Received: from PACIFIC-CARRIER-ANNEX.MIT.EDU by po7.MIT.EDU (5.61/4.7) id AA18752; Mon, 4 Dec 95 01:15:31 EST Received: from SENATOR-BEDFELLOW.MIT.EDU by MIT.EDU with SMTP id AA06936; Mon, 4 Dec 95 01:14:28 EST Received: (from root@localhost) by senator-bedfellow.MIT.EDU (8.6.12/2.3JIK) id BAA06462 for Linux-Development-System-Dist@senator-bedfellow.mit.edu; Mon, 4 Dec 1995 01:13:42 -0500 Message-Id: <199512040613.BAA06462@senator-bedfellow.MIT.EDU> From: Digestifier To: Linux-Development-System@senator-bedfellow.MIT.EDU Reply-To: Linux-Development-System@senator-bedfellow.MIT.EDU Date: Mon, 4 Dec 95 01:13:40 EST Subject: Linux-Development-System Digest #84 Linux-Development-System Digest #84, Volume #2 Mon, 4 Dec 95 01:13:40 EST Contents: Re: Bug(?) with the linux console (liuyu) Re: Standardized Printing for Linux (Mats Andtbacka) APM for ELF compiled kernels? (Tony Robinson) Re: Looking for a Linux-compiled copy of Perl 5 ready-to-use (Douglas Stoun) BSD Sockets ( Buffered? ) (James R. Leu) 1.3.44: isofs continues NOT to work on multi-volume CDs (Dirk Rogalla) Re: HELP! aha1542/dx4-120 problems! (johan Henselmans) Re: `Unable to handle kernel paging request...' (johan Henselmans) Re: why no slackware bugfixes? (Tom Tempero) Re: Searching for ROB NATION (Gregory H. Margo) Re: Kernel 1.3.44 Bug? (Sergei Viznyuk) SVGALIB question (Phil Howard) ---------------------------------------------------------------------------- From: liuyu@PHYS1.PHYSICS.NYU.EDU (liuyu) Subject: Re: Bug(?) with the linux console Date: 30 Nov 1995 18:27:24 -0500 "B. James Phillippe" writes: >Good, I'm not alone. > I was going to post something about it. Happens to me too. When text is >scrolling along, there are occasional "wipes" of text at near light speed along >the right side of the screen. I'm using 1.2.13 and an S3 card. Console is on >80x25 mode. Yup, same here. Kernel 1.2.13 and a ET4000 card. In 80x25 mode. ------------------------------ From: mandtbac@news.abo.fi (Mats Andtbacka) Subject: Re: Standardized Printing for Linux Date: 30 Nov 1995 20:52:15 GMT Reply-To: mandtbac@abo.fi Matthew D. Isleb, in <49iegd$g2h@kirin.wwa.com>: >Has anyone heard anything about stadarizing printing for linux? >I have heard that SCO has some kind of standarization of printing. >As of now, to get any kind of decent printouts you need a postscript printer >and a filter to turn docs into postscript. Or perhaps you have been >lucky enough to find a filter for you specific printer (e.g. HP DeskJet 500) Read the Printing-HOWTO. My own DJ510 is working great printing Unix-style ISO Latin-1 text, DOS style CR/LF text, PostScript through Ghostscript, and .dvi files via dvips and ghostscript. Printing is _simple_, and mindbogglingly customizable, once you understand it. >I find this to be a big drawback for linux. I have been poking around >and learning how to write drivers for linux. Istarted to write one by >stealing the code from a hp deskjet filter program. The way I figure it, >this is how it would work: >a) Drivers would be compiled into the kernel or modularized. Bad idea. Very bad idea. Drivers for what file format, Postscript? Why not dvi? How to print plain ASCII then? And all this in the already bloating kernel, of all places? The kernel is doing great at the things it _should_ be doing, shuffling raw data out the parallel port. Anything beyond that can and should be handled in user mode, by lpd or similar programs. >b) they would be an extention to the lp.c driver (posibly prn.c) >c) This prn driver would register as a new major device. A brand new device for processed output through this driver? It would be needed, of course, but I disagree that such a driver is needed. But even if you did this, wouldn't you need a separate driver for each file format - one for PostScript, one for dvi, one for GIFs, and so on ad nauseam? Each of which would chew their own major (or even minor) device? Not on my root partition, friend. >d) Devices would be created: /dev/prn0, /dev/prn1 >c) Programs could retieve info and capabilities about a specific printer > from the device file or the proc file system. Two paragraphs of Obc.o.l.d.s: Hm, some status reporting through /proc might be neat, but the problem is that the parallel port doesn't allow for much in the way of status reporting. Maybe if lp.c was ever expanded to handle bidirectional high-speed ports you could read incoming data - but that would need a device file in /dev, properly, it doesn't belong in /proc... /proc/lpstat to show what type of parallel port is installed, maybe? Might make some sense. >d) A standard printer configuration utility would be writen to configure > a printer for a program to print. Although applications could be >writen to configure a printer (in a stadardized manner that is). This already exists. It is called /etc/printcap. Seriously, there's nothing wrong with printing through lpr and lpd unless you set them up wrong. When a program wants to print, it just opens a pipe to lpr and asks the user to what printer name (if any) output should be sent; after that, no worries, unless the filters malfunction. If they do, someone has botched. -- " ... got to contaminate to alleviate this loneliness i now know the depths i reach are limitless... " -- nin ------------------------------ From: ajr@eng.cam.ac.uk (Tony Robinson) Subject: APM for ELF compiled kernels? Date: 01 Dec 1995 08:49:54 GMT Has anyone got APM (ftp://tsx-11.mit.edu/pub/linux/packages/laptops/apm) working for an ELF compiled kernel? The obvious problem is symbol reference from assembler that has an underscore appended - removing the underscore allows the patched kernel to compile but it bombs out on booting... Tony Robinson ------------------------------ From: doug@igor.cmr.fsu.edu (Douglas Stoun) Crossposted-To: comp.lang.perl.misc,comp.os.linux.development.apps,comp.os.linux.setup,comp.os.linux.misc Subject: Re: Looking for a Linux-compiled copy of Perl 5 ready-to-use Date: Mon, 04 Dec 1995 03:04:43 GMT Reply-To: doug@igor.cmr.fsu.edu oylerb@tdkt.skypoint.net (Bill A. Oyler) wrote: >Hello, > >I am looking for an FTP site that contains a compiled version of Perl 5 >for the Linux operating system. I simply need a compiled copy of Perl 5 >for use with CGI/HTML programming. > >If you know where I can find a compiled version of Perl 5 for Linux, >please e-mail me! Thanks! The Slackware 3.0 distribution (ELF) has PERL 5 in the 'd' series I believe. ************************************************************* ** Douglas Stoun * doug@igor.cmr.fsu.edu ** ** Center for Music Research * 904-644-5786 ** ** Florida State University * http://www.music.fsu.edu ** ************************************************************* ------------------------------ From: jleu@coredcs.com (James R. Leu) Subject: BSD Sockets ( Buffered? ) Date: 4 Dec 1995 03:26:17 GMT I'm new to BSD sockets, but I've gotten to the point where I can get simple communications to occurs between process with tcp streams. I am having difficulties when trying to communication larger amount of data. For example: The first phase of communication for a program I'm running is to tranfer the command line arguments from the client to the host. The data structure I'm using is simple but here it is: long int arg_max; ( I know I could use a short ) long int arg_num; char arg[ 128 ]; The client can ALWAYS send the first argument without a hitch, the problem comes with any after that.. the problem being they never are recieved/sent. Questions: Do I have to ACK after each write/read? Do I have to buffer the write/reads? Do I have a clue? Quirk: The funny thing about this particular program, is that when I run the client while in gdb it works flawlessly. GO FIGURE! Any help is appreciated, Jim -- James R. Leu Network Administrator CORE Digital Communication Services jleu@coredcs.com ------------------------------ From: dirk@kant.ping.de (Dirk Rogalla) Subject: 1.3.44: isofs continues NOT to work on multi-volume CDs Date: 30 Nov 1995 00:51:29 +0100 Hello Linuxians, the implementation of the iso-9660-filesystem in the kernels (including 1.3.44) continues to refuse to work with mutli-volume CDs. This is somewhat targic, since the Prime Time Freeware folks unfortunately produced their last PTF4UNIX rel. 4.2 CD-set in the multi-volume fashion. Bad Luck for the all Linux-people who bought it. Guess... I bought it (surprise:( The substitute for isofs/inode.c the PTF-folks send has a wrong calling-interface and I am no kernel-intimate. Q: Who else bought PTF4UNIX-4.2 and managed to use it under Linux? .... and .... H O W ? Thanks, Dirk =========================================================================== COPYRIGHT NOTICE: The statements contained in this message are the property of Dirk Rogalla, (C) Copyright 1995. Notwithstanding any express or implied license to the contrary, Microsoft Network is prohibited from distributing, redistributing, posting, duplicating or using this message, or any part thereof, without the author's express written consent. ___________________________________________________________________________ ------------------------------ From: johan Henselmans Crossposted-To: comp.os.linux.setup,comp.os.linux.misc Subject: Re: HELP! aha1542/dx4-120 problems! Date: Fri, 1 Dec 1995 08:31:33 +0000 On Wed, 29 Nov 1995, Fairlight wrote: > > To anyone that can help: > > I've got the following setup: > > aha1542, 5.0mbit/sec rate, 2 scsi drives (internal), and a scsi cdrom > (external), termination all verified correct. I'm running linux 1.3.45. > > My problem is this. I was running a dx33 with this hardware. All ran fine > and without incident. > > However, last night, I upgraded my CPU to a dx4/120. Now, when I do > anything that's very disk intensive (read: more than 1 compile at a time, > or a large "find"), I get errors like: > > scsi0: channel0: resetting for second half of retries > > If I keep going with whatever is running, eventually I see that message a > few more times, and then I start getting scsidisk errors, ext2-fs errors, > and possibly a system crash. > > I've run the adaptec's disk verify, and there is nothing whatsoever wrong > with the media. > > Everything also works fine under DOS...works like a charm. > > I've not changed anything on my scsi bus in months. > > I called the place where I bought the motherboard, they say that chip works > fine in the board I have. > > I called the place I bought the chip from...they say it should work fine. > > I called Adaptec...they say that no matter what chip I put in there, the > card is on the ISA bus, and should work fine. They claim it's not a > hardware issue, but most likely the Linux driver for the aha1542 that > "probably has some incorrect timing loops" was how he put it. He said that > would make it go flaky like that. > > Does anyone know how to correct this problem? > > Thanks in advance... > > mark-> > -- > Fairlight-> ||| fairlite@iglou.com | Fairlight Consulting > __/\__ ||| "I'm talking for free... | http://www.fairlite.com > <__<>__> ||| It's a New Religion..." | info@fairlite.com > \/ ||| PGP Public Key available via finger @iglou, or Key servers > > I had similar problems. I have been told that the 1542CF is very sensitive about cablelength, termination etc. The problems disappeared when I disconnected one of the harddisks, a quamtum empire 2100 with active termination... Regards, Johan Johan Henselmans work: johan@netsense.nl tel: (+31)-(0)20-6261180 http://www.netsense.nl Netsense vof, Nieuwe Herengracht 149 1011 SG Amsterdam home: johan@hsb.nest.nl tel: (+31)-(0)20-6267538 Nieuwe Herengracht 147d 1011 SG Amsterdam ------------------------------ From: johan Henselmans Subject: Re: `Unable to handle kernel paging request...' Date: Fri, 1 Dec 1995 08:39:30 +0000 On 29 Nov 1995, Dr. Chris Nicol wrote: > Hi. I am running Linux kernel version 1.3.25 on a Pentium 100 with 32M > of memory and an ASUS P55TP4XE motherboard. This system has been > functioning fine, and has been `up' for 37 days. Recently, we have > been running some heavy numerical computations on the machine. Last > night, I tried to run a LaTeX programme to generate a dvi file for a > document, and I got an error `unable to handle kernel paging request', > followed by a segmentation fault. The system did not crash, but > continues to run even now. But it will not execute LaTeX programmes, > although it seems to be reasonably happy to run lots of other stuff, > including Netscape. I checked the syslog, and discovered that several > other errors of this type had occurred in the past week, but I ha > not been aware of them till now. Here is the extract from the syslog > log file. Does anyone have any idea what is going on here? Is this a > serious problem? What other debugging information would be of use in > tracking down what this problem is being caused by? Thanks in advance > for any help. Please e-mail me direct at one of the addresses below... > > Chris. > > > Nov 21 23:20:24 dora kernel: <1>Unable to handle kernel paging request at virtual address e066492c > Nov 21 23:20:24 dora kernel: <1>current->tss.cr3 = 00aa9000, hr3 = 00aa9000 > Nov 21 23:20:24 dora kernel: <1>*pde = 00000000 > Nov 21 23:20:24 dora kernel: <6>Oops: 0000 > Nov 21 23:20:24 dora kernel: <6>EIP: 0010:00122f78 > Nov 21 23:20:24 dora kernel: <6>EFLAGS: 00010202 > Nov 21 23:20:24 dora kernel: <6>eax: 00000003 ebx: 0002c363 ecx: 00000301 edx: 20664920 > Nov 21 23:20:24 dora kernel: <6>esi: 00000001 edi: 00000400 ebp: 0002c363 esp: 00b99eac > Nov 21 23:20:24 dora kernel: <6>ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018 > Nov 21 23:20:24 dora kernel: <6>Process rm (pid: 15154, process nr: 47, stackpage=00b99000) > Nov 21 23:20:24 dora kernel: <6>Stack: 00000000 01c40301 00000400 00122fdf 00000301 0002c363 00000400 00000000 > Nov 21 23:20:24 dora kernel: <6> 01c4acdc 00000301 0002c363 001688e9 00000301 0002c363 00000400 01c4ac70 > Nov 21 23:20:24 dora kernel: <6> 01c4ac40 009e5ee0 00000008 009e5ee0 00000008 00f4e006 > Nov 21 23:20:24 dora kernel: <6>Call Trace: 00122fdf 001688e9 0016940f 0016337f 00121d1e 001660c8 0012c470 > Nov 21 23:20:24 dora kernel: <6> 0012c4b6 0010a769 > Nov 21 23:20:24 dora kernel: <6>Code: 66 39 4a 0c 75 32 39 5a 08 75 2d 39 7a 04 75 08 89 d0 5b 5e [snip] In the kernel source directory there is a document describing what you should do with these kind of messages to make them productive for the developers. I had the same problems running a 1.3.45 kernel, it disappeared (for a while) by rebooting the machine. > _______________________________________________________________________________ > Dr. Chris Nicol, Economics Phone: (306) 585-4182; FAX: 585-4815 > University of Regina E-Mail: nicolc@elsie.econ.uregina.ca > Saskatchewan, S4S 0A2, CANADA. nicolc@max.cc.uregina.ca > -- > _______________________________________________________________________________ > Dr. Chris Nicol, Economics Phone: (306) 585-4182; FAX: 585-4815 > University of Regina E-Mail: nicolc@elsie.econ.uregina.ca > Saskatchewan, S4S 0A2, CANADA. nicolc@max.cc.uregina.ca > > Regards, Johan Johan Henselmans work: johan@netsense.nl tel: (+31)-(0)20-6261180 http://www.netsense.nl Netsense vof, Nieuwe Herengracht 149 1011 SG Amsterdam home: johan@hsb.nest.nl tel: (+31)-(0)20-6267538 Nieuwe Herengracht 147d 1011 SG Amsterdam ------------------------------ From: Tom Tempero Subject: Re: why no slackware bugfixes? Date: 3 Dec 1995 20:08:18 GMT HI, IEFBR14 had a method to it's madnessess. IEF refers to the OS utilities (all messages had IEF prefix) as did the utilites. BR14 referred to the branch using address in Register 14 which figured into the subroutine linkage which if memory serves me, the cpu placed the return address. The bug referred to the failure to clear Register 15 which almost guaranteed a non zero return code from the utility. The routine was needed because to allocate datasets (files) for a batch job stream, a dummy program was needed. So much for nostalgia and trivia. -- Cheers.... Thomas Tempero Design Acceleration, Inc tilt@designacc.com Rochester Design Center 1530 Highway 52 N, Suite 1076 Rochester, MN ------------------------------ From: newton@shellx.best.com (Gregory H. Margo) Subject: Re: Searching for ROB NATION Date: 1 Dec 1995 09:18:30 -0800 Reply-To: gmargo@newton.vip.best.com Regarding rclock, rxvt, and fvwm: Rclock is not included with rxvt because they're compeletely unrelated, except for the original author. AFAIK, no one is maintaining rclock, but then it's so simple it doesn't need any maintaining. [Unless of course you've got a mystical fancy rclock version, but I don't know about that, just rumors.] I put up a copy of the old Rob Nation rclock code: ftp://ftp.best.com/pub/newton/rclock/rclock.tar.gz Rxvt is now maintained by Mark Olesen (olesen@me.queensu.ca). The latest version is 2.13 and is archived thusly: ftp://tsx-11.mit.edu/pub/linux/sources/usr.bin.X11/rxvt-2.13.tar.gz ftp://ftp.nuclecu.unam.mx/linux/local/rxvt/rxvt-2.13.tar.gz Fvwm is now maintained by Chuck Hines (chuck_hines@VNET.IBM.COM). The latest beta development version is 2.0.39, available from ftp://ftp.hpc.uh.edu/pubfvwm/version-2/fvwm-2.0.39.tar.gz The Fvwm mailing list is Majordomo based at hpc.uh.edu. Subscribe by sending to majordomo@hpc.uh.edu a message with "subscribe fvwm" and/or "subscribe fvwm-announce" in the body. If you *really* want to talk to Rob Nation, he lurks on the Fvwm mailing list. I hope this answered all the interrelated questions. gm -- +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Gregory H. Margo No free speech restrictions on the Internet! gmargo@newton.vip.best.com ms windows makes me ill newton@best.com ------------------------------ From: sviznyuk@magnus.acs.ohio-state.edu (Sergei Viznyuk) Subject: Re: Kernel 1.3.44 Bug? Date: 4 Dec 1995 04:45:03 GMT In article <30B92D69.3805@bitshop.com>, Steve Radich wrote: >Hello, I am trying to build the Linux kernel for the first time.. I tried >1.3.43 and had compile errors (argh), and now 1.3.44 and am having conflicting >definitions in 8390.c. The problem is as follows (and the course I am following >to try to work aruond it): > > >Line 110: >static void set_multicast_list(struct device *dev); > >line 580: >/* static void set_multicast_list(struct device *dev, int num_addrs, void >*addrs) */ > >I have changed the above to the same as at first (the latter two variables are >not referenced) >static void set_multicast_list(struct device *dev) > > >Being I am not familiar with Linux kernel what-so-ever, is this a known bug (it >won't compile on my system) or should I be changing something else? try to compile gcc 2.7.1 first. I compiled 1.3.43 kernel without a single error and have not seen any bugs so far, except that my bogomips calibration dropped from 33.55 for aout 1.2.10 to 33.28 for ELF 1.3.43 and I do feel the system response is not as crisp as it used to be. Serge >-- >------------------------------------------------------------- > Come see the BitShopping Experience, buying your Christmas > Gifts Online! http://www.bitshop.com >------------------------------------------------------------- ------------------------------ From: phil@zeus.clr.com (Phil Howard) Subject: SVGALIB question Date: 3 Dec 1995 16:01:14 -0600 Although this is for an application, the deep technical level is probably better answered on comp.os.linux.development.system. Will SVGALIB allow an application to set up a video mode and then bypass all its function calls to go to the video ram directly itself, while still supporting the ability to switch virtual consoles and let the application know of the switch? Will the video ram mapping be remapped to a dummy area during away-from-virtual-console (so the app can keep on running if it chooses to)? I have an old program from DOG DAZE (when I used DOS) that I want to rewrite for 32-bit and Linux. But it requires direct video ram access. It will be using a standard VGA video mode that every comptible video card should have (2-color, 8 pix/byte, 320x200). I might also want to extend it to other like modes (8 pix/byte required) but that's not essential. While I'm asking, can SVGALIB be used to also do this in a TEXT mode? Any of the standard VGA modes? Custom register tweak modes based on the old VGA register model if the board/card/chip supports it? -- Phil Howard KA9WGN +-------------------------------------------------+ Unix/Internet/Sys Admin | If these are non-essential federal government | CLR/Fast-Tax | workers, why are they there in the first place? | phil@clr.com +-------------------------------------------------+ ------------------------------ ** 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-System-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.development.system) via: Internet: Linux-Development-System@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-System Digest ******************************