From: Digestifier To: Linux-Misc@senator-bedfellow.mit.edu Reply-To: Linux-Misc@senator-bedfellow.mit.edu Date: Thu, 18 Nov 93 07:13:32 EST Subject: Linux-Misc Digest #311 Linux-Misc Digest #311, Volume #1 Thu, 18 Nov 93 07:13:32 EST Contents: possible bug in select call? (Gautam H. Thaker) Re: Tape Backups H/W (George de Bouter) linux IRQ 5, com3 (Eugene Mondrus) Re: XFREE and OAK Video Board (Lawrence Houston) Re: Why Pascal is not... (was: WANTED: COBOL compiler) (Shun-Chang Tsai) Re: OS/2 and X-Windows (Khan M. Klatt) TLI (Transport Layer Interface) for Linux (EXT Andreas Gerlich) Re: Why Pascal is not... (was: WANTED: COBOL compiler) (Mike Gallo) Re: NIS/YP ?? (Byron A Jeff) Re: New electronic magazine with Linus Torvalds interview (Scott Alfter) intercal & linux (Chris Osborne) Summary of replies: Debugging malloc library available? (Jan Newmarch) ---------------------------------------------------------------------------- From: gthaker@atl.ge.com (Gautam H. Thaker) Subject: possible bug in select call? Date: 17 Nov 1993 19:29:06 GMT The following simple program that calls select() with no file descriptors to wait on but with a valid timeout ranging from 0 to 14000 usec (in 1000 usec) steps does not seem to wait appropriate amount of time. select() seems to just return right away. On SUN OS (and on Intel Paragon) the output is as shown below. #include #include #define NTRIALS 50 void add_timevals(), sub_timevals(); main() { struct timeval start,end,wait; float time; int i,j; char *port; /* "i" is the number of micro seconds we wish to sleep. Select uses microseconds, * isis_sleep_ms() uses ms. * * We sleep for NTRIALS and take an average. */ for (i=0;i< 14000;i+= 1000) { wait.tv_sec = 0; wait.tv_usec = i; gettimeofday(&start,NULL); for(j=0;jtv_usec+=time2->tv_usec; time1->tv_sec +=time2->tv_sec; if (time1->tv_usec >= 1000000) { time1->tv_usec-=1000000; ++time1->tv_sec; } } /* subtract time2 from time1 */ void sub_timevals(time1,time2) register struct timeval *time1,*time2; { time1->tv_usec-=time2->tv_usec; time1->tv_sec -=time2->tv_sec; if (time1->tv_usec < 0) { time1->tv_usec+=1000000; --time1->tv_sec; } if (time1->tv_sec < 0) time1->tv_sec = time1->tv_usec = 0; } /* returns -1,0,1 if time1 if < = > time2 */ int cmp_timevals(time1,time2) register struct timeval *time1,*time2; { if (time1->tv_sec > time2->tv_sec) return 1; if (time1->tv_sec < time2->tv_sec) return -1; if (time1->tv_usec > time2->tv_usec) return 1; if (time1->tv_usec < time2->tv_usec) return -1; return 0; } ============= Output on Linux =========================== Linux> time select_t select timeout: 0 micro_sec - after 50 trials, 0.0292 ms per select select timeout: 1000 micro_sec - after 50 trials, 0.2930 ms per select select timeout: 2000 micro_sec - after 50 trials, 0.3960 ms per select select timeout: 3000 micro_sec - after 50 trials, 0.3961 ms per select select timeout: 4000 micro_sec - after 50 trials, 0.3945 ms per select select timeout: 5000 micro_sec - after 50 trials, 0.3974 ms per select select timeout: 6000 micro_sec - after 50 trials, 0.3957 ms per select select timeout: 7000 micro_sec - after 50 trials, 0.3973 ms per select select timeout: 8000 micro_sec - after 50 trials, 0.3957 ms per select select timeout: 9000 micro_sec - after 50 trials, 0.3974 ms per select select timeout: 10000 micro_sec - after 50 trials, 0.3957 ms per select select timeout: 11000 micro_sec - after 50 trials, 0.5965 ms per select select timeout: 12000 micro_sec - after 50 trials, 0.5964 ms per select select timeout: 13000 micro_sec - after 50 trials, 0.5966 ms per select real 1.0 user 0.0 sys 0.0 ====================== output on SUNOS 4.1.3 ==================== SunOS> time select_t select timeout: 0 micro_sec - after 50 trials, 0.0774 ms per select select timeout: 1000 micro_sec - after 50 trials, 9.9269 ms per select select timeout: 2000 micro_sec - after 50 trials, 9.9901 ms per select select timeout: 3000 micro_sec - after 50 trials, 9.9930 ms per select select timeout: 4000 micro_sec - after 50 trials, 9.9398 ms per select select timeout: 5000 micro_sec - after 50 trials, 9.9979 ms per select select timeout: 6000 micro_sec - after 50 trials, 9.9608 ms per select select timeout: 7000 micro_sec - after 50 trials, 10.0299 ms per select select timeout: 8000 micro_sec - after 50 trials, 9.8313 ms per select select timeout: 9000 micro_sec - after 50 trials, 9.9955 ms per select select timeout: 10000 micro_sec - after 50 trials, 9.8548 ms per select select timeout: 11000 micro_sec - after 50 trials, 19.9973 ms per select select timeout: 12000 micro_sec - after 50 trials, 19.8562 ms per select select timeout: 13000 micro_sec - after 50 trials, 19.8612 ms per select 8.4 real 0.0 user 0.0 sys -- Gautam H. Thaker (gthaker@atl.ge.com) 609-866-6412 (fax x6397. Dialcom 8-777) Martin Adv. Tech. Lab., MS 145-2; Route 38; Moorestown, NJ 08057. 767-4396 (H) ------------------------------ Crossposted-To: comp.sys.ibm.pc.hardware,sci.electronics From: s414557@kub.nl (George de Bouter) Subject: Re: Tape Backups H/W Date: Wed, 17 Nov 93 16:38:52 GMT In article <2cbtmi$lv5@crl.crl.com> rjust@crl.com (Randy Just) writes: >Thomas Monk (tmonk@cse.dnd.ca) wrote: [...] >: I was wondering if anybody would know if there is a piece >: of hardware for a PC that can be used for accessing a VCR, for tape >: back ups. >: Or if someone would know where I could find a schematic for >: building one? [...] >This project can be found in the Circuit Cellar Project file, Volume One. It And can someone tell me what it that is and where I can find it? >included a schematic and source code (assembler and C). It is designed to >work off the expansion port. As to how fast it backs up... Well, you might >time it with a calendar I would guess. > >Have fun. Well, it's a cheap and safe way. ------------------------------ From: em119@merhaba.cc.columbia.edu (Eugene Mondrus) Subject: linux IRQ 5, com3 Date: 16 Nov 93 01:10:58 GMT I have installed Linux on CD-Rom in the past couple of days, and have had some problems. I have not been able to connect to my modem on the Com3 port it uses IRQ5 and the standard IO address for com 3. However when Linux boots it only recognizes COM3 on IRQ 4. I have tried to recompile the kernel, butwith no sucess. First I kept getting kicked out when it it tried to access the domainname file. And then after I remed that part out and recomplied my dependencies it kicked me out with a complie error in one of the scripts. HELP !!!!!!!!!!!!!!!! ------------------------------ From: houston@norton.geog.mcgill.ca (Lawrence Houston) Subject: Re: XFREE and OAK Video Board Date: 16 Nov 93 07:39:52 GMT From article <1993Nov15.183020.22773@matuc2.mat.uc.pt>, by arsenio@mobix.uc.pt (): > I 've installed Linux and i am configuring XFREE, but i can't > get my videoboard chipset right, i 've a OAK videoboard with > 1 Mb of memory. Could someone tell me what chipset to > use in the Xconfig file, email me any help. Thanx. Arsenio: If you have XFree86 2.0, then support for the OAK Chipset is built into XF86_SVGA and should be identified as something like oit067 or oit077. If you are running XFree86 1.3 then you will need the XF86_SVGA Server which Steve Goldman has kindly made available on ns.encore.com (/pub/development/languages/C), to which he has added support for the OAK Chipsets. In fact it was his support for the OAK Chipset which was added to the XFree86 2.0 Distribution! Lawrence Houston - (houston@norton.geog.mcgill.ca) ------------------------------ Crossposted-To: comp.lang.misc Subject: Re: Why Pascal is not... (was: WANTED: COBOL compiler) From: stsai@scws1.harvard.edu (Shun-Chang Tsai) Date: 17 Nov 93 16:58:09 GMT You can ftp Kerbighan's "Why Pascal is Not My Favorite Language" from quartz.rutgers.edu:/pub/computer/langauges. ------------------------------ Crossposted-To: comp.sys.ibm.pc.hardware,comp.os.os2.misc,comp.unix.misc From: n9044144@gonzo.cc.wwu.edu (Khan M. Klatt) Subject: Re: OS/2 and X-Windows Date: 18 Nov 93 00:07:53 GMT coult@magellan (COULT NICHOLAS ASHTON) writes: >I have heard that, with the proper software, it is possible under OS/2 to run an X-server >which serves applications remotely over a modem connection. >I have some questions regarding this. >1. Is it really possible? I've never heard of this being possible... if so I, too, would appreciate answers to the following questions. >2. What software is needed to do this? >3. How fast would the modem have to be for this to be useable? >4. Is there any special software/hardware which would be needed on the other >end of the connection? >Any help is greatly appreciated. If you know of a good place to look for this >information, please let me know. Indeed, me too. >-Nick Khan n9044144@henson.cc.wwu.edu -- -Khan M. Klatt---n9044144@henson.cc.wwu.edu---Western Washington University "It takes about a month for women to get used to me. Kinda like a skittish animal.. they have to sniff you for a while, walk around you until they get comfortable." -Andrew Ghali ------------------------------ From: gerlich@felix.rz.fh-ulm.de (EXT Andreas Gerlich) Subject: TLI (Transport Layer Interface) for Linux Date: 17 Nov 1993 22:59:30 GMT Reply-To: gerlich@felix.rz.fh-ulm.de Hello Linux-Users, I am looking for TLI (Transport Layer Interface) library for Linux. Do you have any kind idea where it can be found? Andreas -- Andreas Gerlich, FH-Ulm, Germany, email: gerlich@fh-ulm.de ------------------------------ From: mikeg@psg.com (Mike Gallo) Crossposted-To: comp.lang.misc Subject: Re: Why Pascal is not... (was: WANTED: COBOL compiler) Date: 17 Nov 93 10:13:05 GMT In article <2cb2sk$fp7@nz12.rz.uni-karlsruhe.de> ig25@fg70.rz.uni-karlsruhe.de (Thomas Koenig) writes: > >You're quite right there. It also would seem that Nikolaus Wirth >agreed, as well; the deficiencies pointed out by Kernighan are all fixed >in Modula-2, especially the "There is no escape" clause. Using the >SYSTEM.ADDRESS type, you can convert from and to pointers as freely as >in C; it just takes more verbiage, and effort. Indeed, Prof. Wirth never intended for Pascal to be a "real world" programming language. It was intended to be a teaching tool only. That Pascal has spread so far beyond it's original domain reflects very well upon its clean design, IMHO. -- A penny saved is a Congressional oversight. ------------------------------ From: byron@cc.gatech.edu (Byron A Jeff) Subject: Re: NIS/YP ?? Date: Thu, 18 Nov 1993 01:05:44 GMT In article , Jocke Berglund wrote: >Do anyone out there hav nis or yp for linux ?? >I'm setting up a network with linux boxes connected to a sun server and >it would be very nice if i could use nis/yp for the passwd.. ====== Begin included text from a couple of saved files =========== Yes. There is a port of NIS to Linux on ftp.uni-paderborn.de pcsoft2/linux-local/yp/yp-linux.tar.gz. I use it for over a month by now without any major problems. I did find a bug in ypset (which is rearely used anyway) but I don't know if it was integrated into the release. I heard (and tried to help to) people with troubles running this package. I suspect that the main problem might stam from the binaries/libraries compiled with somewhat old versions. I will be glad to hear what is up with this package from its author. Cheers, --Amos -- --Amos Shapira (Jumper Extraordinaire) | "It is true that power corrupts, ****** I am running a number of Linux workstations (SLS1.03 PL12 kernel) with (NIS) yp on them and an IBM RS/6000 running as an NIS server. The NIS (yp) I obtained came from: ftp.huji.ac.il /pub/linux/packages/yp ============================= End of included text ======================= So it does exist. However when I pulled it down I found to my horror of horrors that the libraries are 4.4.2. My slick new Slackware 1.1.10 has 4.4.4. I'm afraid that if I downgrade my libraries that other stuff (like X 2.0) will start to break. So my question is: Is ther a NIS compiled with lib 4.4.4? Barring that is there anything I'd have to worry about downgraing to lib 4.4.2? Thanks BAJ --- Another random extraction from the mental bit stream of... Byron A. Jeff - PhD student operating in parallel! Georgia Tech, Atlanta GA 30332 Internet: byron@cc.gatech.edu ------------------------------ From: sknkwrks@blind-lemon.cs.unlv.edu (Scott Alfter) Subject: Re: New electronic magazine with Linus Torvalds interview Date: 18 Nov 1993 01:15:32 GMT In article <2c5qqn$h4a@blaze.cs.jhu.edu> bogstad@blaze.cs.jhu.edu (Bill Bogstad) writes: >In article , >Steve VanDevender wrote: >>An interesting note: Ghostview 1.5 with Ghostscript 2.6.1 will >>only see the first four pages of the Meta magazine Postscript >>document. Ghostview 1.4.1 sees the entire document. > > I had the same problem using Ghostview on a Sun. Looks like >it is not specific to Linux... Trying to read Meta with Preview on a NeXT is even worse...it crashes the system! Ghostscript 2.6.1 on the same NeXT (or on a Sun) works fine...or at least it works fine here. _/_ Scott Alfter (sknkwrks@cs.unlv.edu) Ask me about SoftDAC--digital / v \ Call the Skunk Works BBS today! audio for your Apple IIe/IIc! (IIGS( (702) 894-9619 300-14400 V.32bis 1:209/263 Apple II, IBM, Trek, & more! \_^_/ ---===#### Why be politically correct when you can be RIGHT? ####===--- ------------------------------ Crossposted-To: alt.lang.intercal,connect.audit From: boris@ibmpcug.co.uk (Chris Osborne) Subject: intercal & linux Date: Wed, 17 Nov 1993 21:29:22 GMT To all linux users, INTERCAL is arguably the most important language compiler to have on your system. It offers a mature software development paradigm together with mental stimulation for both you and your processor. Unfortunately the standard release of C-intercal ( version 0.9 ) will not operate correctly under linux. This is mainly because of differences between BISON & FLEX and the standard yacc/lex used in its implementation. The following differences should allow YOU to compile intercal & reap the attendant benefits. These diffs have been extensively tested ( over 3000 man- seconds of testing effort ) So you can see how reliable they must be. Intercal0.9 is available from rochester.edu ( ftp.lle.rochester.edu I think, but have lost my original reference ). After loading the 0.9 release, go to the src subdirectory, use patch to apply the diffs: patch < thisdoc ( whatever you saved this post as ) edit the Makefile & away you go.... N.B. don't worry about the apparently reversed dates below, I did this a while ago & had to re-load the original stuff to calculate the diffs. Chris Osborne boris@ibmpcug.co.uk Warning! this file contains invisible smileys. ........................................................................... *** ick.y Tue Nov 16 18:40:35 1993 --- ick.y.new Mon Sep 13 18:26:34 1993 *************** *** 76,89 **** /* A program description consists of a sequence of statements */ program : /* EMPTY */ ! | program statemnt ; - statemnt: command - | error - {splat(); lose(yylineno,E017);} - ; - /* * Each command consists of an optional label, followed by a preamble, * followed by an optional probability, followed by the statement body. --- 76,84 ---- /* A program description consists of a sequence of statements */ program : /* EMPTY */ ! | program command ; /* * Each command consists of an optional label, followed by a preamble, * followed by an optional probability, followed by the statement body. *************** *** 103,108 **** --- 98,104 ---- /* There are two forms of preamble returned by the lexer */ please : DO {$$ = 1;} | DO NOT {$$ = -1;} + | error {lose( E017,yylineno,(char *)0 );} ; /* Here's how to parse statement bodies */ *************** *** 123,128 **** --- 119,125 ---- | READ_OUT outlist {ACTION($$, READ_OUT, $2);} | GIVE_UP {ACTION($$, GIVE_UP, 0);} | COME_FROM LABEL {NEWFANGLED {TARGET($$,COME_FROM,$2)}} + | BADCHAR {$$=splat();} | error {$$=splat();} ; *** lexer.l Tue Nov 16 18:40:58 1993 --- lexer.l.new Wed Sep 8 17:40:27 1993 *************** *** 1,10 **** --- 1,25 ---- %{ #include #include + #include #include "ick.h" #include "y.tab.h" #include "lose.h" + #ifdef FLEX_SCANNER + #define YYLMAX 1024 + #undef yywrap + extern FILE *yyin; + int yylineno = 1; + #undef YY_INPUT + #undef getc + #define YY_INPUT( buf, result, max ) \ + {\ + int c = getc( stdin );\ + result = ( c == EOF ) ? YY_NULL : ( buf[0] = c, 1 );\ + } + #endif /* FLEX_SCANNER */ + #ifdef MAIN YYSTYPE yylval; #endif /* MAIN */ *************** *** 88,93 **** --- 103,111 ---- { do { c = fgetc(fp); + #ifdef FLEX_SCANNER + if( c == '\n' ) yylineno++; + #endif /* FLEX_SCANNER */ } while (c != '\\' && isspace(c)); } *************** *** 107,117 **** } else if (c == '\n') { - extern char *strcpy(),*malloc(); - *lineptr = '\0'; lineptr = linebuf; textlines[yylineno] = strcpy(malloc(1+strlen(linebuf)),linebuf); return('\n'); } else --- 125,138 ---- } else if (c == '\n') { *lineptr = '\0'; lineptr = linebuf; + #ifdef FLEX_SCANNER + textlines[yylineno++] = + strcpy(malloc(1+strlen(linebuf)),linebuf); + #else /* FLEX_SCANNER */ textlines[yylineno] = strcpy(malloc(1+strlen(linebuf)),linebuf); + #endif /* FLEX_SCANNER */ return('\n'); } else *************** *** 221,227 **** } #endif /* MAIN */ ! static int yywrap() { return(EOF); } --- 242,248 ---- } #endif /* MAIN */ ! int yywrap() { return(EOF); } *** lose.c Tue Nov 16 18:41:28 1993 --- lose.c.new Wed Sep 8 16:12:13 1993 *************** *** 16,22 **** (void) fprintf(stderr, m + 4, s, n); else (void) fprintf(stderr, m + 4, n); ! (void) fprintf(stderr, " CORRECT SOURCE AND RESUBNIT\n"); exit(atoi(m)); } --- 16,22 ---- (void) fprintf(stderr, m + 4, s, n); else (void) fprintf(stderr, m + 4, n); ! (void) fprintf(stderr, " CORRECT SOURCE AND RESUBMIT\n"); exit(atoi(m)); } *** perpetrate.c Tue Nov 16 18:42:25 1993 --- perpetrate.c.new Wed Sep 8 16:26:29 1993 *************** *** 115,121 **** } (void) signal(SIGSEGV, abend); ! (void) signal(SIGBUS, abend); (void) sprintf(buf2,"%s/%s",includedir,SKELETON); --- 115,121 ---- } (void) signal(SIGSEGV, abend); ! /* (void) signal(SIGBUS, abend); */ (void) sprintf(buf2,"%s/%s",includedir,SKELETON); *************** *** 321,327 **** (void) fclose(ifp); } ! static void print_usage(prog,options) char *prog, *options; { --- 321,327 ---- (void) fclose(ifp); } ! void print_usage(prog,options) char *prog, *options; { ------------------------------ From: jan@pandonia.canberra.edu.au (Jan Newmarch) Subject: Summary of replies: Debugging malloc library available? Date: 18 Nov 93 00:28:30 GMT In jan@pandonia.canberra.edu.au (Jan Newmarch) writes: >I have been using the dbmalloc library from Conor P. Cahill on a Sun >quite successfully. I tried to build it on Linux and ran into all sorts >of problems. Has anyone successfully built this library or any other >debugging malloc one? Johannes Grosen suggested using a malloc package called Checker that is on sunsite under "devel". This looked pretty nice (a GNU Purify) but had dificulties with multiple libraries, not seeming to recognise the -L option and other problems. Gustaf Neumann suggested a malloc library from ftp.cs.toronto.edu under the /pub/moraes directory. This worked fine (after fixing one type error in the source) and found my memory bug. Thanks! Jan -- Jan Newmarch, Information Science and Engineering, University of Canberra, PO Box 1, Belconnen, Act 2616 Australia. Tel: (Aust) 6-2012422. Fax: (Aust) 6-2015041 AARnet: jan@ise.canberra.edu.au ------------------------------ ** 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-Misc-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.misc) via: Internet: Linux-Misc@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-Misc Digest ******************************