Received: from SOUTH-STATION-ANNEX.MIT.EDU by po7.MIT.EDU (5.61/4.7) id AA25370; Fri, 23 Feb 96 15:20:13 EST Received: from SENATOR-BEDFELLOW.MIT.EDU by MIT.EDU with SMTP id AA22092; Fri, 23 Feb 96 15:19:48 EST Received: (from root@localhost) by senator-bedfellow.MIT.EDU (8.6.13/2.3JIK) id PAA03119 for Linux-Development-System-Dist@senator-bedfellow.mit.edu; Fri, 23 Feb 1996 15:14:19 -0500 Message-Id: <199602232014.PAA03119@senator-bedfellow.MIT.EDU> From: Digestifier To: Linux-Development-System@senator-bedfellow.MIT.EDU Reply-To: Linux-Development-System@senator-bedfellow.MIT.EDU Delivered-To: Linux-Development-System@senator-bedfellow.mit.edu Precedence: bulk Date: Fri, 23 Feb 96 15:14:15 EST Subject: Linux-Development-System Digest #415 Linux-Development-System Digest #415, Volume #2 Fri, 23 Feb 96 15:14:15 EST Contents: Maintainer of net-tools? (Ian Chard) Re: any DHCP server port to linux ??? (Paul Slootman) Re: Is majordomo@vger.rutgers.edu up? (Paul Slootman) Re: kerneld problem? (Bjorn Ekwall) Re: third party kernel drivers (Bjorn Ekwall) pty programming and vhangup() (Henrik Nordstrom) Re: Initialization failure with modprobe... (Bjorn Ekwall) Re: OSF Mach kernel (Francois BARBOU DES PLACES) Re: Segmentation fault!!! Q (Paul Slootman) Re: 1.3.66 http socket problems? (Studnitzky Boaz) Re: need secure OS to entrust millions to (Steve Gailey) Pascal to C translator? (Christopher Thomas) memory for cursor pixmap ?? (Nicholas Mc Guire) Initialization failure with modprobe... (BLue) ? 33.6 MODEMS ? (coe9222057@dlsu.edu.ph) Passing boot parameters to modules, how? (Changaroth M Anup) Re: Massive bug in GCC 2.7.2 (Chadley Kaye Dawson) ---------------------------------------------------------------------------- From: ian@tanagra.demon.co.uk (Ian Chard) Subject: Maintainer of net-tools? Date: Fri, 9 Feb 1996 22:10:29 GMT Hi, Can someone email me with the address of whoever maintains the net-tools package? I've been told it's Bernd Eckfels, but I can't seem to find his address anywhere. Thanks Ian -- [ Ian Chard, back in Manchester | "They'll be Stormtroopers over ] [ Email: ian@tanagra.demon.co.uk | the White Cliffs of Dover..." ] [ NTS: G7OMZ @ GB7VRB.#38.GBR.EU | ] [ Phone: +44 161 434 6492 | -- Me ] ------------------------------ Crossposted-To: comp.os.linux.networking From: paul@wau.mis.ah.nl (Paul Slootman) Subject: Re: any DHCP server port to linux ??? Date: Thu, 22 Feb 1996 19:23:19 GMT Christoph Lameter wrote: > >Move from one subnet to the next? That can be simply brought about by >recognizing the net the request came from and assigning addresses based >on the origin. Ah! Pray tell, how do I recognize from which net the request came from? I'm working on the bootp-2.4.3DD-whatever server at the moment, and am having trouble doing exactly that. The request comes in from the client as a broadcast, and does not have any IP address yet (that's why the client's doing the DHCP in the first place!). I have not found any way of determining on which net the broadcast originated (or to be more precise, on which NIC the broadcast was received on...) I must confess to not knowing the gory details about this type of network stuff, although I can find my way round quite reasonably... >: I am not saying that what you have is bad. It just not 'good enough' for >: sites which need a true DHCP server. Yours is not. >Ok. Then lets fix it. It contains over 90% of the functionality already. That was my idea. I'm a bit put off by the existing code, however (it's been hacked on by about 10 different people, as far as I can tell, apparently without any real coordination. I've spent some time cleaning up stuff when I come across it (for example, stat(filename, &st) followed by a test (st_nlink != 0) !! As far as I can tell, if you can stat() a filename, it's not been unlinked...) >: That has nothing to do with DHCP. The WIDE server supports bootp so your >: objection is moot. I once looked at the WIDE server, but that was horribly non-portable, at a quick look (I didn't waste hours on it, so perhaps it's less bad than I thought?) >: In my opinion, it is far easier to add bootp as an infinate lease to DHCP >: than it is to try to make a bootp server support non-infinate leases of a >: DHCP server. >What could be that difficult? The only thing needed is to configure the >durations. That is supported by "my bootpd"! What else do you wish for? The implementation does _not_ make it simple to revoke a dynamically assigned address after a specified time... Paul Slootman -- Murphy Software, Enschede, The Netherlands work: paul@ahwau.ahold.nl / paul@murphy.nl home: paul@wurtel.hobby.nl ------------------------------ From: paul@wau.mis.ah.nl (Paul Slootman) Subject: Re: Is majordomo@vger.rutgers.edu up? Date: Thu, 22 Feb 1996 19:27:29 GMT Michael J Gebis wrote: >Most people running the beta kernels SHOULDN'T be on the linux-kernel >mailing list. They SHOULD be on a read-only digestified version of >the linux-kernel list. Quite frankly, most people (myself included) I agree. However, I gather the digest version of linux-kernel is dead, as in not-working, and apparently it's been that way for quite some time. Paul ------------------------------ From: Bjorn Ekwall Subject: Re: kerneld problem? Date: 22 Feb 1996 08:07:58 GMT Georg List wrote: >Who can tell me whats wrong here: >I am using linux-1.3.67 and modules-1.3.57. >I configured the kernel to autoload modules. >When I try to call smbmount the following error message appears: > >-- Cannot open /lib/modules/1.3.67/smbfs.o >Error: Unable to start smbfs, exiting... > >Why is kerneld searching in /lib/modules/1.3.67/ >instead of /lib/modules/1.3.67/fs (where smbfs.o resides). >When I copy smbfs.o to /lib/modules/1.3.67/ everything works properly. >All the other modules can be loaded without problems (ppp, hpfs, ncpfs). Actually it's not kerneld that fails, it is smbmount! It seems that smbmount tries to "outsmart" kerneld by attempting to load the module itself. Unfortunately it is looking in the wrong place... Take a look at the sources for smbmount and _remove_ the piece of source that tries to load the smbfs module, and everything should be all right again. Bjorn > >Thanks for any help you can provide. > >-Georg > >******************************************************** >* Georg List --- Wirtschaftsuniversitaet Wien * >* Department of Industrial Information Processing * >* Vienna University of Economics & Business Admin. * >* Pappenheimgasse 35/3/5, 1200 Vienna, Austria * >* TEL.: +43 (1) 3303636-31 (ext.) FAX -10 (ext.) * >* E-Mail: Georg.List@wu-wien.ac.at * >******************************************************** ------------------------------ From: Bjorn Ekwall Subject: Re: third party kernel drivers Date: 22 Feb 1996 08:15:32 GMT Alexandre Maret wrote: >Hi > >I would like to understand why there is so much problems >with writing drivers for Linux (unix ?). Why is it hard >for developpers to write a kernel driver without >distributing the sources ? I've heard about "name space" >and I don't know a lot about it. What is it exactly ? > >Has it something to do with the fact I can't load >some modules twice (sound) ? Since an external driver can be seen as a part of the kernel, and since the kernel is distributed under the GPL, the modules are required to be distributed under the GPL as well, i.e. with source. The thing about not being able to load the same module twice under the same name is just simple self-protection. If you _really_ want to have the same module code loaded more than once, use the "-o" option to insmod/modprobe (it's mentioned in the insmod man page). Bjorn ------------------------------ From: henrik@ida.his.se (Henrik Nordstrom) Subject: pty programming and vhangup() Date: 23 Feb 1996 14:06:56 GMT I am trying to write a access server, a lot like telnetd but very specialized. Linux-1.3.58 I need to open a pty and run some programs on this, and to make sure that no other process are using the same pty i think that I have to use vhangup(). I tried something like this: master=open(masterside,O_RDWR|O_NOCTTY); if(!fork()) { setsid(); setpgrp(); slave=open(slavesite,O_RDWR); /* This also sets our controlling tty */ signal(SIGHUP,SIG_IGN); /* Dont kill our selves */ vhangup(); /* Kill all other */ signal(SIGHUP,SIG_DFL); slave-open(slaveside,O_RDWR); /* Reopen slave side */ .... But the vhangup() doesn't work unless I insert a sleep somewhere after the open of the master side (before or after fork() doesn't matter). Is there any explanation to this? Does anybody know how and when vhangup() is supposed to work? Is there any better way to do this? --- Henrik Nordstrom University of Skoevde Sweden ------------------------------ From: Bjorn Ekwall Subject: Re: Initialization failure with modprobe... Date: 23 Feb 1996 06:45:45 GMT anupmc@asiaonline.net (BLue) wrote: >I'm having an "Initialization of aha152x failure" when I try to >modprobe the lower level SCSI module (after having successfully loaded >scsi_mod). Just to be on the safe side: have you run "depmod -a"? There might be a need for some start-up options for the module as well (just guessing), in which case you can have modprobe send these automatically to insmod by just adding the relevant "options aha152x whatever=some_value" to your "/etc/conf.modules". > [...] > >My dummy net driver seems to have problems as well (although >they may be unrelated ;). modprobe of dummy works just fine >but if I try and ifconfig dummy x.y.z.t up, I get SIOCSIADDR: >No Such Device. > >Any clue anyone? It's "dummy0", not "dummy" Cheers, Bjorn URL: ------------------------------ From: Francois BARBOU DES PLACES Subject: Re: OSF Mach kernel Date: Tue, 20 Feb 1996 11:27:29 +0100 Lam Dang wrote: > > > I'm getting a kernel panic a few seconds after "Going multiuser..." > > Warning: dev (4,3) tty->count(1) != #fd's(2) in tty_open > Kernel panic: osfmach3_tty_asym_read: not implemented > > When I entered "cont" to the debugger, it either rebooted or came > back with the same warning and panic, except for "dev (4,5)". If > I entered "cont" again it would reboot. Your VGA board wasn't recognized by the microkernel, so the Linux server couldn't map the VGA memory. It won't be able to use virtual consoles, gpm, sdoom, etc... The fall-back code (the dumb console) has a bug that makes it panic when more than one virtual console is activated. Just comment out the "c2", "c3", etc... entries in /etc/inittab (but leave "c1" untouched) and the panics should disappear. [ Francois ] -- +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ | Francois BARBOU DES PLACES | | e-mail : barbou@gr.osf.org | | OSF Research Institute | | or : barbou@ri.osf.fr | | 2 avenue de Vignate | | phone : (+33) 76.63.48.74 | | 38610 GIERES FRANCE | | fax : (+33) 76.51.05.32 | +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ +=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+=+ ------------------------------ From: paul@wau.mis.ah.nl (Paul Slootman) Subject: Re: Segmentation fault!!! Q Date: Fri, 23 Feb 1996 09:20:16 GMT >Jeremy Ehmke wrote: >> >> Alan K. Adamson (ne1h@ne1h.radio.org) wrote: >> : I have the following. >> >> : #include >> : #include >> : #include >> >> : char *a; >> >> : void main(void) >> : { >> : char *t = "akaXXXXXX"; >> >> : a = mktemp(t); >> >> : printf ("%s %s/n", t , a); >> : } >> >> : I get a seg fault everytime on the call to mktemp. >> >> I'm pretty sure it has nothing to do with your version of libc, etc. That's right. This is a C / gcc thing. When you say char *t = "akaXXXXXX"; you let pointer t point to the (static!) string "akaXXXXXX". Next, you want mktemp() to change that pointer. Gcc by default puts static strings into the text segment; as a result, they're read-only. If you want to be able to overwrite these strings, add the option -fwritable-strings to the gcc command line. That should do the trick. However, it may be "better" to declare t as follows: char t[10]; and then copy your template into it: strcpy(t, "akaXXXXXX"); That will work without the need for any gcc options. Paul Slootman -- Murphy Software, Enschede, The Netherlands work: paul@ahwau.ahold.nl / paul@murphy.nl home: paul@wurtel.hobby.nl ------------------------------ From: s3176015@techst02.technion.ac.il (Studnitzky Boaz ) Subject: Re: 1.3.66 http socket problems? Date: Thu, 22 Feb 1996 08:44:32 GMT I am using 1.3.66 and ever since 1.3.61 (or 1.3.62) something went broke with the sockets. I'm not referring to the CLOSE bug whatever that is, my problem is with the ftpd (wu.ftpd). It seems that no one can ftp to my machine. ftpd starts, opens the connection and stuff, than closes. After looking at the syslog i had the following message (more or less): ftpd: setsockopt: trying to set SOCKET option on a non-socket. Does this help anyone? Could it be because i'm using the pentium optimizing gcc (which is still experimental though working)? a note about pentium gcc: It seems to align data to 32 bit boundries even inside structures! so that: struct { short i; long j; }; is 8 bytes in size! could this be it? Boaz Studnitzky s3176015@t2.technion.ac.il ------------------------------ From: Steve Gailey Crossposted-To: comp.os.linux.misc,comp.os.linux.networking,comp.unix.bsd.freebsd.misc,comp.unix.bsd.netbsd.misc,comp.unix.bsd.bsdi.misc Subject: Re: need secure OS to entrust millions to Date: Fri, 23 Feb 1996 09:31:21 +0000 Bryce wrote: > I'm writing documentation which advises banks on how to > setup an electronic banking software package on a > Net-connected, firewall-protected Intel box. Some of the > most important banks in the world will be reading this > documentation very soon. The current version of the > documentation, which I inherited, advises them to run > FreeBSD or BSDI. I'm considering changing this > recommendation to Linux. Whilst I am an ardent advocate of FreeBSD and use it in a number of commercial situations for many of my customers, I suspect that these banks are not going to take you too seriously. I am currently installing both Firewalls and various Internet resources on behalf of one of my clients (The largest Open Systems company in the world) and a number of these installations are banks. Without exception they are NOT Interested in even discussing free tools, including operating systems. They have money to spend and are very distrustfull of anything they get for free. I am often ammused when they would rather pay double for the services of my company when purchased through a "Blue Chip" computer company, rather than directly. I presume that I must work harder when I sub-contract. :-) Whilst some of them are comeing around to think that Intel platforms may be of some use. Most of them are still going for RISC based workstations and servers. I don't mean to put you off, I would love to have FreeBSD taken more seriously. ========================================================== Steve Gailey steveg@metrosol.demon.co.uk Metronome Solutions Ltd. Tel: +44 973 563 239 Fax: +44 1732 460 742 ========================================================== ------------------------------ From: Christopher Thomas Subject: Pascal to C translator? Date: Thu, 22 Feb 1996 23:15:03 -0800 I have recently moved over to Linux from MS Windows and am now looking to port my applications that I wrote for Windows to Linux and X Windows. My applications are financial analysis apps with advanced GUI's. My apps for Windows were written using the visual development package Borland Delphi, and I now wish to port these to C++ in Linux. Has anybody got any suggestions other than just physically translating all the code? I have a Pascal to 'C' translator but I doubt if this will begin to work with "Object Pascal" which is what Delphi uses. Any suggestions would be greatly appreciated. Thanks Christopher Thomas ------------------------------ From: root@tuwien.ac.at (Nicholas Mc Guire) Subject: memory for cursor pixmap ?? Date: 23 Feb 1996 10:19:27 GMT Hi out there ! with xdpyinfo I get a maximum pixel size for the cursor of 64x64 . I wonder if there is a simple way to change this to say 256x256 ?? Linux 1.2.8 (Slackware) XFree 3.1.2 ( hope thats corect X11R6 ) runing an accelerated server ( XF86_S3 ) if I need to change source code , maby someone can point me to a file in the Server source where to get started..... thanks ! Nicholas Mc Guire *------------------------------------------------------------------------* *THERE {__} Tech. Univ. Vienna Austria * *IS oo ) Nicholas * *LIFE O_ `_-------. Mc Guire * *IN (-. ,-\ root@mdmtpc13.tuwien.ac.at * *THE || )---< ) nmg@mdmtws3.tuwien.ac.at * *NET ! ||| | | nmg@diddl.tuwien.ac.at * =========================================================================* ------------------------------ From: anupmc@asiaonline.net (BLue) Subject: Initialization failure with modprobe... Date: 20 Feb 1996 10:44:54 GMT I'm having an "Initialization of aha152x failure" when I try to modprobe the lower level SCSI module (after having successfully loaded scsi_mod). Kernel 1.3.66 and modules-1.3.57 (clean), anyone else seeing something similar ? If the kernel is built with aha152x builtin, everything works fine and I can access my SCSI devices, but with modules, aha152x seems to keep failing "Initialization.." My dummy net driver seems to have problems as well (although they may be unrelated ;). modprobe of dummy works just fine but if I try and ifconfig dummy x.y.z.t up, I get SIOCSIADDR: No Such Device. Any clue anyone? BLue ------------------------------ From: coe9222057@dlsu.edu.ph Subject: ? 33.6 MODEMS ? Date: 22 Feb 96 14:53:22 +0800 Hi ! I'll be purchasing a couple of Modems soon .. I came across an article in PC magazine , about Hayes 28.8 modems outperforming some 33.6 modems . It was also stated that the Hayes 28.8 OPTIMA (230,000 bps throughput) was attached to a Hayes ESP COMMUNICATION ACCELERATOR (900,000 bps throughput). Question : 1.) What if these 33.6 Modems where attached to Hayes' ESP COMMUNICATION ACCELERATOR , will it make a difference ? 2.) Are 33.6 Modems compatible with LINUX ? 3.) Can Comtrol Rocketport support these 33.6 modems ? Thanks in advance for info ! Ken (coe9222057@vax6.dlsu.edu.ph) ------------------------------ From: anupmc@asiaonline.net (Changaroth M Anup) Subject: Passing boot parameters to modules, how? Date: 21 Feb 1996 04:00:12 GMT Anyone knows how to pass typical boot parameters to a module? ie. if I had append="aha152x=0x340,11,7,1" in lilo.conf, how would I pass these as symbol= options in conf.modules ?? (This is related to my ealier post of Initialization failure whenattempting to load aha152x module). ------------------------------ From: Chadley Kaye Dawson Crossposted-To: comp.os.linux.development.apps Subject: Re: Massive bug in GCC 2.7.2 Date: Fri, 23 Feb 1996 10:41:24 -0500 > I beleive I have discovered a nice bug in the intel GCC 2.7.2. Try to > compile a C++ program that uses STDIO functions, like printf, etc... The > thing will segfault when you try to write to stdout. this is not true of > GCC 2.7.0 and is independant of libc and libg++/libstdc++ -- Watch out > for this one!!! Although I am having some trouble of my own with the g++ libiostream library, simple I/O to sdtout seems to be working ok: ======================================== > more test.cc #include main(){ printf("Hello World\n"); } > g++ test.cc -o test > ./test Hello World > g++ -v gcc -v Reading specs from /usr/lib/gcc-lib/i486-linux/2.7.2/specs gcc version 2.7.2 > ldd ./test libm.so.5 => /lib/libm.so.5.0.4 libc.so.5 => /lib/libc.so.5.2.16 > > uname -a Linux chiba 1.3.45 #2 Sun Dec 3 23:38:06 EST 1995 i586 > ======================================== ------------------------------ ** 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 ******************************