Received: from SOUTH-STATION-ANNEX.MIT.EDU by po7.MIT.EDU (5.61/4.7) id AA05066; Thu, 14 Mar 96 13:36:24 EST Received: from SENATOR-BEDFELLOW.MIT.EDU by MIT.EDU with SMTP id AA17020; Thu, 14 Mar 96 13:35:36 EST Received: (from root@localhost) by senator-bedfellow.MIT.EDU (8.6.13/2.3JIK) id NAA00402 for Linux-Development-System-Dist@senator-bedfellow.mit.edu; Thu, 14 Mar 1996 13:14:35 -0500 Message-Id: <199603141814.NAA00402@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: Thu, 14 Mar 96 13:14:29 EST Subject: Linux-Development-System Digest #504 Linux-Development-System Digest #504, Volume #2 Thu, 14 Mar 96 13:14:29 EST Contents: Re: linux-1.2.13 needs 1.2.9.elf patch? (Jonathan) Re: locale support for ISO-8859-1 (Martin Forssen) Re: tape backup onto vhs video recorder (Rob Janssen) Re: memmory mapping (Rob Janssen) eexpress.c: MAX_EEXP_CARDS ? (Chun-Yu Lee) Re: X11 and Trident 9440 PCI (Robert Schaefer) Re: 2000 IS a leap year. [WAS Re: Date bug? Seems rather important.] (David Dallmann) Re: [Help] ELF ? (Chen Qian) Re: How to input a character w/out echoing (Nick Kralevich) Re: 1.3.72 and ppp-2.2.0e - Am I the only one? (Richard Ketchersid) Re: Socket connection between SUN and Linux (Dr. Charles E. Campbell Jr.) Slow localhost messaging speed ? (Peter Choynowski) Re: Moo-Tiff (& unable to resolve symbol stat) (Marko Siladin) ---------------------------------------------------------------------------- From: jwalther@smartt.com (Jonathan) Subject: Re: linux-1.2.13 needs 1.2.9.elf patch? Date: Thu, 14 Mar 1996 06:26:50 GMT On 13 Mar 1996 21:47:38 GMT, crosson@cam.nist.gov (Bob Crosson) wrote: >Does the Linux-1.2.13 kernel source need the linux-1.2.9.elf.diff >patch to allow it to be compiled with ELF libraries? I also need the answer to this. Someone please post? *How* do you get 1.2.13 to compile with ELF support anyways? I always get errors... Jonathan ------------------------------ From: maf@math.chalmers.se (Martin Forssen) Subject: Re: locale support for ISO-8859-1 Date: 14 Mar 1996 07:29:53 GMT Isaac Miller writes: >I assume that in the newer libc, the non standard (C/POSIX) locale >tables >are externally loaded rather then taken hard-coded, but I cannot find >where >this external data should reside, and more over, how it can be produced. >Can someone shed some light from his/her experience ? I have produced the needed files and combined into tarfiles. The whole thing can be retrieved over WWW at http://www.dtek.chalmers.se/~maf/i18n /MaF ------------------------------ From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: tape backup onto vhs video recorder Reply-To: pe1chl@wab-tis.rabobank.nl Date: Wed, 13 Mar 1996 21:08:15 GMT In <4i4goq$gqu@hermes.synopsys.com> jbuck@synopsys.com (Joe Buck) writes: >Kai Leibrandt writes: >>Well, the very fact it uses a dsp (to convert digital to analog data) >>causes my neckhairs to stand up. VHS is an analog medium and as such I >>don't think it has any kind of right as a backup medium for computers. >You'd better throw away your hard disk them. The more advanced hard disks >also use DSP's, to convert between digital data and analog information >on the disk surface (the magnetic flux level your disk head reads is >an analog signal). Actually, the DSP on most harddisks is used to control the head positioning, forming a servo loop using the control signals that come back from the read/write heads. But you are right, the information on the disk is of course analog. 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: memmory mapping Reply-To: pe1chl@wab-tis.rabobank.nl Date: Wed, 13 Mar 1996 21:10:53 GMT In <4i2d1d$98v@news.fsu.edu> lamm@huey.met.fsu.edu (James Lamm) writes: >I am writing a kernel module which will reqeust that a >small routine will run (on IRQ 15) which will copy meg >4 of system memory to a disk file. How do I access bytes >4*(1024*1024) - 5*(1024*1024) in kernel code. I see that >mmap would do it in user space, will it work in kernel code >or can I just do a memcpy with a char * set to 4*(1024*1024). Yes, you can do that. Whether it is a sensible thing to copy 1 meg inside an interrupt handler is another thing... 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: d791013@ce.ntu.edu.tw (Chun-Yu Lee) Subject: eexpress.c: MAX_EEXP_CARDS ? Date: 14 Mar 1996 12:54:45 GMT In eexpress.c when compiled as module, the variable MAX_EEXP_CARDS seems undefined. (or EEXP_MAX_CARDS?) Chun-Yu Lee d791013@ce.ntu.edu.tw ------------------------------ From: rjsc@home.winc.com (Robert Schaefer) Subject: Re: X11 and Trident 9440 PCI Date: 14 Mar 1996 08:05:10 -0500 In article <314410AC.1A08@isoft.de>, Stefan Bremer wrote: >My Problem : I've bought a PC with Trident 9440 but the Linux kernel >doesn't support this graphic-card. >Now I've tried to setup X11 with standard Trident(8900) and with SVGA but >my monitor (15" 60kHz) can't sync any frequency. > >Has anybody the correct timings for mode 800x600? > >Can anybody help me? > >Thanx. Don't bother. If you go to ftp.xfree86.org you'll find a directory called pub/XFree86/beta which contains the latest public beta software for XFree86. This latest driver is a full distribution, so you need to download the libraries and binaries, in addition to the server. But you'll then find that the 9440 is now supported. Unfortunately, its non-accelerated at this point, but the SVGA server will recognize it and use it nicely. Also, xf86config should now help you out with configuring your system. So give 3.1.2D a try and let me know if you have any problems/questions. Bob Schaefer rjsc@winc.com rjsc@lubrizol.com ------------------------------ From: dallm931@cs.uidaho.edu (David Dallmann) Crossposted-To: comp.os.linux.setup,comp.os.linux.misc Subject: Re: 2000 IS a leap year. [WAS Re: Date bug? Seems rather important.] Date: 14 Mar 1996 03:54:35 GMT : In addition, is : a summary of the International Date and Time Standard (ISO 8601), : which also specifies leap years. As you can read there, too, 2000 is a : leap year, because the number of the year is divisible by 400. This is correct: if (year % 400 == 0) printf("leap year.\n"); else if (year % 100 == 0) printf("no leap year.\n"); else if (year % 4 ==0) printf("leap year.\n"); ,-----------------------------------------------------------------------------, | dallm931@cs.uidaho.edu http://www.cs.uidaho.edu/~dallm931 | `-----------------------------------------------------------------------------` Experts say that one out of four people suffer from paranoia. Check three of your friends, and if they seem ok, you must be it.. ------------------------------ From: chen@is.aist-nara.ac.jp (Chen Qian) Subject: Re: [Help] ELF ? Date: 13 Mar 1996 07:53:02 GMT >> >> In article <4h2793$54g@twix.cs.umd.edu> sjw@cs.umd.edu (Shuo-jen Wu) writes: >> -> I have a libpq.a library. It is not for ELF system. I upgrade my linux >> -> to ELF. I can't use that library any more. Is there any way that I can use it ? >> -> >> -> I got linking error. Should I add any special option ? >> >> If the lib is in a.out-format obviously you can't use it within >> ELF-programs. >> >> Either get the lib in ELF or compile those programs, that use this >> lib, in old a.out-format. >> There is a tool that can convert a.out-format lib??.a to elf-format lib??.a. Get binutils-2.6.0.10 and file "release.binutils-2.6.0.10", read it carefully. -- Qian CHEN Graduate School of Information Science Nara Institute of Science & Technology 8916-5 Takayama, Ikoma, Nara 630-01, JAPAN Tel : +81-7437-2-5272 FAX : +81-7437-2-5279 Email: chen@is.aist-nara.ac.jp ------------------------------ From: nickkral@america.CS.Berkeley.EDU (Nick Kralevich) Crossposted-To: comp.os.linux.development.apps Subject: Re: How to input a character w/out echoing Date: 11 Mar 1996 17:51:23 GMT In article , pretzel wrote: >How does one input a character or string without echoing to the console? >Not that complicated... in DOS it's "getch()" but as I recently found >out, not so in Linux.... Hmmm, this seems to do what your looking for: ===== Begin ===== NAME getpass - get a password SYNOPSIS #include char *getpass( const char * prompt ); DESCRIPTION The getpass function displays a prompt to, and reads in a password from, /dev/tty. If this file is not accessible, getpass displays the prompt on the standard error output and reads from the standard input. The password may be up to _PASSWORD_LEN (currently 128) characters in length. Any additional characters and the terminating newline character are discarded. (This may be different in Linux.) Getpass turns off character echoing while reading the password. RETURN VALUES Getpass returns a pointer to the null terminated password. FILES /dev/tty SEE ALSO crypt(3) HISTORY A getpass function appeared in Version 7 AT&T UNIX. BUGS The getpass function leaves its result in an internal static object and returns a pointer to that object. Sub- sequent calls to getpass will modify the same object. The calling process should zero the password as soon as possible to avoid leaving the cleartext password visible in the process's address space. ===== End ===== (cc'd to author) Take care, -- Nick Kralevich nickkral@cory.eecs.berkeley.edu ------------------------------ From: Richard Ketchersid Subject: Re: 1.3.72 and ppp-2.2.0e - Am I the only one? Date: Wed, 13 Mar 1996 19:01:32 -0800 root wrote: > > In article <4i7682$2ga@bias.ipc.uni-tuebingen.de>, Harald Koenig wrote: > >In <4i487f$9d1@newsflash.hol.gr> devmorfo@patmos.hol.gr (Dimitris Evmorfopoulos) writes: > > > >> root (root@chaos.anarch.org) wrote: > >> : Well, on my system, ppp stopped working when I booted 1.3.72. Yes, I > > > >> I'm using ppp with 1.3.72 right now, and I have no problems whatsoever. > Thanks for your responses, shaniac@primenet.com -- I have problems with diald running pppd, but if you run pppd on its own things work ok. -- Conclusion: The problem is with SLIP. There have benn other signs of problems with SLIP in recent kernels.( btw -- diald uses SLIP. ) Richard Ketchersid ketchers@math.berkeley.edu ------------------------------ From: cec@gryphon.gsfc.nasa.gov (Dr. Charles E. Campbell Jr.) Crossposted-To: comp.unix.programmer Subject: Re: Socket connection between SUN and Linux Date: 14 Mar 1996 14:39:00 GMT Better yet, use saddr.sin_port= INADDR_ANY and get a port that's currently available rather than hardcoding a port that may conflict with someone else's selection. See SSL below -- that's how it does it! /* --------------------------------------------------------------------- */ In article <4i8f8q$3b7@aztec.co.za> soren@aztec.co.za (Soren Aalto) writes: >Cantel Sebastien (fly@mime.univ-paris8.fr) wrote: >: Hi, > >: I trying to write an ftp program. It work very well between 2 Solaris >: systems and between 2 Linux systems. But when I want to run it between >: Solaris and Linux, the client return an error on the connect function. >: The errno code error is a "connection refused" ! > >I would guess that your problem has to do with byte ordering, >specifically the byte ordering of port numbers used in your >connection. > >You probably have some code like > > saddr.sin_port = 2000 ; > >to set the port number where you listen for connections. >Make sure to ntohs these values--I always find I have >problems with byte ordering. The above code should >read > > saddr.sin_port = ntohs (2000) ; /* --------------------------------------------------------------------- */ The Simple Sockets Library makes sockets easy to use! And, it comes as public domain source code: * Resembles C's file i/o Sclose Sgets Sopen Sprintf Sputs Sread Sscanf Swrite Svprintf * Uses currently available ports only, so if a server can be brought up, it will be * Named sockets -- servers have names, not fixed port addresses! (no more REUSEADDR problems, linger setting, zombie sockets!) * Streaming/blocking sockets used * Polling supported (ie. avoids blocking): Stest Smasktest Speek * Numerous functions to handle multiple concurrent clients Smaskfdset Smaskpush Smasktime Smaskunset Smaskwait Smaskget Smaskset Smaskunfdset Smaskuse Sprtmask Smaskpop Smasktest * Timeout control Stimeoutwait Smasktime * The PortMaster: Maintains a distributed port/server-name database Each machine runs a PortMaster which maps servers by name to their ports (transparent to programmers and users!) * Firewall Support The PortMaster can (optionally) only "put through" connections from approved machine sites * Supports "path"ing for sockets: one can use an environment variable to track down a server, by name, on any machine in a group Sopenv * Doesn't need root privileges (that's Unix lingo...) It has been tested on: Unix (SGI, DecStation, AIX, Sun 3, Sparcstation; version 2.02+: Solaris 2.1, SCO), VMS, and MSDOS. It is provided in source code form, of course, and sits atop Berkeley sockets and tcp/ip. You can order the "Simple Sockets Library" from Austin Code Works 11100 Leafwood Lane Austin, TX 78750-3464 USA Phone (512) 258-0785 Fax (512) 258-1342 Inet info@acw.com Ask for the "SSL - The Simple Sockets Library". Last I checked, they were asking $20 US for it. You can also order the "Simple Sockets Library" from COSMIC The University of Georgia 382 East Broad Street Athens, Ga. 30602 USA Phone (706) 542-3265 Fax (706) 542-4807 Ask for the "SSL - The Simple Sockets Library", GSC-13520. COSMIC is asking $400 US for the SSL and $14 for the manual (again, the last I checked). In addition, the Simple Sockets Library has been made freely available at http://www.interlog.com/~vic/sock-faq/ssl.tar.gz by Vic Metcalfe. It was developed at NASA by Dr. Charles E. Campbell, Jr and Terry McRoberts, and is in the public domain (and yes, these two sources do distribute the same software!). The manual is a TeX document, and a PostScript version is also supplied. The SSL has been used to do real-time control of robots and to do real-time sensor reading (we use a bridge to keep down local traffic), so experience shows that the SSL provides a low-overhead interface to Berkeley sockets. EXAMPLES: A) A Server/Accept #include "sockets.h" Socket *srvr; Socket *skt; srvr= Sopen("srvrname","s"); /* open a server called "srvrname" */ skt = Saccept(srvr); /* accept a client on the given server */ Sputs("hello client",skt); /* send a string to the client */ Sclose(skt); /* close the accept Socket */ Sclose(srvr); /* close the server */ B) A Client #include "sockets.h" char buf[BUFSIZE]; Socket *client; client= Sopen("srvrname","c"); /* open client to the "srvrname" server */ Sgets(buf,BUFSIZE,client); /* get a string from the server */ printf("server said <%s>\n",buf);/* print out what the client got */ Sclose(client); /* close the client Socket */ Good Luck! -- Charles E Campbell, Jr, PhD _ __ __ Goddard Space Flight Center / /_/\_\_/ / cec@gryphon.gsfc.nasa.gov /_/ \/_//_/ finger cec@abacus.gsfc.nasa.gov for PGP public key `-( ..... ------------------------------ From: pkc@scs.carleton.ca (Peter Choynowski) Subject: Slow localhost messaging speed ? Date: 11 Mar 1996 17:35:58 GMT I have been measuring localhost messaging performance for work with MPI, and need some advice if my numbers are right, and if they can be improved. Using ttcp, and running both ends on the same host ( P5/133 1.3.{69,72} ) I am getting about 7858 KB/s ( 1.3.69 ), and 8225 KB/s ( 1.3.72 ). The numbers don't look too bad, but compared to a Sun SS20/71 at 17111 KB/s it looks slow. As we all know, most Linux systems can run circles around Suns, so as you can imagine I was not very happy to get this result :) Any pointers or ideas if and when Linux will beat Solaris 2.4 ? TIA, Peter ------------------------------ From: Marko Siladin Crossposted-To: comp.os.linux.development.apps,comp.os.linux.misc Subject: Re: Moo-Tiff (& unable to resolve symbol stat) Date: Fri, 15 Mar 1996 11:03:44 -0500 Marco van Os wrote: > > Jason H. Heiss (jheiss@cco.caltech.edu) wrote: > : marco@strange.iaf.nl (Marco van Os) writes: > > : >Has anyone here used Moo-Tiff? Is it any good? And if I buy it, will > : >I be able to use it with the future releases of Linux? > > : >Marco > > : >-- > : >Marco H.J.W. van Os > > : >But love is not a victory march, it's a cold and broken hallelujah! > : > - LC > > : I couldn't get it to work properly and have had a difficult time getting > : support from Infomagic. I would recommend a different company. YMMV. > > Hm, I've heard a lot other stories too. I got a lot of e-mails telling me that > it's a pleasant environment to program in. I've been usin Moo-Tiff for almost a year now and had little trouble with the a.out version (March 1995 release) I recently upgraded to Slackware 3.0 and Moo-Tiff (august 1995). I downloaded the binary fixes for Slackware from their ftp site. The window manager works fine and I was able to compile some of the demos. BUt when I tried to compile my application I get some "strage" errors. 1. If I use my makefile (which I successfuly used for the old version I get an internal compiler error: ld got signal 11. I traced that down to the fact that I included -L/usr/lib on my gcc line in the make file. When linking gcc tried to link /usr/lib/libc.so.1 -> which has a symbolic link to the gcc libc. When the linker tried to link that library it gave a segmentation fault. Could someone tell me why? 2. After I removed the -L/usr/lib entry the program compiled fine. (not a single complaint). But then when I tried running it, it gave me a runntime error - unable to resolve symbol 'stat'. I don't use stat in my program. I read in the GCC-FAQ that ELF converted stat into a macro rather than being a function. Could it be that the Moo-Tiff libraries depend on the stat function? Did anyone else run into a problem like this? Thanks in advance. -- ...Marko ==================///////////=\\\\\\\\\\\======================= (O) ^ ^ (O) Marko Siladin | | | VCU, Dept. of Math. Science msiladin@athos.mas.vcu.edu \ / Division of Computer Science http://athos.mas.vcu.edu:8080/grad/marko/home.html ==================\\\\\\\\\\\=///////////======================= ------------------------------ ** 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 ******************************