Received: from SOUTH-STATION-ANNEX.MIT.EDU by po7.MIT.EDU (5.61/4.7) id AA00612; Sun, 31 Dec 95 17:17:04 EST Received: from SENATOR-BEDFELLOW.MIT.EDU by MIT.EDU with SMTP id AA03980; Sun, 31 Dec 95 17:16:56 EST Received: (from root@localhost) by senator-bedfellow.MIT.EDU (8.6.12/2.3JIK) id RAA19888 for Linux-Development-System-Dist@senator-bedfellow.mit.edu; Sun, 31 Dec 1995 17:13:40 -0500 Message-Id: <199512312213.RAA19888@senator-bedfellow.MIT.EDU> From: Digestifier To: Linux-Development-System@senator-bedfellow.MIT.EDU Reply-To: Linux-Development-System@senator-bedfellow.MIT.EDU Date: Sun, 31 Dec 95 17:13:38 EST Subject: Linux-Development-System Digest #169 Linux-Development-System Digest #169, Volume #2 Sun, 31 Dec 95 17:13:38 EST Contents: Re: Info on APM feature in 1.3.? (Luca Lizzeri) Re: Fix for 1.3.52 register_netdev error. (Michael Monson) Re: 1.3.50; why is bdflush reported to be huge by 'ps' (Aurel Balmosan) reproducible bug: eth0: bogus packet: status=0x0 nxpg=0x9 size=1518 (Molnar Ingo) Re: Error 'buffer botch' on cd-rom 'Goldstar R540B' (Rob Janssen) Re: Elf Sluggish - interactive perf measure wanted. (Rob Janssen) Re: insmod and device drivers (Rasmus Lerdorf) Re: kerneld - was: modprobe.c compile error (Robert Millner) Re: 1.3.50; why is bdflush reported to be huge by 'ps' (Miquel van Smoorenburg) Re: Please Help with aha2940 (Mark Swanson) Where are Kernel Change Summaries? (Paul J. Solomon) WANTED: A secure kernel `unlink' function. (Ronald F. Guilmette) applying kernel patches (Eric Berkowitz) Re: Bug in Lance network driver in 1.3 kernel (Eric Wampner) Re: applying kernel patches (Jason Mcmullan) ---------------------------------------------------------------------------- From: ll@niche.telnetwork.it (Luca Lizzeri) Crossposted-To: comp.os.linux.hardware,comp.os.linux.misc Subject: Re: Info on APM feature in 1.3.? Date: Sat, 30 Dec 1995 15:48:14 GMT nelson@seahunt.imat.com (Michael Nelson) wrote: >In article <4ban0p$a6t@Holly.aa.net>, "B. James Phillippe" writes: >|> Hello, >|> >|> I have read the documentation for 1.3.? kernels, but was unable to find any >|> documentation on the APM features. Where can I go to get more information on >|> what each option does and how to decide which to set? > Read the comments in /usr/src/linux/drivers/char/apm_bios.c > > Michael Also in 1.3.52 configuration help has been added (choose ? at the relevant items during 'make config'). ------------------------------ From: sl18g@cc.usu.edu (Michael Monson) Subject: Re: Fix for 1.3.52 register_netdev error. Date: 31 Dec 95 01:42:45 MDT James MacKinnon writes: > On 29 Dec 1995, Michael Monson wrote: >> If you're having this problem, the easiest fix is to simply turn modules on >> with make config. > > Configure has changed. You want to answer 'Yes' to the first config question. I was, of course, aware of that. However, I have no use for modules at the moment, so I prefer to leave them turned off, as a minor convenience. >> A better way might be to comment out these lines in drivers/net/net_init.c: > > NOT a good idea... modules are becoming an integral part of the kernel. You're right. However, at the moment, the routines enclosed by the #ifdef are still needed in dozens of places within the kernel. Until they are no longer needed (unless modules are enabled) perhaps the #ifdef should be commented out, or CONFIG_MODULES should be hard defined somewhere. The config script probably shouldn't have an option to break compiles. :-) > You don't want to be editing all future release files this way :-) I'm posting here so in the future I won't! :-) > James S. MacKinnon Office: P-139 Avahd-Bhatia Physics Lab (sl18g@cc.usu.edu) ------------------------------ From: aurel@xylo.owl.de (Aurel Balmosan) Subject: Re: 1.3.50; why is bdflush reported to be huge by 'ps' Date: Sun, 31 Dec 1995 07:33:41 GMT Michael Nelson (nelson@seahunt.imat.com) wrote: : In article <4c4igi$amm@ccnet3.ccnet.com>, scall@ccnet.com (Scott Call) writes: : |> Sorry if this is a lame or FAQish question..but : |> : |> I installed 1.3.50 on my Caldera Preview 2 (redhat based) system. : |> : |> Now all the rc files starting system calls report an error getting sid : |> from process 2. : |> : |> This doesn't happen on my slackware 3 based pentium at work, but it does : |> here on my 486 This is not a problem of sysvinit (I think). The process structure of the process 2 is not fully initialised. I have looked into the newest patches -1.3.51 -1.3.52 and there the setup of the process structure for process 2 is extended. So get the new kernel patches. -- ================================================================ Aurel Balmosan aurel@xylo.owl.de ------------------------------ From: mingo@teto.sch.bme.hu (Molnar Ingo) Subject: reproducible bug: eth0: bogus packet: status=0x0 nxpg=0x9 size=1518 Date: 31 Dec 1995 11:54:44 GMT Has anyone else seen something similar? I was doing the following: ftp> get linux-1.3.52.tar xxx.xxx 200 PORT command successful. 150 Opening BINARY mode data connection for linux-1.3.52.tar (19066880 bytes). eth0: bogus packet: status=0x0 nxpg=0x16 size=1518 eth0: bogus packet: status=0x0 nxpg=0xa size=1518 226 Transfer complete. 19066880 bytes received in 57.3 secs (3.3e+02 Kbytes/sec) ftp> Running 1.3.52 on both sites, Slackware 3.0 default installation, WD8013 and NE2000 noname network cards. 486 kernel, no other unusual kernel option. The problem came after i made a 1.3.45 -> 1.3.52 upgrade on the receiving host. The bug is reproducible, and it goes away when i switch back the receiving host (it has the WD card) to 1.3.45 . -- -- Copyright 1995. Ingo Molnar, mingo@hercules.elte.hu, Microsoft Network is prohibited from redistributing this work in any form, in whole or in part without license. License to distribute this work is available to Microsoft at $500. Transmission without permission constitutes an agreement to these terms. Make money fast! Copy this signature :) ------------------------------ Crossposted-To: comp.os.linux.misc From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: Error 'buffer botch' on cd-rom 'Goldstar R540B' Reply-To: pe1chl@wab-tis.rabobank.nl Date: Sun, 31 Dec 1995 09:00:41 GMT In jmathijs@solair1.inter.NL.net (J. Mathijssen) writes: >Repeating 'ls' finally gives me the full directory. I tried the new >kernel 1.3.50, which solves this CD-ROM problem, but gives other >problem in return: 'make' doesn't work because of some error in 'dirent'. It is not because of some error in 'dirent', but because of some error in 'make'... When you update your version of 'make' it will work, and you can run 1.3.50 (and above) Rob -- +------------------------------------+--------------------------------------+ | Rob Janssen rob@knoware.nl | BBS: +31-302870036 (2300-0730 local) | | AMPRnet: rob@pe1chl.ampr.org | AX.25 BBS: PE1CHL@PI8WNO.#UTR.NLD.EU | +------------------------------------+--------------------------------------+ ------------------------------ From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: Elf Sluggish - interactive perf measure wanted. Reply-To: pe1chl@wab-tis.rabobank.nl Date: Sun, 31 Dec 1995 09:03:32 GMT In <4c38vo$7kl@kala.cc.jyu.fi> eloranta@voimax.voima.jkl.fi (Jussi Eloranta) writes: >One thing that could be related to the ELF being slow... Someone posted >a test which tested, hmm.. now was it libc atof or something and I tested >it for both dll and elf and dll was meny times faster than elf version >for some unknown reason. Ah.. I found the original article... Here it is: >---- clip ---- It was already pointed out that the test was comparing apples and oranges. The "replacement code" was not at all equivalent to the library code it was compared with. The only interesting comparison would be between a.out and ELF shared libraries. Rob -- +------------------------------------+--------------------------------------+ | Rob Janssen rob@knoware.nl | BBS: +31-302870036 (2300-0730 local) | | AMPRnet: rob@pe1chl.ampr.org | AX.25 BBS: PE1CHL@PI8WNO.#UTR.NLD.EU | +------------------------------------+--------------------------------------+ ------------------------------ From: rasmus@krone.dialin.utoronto.ca (Rasmus Lerdorf) Subject: Re: insmod and device drivers Reply-To: rasmus@io.org Date: Sun, 31 Dec 1995 15:24:33 GMT On Sat, 30 Dec 1995 19:14:06 GMT, Eric Hill wrote: >Recently, I modified a device driver originally for SCO unix for a linux >system. I compiled it to an object file, copied it to the /boot >directory, and tried to use insmod so that I could actually use it. >Insmod keeps on returning the same error "Kernel_Version Undefined". How >do I set the "kernel version" for the driver? Any help would be greatly >appreciated. Try using the insmod that comes with DosEmu. Right from the README file from that package: This is a modified insmod for development purpose, which can load a module even if some syms are undefined in ksyms. The addresses then come out of /usr/src/(z)System.map. To avoid reading the wrong map I compare the known addresses from get_kernel_sym() against the ones in (z)System.map. So there is *very* low propability of crashes. It would seem to me that this undefined Kernel_Version symbol would be in System.map and by using this modified insmod, you should be able to pick it up. -- Rasmus Lerdorf rasmus@io.org ------------------------------ From: millner@millner.bevc.blacksburg.va.us (Robert Millner) Subject: Re: kerneld - was: modprobe.c compile error Date: 31 Dec 1995 15:38:50 GMT "MS" == Mark Shadley writes: MS> Dec 20 23:22:34 shadow modprobe: Can't locate module block-major-13 MS> Dec 20 23:22:35 shadow modprobe: Can't locate module block-major-13 MS> Dec 21 06:01:01 shadow modprobe: Can't locate module binfmt-30821 MS> Dec 21 06:01:02 shadow modprobe: Can't locate module binfmt-30821 MS> The binfmt-xxxxx messages showed up during every cron job. I have been using it here on my system as well with 1.2.11-13 and now 1.3.49. I have encountered no problems with it on those kernels except one part of the patch[1] I had to do manually for 1.3.49. The block-major-13's are IBM xt disk devices. I don't know about the binfmt's but I suspect that it is trying to glean a magic number off of your script and tell the kernal to execute that kind of a program. Kerneld understands a "block-major-*" to module assignment in /etc/conf.modules. I have found that useful in making the sound driver dynamically load among others. Rob [1] The patch in the kernal mailing list, not the one with the modutils package. ------------------------------ From: miquels@drinkel.ow.org (Miquel van Smoorenburg) Subject: Re: 1.3.50; why is bdflush reported to be huge by 'ps' Date: Sun, 31 Dec 1995 18:13:50 +0100 (MET) In article <4c51hi$kbc@myrddin.imat.com>, Michael Nelson wrote: >In article <4c4igi$amm@ccnet3.ccnet.com>, scall@ccnet.com (Scott Call) writes: >|> Now all the rc files starting system calls report an error getting sid >|> from process 2. > > I had the same problem on my RedHat 2.1 system with 1.3.5[0-2]. >I updated the SysVInit package to 2.58 and the problem went away. > > You can get 2.58 from: > > ftp.cistron.nl:pub/People/miquels/sysvinit-2.58-1.tar.gz ... but please don't put it on other ftp sites, as this is not yet the "final" one. I need two or three more days. Actually, probably just an hour or two but I'm going off to the new years parties right now :) Mike. -- + Miquel van Smoorenburg + Cistron Internet Services + Living is a | | miquels@cistron.nl (SP5) | Independent Dutch ISP | horizontal | + miquels@drinkel.ow.org + http://www.cistron.nl/ + fall + ------------------------------ From: mark@hunter.mas.org (Mark Swanson) Subject: Re: Please Help with aha2940 Date: 31 Dec 1995 16:13:22 GMT In article <4bdge4$k25@host.di.fct.unl.pt>, Luis Miguel Flores wrote: >I' ve an aha2940(SCSI/PCI) and I can't get Linux to detect my hard drive >I've already tried the kernel aha2940 at nic.funet.fi. >Please somebody help me. I'm really in need. >-- If you have tried the aha2940 kernel and it doesn't work, you may want to try and change the interrupt your 2940 is using through your BIOS setup utility. (If you BIOS can do that.) -- ===================================================================== Mark Swanson Tel: (613)565-0093 DigiCrypt Technologies, Inc. Fax: (613)729-5908 1318 Wellington Street Internet: mark@luckman.com Box 36041 Ottawa, Ontario K1Y 4V3, Canada ===================================================================== ------------------------------ From: solomon@interaccess.com (Paul J. Solomon) Crossposted-To: comp.os.linux.help,comp.os.linux.misc Subject: Where are Kernel Change Summaries? Date: 31 Dec 1995 19:09:41 GMT Do any summaries of Linux Kernel changes exist, preferably in summary form: i.e. I don't care about which exact files were changed, but what new functionality, bug fixes, etc. were added to a specific version of a kernel. -paul. -- ____ __ / __ \ ____ _ __ __ / / ------------------------------------ / /_/ // __ `// / / / / / Paul J. Solomon / ____// /_/ // /_/ / / / solomon@interaccess.com /_/ \__,_/ \__,_/ /_/ 312/474-7688w 312/883-4662h ================================================================= "And thus begins the WEB..." - Marillion ------------------------------ From: rfg@ccnet.com (Ronald F. Guilmette) Subject: WANTED: A secure kernel `unlink' function. Date: 31 Dec 1995 12:40:49 -0800 I'd like to have patch for the kernel... a relatively simple one. (If any of you weekend hackers happen to have some time on your hands, maybe you would consider implementing this.) In a nutshell, I'd like to have a kernel in which the `unlink' function actually overwrites (e.g. with zeros) the file being unlinked. Seems to me that security-minded people who are operating multi-user systems would appreciate this. Of course there are bound to be naysayers who will immediately scream that this would make the system too slow... but I have a suggestion which could possibly reduce that problem. Here's the idea.... Implement a kernel `unlink' function which either does or doesn't do erasing depending upon the setting of some environment variable in the calling process. I can imagine two different way to implement this, depending upon what the sysadmin wants to be the default behavior. For the paranoid sysadmin, the default behavior would be to erase, and the kernel would check for NO_ERASE defined in the environment of the calling process (and if it found that symbol defined, it would NOT actually do the erasing). For the relaxed sysadmin, the default would be not to erase files when unlinking them, but if ERASE is set in the environment of the calling process, files would be erased as they are unlinked. Of course, the selection of which default to use (paranoid or relaxed) would be a kernel build-time option. In practice, I would guess that the default that most people would select would be the relaxed default, thus causing erasure of unlinked files only when specifically requested by the user issuing the `unlink' request. (Paranoid users would just set ERASE in their .profile or their .cshrc files, and then leave it set all of the time.) Can this sort of stuff be done _outside_ the kernel? Yes, of course. But it cannot be done _transparently_ (e.g. to pre-existing binaries of pre-existing programs) unless it is implemented directly within the kernel implementation of `unlink'. (I think that efficiency concerns would also dictate that this would be best done within the kernel too, but that is a separate issue.) If you have comments on this idea, and if you would like me to see them, please send them to me via E-mail, as I do not normally read this newsgroup. Thanks. P.S. I just realized that (to be _really_ effective) this whole idea should really be applied not only to `unlink', but also to `truncate' and `ftruncate' also. -- -- Ron Guilmette, Roseville, CA -------- Infinite Monkeys & Co. ------------ ---- E-mail: rfg@monkeys.com ----------- Purveyors of Compiler Test Suites - ============================================================================ ------------------------------ From: ueberkow@uxa.ecn.bgu.edu (Eric Berkowitz) Subject: applying kernel patches Date: 31 Dec 1995 15:15:38 -0600 I have been trying to keep up with the development of the 1.3.xx keernels and with the speed that new versions come out I have a question. Do the kernel patches need to be applied sequestialy or are they compunded from a certain level so that I can skip some. That is, if I have kernel 1.3.50 do I need to apply patch 1.3.51 and the 1.3.52 or can I apply 1.3.52 directly? Thank you for your help Eric ------------------------------ From: eww@grebble.oau.org (Eric Wampner) Crossposted-To: comp.os.linux.networking Subject: Re: Bug in Lance network driver in 1.3 kernel Date: 31 Dec 1995 07:44:48 -0500 Andrew Ruthven (andrew@cerberus.nui-wgtn.gen.nz) wrote: : According to Klaus Liedl : : >Andrew Ruthven (andrew@cerberus.nui-wgtn.gen.nz) wrote: : > : >: I've had a few problems with using the Lance driver in the 1.3.* kernel. : >: When I configure it with ifconfig and specify the broadcast address it hangs : >: the whole machine. : > : >I have similar problems here. : >However e.g. 1.3.20 or 1.3.21 works fine. : >All the 1.3.4x kernels show the problems you mention above. : >(several things changed in the lance driver inbetween ...) : It's good to know it isn't just by machine. Let's just hope that these : small problems get fixed soon. :) I've found a temporary solution is to : not give it a broadcast mask. The interesting thing is that it would : sometimes work fine. Strange. : Another thing with the Lance driver is that it still has the same version : number as from the 1.2 kernel. :) I am running 1.3.37 and for a little while ran 1.4.41, and I have a Zeos Pantera 90 with AMD SCSI+Net, uses the lance driver. I have never tried specifying the broadcast address, I use the netmask instead. If you are looking for a work-around, that will do it. I don't think you loose any functionality over the broadcast (unless you are doing something silly like using 0 for broadcast. :-) eric -- --- Eric Wampner eww@grebble.oau.org "Valkyrie..." eww@seminole.iag.net SNAIL 2618 Adela,32826-3253 ***G'Kar: "Weep for the chicken, Na'Toth. Weep for us all"*** ------------------------------ From: jmcc@m5.vi.ri.cmu.edu (Jason Mcmullan) Subject: Re: applying kernel patches Date: 31 Dec 1995 21:48:09 GMT Eric Berkowitz (ueberkow@uxa.ecn.bgu.edu) wrote: : I have been trying to keep up with the development of the 1.3.xx : keernels and with the speed that new versions come out I have a : question. : Do the kernel patches need to be applied sequestialy or are they : compunded from a certain level so that I can skip some. That is, if I : have kernel 1.3.50 do I need to apply patch 1.3.51 and the 1.3.52 or can : I apply 1.3.52 directly? All patches must be applied sequentially, ie: 1.3.48 -> 1.3.49 -> 1.3.50 -> 1.3.51 -> 1.3.52 -> ... Here's a shell script I wrote that takes care of most of it for me. It takes as parameters the kernel version and the patches you wish to apply, ie: If I had a v1.3.48 kernel, and I want to upgrade to v1.3.52: % patchwith 1.3 49 50 51 52 This patcher checks for .rej files, and stops patching if any are found. All .orig files are retained after patching so you can do a "find /usr/src/linux -name "*.orig" " after patching to see which files have been changed. ======================= CUT HERE ============================== #! /bin/sh # GNU Copylefted by Jason McMullan # Set these to reflect where you keep your kernel # and where you keep your patches. I currently have # the symlink /usr/src/linux -> /usr/src/linux_1.3 # and keep my patch-1.3.xx.gz files in /usr/depot/Kernel LINUXDIR=/usr/src/linux_ # <- note that the version (ie "1.3") is after the _ PATCHDIR=/usr/depot/Kernel if [ -z "$1" -o -z "$2" ]; then echo "Usage:" echo "$0 x.yy zz [zz zz ...]" echo echo " Where 'x.yy' is the kernel version," echo "and 'zz' is the patch number." exit 1 fi if [ -z "$EDITOR" ]; then EDITOR=/bin/vi fi version=$1 shift if [ ! -d $LINUXDIR$version ]; then echo "ERROR: can't find directory $LINUXDIR$version" exit 1 fi cd $LINUXDIR$version for pat in $*; do shift patfile=$PATCHDIR/patch-$version.$pat.gz if [ ! -f $patfile ]; then echo "Patch file $patfile now found!" echo "Aborted..." exit 2 fi echo "Patching with patch-$version.$pat..." gzip -dc $patfile | patch -p1 rejects=`find . -name "*.rej" -print` if [ -z "$rejects" ]; then echo "No rejects found." else echo "Rejects detected:" for rej in $rejects; do echo " $rej" done echo "Patches not applied: $*" exit 3 fi if [ ! -z "$*" ]; then echo "Patches remaining: $*" echo "Press ENTER to continue." read i fi done ================= CUT HERE =========================== -- Copyright 1995 Jason McMullan; all rights reserved; license for the Microsoft Network to distribute this text for US $100,000 per copy is hereby granted; all other use of this text by the Microsoft Network, including storing, relaying, or reuse is expressly prohibited; distribution by the Microsoft Network indicates full acceptance of the licensing terms. MICKEY# 4729-304-8732 ------------------------------ ** 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 ******************************