From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Tue, 28 Sep 93 09:25:12 EDT Subject: Linux-Admin Digest #80 Linux-Admin Digest #80, Volume #1 Tue, 28 Sep 93 09:25:12 EDT Contents: X Setup... Half-solved. (Rick Miller) Re: [Q} HELP MULTIPORT SERIAL NOT WORKING (rich@mulvey.com) dial-in was working and now doesn't (Harry C. Johnson) [Q]: SLIP problem. (Yi-Tsun Chang) [Q]: How do I make SLIP connection with BOOTP under linux (Yi-Tsun Chang) Disk corrupted during Slackware install (Charles Kowal) Re: Linux Administration books (A Warren Pratten) SECURITY HOLE in SLS (anoyone can shutdown!) (Mike Jagdis) Re: [Not] enough SLS bashing (Eberhard Moenkeberg) options hardwired in kernel? (Maurizio Codogno) Re: Memory LEAKING!*=--.._ (Marino Ladavac) Re: SECURITY HOLE in SLS (anyone can shutdown!) [FIX] (Alan Cox) Re: ext2fs INODES (Alan Cox) Re: SMAIL: problems with MXed sites in SLS (Alan Cox) ---------------------------------------------------------------------------- From: rick@ee.uwm.edu (Rick Miller) Crossposted-To: comp.os.linux.help Subject: X Setup... Half-solved. Date: 27 Sep 1993 12:46:20 GMT Folks are *still* trying to tell me it's a permissions problem... It ISN'T. Others are *still* trying to tell me it's because my .profile doesn't get executed by xdm, but it DOES (I included it in every shell-script there was). It seems that no matter where I define LD_LIBRARY_PATH (and export it), the programs JUST DON'T SEEM TO BE SEARCHING THAT PATH! Why? Well, I solved the "more" problem by adding a line to my .profile: unset TERMCAP For some reason, the TERMCAP variable gets set to a bad value when I come in through xdm. I don't know why. I suspect that this has something to do with the fact that LD_LIBRARY_PATH isn't being searched. I've hacked a quick workaround by linking the shared X libraries into the root directory, but the fact that the binaries are looking for "//libXaw.so.3" makes me think the LD_LIBRARY_PATH variable isn't getting used properly. I suspect this was the reason the TERMCAP variable was getting set wrong since the /usr/lib/terminfo file is *not* linked up to root. I do *NOT* want to link all my libraries into my root directory. Can someone give me a clue why TAMU.99.12+ would be behaving this way? RICK MILLER Voice: +1 414 221-3403 P.O. BOX 1759 FAX: +1 414 221-4744 MILWAUKEE, WI Send a postcard and I'll send one back. 53150-1759 USA Sendu bildkarton kaj mi retrosendos unu. ------------------------------ Crossposted-To: comp.os.linux.help From: rich@mulvey.com Subject: Re: [Q} HELP MULTIPORT SERIAL NOT WORKING Date: Mon, 27 Sep 1993 14:45:01 GMT sysop (root@suburbia.apana.org.au) wrote: : Please help. : I have 2 four port cards in my linux machine, running : as public access unix machine. : Users connect fine, but if they don't logout proberly, : ie, turn their modem off mid session, then their jobs : don't get hung up ( ie, their process hang around. ) : I am running MCC pl10+. I got pl13, and installed that. : I am including relevant files below, tell me if you think : I am doing anything wrong. Thank you for posting thousands of lines of your configuration files, as well as the source to setserial.c. It was immensely helpful. Especially when I'm paying for my feed. I *LOVE* paying for stuff like that. Do I think you're doing anything wrong? Yes. How about showing a little bit of awareness of proper posting techniques? The one in question says that you *DON'T* post huge copies of source/configuration files. You *MAIL* them to the .00005% of the people who respond to you and say that they might have solutions, instead of clogging the net with otherwise useless data. Please keep this in mind in the future. - Rich -- Rich Mulvey Amateur Radio: N2VDS Rochester, NY rich@mulvey.com "Ignorance should be painful." ------------------------------ From: hcj@gull.uncc.edu (Harry C. Johnson) Subject: dial-in was working and now doesn't Date: Mon, 27 Sep 1993 17:40:04 GMT I am using uugetty(getty_ps?) and have the ringback feature enabled. We have a USR Internal Sportster 14.4 modem, and everything worked fine for a while. After restarting my computer however I could not get the modem to answer. I set the appropriate DIP switch so that ATZ would reset the modem to its factory defaults instead of the Non-volatile ram setups, and now the modem answers fine, but screws up in the getlogname() call. I have the debug option turned on in my uugetty.ttyS1 file, and the debug output doesn't report any errors. After the modems connect though I get a line of about 30 characters of garbage (though all the characters are the same). Anything I type is echoed as garbage. My uugetty.ttyS1 file is as follows, and worked for some time before the reboot. SYSTEM=nmrlab-pc DEBUG=777 VERSION=/proc/version LOGIN=/bin/login ISSUE=/etc/issue CLEAR=NO HANGUP=YES ALTLOCK=cua1 INITLINE=cua1 INIT="" ATZ\r OK TIMEOUT=60 RINGBACK=YES INTERRING=6 WAITFOR=RING DELAY=3 CONNECT="" ATA\r CONNECT "" \r \d\d ALTLINE=cua1 Any help would be greatly appreciated. One of our undergrads has been diagnosed with cancer but would like to continue with her research using her computer from home. The 14.4 line is to be her access to the campus. Thank you very much for your time. -Harry -- ============================================================================= Harry C. Johnson IV Phone: (704)547-4339 E-Mail: hcj@gull.uncc.edu Department of Chemistry University of North Carolina at Charlotte Charlotte, North Carolina 28223 "Seems to me, its all just chemistry..." -Neil Peart ============================================================================= ------------------------------ From: ytchang@magnus.acs.ohio-state.edu (Yi-Tsun Chang) Subject: [Q]: SLIP problem. Date: 27 Sep 1993 18:12:42 GMT Thanks in advance. Alex ------------------------------ From: ytchang@magnus.acs.ohio-state.edu (Yi-Tsun Chang) Subject: [Q]: How do I make SLIP connection with BOOTP under linux Date: 27 Sep 1993 19:03:45 GMT Here is the problem : I trid to use SLIP of my linux box to connect to the SLIP server of my school. After the connection, and entering the server's SLIP mode, the server only gives me my ip number but NO server's IP. Therefore, I can not use 'remote' in dip to set the remote's ip. This SLIP connection works well under DOS using NCSA telnet with BOOTP. How do I make a SLIP connection with BOOTP under linux. Thanks in advance. ------------------------------ Crossposted-To: comp.os.linux.help From: kowal@stsci.edu (Charles Kowal) Subject: Disk corrupted during Slackware install Date: Mon, 27 Sep 1993 18:37:54 GMT A few days ago, I tried to install the Slackware 1.0.3 distribution using the "test" 5.25 boot-root disks from ftp.cdrom.com. I wanted to put Linux on my Western Digital Caviar 2340 drive on /dev/hdb1. (/dev/hda is a WD 2220 which I use for MSDOS). The disk was already partitioned with the ext2fs file system from an earlier SLS installation (which was working okay). The Slackware fdisk said that I had "Native Linux" on the partitions. Everything went well until the end of the installation. I chose to install LILO, and there were no complaints, but I did not yet understand how to make a boot disk. When I rebooted the machine from the original boot disk, LILO did not appear. After "messing around" with LILO for a while, I decided to run fdisk and repartition the disk. After I did that, fdisk produced the message: "Partition 1 does not start on cylinder boundary: phys=(0,0,2) should be (0,1,1)" What does this mean, and what can I do about it? I re-partitioned the drive under MS-DOS, and it seemed to be fine. So I don't think that this is a hardware problem. I also booted from an old SLS disk and fdisk produced no warnings, but I am unable to re-install SLS. Back to Slackware - same warning message, no LILO, and unable to run 'dd' to make a boot disk. Could LILO have caused the problem? Is Slackware's fdisk at fault? And how can I *fix* it?!! I would appreciate any help. -Charles Kowal kowal@stsci.edu ------------------------------ From: warren@csd.uwo.ca (A Warren Pratten) Subject: Re: Linux Administration books Date: Mon, 27 Sep 1993 20:56:11 GMT In article , stock@dutsh7.tudelft.nl (Robert Stockmann) writes: |> |> A Good book is: |> "Unix System administration handbook" by Evi Nemeth, Garth Snyder and Scott |> Seebass. Prentice Hall , Englewood Cliffs, 1989 |> ISBN 0-13-933441-6 |> |> Robert Stockmann I agree completely. I bought this book about 4 months ago and have found it extremely useful for administrating Linux (as well as several other platforms). -- A Warren Pratten Warren@csd.uwo.ca Sender: warren@chaplin (A Warren Pratten) From: warren@chaplin.uwo.ca (A Warren Pratten) Path: chaplin!warren Newsgroups: comp.os.linux.admin Distribution: world Followup-To: References: <27vkn9$5k8@whale.st.usm.edu> Organization: Computer Science Department - Univ. of Western Ontario Subject: Re: Linux Administration books Keywords: UNIX administration, books In article , stock@dutsh7.tudelft.nl (Robert Stockmann) writes: |> |> A Good book is: |> "Unix System administration handbook" by Evi Nemeth, Garth Snyder and Scott |> Seebass. Prentice Hall , Englewood Cliffs, 1989 |> ISBN 0-13-933441-6 |> |> Robert Stockmann I agree completely. I bought this book about 4 months ago and have found it extremely useful for administrating Linux (as well as several other platforms). -- A Warren Pratten Warren@csd.uwo.ca -- A Warren Pratten Warren@csd.uwo.ca ------------------------------ From: jaggy@purplet.demon.co.uk (Mike Jagdis) Subject: SECURITY HOLE in SLS (anoyone can shutdown!) Date: Sun, 26 Sep 1993 15:42:00 +0000 * In message <27v8k9$4c2@winx03.informatik.uni-wuerzburg.de> Werner Gold said: WG> Even though I had a root password and a shadow entry the WG> passwd entry looked like this: WG> root::0:0::/root:/bin/sh WG> I think the passwd executable should check the contents of WG> the second column and add WG> an asterisk automatically. (I dont know who writes the WG> shadow routines - please think about it) The shadow suite is correct. It *does* create accounts which are locked as far as /etc/passwd is concerned. WG> Peter is not to blame for that, because the missing entry is WG> neccessary for making the installation without a root-password. Yes, Peter damn well *IS* to blame for that! When SLS first went shadow he deliberately broke it and has *never* *EVER* *bothered* to fix this glaringly stupid security hole despite the seemingly endless traffic it generates from people who blame every package under the sun - but *NOT* SLS itself for being terminally brain damaged! WG> This might be a stupid U*ix problem that anyone who read any WG> sysadmin book should know, WG> but how many machines are out there with a passwd entry like WG> above... It isn't a stupid Unix problem. It's a stupid SLS problem (and anything based on SLS - I see Slackware recently "discovered" it). I agree that any half competent Unix sysadmin should spot something like that. The fact that it has been in SLS so long should sound more than just the odd warning bell. Yeah, this is a flame against SLS :-). A significant proportion of traffic on the Linux newsgroups seems to be because SLS can't, or won't, make the effort to get it right. I can only assume that that is because they take more notice of their paying customers. i.e. people who don't have access to the net so can't download SLS for free and therefore aren't aware of the major security problems they have... Mike ------------------------------ Date: Wed, 22 Sep 93 08:53:51 +0100 From: Eberhard_Moenkeberg@p27.rollo.central.de (Eberhard Moenkeberg) Subject: Re: [Not] enough SLS bashing Hello Patrick and all others, on 20.09.93 Patrick J. Volkerding wrote to All in USENET.COMP.OS.LINUX.ADMIN: PJ> In the first place, I'm not required to improve anything until the day PJ> you start writing out my paychecks. It seems to be a law of "nature" that our maintainers of the really big distributions a) are a "first choice" target of stupid annoying b) can't avoid feeling that and reacting accordingly. Patrick, please do not get as upset as pmacdona seems to be. :-> It more and more seems to me that SLS and Slackware (and mcc and tamu too, if more people get "interested") are synonyms for PBS ("pearls before swine"). Obey the pearls before the swine. Greetings ... Eberhard ------------------------------ From: mcodogno@nyx.cs.du.edu (Maurizio Codogno) Subject: options hardwired in kernel? Date: Tue, 28 Sep 93 09:28:17 GMT I was trying to recompile the kernel under SLS 1.03, to get rid of some option I don't plan to ever use (I always do the same in office with my Suns). Unfortunately I noticed that sbpcd.c *wants* that the SoundBlaster CD-ROM option is enabled (eventually I deleted all references to the file :-) ) and when building kernel.o it complains for lack of _sys_quotactl symbol or something like that (and I had not yet found a workaround) Is it just a problem with SLS, or quota should be present in all kernels? I prefer to get rid of it, if else for philosophical reasons - I am the only one who use my machine. -- ciao! .mau. ===== Maurizio Codogno - CSELT UF/DU dept. - Torino Italy "home" email: mau@beatles.cselt.stet.it ------------------------------ Crossposted-To: comp.os.linux.help,comp.os.linux From: lan_lada@rcvie.co.at (Marino Ladavac) Subject: Re: Memory LEAKING!*=--.._ Date: Tue, 28 Sep 1993 10:15:45 GMT : jP@hpacv.com writes: : : >Hello! : > Just installed SLS 0.99.12 and hooked that baby right up to the : >net. All went great and life was good UNTIL I did a top or a free. : > Here output RIGHT after bootup! HELP! I'm missing 15 meg! : > Check this out........... : : >Here's the free output: : : > total used free shared buffers : >Mem: 14964 13792 1172 1856 10696 : >Swap: 0 0 0 Slightly on a tangent, free reports only 13800 K total on my 16 M machine. Its .99pl10, btw. Does anyone know where did the rest go (I know that 1 meg is kernel text plus IO space; I'm still missing at least a megabyte?) Included in hte kernel: SYSV IPC, ext2, minix, dos, no SCSI, inet, no soundcards. Thanks, /Alby. BTW, I read I have to enable IPC during boot. How? -- Proof by Intimidation: "I'm bigger, therefore I'm right." noone@nowhere.in.particular ------------------------------ From: iiitac@swan.pyr (Alan Cox) Subject: Re: SECURITY HOLE in SLS (anyone can shutdown!) [FIX] Date: Tue, 28 Sep 1993 12:29:09 GMT In article clau@acs.ucalgary.ca (Christopher Lau) writes: >> >> Ouch! This worked with Slackware, too. I don't know about other >> distributions. >> > >Hmm.. this is a bug in NET2 that doesn't exist in NET1 (or whatever the >pre-pl12 networking utilities are known as..).. Another reason not to >use NET2 for at least another revision or two.. > Garbage. It's because tbe people who put SLS and Slackware together 1) Didnt set the system up with all shadowed passwords as :x: instead of :: 2) Used an in.rshd and in.rlogind built for non shadow password systems. For once 8-) the network software is fine. Alan ------------------------------ From: iiitac@swan.pyr (Alan Cox) Subject: Re: ext2fs INODES Date: Tue, 28 Sep 1993 12:33:47 GMT In article <1993Sep27.012713.628@bhhome.ci.net> bill@bhhome.ci.net (Bill Heiser) writes: >Is it just 'cpio' that is broken, ie. can't handle ext2fs's inode >numbers? It seems to me that it probably doesn't matter what CPIO >does with inode numbers ... since if I use cpio to Extract a file >(or files) from the backup, it will generate a new inode number when >it restores the file to the filesystem, no? CPIO is an old braindead format that uses 16 bits for inode numbers. Your machine has over 65535 inodes so CPIO can't store inode numbers correctly. ALan iiitac@pyr.swan.ac.uk ------------------------------ From: iiitac@swan.pyr (Alan Cox) Subject: Re: SMAIL: problems with MXed sites in SLS Date: Tue, 28 Sep 1993 12:32:09 GMT In article vandi@garfield.catt.ncsu.edu (Brian Vandivier) writes: >I'm having the same problem in which I get mail back saying that my >smart host can not be found. I have no problem pinging the smart host. >I've tried several different hosts and all return the same thing. I can >send mail to machines, but not to domains names. Could someone post >about a solution to this. I'm using the smail that comes with SLS 1.03 > For internet connected sites the best answer is to install BSD sendmail instead of smail. Alan iiitac@pyr.swan.ac.uk ------------------------------ ** 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-Admin-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.admin) via: Internet: Linux-Admin@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-Admin Digest ******************************