From: Digestifier To: Linux-Admin@senator-bedfellow.mit.edu Reply-To: Linux-Admin@senator-bedfellow.mit.edu Date: Sat, 10 Sep 94 19:13:25 EDT Subject: Linux-Admin Digest #47 Linux-Admin Digest #47, Volume #2 Sat, 10 Sep 94 19:13:25 EDT Contents: Re: [ALERT] Password problem with Linux (Mark A. Horton KA4YBR) Re: Drives with too many heads (Mark A. Horton KA4YBR) Doom HAS no pixel doubling (Byron Faber) Re: Newbie Inetd.conf & telnetd question (Miguel de Icaza) Setting up Sound Blaster Pro (Paul Julie) How send a break on a serial line? (Hans-Georg von Zezschwitz) Re: Doom HAS no pixel doubling (Christopher Wiles) Re: fsck during boot (Oz Dror) Just Getting Started (Richard Manganel) Re: DOMM 4 Linux /X is OUT !!! (Jan Fure) Re: 20 Mb RAM: Linux sees 16Mb (Robert Stockmann) Re: Multiple xterms in Seyon? (Mike Gaffney) Re: FTAPE...Im' soo close, yet so far???? (William Huang) ---------------------------------------------------------------------------- From: mah@ka4ybr.com (Mark A. Horton KA4YBR) Subject: Re: [ALERT] Password problem with Linux Date: Sat, 10 Sep 1994 10:05:07 GMT Craig Woodward (cdw@cci.com) wrote: : In article <34icuj$6gm@pandora.sdsu.edu>, : Tracy R. Reed wrote: : >James Beckwith (beckwith@io.org) wrote: : >: Any suggestions about using a cracker to check out new passwords before : >: saving them? : > : >There are a couple of utilities on sunsite that do this. One is called : >Crack, I don't remember what the other is called. : Using Crack in this manner is kind of akin to killing ants with : a hammer. Look for programs like uva.passwd. It may take a little patching : to work but it should be happy. I *think* I put my linux-patched version : of patch on ftp.csh.rit.edu:/pub/linux/system/? But it's not that big of a : switch. It checks for simple user info and scans for dictionary (and partial : match) passwords. IT takes a few seconds longer to change a password, but : it's happy. :) : -Woody Good suggestion, Woody... It's also a good idea to occasionally run crack and the cops programs to hunt for system security weaknesses that may been induced accidentally during normal administration. It takes a few iterations at first to get things all cleaned up, but then running it via cron every week or so will keep you on top of things. (hopefully BEFORE the weasels find the holes in the fences!) :) -- mark -- "Ask me about my OTHER operating systems ... SunOS, Solaris, SCO, AIX... On second thought, WHY?" ============================================================ Mark A. Horton ka4ybr mah@ka4ybr.atl.ga.us P.O. Box 747 Decatur GA US 30031-0747 mah@ka4ybr.com +1.404.371.0291 33 45 31 N / 084 16 59 W ------------------------------ From: mah@ka4ybr.com (Mark A. Horton KA4YBR) Subject: Re: Drives with too many heads Date: Sat, 10 Sep 1994 10:45:54 GMT Stephen Vance (srvance@unix.secs.oakland.edu) wrote: : I hate to post this subject again, and heartily apologize in advance; I did : not think that it would apply to me. I am trying to install Linux on a new : laptop (Slackware 1.2). Using the 'bare' boot disk, I get the message : hd.c: ST-506 interface disk with more than 16 heads detected : probably due to non-standard sector translation. Giving up. : (disk 1: cyl=524; sect=63; head=32) : I looked at the source and there is no explanation. This leaves me unable : to fdisk. Could someone *please* be kind enough to post or e-mail what I : need to do to get around this??? : Thanks in advance, : Steve Hello Steve, This is becoming quite common with the larger IDE drives. You have two options, (well, more than that really, but it gets messier!). First you can make Linux believe the DOSish CMOS description of the drive (a little complicated) or lose 12 MB of your disk and only use the first (REAL) 1024 cylinders (SIMPLE). The simple option first: SIMPLE: Boot the machine and interrupt the boot to go to CMOS configuration. On the (usually) base configuration screen you will have the disk types (hard drives) listed... yours may be called "user" or 47 for the type. This is where you will find the cylinders, heads, and sectors of the device listed. Only the number of sectors is sacred and MUST reflect reality! The drive you have is, in reality, a 1048 cylinder, 16 head, 63 sector device. The cylinders and heads have been remapped to satisfy the boot and DOS views of reality which cannot handle more than 1024 cylinders on a device. (10 bits for addressing... 2^10 = 1024) The simple fix is to set the cylinders to 1024 and the heads to 16, leaving sectors as 63. This will cost you 24 cylinders [or 24*(540/1048) MB], but will insure that everyone is looking at the same geometry. From then on, the install is simple and straightforward. NOT_SO_SIMPLE: Leave your CMOS set at cyl=524; sect=63; head=32. Boot the Linux boot kernel you've made. At the FIRST prompt, enter the REAL disk geometry (NOT what CMOS has set!). In this case, you would type: ramdisk hd=1048,16,63 Continue the boot process. When told to, replace the boot diskette with the root diskette you've chosen. After Linux comes up, login as root and invoke fdisk. Assuming this is the first IDE drive you are using, type in: fdisk /dev/hda When fdisk starts, press "p" to see the existing partitions. Delete any you DON'T need (hopefully in this case all of them!) Next go to expert mode (enter an "x") From the expert menu, change the cylinders (type in a "c") to the CMOS setting. Likewise for the number of heads (type in an "h"). Then return to the main menu and define your partitions as you would normally... with the parameters you"ve set fdisk to use as far as number of heads and cylinders. When done, save the new table via the "w" option. After the partition table has been updated, continue the install using "setup" and all should be fine. Note that when you wish to use fdisk again, you'll have to go through this drill all over to force it to buy what it sees. But now Linux, DOS, and CMOS are all seeing the same (distorted) view of reality! :) Please note that the above techniques work with ANY drive with more than 1024 cylinders, just adjust the numbers for cylinders and heads as appropriate... i.e. Reduce the number of cylinders via a factor to get it below 1024 and then multiply the number of heads by this same factor. DO NOT MESS WITH SECTORS! This is the one piece that actually has to be real! -- Mark -- "Linux! Guerrilla UNIX Development Venimus, Vidimus, Dolavimus." ============================================================ Mark A. Horton ka4ybr mah@ka4ybr.atl.ga.us P.O. Box 747 Decatur GA US 30031-0747 mah@ka4ybr.com +1.404.371.0291 33 45 31 N / 084 16 59 W ------------------------------ From: btf57346@uxa.cso.uiuc.edu (Byron Faber) Subject: Doom HAS no pixel doubling Date: 10 Sep 1994 15:41:59 GMT I was told by a friend (rumors rumors) that the pixel doubling, etc was NOT included because XFree can change resolutions on the fly. So for all you FOOOOOLS out there bitching. TURN DOWN YOUR RESOLUTIONS DUMMIES. public service announcement, disregard if you have any intelligence. bf -- Real programmers don't comment their code. It was hard to write, it should be hard to understand. b-faber@uiuc.edu & http://www.cen.uiuc.edu/~bf11620 ------------------------------ From: miguel@sphinx.nuclecu.unam.mx (Miguel de Icaza) Crossposted-To: comp.unix.admin Subject: Re: Newbie Inetd.conf & telnetd question Date: 08 Sep 1994 21:16:47 GMT : I am trying to setup our internet server to accept telnet connections at : a specific port and directly login to an account (so that the user does : not see a login prompt). Any ideas on how i go about doing this ? : I have seen it done, just don't see how to do it, have looked at the : inetd.conf file trying to figure it out... You need to: 1. Add a new entry in the /etc/services for the new service, like: my-telnet 9000/tcp 2. Add a new entry in the /etc/inetd.conf for the daemon that will receive the connection: my-telnet stream tcp nowait root /usr/sbin/mtel mtel Now, the mtel program is a modified in.telnetd that runs your desired login program instead of /bin/login. Hope this helps, Miguel. ------------------------------ From: pjulie@tse.com (Paul Julie) Subject: Setting up Sound Blaster Pro Date: Fri, 9 Sep 1994 19:26:39 GMT Thanks for all of your help....I finally got my swap file activated. (Thanks Joerg F, Mikey..etc) However, I have a problem configuring my sound card. I have read all of the doc to no avail. I started configuring my sound blaster for SB pro, did a make depend then make config. It suddently halted when it started compiling for sound!!!! Then I decided to compile the sound stuff myself.... and I had the following errors: darkstar:/usr/src/linux-1.0.9/drivers/sound# make cc -c soundcard.c In file included from sound_config.h:226, from soundcard.c:31: sound_calls.h:201: warning: `struct patmgr_info' declared inside parameter lis sound_calls.h:201: warning: its scope is only this definition or declaration, sound_calls.h:201: warning: which is probably not what you want. In file included from sound_config.h:227, from soundcard.c:31: dev_table.h:99: warning: `struct patmgr_info' declared inside parameter list dev_table.h:103: field `info' has incomplete type dev_table.h:121: field `info' has incomplete type make: *** [soundcard.o] Error 1 darkstar:/usr/src/linux-1.0.9/drivers/sound# IN fact the ultrasound.h and soundcard.h contained the following #include So I when into linux/soundcard.h and found this!!!! #include :-( Something wrong here, does anybody have the 2 correct files handy, they could send me. As an attachment or something. I really don't want to sound load the whole thing again. Or if I had too could you give me the filename and ftp site I could get only the sound stuff.(up-to-date of course) I need ultrasound.h and soundcard.h. I am using SlackWare 2.0 (FYI on diskette) Thanks again for all of your help, I am slowing getting this going!!! -- =================================================== The Toronto Stock Exchange Paul Julie - Applications Specialist Telephone: 1-416-947-4569 E-mail: pjulie@tse.com =================================================== ------------------------------ From: 1zezschw@rzdspc2.informatik.uni-hamburg.de (Hans-Georg von Zezschwitz) Subject: How send a break on a serial line? Date: 10 Sep 94 17:15:45 GMT Hello, I think I read most of the standart (RT)FM, but I cant find a answer to this problem: How can I send a break signal on a serial line? (in a C-program) And probably, somebody can tell me where to find information like this and more (e.g. the detailed use of fcntl and ioctl for programming the serial interface). My man-pages for those functions are very short, probably there is more detailed version out now. I'm trying to communicate with a medical monitor via Linux - you will not directly save lifes by helping me, but make me happy to continue my work. Thanks, Georg Zeschwitz ============================================================================== H. Georg v.Zezschwitz eMail: zeschwitz@uke.uni-hamburg.de Muenzstr. 10 1zezschw@informatik.uni-hamburg.de 20097 Hamburg THE DAY YOU BRING Voice: +49 40 2369342 OUT LINUX FOR WINDOWS I'LL BELIEVE IN YOU, BILL ------------------------------ From: a0017097@wsuaix.csc.wsu.edu (Christopher Wiles) Subject: Re: Doom HAS no pixel doubling Date: Sat, 10 Sep 1994 17:05:26 GMT btf57346@uxa.cso.uiuc.edu (Byron Faber) writes: : I was told by a friend (rumors rumors) that the pixel doubling, etc : was NOT included because XFree can change resolutions on the fly. Bullshit, sir. Check the README. : So for all you FOOOOOLS out there bitching. TURN DOWN YOUR RESOLUTIONS : DUMMIES. Know whereof you speak before you post such. : public service announcement, disregard if you have any intelligence. Disregard your advice if one has any intelligence? That's a thought ... -- Chris a0017097@wsuaix.csc.wsu.edu wileyc@halcyon.com wileyc@quark.chs.wa.com "... but I want to use all eight comm ports SIMULTANEOUSLY!" PGP 2.6 public key available by finger for the clinically paranoid. ------------------------------ From: dror@netcom.com (Oz Dror) Subject: Re: fsck during boot Date: Sat, 10 Sep 1994 18:22:00 GMT looking at the script in /etc/rc.d/rc.S ... # Test to see if the root partition is read-only, like it ought to be. READWRITE=no if echo -n >> "Testing filesystem status"; then rm -f "Testing filesystem status" READWRITE=yes fi # Check the integrity of all filesystems if [ ! $READWRITE = yes ]; then /sbin/fsck -A -a ... It seems that fsck -a -A is performed. If everything is as it should be i.e., the root file system is read-only. regardless of any boot or shutdown problem. -Why? -Oz -- NAME Oz Dror, Los Angeles, California SMAIL dror@netcom.com PHONE (213) 874-7978 Fax (213) 874-7965 ------------------------------ From: manganel@gate.net (Richard Manganel) Subject: Just Getting Started Date: 10 Sep 1994 21:42:08 GMT I've read enough about Linux to pique my interest but don't really know how to get started. I've downloaded some stuff but it all appears to be text material. Shouldn't there be a .com or .exe file like there is with Dos or Windows? Since I spend some time on a Unix network, I'm somewhat familiar with it. I do get a little confused about some of the files I see listed when I go to download from ftp site (tar, Z, etc.). Are they all handled by gunzip? Anything you can suggest to get Linux installed on my 386-33 would be appreciated. ------------------------------ From: fure@owlnet.rice.edu (Jan Fure) Subject: Re: DOMM 4 Linux /X is OUT !!! Date: 10 Sep 1994 19:49:51 GMT Hi: Thanks for many helpful replies pointing out that I probably downloaded linxdoom.tgz in ascii format instead of binary! This proved to be the case, so sorry for the wasted bandwith, and again, thanks to you who replied! (I guess I'm tainted as an ftp-newbie for life now :-) Jan ------------------------------ From: stock@dutsh7.tudelft.nl (Robert Stockmann) Subject: Re: 20 Mb RAM: Linux sees 16Mb Date: Sat, 10 Sep 1994 20:42:01 GMT In , Robert Stockmann (stock@dutsh7.tudelft.nl) wrote: : Now when I boot Linux (32-bit OS) I still get 'only' 16 Mb RAM : instead of 20Mb. YES I have disabled the 16 Mb RAM limit in the kernel: : bool 'Limit memory to low 16MB' CONFIG_MAX_16M n : What is happening here??? : My Machine: : HP 486 EISA 25MHz, with 20 Mbyte RAM, cache in 486 chip, : memory burstmode accessed by HP-designed onboard chip. : Adaptec Fast SCSI-2 1740 EISA controller : Fujitsu Fast SCSI-2 harddisk M2624F-512 : Toshiba Fast SCSI-2 CDROM drive XM-3401TA : ESDI harddisk controller with 150MB Micropolis drive : ATI Graphics Ultra PRO EISA (Mach 32) : Soundblaster PRO sound card : My OS: : Linux SLackware PRO 2.0 (kernel 1.0.9) XFree 2.1.1 Well I couldn't live with this! Spent mony for 4Mb for nothing? Noway!, So I thought lets look in the kernel sources, and it looked quit easy to hack it ....I added the following line in init/main.c: memory_end = 20*1024*1024; so that the resulting code looks like this: asmlinkage void start_kernel(void) { /* * Interrupts are still disabled. Do necessary setups, then * enable them */ set_call_gate(&default_ldt,lcall7); ROOT_DEV = ORIG_ROOT_DEV; drive_info = DRIVE_INFO; screen_info = SCREEN_INFO; aux_device_present = AUX_DEVICE_INFO; memory_end = (1<<20) + (EXT_MEM_K<<10); memory_end &= PAGE_MASK; ramdisk_size = RAMDISK_SIZE; copy_options(command_line,COMMAND_LINE); #ifdef CONFIG_MAX_16M if (memory_end > 16*1024*1024) memory_end = 16*1024*1024; #endif memory_end = 20*1024*1024; /* give me 20! */ if (MOUNT_ROOT_RDONLY) root_mountflags |= MS_RDONLY; if ((unsigned long)&end >= (1024*1024)) { memory_start = (unsigned long) &end; low_memory_start = PAGE_SIZE; } else { memory_start = 1024*1024; low_memory_start = (unsigned long) &end; } low_memory_start = PAGE_ALIGN(low_memory_start); memory_start = paging_init(memory_start,memory_end); if (strncmp((char*)0x0FFFD9, "EISA", 4) == 0) EISA_bus = 1; . . . . etc.. kernel boot-up messages are now: Console: colour EGA+ 80x25, 8 virtual consoles Serial driver version 3.99a with no serial options enabled tty00 at 0x03f8 (irq = 4) is a 16550A tty01 at 0x02f8 (irq = 3) is a 16550A lp_init: lp1 exists, using polling driver snd2 at 0x220 irq 5 drq 1 snd1 at 0x388 irq 0 drq 0 Calibrating delay loop.. ok - 12.42 BogoMips Configuring Adaptec at IO:4c80, IRQ 11 scsi0 : Adaptec 1740 scsi : 1 hosts. Vendor: FUJITSU Model: M2624F-512 Rev: 0405 Type: Direct-Access ANSI SCSI revision: 02 Detected scsi disk sda at scsi0, id 0, lun 0 Vendor: TOSHIBA Model: CD-ROM XM-3401TA Rev: 0283 Type: CD-ROM ANSI SCSI revision: 02 Detected scsi CD-ROM sr0 at scsi0, id 4, lun 0 scsi : detected 1 SCSI disk 0 tapes 1 CD-ROM drive total. Scd sectorsize = 2048 bytes Memory: 19196k/20480k available (560k kernel code, 384k reserved, 340k data) This processor honours the WP bit even when in supervisor mode. Good. Floppy drive(s): fd0 is 1.44M, fd1 is 1.2M Swansea University Computer Society Net2Debugged [1.30] IP Protocols: ICMP, UDP, TCP SLIP: version 0.7.5 (4 channels) CSLIP: code copyright 1989 Regents of the University of California Checking 386/387 coupling... Ok, fpu using exception 16 error reporting. Linux version 1.0.9 (root@mymachine) #4 Fri Sep 9 21:29:34 MET DST 1994 Partition check: sda: sda1 sda2 sda3 hda: hda1 VFS: Mounted root (ext2 filesystem). Adding Swap: 10236k swap-space However I still think its weird that the boot/setup.S routine can't get my full 20 Mb RAM... Robert Stockmann stock@dutsh7.tudelft.nl ------------------------------ From: mgaffney@newt.mrj.com (Mike Gaffney) Subject: Re: Multiple xterms in Seyon? Date: 10 Sep 1994 14:01:09 -0400 Term is definately the way to go. You will still need seyon to get the initial connection. After you are connected, just you will kick off term on the remote machine, suspend the seyon program with the "suspend" button, kick off term on your Linux box, and then start as many trsh sessions as you want. It's great! I highly recommend Mosaic+term too. Once I got term itself up Mosaic+term went fine the first time. MJG -- ============================================================================== These neural nuggets are mine, all mine. I wouldn't let my company take credit for them even if they _wanted_ to. ------------------------------ From: wyhuang@sdcc15.ucsd.edu (William Huang) Crossposted-To: comp.os.linux.misc Subject: Re: FTAPE...Im' soo close, yet so far???? Date: 9 Sep 94 02:51:44 GMT .... OK .... this must be a stupid question but after working for an hour I give up ... I just installed Slackware 2.0, kernel version 1.0.9, ftape version 0.9.10 (I think ... theres a *0.9.10.patch file in my /boot directory). It would not work unless I'm root. Then I chmod a+rw /dev/ftape from root. It then worked once, and then no more. There must be an obvious answer to this .... thanks in advance for your help. ------------------------------ ** 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 ******************************