you have the time and ability to help us trace this, we can give you > info on who is doing it and maybe try to pressure HIS site to drop him. > >Now, lets be PC in this world. And I am not talking about Personal computers, >or Providence College. It should say IT'S. Hmmm. I would say let's be PC about our grammar and say ITS. ^^^^^ ^^^ Forgive me for this but even though I am NOT a native English speaker, I am sick and tired of seeing those who are make a mess of this beautiful language. Viktor ------------------------------ From: hm@seneca.ix.de (Harald Milz) Subject: Re: [?] DIP with auto-redial? Date: Wed, 28 Sep 1994 20:35:39 GMT Reply-To: hm@seneca.ix.de I posted recently that I found a way to cleanly return exitcodes from dip. I mirror some FTP directories via a temporary SLIP link every morning and like to start the SLIP/mirror process via a crontab entry, so returning a unique exitcode is mandatory to decide whether re-dialing is appropriate or not. I received a couple of requests to send the patch, and here's how I did it. It's fairly simple, though. Here's the patch to dip-3.3.7/command.c: ============================= snip ============================= --- command.c.old Fri May 6 08:03:56 1994 +++ command.c Fri May 6 08:31:19 1994 @@ -873,6 +874,7 @@ char *argv[32]; int argc, i; int running; + int exitcode=0; register char *sp; /* Initialize the command level module. */ @@ -913,9 +915,12 @@ if (commands[i].name != (char *)NULL) { errlevel = (*commands[i].func)(argc, argv); } else printf("? Unknown command (%s).\n", argv[0]); - } else running = 0; + } else { + running = 0; + exitcode= argc == 1 ? 0 : atoi(argv[1]); + } } while(running); /*NOTREACHED*/ - exit(0); + exit(exitcode); } ============================= snip ============================= The patch is relative to dip-3.3.7 which is not the hottest, but I found at least uri-dip-337d looks quite similar. The line numbers probably vary, but it's simple enough to apply manually. In the dip script, I now can say, ============================= snip ============================= [...] # Prepare for dialing. send ATZ\r\n wait OK 2 if $errlvl != 0 goto error1 send ATL0DTphone_number_1\r\n if $errlvl != 0 goto error2 wait CONNECT 60 if $errlvl == 0 goto login send ATL0DTphone_number_2\r\n if $errlvl != 0 goto error2 wait CONNECT 60 if $errlvl != 0 goto error3 # We are connected. Login to the system. login: wait ogin: 10 if $errlvl != 0 goto error4 send LOGIN\r\n wait ssword: 20 if $errlvl != 0 goto error5 send PASSWORD\r\n sleep 5 done: print CONNECTED to $remote with address $local mtu $mtu mode SLIP exit 0 error1: print Modem does not respond exit 1 error2: print dial failed exit 2 error3: print no CONNECT message within timeout exit 3 error4: print no login: message exit 4 error5: print no password: message exit 5 error6: print no SLIP string exit 6 ============================= snip ============================= In each of the cases, dip returns the respective exitcode to the calling shell. Re-dialing is only worth it if the exitcode is 3 which is also returned if the line is busy. In all other cases, I chose not to re-dial. The shell script calling mirror-2.1 looks like follows: ============================= snip ============================= #!/bin/sh # call mirror via SLIP to nathan MIRROR="/usr/local/lib/mirror/mirror" DIP="/usr/sbin/dip" DIP_PID="/etc/dip.pid" DIPSCRIPTS="/conf/net/slip" LOG="/tmp/mirror.SLIP.log" LOCK="/usr/local/lib/mirror/mirror.lock" PACKAGES="" SYSTEM=nathan CMDLINE="$0 $*" DATE() { date +"%d.%m.%y %H:%M:%S" } log() { echo $* echo `DATE` $* >> $LOG } [...] # call dip and evaluate exitcode; re-run via at(1) if necessary $DIP $DIPSCRIPTS/$SYSTEM.dip RESULT="$?" case "$RESULT" in "0") ;; "1") log "modem does not respond" exit $RESULT ;; "2") log "dial failed" exit $RESULT ;; # the next one applies for BUSY as well ... "3") echo "$CMDLINE" | at now + 2 minutes > /dev/null 2>&1 log "no CONNECT message within timeout" log "setting up at job \"$CMDLINE\"" exit $RESULT ;; "4") log "no login: message" exit $RESULT ;; "5") log "no password: message" exit $RESULT ;; "6") log "no SLIP string" exit $RESULT ;; *) log "unknown exitcode from dip: $RESULT" exit $RESULT ;; esac # we're connected now -> issue mirror command [...] ============================= snip ============================= That's it, folks. Ciao, hm -- Harald Milz (hm@seneca.ix.de) iX Multiuser Multitasking Magazine ------------------------------ ** FOR YOUR REFERENCE ** The service address, to which questions about the list itself and requests to be added to or deleted from it should be directed, is: Internet: Linux-Development-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.development) via: Internet: Linux-Development@NEWS-DIGESTS.MIT.EDU Linux may be obtained via one of these FTP sites: nic.funet.fi pub/OS/Linux tsx-11.mit.edu pub/linux sunsite.unc.edu pub/Linux End of Linux-Development Digest ****************************** .