From: Digestifier To: Linux-Activists@senator-bedfellow.mit.edu Reply-To: Linux-Activists@senator-bedfellow.mit.edu Date: Sun, 15 Aug 93 20:13:12 EDT Subject: Linux-Activists Digest #120 Linux-Activists Digest #120, Volume #6 Sun, 15 Aug 93 20:13:12 EDT Contents: Re: why Linux does not like SIGBUS (Yonik Christopher Seeley) Re: Curses:where are nodelay() and keypad()? (Zeyd M. Ben-Halim) 99p12 - SLIP and DOS-ka9q anyone? (Michael Will) Disk full unless root login...? (Maz H. Spork) xfree (Rebel Lion) Re: NetBSD's ash as /bin/sh substitute on Linux (Arjan de Vet) linux patch level 10 (William J. Bitner) Problem compiling wih gcc from SLS 1.03 (Mark Hessling) Re: Disk full unless root login...? (Michael Griffith) is ne2100 compatible supported? (Radivoj Suvacarov) Wouldn't it be nice..... (Berk Walker) My boot floppy doesn't work (yes, I read the FAQ) (Dave A. B. Dobson) Re: How to set IRQ in Linux (kalle@dg8lav.toppoint.de) Re: why Linux does not like SIGBUS (Maurice S Barnum) Re: Jove -- Johnathan Own Version of Emacs? (Jim Nelson) disk corruption with e2fs (Jason Haar) Re: Reading and writing M (Simon Hobson) xdaliclock on Linux (Andrew Tefft) stale LCK file after uucp session (JL Gomez) ---------------------------------------------------------------------------- From: yseeley@leland.Stanford.EDU (Yonik Christopher Seeley) Subject: Re: why Linux does not like SIGBUS Date: Sun, 15 Aug 93 20:07:48 GMT In article <93227.204302K111114@ALIJKU11.BITNET> writes: > >the 386/486 CPU does! have an option to enable alignment checking, this is >set in the CPU-flags, it is the AC bit. the reason, why one would like to do >alignment checking is the speed of load operations, Don't you mean that aligning data on word boundaries will make the data access faster? I don't see why *checking* for this alignment would speed things up. Seems like that there is a choice to either align your data or not, but why enforce one way? If you want it to be faster than align! I think that gcc does align on word boundaries in all of the normal cases. > loading a long on a non >byte boundary takes additional memory cycles. I'm curious if Linux does >alignment checking too? (can someone answer this?), and just has no SIGBUS, >or if it does no a.c., then I'd be interested how many apps. would die :-) > >a very quick and dirty fix I use for the SIGBUS is to say: -DSIGBUS=SIGKILL >when compiling sources. (i try have sources unmodified...) > >regards, herp - Yonik Seeley yseeley@cs.stanford.edu ------------------------------ From: zmbenhal@netcom.com (Zeyd M. Ben-Halim) Subject: Re: Curses:where are nodelay() and keypad()? Date: Sun, 15 Aug 1993 20:04:31 GMT In article <1993Aug14.224150.705@tctube.spb.su> mbravo@tctube.spb.su (Michael E. Bravo) writes: >erc@apple.com (Ed Carp) writes: > >>are SYSV'isms). On the other hand, ncurses supports both routines. >>Available at the linux FTP server nearest you. :) > >Well, just had a problem related to this thread... > >I have ncurses which came with SLS 1.01 >With it, I made a small program which does only initscr(), refresh() and >then closes curses and exits. On console or VCs, it runs just fine. On a >terminal connected to serial port (VT100 clone) it coredumps. Some tinkering >with gdb showed that it crashes in refresh, trying to execute >mvcur(0,0,0,0). Any ideas? The version supplied by SLS is old, plus it looks like it doesn't have terminfo entry for vt100 (which is the cause of your problem). You can try getting 0.7.3 from netcom.com:pub/zmbenhal or wait a week or so for 1.8 to come out officially. Zeyd >-- >Michael E. Bravo AKA /\/\ike 7 812 231 3951 (home) > The Communication Tube mbravo@tctube.spb.su ------------------------------ From: michaelw@desaster.hanse.de (Michael Will) Subject: 99p12 - SLIP and DOS-ka9q anyone? Date: Fri, 13 Aug 1993 23:37:00 GMT Hello, I tried today to connect my linux via SLIP to an i286 with ka9q, but somehow I did not succeed. I do not know to much about ka9q, most of the documentation refered to the packet-radio-stuff, and so... I managed to get something onto the wire from both sides, I tested it with terminalprograms each. (connected via nullmodem) Only a ping never got ansered as far as I could see, and ofcourse nothing else. On ka9q I took pk0 as device, on linux it is sl0. On linux I did: dip -t remote 1.2.3.4 port cua1 speed 38400 mode SLIP On ka9q - as far as I remember - I did an attach asy 0x03f8 4 pk0 2048 1500 4096 38400 (I do not remember the exact syntax). Doing a ping on ka9q did come back immediatly without answer - only the local ip got a positive res=55 or something... What Do I have to do to get the beasts running together? What have I done utterly wrong? Every help welcome, please mail me. Cheers, Michael Will -- Michael Will Linux - share and enjoy :-) Life is not there if you can't share it... Hazel'O'Connor Breaking Glass Happily using Linux 0.99p12 with X11R5, \LaTeX, cnews/nn/uucp and: PGP! >>> Ask for Linux and / or pgp-Information <<< ------------------------------ From: halgrim@diku.dk (Maz H. Spork) Subject: Disk full unless root login...? Date: Sun, 15 Aug 1993 20:37:51 GMT Except when logged in as root, no disk writes are acceptable on my Linux .99. Any clues? Maz (halgrim@diku.dk) ------------------------------ From: rlion@access.digex.net (Rebel Lion) Subject: xfree Date: 15 Aug 1993 16:45:49 -0400 whats the group to post to with xfree86 questions? rl ------------------------------ From: devet@adv.win.tue.nl (Arjan de Vet) Crossposted-To: comp.unix.shell Subject: Re: NetBSD's ash as /bin/sh substitute on Linux Date: 15 Aug 1993 21:47:18 +0200 In article , Kevin Brown wrote: >In article <24bucm$43@adv.win.tue.nl> devet@adv.win.tue.nl (Arjan de Vet) >writes: >>Sunando send me his sources and I must say this version of ash works quite >>well under Linux after using Sunando's new Makefile (the original Makefile >>was absolutely unusable under Linux). > >You need Berkeley's "pmake" to use the supplied makefile. I ported pmake >to System V a while back, so it wasn't a problem for me... I see it has just been ported to Linux :-) >>It runs indeed configure scripts very well except it cannot do `cd .' :-) > >Interesting. I hadn't tried that...:-) Not the first test you think of indeed :-) Seems to be a Linux-only bug. >>It has however some serious bug (I think): many shell scripts from INN and >>smail use commands in backquotes (`date`). When running these scripts from >>the command line, they work fine, but when run from crond they hang at the >>first `...` command, consuming 100% CPU time. This is also the case for >>/etc/rc scripts. I started using debugging traces but haven't been able >>yet to find the problem. > >I found *something* that behaved like this very quickly: I sic'ed C-news >on ash. :-) C-news has this wonderful ability to expose the bugs in >any Bourne-style shell you care to name. :-) From experiences with bash 1.12 I know that also Configure scripts (e.g., the one from trn) and INN's shell scripts are also *extremely* good in exercising Bourne-compatible shells. >>When this bug has been fixed, ash will be a good (and small: 66K) /bin/sh >>substitute for Linux. > >Indeed. I wonder if it's the only bug, though. Probably not... > >Does *anyone* know where there's a comprehensive test suite for Bourne- >compatible shells? Cnews/Configure/INN. But a real test would be very nice. Arjan -- Arjan de Vet (home) Eindhoven University of Technology, the Netherlands (work) ------------------------------ From: bitner@clear.com (William J. Bitner) Subject: linux patch level 10 Reply-To: bitner@clear.com (William J. Bitner) Date: Fri, 13 Aug 1993 19:49:06 GMT I'm hoping someone can help by e-mailing the patch level 10 for linux to me. I'd like to upgrade compilers, (from 2.3.3) but the compiler FAQ say I need the newer library which in turn the FAQ says I need linux.99pl10. I only have access to uunet archives and my most current listings only reveal available patches for p9 and p11. SO, I'd really like p10. If some nice soul could uuencode and e-mail p10 (only the patches), I'd really appreciate it. Email to bitner@clear.com Bill Bitner ------------------------------ From: mark@snark.itc.gu.edu.au (Mark Hessling) Subject: Problem compiling wih gcc from SLS 1.03 Date: 15 Aug 93 21:11:05 GMT I am having problems with gcc 2.4.5 installed from SLS 1.03. The compiler complains about not being able to include stddef.h from within stdlib.h. stddef.h resides in /usr/include/linux, but the precompiler does not include this path when searching for header files. If I then specify -I/usr/include/linux on the compile line, stddef.h does get included correctly. The next error is that float.h does not exist anywhere. This is also included from within stdlib.h. If I create a dummy float.h and try to recompile, I get a syntax error in stdlib.h complaining about wchar_t being undefined. This is supposed to be defined in stddef.h but isn't. I have reinstalled the full c series to see if that fixed the problem; it didn't. Have I done something silly, or has anyone else had the same problems ? 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: grif@ucrengr.ucr.edu (Michael Griffith) Subject: Re: Disk full unless root login...? Date: 15 Aug 93 21:18:21 GMT In article <1993Aug15.203751.27724@odin.diku.dk> halgrim@diku.dk (Maz H. Spork) writes: > >Except when logged in as root, no disk writes are acceptable >on my Linux .99. > >Any clues? Under UNIX, usually the last 10% of the disk is reserved for root. On most filesystems, performance suffers considerably if it is over 90% full. So, when df reports 100%, that is 100% of the user portion, not of the whole filesystem. As root, you can write to the last 10%, if you really have to. It is advised that you delete or compress something, since using up the reserved portion of a filesystem is a risky practice. Followups to comp.os.linux.help. -- Michael A. Griffith grif@cs.ucr.edu ------------------------------ From: suvac@news.cs.andrews.edu (Radivoj Suvacarov) Subject: is ne2100 compatible supported? Date: Sun, 15 Aug 1993 20:40:23 GMT The title says it all. I need to know if ne2100 ethernet boards are supported on Linux. Regards, Radivoj -- ***************************************************************** * Radivoj Suvacarov * (616) 471-6882 (home) * * 500 Garland C-4 * 471-3251 (office) * * Berrien Springs MI 49103 * suvac@andrews.edu * ------------------------------ From: berk@techsys.mcws.fidonet.org (Berk Walker) Subject: Wouldn't it be nice..... Date: Sun, 15 Aug 93 11:10:52 PDT ... if I could convert my Xenix _drivers_ to work under Linux? Like my multiport boards, etc? Maybe when 1.0 or so arrives, some soul will look towards that. A note to those whose SLS had no man pages on e2fsck and have their boot/root on a partition that needs it.... run "shutdown now" (placing you in single user mode) and then "e2fsck -av /dev/hda1"<-1 in my case.. your part. may vary. There are probably wiser switches, but it seems to work. bbw -- Berk Walker berk@techsys.mcws.fidonet.org Member: bill nelson Fan Club ------------------------------ From: dave@phonon.uwaterloo.ca (Dave A. B. Dobson) Subject: My boot floppy doesn't work (yes, I read the FAQ) Date: Sun, 15 Aug 1993 21:56:16 GMT I have SLS 1.01 and it installed fine. I have been trying to upgrade to linux-0.99.pl11 (and at the same time libc-4.4.1 and gcc-2.4.5). Yesterday I got everything compiled ok, and did a make zdisk to copy the pl11 kernel to a floppy. I tried rebooting using this floppy and I got the message: VFS: Insert root floppy and press ENTER So, I pressed enter and got: kernel panic: VFS: Unable to mount root Using HJ Lu's rootdisk, I could boot linux and mount my hard disk. I looked at the zImage of the new kernel and it specified the root disk as the floppy (that's what I specifed as ROOTDEV in the Makefile). I tried changing it using rdev to my hard disk (/dev/hda2) and then copying that to the floppy, but all it did was reboot the computer right after Loading..... Apparently I'm missing something, and section III.18 of the FAQ is not any help. BTW: I tried using the old /Image but that didn't work either. Could someone tell me how to boot from a floppy to Linux on my /dev/hda2 ? Thanks, Dave -- =================================================================== Dave Dobson dave@phonon.uwaterloo.ca or dave@eos.uwaterloo.ca University of Waterloo, Electrical & Computer Engineering, Silicon Devices and Integrated Circuits Research Group (SiDIC) ------------------------------ From: kalle@dg8lav.toppoint.de Subject: Re: How to set IRQ in Linux Date: Sun, 15 Aug 1993 21:02:16 GMT Tobias Thompson (odysee@csd.uu.se) wrote: : My problem is that I must have com1-3 in DOS/WINDOWS/Linux and have put : my modem (COM3) at irq 5. Is there a simple solution or do I have to do : some programming? Any advice is useful. Use SETSERIAL 2.x for this job. I use it all the time with 4 serial and com3 on irq 5, com4 on 2. You will find this small programm on any host. (hopefully) .-) Kalle -- * - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - * | Karlheinz Hagen, | IP-Addr: 44.130.5.11 | | Katenweg 8 | Email : kalle@dg8lav.toppoint.de | | 24867 Dannewerk | AmprNet: dg8lav%db0sef@db0hhn.ampr.org| ------------------------------ From: msb@cats.ucsc.edu (Maurice S Barnum) Subject: Re: why Linux does not like SIGBUS Date: 15 Aug 1993 22:22:47 GMT In <93227.204302K111114@ALIJKU11.BITNET> writes: >a very quick and dirty fix I use for the SIGBUS is to say: -DSIGBUS=SIGKILL >when compiling sources. (i try have sources unmodified...) I use -DSIGBUS=SIGUNUSED, personally. And I'm thinking of modifying to have that for me. -- Maurice S. Barnum The errors of great men are venerable msb@cats.ucsc.edu becauxse they are more fruitful than mbarnum@eis.calstate.edu the truths of little men.... mbarnum@nyx.cs.du.edu --- F. Nietzsche PGP key: finger mbarnum@nyx.cs.du.edu or msb@ucscb.ucsc.edu ------------------------------ From: jnelson@plains.NoDak.edu (Jim Nelson) Subject: Re: Jove -- Johnathan Own Version of Emacs? Date: Sun, 15 Aug 1993 22:25:18 GMT In article ellis@nova.gmi.edu (R. Stewart Ellis) writes: #arcardw@indsvax1.indstate.edu (Paul Cardwell) writes: # # #> Has anyone ported this to Linux (or tried to compile it?)? I use it on #>our unix machine here at the University -- nice! If you have any other good #>editors (something smaller than Emacs), please respond to my email. #The jove code is being reorganized to make it more portable. The latest #released version is I believe 4.14.10, at relay.cs.toronto.edu:/pub/moraes. If don't remember where i go it, but I got jove 4.14.7 to compile under patchlevel 9 and 11 with almost no troubles. I haven't tried any newer versions yet. -- Jim, in the Land of the Lost. |Disclaimer: These are probably ObQuote: Do Not Meddle in the Affairs of | opinions. I'm probably not Wizards, For You are Crunchy, | supposed to have any. and Good with Ketchup. | Blessed Be! ------------------------------ From: Jason Haar Subject: disk corruption with e2fs Date: Sun, 15 Aug 1993 22:58:22 GMT Over the weekend my Linux system hung :-( As in completely - it lost its ethernet connectivity, I couldn't log in on any of the consoles, etc. In the end I closed my eyes and hit the Reset button (owch!). Upon coming up, it screamed about having lost a tonne of files, and threw them into /lost+found (it was the root partition that was hit). Over 30Megs to be precise <:-((. That's what's odd - there is _no way_ most of those files were open (even for read) when the crash occured - it's my area where I compile things - and if I wasn't using it, then it wasn't in use. Anyway, things look really messy now, my /etc/fstab file has now become /etc/fstab/ (a directory!), and when I re-created a lost directory, everything in it got mapped into lost+found as well! (as in when I started deleting files in /lost+found, it started deleting files create _since_ the crash). So it looks like I'll have to remake that partition. (BTW, my system does a "e2fsck -a" at boot - like several people have suggested - but the man page says "use with caution" - should I really be doing something else?). Anyway, my question is - is there an "official" way to deal with such occurences. If anyone has come up with a half sure-fire way of recovering from a crash without going through what I did, could they Email me the details? I'll summerize for the Net as I'm sure we'd all like to know :-) SLS 1.03, Linux.00p11, default kernel, Micropolis 560HD with UltraStore 34F SCSI 2 controller. -- Cheers Jason Haar, Network Consultant ------------------------------ From: Simon.Hobson@purplet.demon.co.uk (Simon Hobson) Subject: Re: Reading and writing M Date: Sun, 15 Aug 1993 12:13:00 +0000 In a message dated 08-14-93 09:07 MIKE PARK wrote to ALL: MP> Does anyone have crufty (or otherwise) source code to read (and even MP> better write) macintosh disks on non-mac hardware? I'm willing to MP> spend money to obtain this source. MP> MP> On a related question is the floppy format public knowledge? MP> Maybe described in Inside Macintosh? MP> MP> Would it be tough to write a basic set of routines to read their MP> format? Forget about trying to read Mac 800k disks, they use Group Code Recording and variable speed which a PC controller/drive cannot handle. The 1.4M disks use the same MFM coding and sector structure as used on PCs. I gather that there are several packages around for DOS that will do exactly what you want (ie read & write Mac disks), but I don't suppose you will find source code ;-) I am sure that the format for Mac floppies is in Inside Mac, I only have vols 1-3 which of course do not include HFS (heirarchical File System) disks ! What you need of course is the description and data stuctures for HFS disks. I would guess that HFS is described in IM vol 4 (the Mac Plus update) where HFS first appears. There may be something in vol 5 (SE & II family) or Vol 6 (System 7). One other thing, I think Apple have now discontinued Inside Mac (which was getting somewhat BIG) in favour of a collection of volumes covering specific areas - like 'everything to do with file systems and files in one volume' - but I can't remember what they are called. The source for Apple manuals etc is Apple Programmers and Developers Association (APDA). They can be reached on : Fax : (408) 562-3971 Phone : 800-282-2732 (US), 800-637-0027 (Canada), (408) 562-3910 (International) AppleLink : APDA Genie : A.DEVELOPER3 Compuserve : 76666,2405 Internet : APDA@applelink.apple.com MacNet : APDA MCI : POSTROM Hope this helps !, Simon -> Alice4Mac Eval:18May93 ------------------------------ From: Andrew Tefft Subject: xdaliclock on Linux Date: 15 Aug 1993 20:07:51 -0400 Reply-To: teffta@engr.dnet.ge.com Has anyone tried to compile xdaliclock under Linux? I got v. 1.05 and compiled it; it just apparently did nothing. I started it up in gdb and hit control-c and it appeared to be "stuck" at a select() call. So I used -DNO_SELECT as suggested in the Imakefile, same result. I wanted the debugging symbols so I compiled with -g... and it worked! I seem to remember that this forces static linking so I tried -static instead of -g and it continued to work. While I should be satisfied that it works, it is about 10x the size it should be. So, has anyone else run into something like this? I'm using gcc 2.4.3, kernel pl11, libc 4.4.1. It links with libX11.so.3.0.1 and libXext.a. ------------------------------ Crossposted-To: comp.os.linux.admin From: gomez@enuxsa.eas.asu.edu (JL Gomez) Subject: stale LCK file after uucp session Date: Sun, 15 Aug 1993 23:59:43 GMT I have a remote site that calls my machine for e-mail delivery. I never call this machine. However, uucico leaves a stale LCK file for this machine, even when the session is over. I'm running Slackware v1.01 with uucp1.04 binaries. Thanks for any info! -- gomez@enuxsa.eas.asu.edu ------------------------------ ** 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 ******************************