From: Digestifier To: Linux-Misc@senator-bedfellow.mit.edu Reply-To: Linux-Misc@senator-bedfellow.mit.edu Date: Sun, 21 Nov 93 09:13:08 EST Subject: Linux-Misc Digest #325 Linux-Misc Digest #325, Volume #1 Sun, 21 Nov 93 09:13:08 EST Contents: Re: Mosaic-2.0+term, no luck with the patches (Scott A. Laird) HELP! Anyone using ATI GUP with Linux? (Uday Menon) moving to linux (Cheng Howard Chi Ho) Re: ALPHA99.13r (Mike Horwath) daemons, daemons, gimme them daemons! (Christopher Stevens) Oleo 1.5 binaries (Dragon Slayer) [term compilation error under school's sun workstation (Mike Lee) *** PLEASE READ THIS BEFORE POSTING *** (misc-2.07) (Ian Jackson) Re: how fast is linux? (jacobsd@heart.cor.epa.gov) [accidently "make" everything and drive full (Mike Lee) Re: Strange behaviour of malloc!! Is it (gasp) broken? (jacobsd@heart.cor.epa.gov) Re: Problem with ext2fs (Scott Barker) Which file system do you use and why? (Scott Barker) Re: [term compilation error under school's sun workstation (R. Stewart Ellis) ---------------------------------------------------------------------------- Crossposted-To: comp.infosystems.www,comp.os.linux.help From: lair@kimbark.uchicago.edu (Scott A. Laird) Subject: Re: Mosaic-2.0+term, no luck with the patches Reply-To: lair@midway.uchicago.edu Date: Sun, 21 Nov 1993 05:46:59 GMT They didn't work for me, but I haven't had time to do more than install them and rebuild. I'll probably get around to looking at them closer next week at tehe earliest. Scott. -- Scott A. Laird | "But this goes to 18,446,744,073,709,551,615" lair@midway.uchicago.edu | - Nigel on his 64-bit computer ------------------------------ From: ummenon@nyx10.cs.du.edu (Uday Menon) Subject: HELP! Anyone using ATI GUP with Linux? Date: Sun, 21 Nov 93 06:12:10 GMT Hello there! Are you using a ATI Graphics Ultra Pro VLB video card under Linux ? If so ... Could you please tell me what version of Xwindows you are running ? What version of Linux are you running ? Could you e-mail me a copy of your 'Xconfig' file ? Is it true that Xfree86 1.3 does not support the GUP's chipset ? PLEASE ... e-mail me a response ..... Thanks. MAX ------------------------------ From: cs93156@assn015.cs.ualberta.ca (Cheng Howard Chi Ho) Subject: moving to linux Date: 20 Nov 93 18:41:21 GMT I am considering a move to Linux. I have a 486-33, ATI wonder/XL, 215MB hard drive (400 with stacker), 8MB RAM, USR 14.4 Fax/Modem. My questions: 1. Can the above system use Linux? 2. Can I still use my DOS and Windows applications? 3. What about xWindows? 4. Are there any fax software for Linux? 5. How big a hard drive do I need to have both Windows and Linux? 6. How much RAM do I need? Thanks in advance -- Howard Cheng cs93156@assn001.cs.ualberta.ca hcheng@gpu.srv.ualberta.ca ------------------------------ From: root@jacobs.mn.org (Mike Horwath) Subject: Re: ALPHA99.13r Date: 20 Nov 1993 19:47:30 GMT Defilippo Francesco (clint@hal9000.unipv.it) wrote: : Hello I'v tryed to compile the alpha version of linux 99.13r with new : gcc 2.5.3 but it have fount a lot of error. : | Francesco Defilippo | : | clint@hal9000.unipv.it | wow, this helps us alot there. Please, tell us what the errors were and maybe we can help, and this article should maybe go to comp.os.linux.help instead, but that is left for the original poster to move :) -- Mike Horwath IRC: Drechsau BBS: Drechsau LIFE: lover root@jacobs.mn.org drechsau@jacobs.mn.org Jacob's Ladder 612-588-0201 UUCP, UseNet, Linux files, BBS ------------------------------ From: albert@albert.stu.rpi.edu (Christopher Stevens) Subject: daemons, daemons, gimme them daemons! Date: 21 Nov 1993 09:20:02 GMT As I've looked through various inet files, I've noticed some daemons which I don't have which I would like to have if they exist. I would like any misc. daemon, but I would especially like to know if any of the following exist: timed (is this ntpd/xntpd?) rpc.* (I already posted to c.o.l.help on this) r-daemons like rlogind, rshd, rexecd, rstatd, rusersd, rcpd, rwhod routed (for routing/gatewaying/bridging?) Other daemon-related stuff: pop-2 or pop-3 client rlogin, rsh, rexec, rstat, rusers, rcp, rwho, ruptime Other unrelated stuff: ping that output's "hostname is alive" rather than the detailed output Thanks in advance to anyone who can help me with the location of source for any of these daemons and other things. -- Christopher Stevens stevec2@rpi.edu stevens@acm.rpi.edu albert@albert.stu.rpi.edu root@albert.stu.rpi.edu ------------------------------ From: ds2@dragonslair.caltech.edu (Dragon Slayer) Subject: Oleo 1.5 binaries Date: 21 Nov 1993 09:00:02 GMT Does anyone know where I can get a compiled copy of Oleo, the GNU spreadsheet? There is an old alpha version on sunsite, but it is statically like with some old libs. I can't seem to get it correctly compiled because it seg faults when I enter data. FWIW: My system: gcc 2.4.5 libc 4.4.4 AMD486DX-40 linux 0.99pl13q Please email me directly. I'll post a summary if others are interested. Thanks in advance. -- ds2@dragonslair.caltech.edu Dragon's Lair, USA ------------------------------ From: mlee@eng.sdsu.edu (Mike Lee) Subject: [term compilation error under school's sun workstation Date: 21 Nov 93 10:06:37 GMT hello, I am trying to compile term under school's unix and got an error when I typed "make." The error is "Command faild for file "client.a"" What is causing this and how do I remedy this? I am trying to get term to run so I can have virtual school site at my machine. Thanks...Mike ------------------------------ From: ijackson@nyx.cs.du.edu (Ian Jackson) Subject: *** PLEASE READ THIS BEFORE POSTING *** (misc-2.07) Date: Sun, 21 Nov 1993 11:03:01 GMT Please do not post questions to comp.os.linux.misc - read on for details of which groups you should read and post to. Please do not crosspost anything between different groups of the comp.os.linux hierarchy. See Matt Welsh's introduction to the hierarchy, posted weekly. If you have a question about Linux you should get and read the Linux Frequently Asked Questions with Answers list from sunsite.unc.edu, in /pub/Linux/docs, or from another Linux FTP site. It is also posted periodically to c.o.l.announce. In particular, read the question `You still haven't answered my question!' The FAQ will refer you to the Linux HOWTOs (more detailed descriptions of particular topics) found in the HOWTO directory in the same place. Then you should consider posting to comp.os.linux.help - not comp.os.linux.misc. Note that X Windows related questions should go to comp.windows.x.i386unix, and that non-Linux-specific Unix questions should go to comp.unix.questions. Please read the FAQs for these groups before posting - look on rtfm.mit.edu in /pub/usenet/news.answers/Intel-Unix-X-faq and .../unix-faq. Only if you have a posting that is not more appropriate for one of the other Linux groups - ie it is not a question, not about the future development of Linux, not an announcement or bug report and not about system administration - should you post to comp.os.linux.misc. Comments on this posting are welcomed - please email me ! -- Ian Jackson (urgent email: iwj10@phx.cam.ac.uk) 2 Lexington Close, Cambridge, CB4 3LS, England; phone: +44 223 64238 ------------------------------ From: jacobsd@heart.cor.epa.gov Subject: Re: how fast is linux? Date: 21 Nov 1993 12:00:11 GMT In <1993Nov19.154931.6921@seq.uncwil.edu> cosper@seq.uncwil.edu (Kit Cosper) writes: >clark@ist.flinders.edu.au (Steven R Clark) writes: >>In article i29@Tut.MsState.Edu, mhp1@Isis.MsState.Edu (Michael H Price II) writes: >>>I am thinking about upgrading to linux but a friend told me it ran slow. How >>>fast/slow would it run on a 386DX-40 with 8megs RAM? >>Compared to what? Doing what? >>Give it a go and decide for yourself. > Agreed. Jus fooling around yesterday I did a VERY UN-Scientific >experiment (but it was fun!! :-) I compiled some code I found >that computes PI up to 15K decimal digits and hard wired it to >compute to 5K digits. I then ran it on my Linux Box (486/66 16MB) >and TIME gave me somewhere around 37 seconds real time. First, I don't think we should be comparing hardware speeds -- after all, a CM5/1024 running FEM code will be awfully darn fast, but what does that have to do with Linux? Having said that... I, in my suite of miscellaneous benchmark codes, have a Spigot algorithm PI generator. It pops out 10000 digits of PI. Why anyone would ever want to really do this other than as a programming exercise is somewhat beyond me, but anyway: Sun Sparc 10/51 77.5 seconds Sun Sparc 10/41 96.1 Sun Sparc 10/30 109.9 HP 9000/730 148.4 IBM RS/6000 320 179.1 486/66V (Linux) 185.5 HP 9000/710 203.4 Sun Sparc LX 206.0 HP 9000/705 295.8 Sun Sparc 2 551.3 Sun Sparc 1+ 962.8 Did you compile your code on the Sun with "cc -O", or did you use the far superior "gcc -O2 -mv8"? Sun hasn't changed it's bundled compiler in many years, but gcc compiles and installs with about 5 commands and a couple hours (and about 60 Meg of disk). You could also purchase their unbundled C compiler which also makes fast code. On this example, something is really wrong with their bundled compiler, because it's about 4-5 times slower than gcc. Normally the gap isn't anywhere near so wide -- probably because the code does a lot of integer divides, which SPARC V7 doesn't have. Adding -mv8 to gcc makes it include this instruction. Knowing the proper compile options can make a big difference. >Your mileage may vary, but I say I DEFINITELY have much more bang/$ Well, no argument there. But the Suns aren't exactly the fastest number crunching workstations out there. They do have other advantages that make them quite nice. If you just want raw CPU/dollar, then most workstations are at a huge disadvantage dince they include so much extrenous hardware. A used Honda Civic gets a lot better acceleration per dollar than a new Porsche 911, but you can line up 50 of them and the Porsche will still get there faster. Performance to price curves tend to be exponential. It is surprising how well the 486/66 does. On the other hand, this is integer code, calculating pi notwithstanding. It uses longs only. On something like gaussian elimination (generic code), it manages about 2.6 MFLOPS, where a Sparc 2 (a 1990 machine) gets 3.1, and a Sparc 10/30 gets 7.3. Some people will pay 5 times as much to get 3 times the speed. Our lab does. We also like the MP ability and the >128MB of RAM. To get the same capabilities in a x86 box would raise the price a lot. Anyway, back to Linux! It's fast. Most of that feeling comes from the disk access -- I untar a file under Linux and it's near instant. That same process on a Sun makes a noticible pause. Compiling is almost twice as fast as a Sparc 2. This is much more than the hardware should account for. I would guess that if synchronous writes were turned on, my Linux machine would feel a lot slower. I can't provide any comparisons to *BSD, Mach, Amoeba, or Solaris x86 running on the same hardware, which is probably more to the point of the original poster's question. Overall, Linux seems to get most of what the hardware is capable of though. I've heard that some parts of the networking aren't quite debugged, such as NFS being much slower than it should. Soon though. Maybe someone will get real ambitious and update it to NFS 3 while they're at it... -- Dana Jacobsen jacobsd@solar.cor2.epa.gov Computer Sciences Corp. "I am completely impartial, Dr. Billings. I merely see to it that you teachers say nothing which might threaten our freedom of speech!" -- The Forever Machine, by Mark Clifton and Frank Riley ------------------------------ From: mlee@eng.sdsu.edu (Mike Lee) Subject: [accidently "make" everything and drive full Date: 21 Nov 1993 12:04:48 GMT hello, I was trying to compile Aspirine on my linux and I accidently used the program "MakeDistribution" and it did something horrible to my system. I am not sure exactly what it did, but now I have no room for any writes to the harddrive, cannot even use vi. When I ran the file, it seems to go into everyfile in my directories to do something to every file. I thought it was normal and sat back and watch. Soon, there was an error saying disk write error, full or something like that. I went to the root directory and found a compressed files that just got created and deleted it. I went to several other directories to delete some unwated files and nothing happened. I still could not write to the disk. What happend, what can I do, Thanks for any advise...Mike ------------------------------ From: jacobsd@heart.cor.epa.gov Subject: Re: Strange behaviour of malloc!! Is it (gasp) broken? Date: 21 Nov 93 12:10:19 GMT In hm@seneca.ix.de (Harald Milz) writes: >Several OS's malloc() routines allow for malloc'ing uninitialized memory to >high extents. You start getting problems when you touch (i.e. initialize) the >malloc'ed memory because (I think, please correct me if I'm wrong) the memory >manager releases `physical' memory upon initialization. calloc() fills the >pages with NULLs, as I remember. Thus, the pages get touched immediately >and are spiited out by the memory manager. >Certain OS's even allow for malloc'ing much memory, killing the biggest >process when out of physical (including swap) memory. This OS is called >AIX (yup!). IBM received so many complaints about this that it made the default malloc be the POSIX compilant version, which acts as all the mallocs we know and love do. The old one (lazy allocation, shotgun conflict resolution) is available by linking to a special library. At least that's what a recent press release said. I'm having some problems with my Linux machine doing the same thing though. I have a program that statically allocates a large array, 9 meg to be exact. When I run multiple copies of this, ps shows that the memory is not actually allocated until it is used. When I run enough copies so that it uses more swap space than exists, they slowly allocate the last few pages, then the machine locks up for 5 minutes to 2+ hours (5 minutes with 16 MB RAM, no swap, no X, 2 copies running; with 40MB swap, X running, 6 copies, I pressed the restart button after waiting two hours). It then kills off any process that asks for more memory until it's happy. This, in the small case, meant that a root shell got killed when I, as a normal user, ran my program. This is with Slackware 1.0.2, Linux 0.99.13p. Anyone else see this? I don't really mind it's behavior so much as I don't like the large timeout. I read through the kernel memory routines, and I see where it runs out of memory and kills off the process, but I don't see why it would wait so long. Any clues for turning on kernel debugging? -- Dana Jacobsen jacobsd@solar.cor2.epa.gov Computer Sciences Corp. "But it is infantile to think that God created in order that His creation bow and scrape from morning till night, in order that He be loved, in advance installments, for what will come Yonder, if Here happens not to be to one's liking..." -- The Twenty-First Voyage, in the Star Diaries, by Stanislaw Lem ------------------------------ Crossposted-To: comp.os.linux.help From: barker@enel.ucalgary.ca (Scott Barker) Subject: Re: Problem with ext2fs Date: Sun, 21 Nov 1993 12:51:48 GMT > I just discovered a problem with ext2fs, I think. I'm running on a 50 Mhz > 486DX/2 with EISA bus, ADAPTEC 1542b SCSI controller, and a SCSI hard drive. > Anyway, I noticed that when I manipulate large files - whether copying them > from another partition or creating them by gunzipping smaller files, or by > copying them from tape - the file system becomes corrupted. Has anyone else > had this problem with large files and ext2fs? By large, I mean anything over > about 7 Meg. Is this perhaps due to some limitation in the ext2fs? The problem is definitely in ext2fs. I have tried to repeat the problem with xiafs and msdos file systems, and cannot. I only have the problem with ext2fs. Anybody have any ideas what the problem might be? -- Scott Barker barker@enel.ucalgary.ca "Any fool knows men and women think differently at times, but the biggest difference is this. Men forget, but never forgive; women forgive, but never forget." - Thom Merrilin, "The Dragon Reborn" (by Robert Jordan) ------------------------------ Crossposted-To: comp.os.linux.help From: barker@enel.ucalgary.ca (Scott Barker) Subject: Which file system do you use and why? Date: Sun, 21 Nov 1993 12:56:24 GMT I guess some of you may have noticed my recent postings lately. Getting to be a lot of them. I've been dealing with linux since 0.99pl6, and was using pl6 until recently. With no problems. Now, all of a sudden, the newer versions are causing me trouble (pl12 and pl13). Anyway, on to the topic. I was wondering what file system (ext2, xia, minix) everyone was using, and why. Disadvantages, advantages, etc. I'll summarize the results and post them. It's mostly for my own curiosity, but I'm sure others are curious to. E-mail me at the address below with your responses. Thanks in advance. -- Scott Barker barker@enel.ucalgary.ca "Any fool knows men and women think differently at times, but the biggest difference is this. Men forget, but never forgive; women forgive, but never forget." - Thom Merrilin, "The Dragon Reborn" (by Robert Jordan) ------------------------------ From: ellis@nova.gmi.edu (R. Stewart Ellis) Subject: Re: [term compilation error under school's sun workstation Date: 21 Nov 93 13:29:08 GMT mlee@eng.sdsu.edu (Mike Lee) writes: >hello, I am trying to compile term under school's unix and got an error >when I typed "make." The error is "Command faild for file "client.a"" >What is causing this and how do I remedy this? I am trying to get term >to run so I can have virtual school site at my machine. >Thanks...Mike What model of sun workstation is it? Sun 2? Sun 3? 386i? What version of SunOS? 3.5? 4.0.x? 4.1.x? 5.x? Which compiler? bundled? unbundled? gcc? (which version of gcc?) ALWAYS tell people who are being asked for help enough so they can be reasonably confident of their diagnoses. Also give complete error messages. There are way too many variables that you do not specify to be sure, but my guess is that you were using the bundled C compiler on SunOS 4.1.x, which will not work because term requires an ANSI compliant compiler. See if gcc is installed on your sun computer, or use the unbundled C compiler if gcc is not. If there is not an ANSI compiler on your system there are other ways of approaching it, but they are way too difficult for you if you are having trouble with this. IF you have an ANSI compiler, specify it in the Makefile where it says: CC=cc Comment that out and put: CC=gcc #or whatever Hope this helps. -- R.Stewart(Stew) Ellis, Assoc.Prof., (Off)313-762-9765 ___________________ Humanities & Social Science, GMI Eng.& Mgmt. Inst. / _____ ______ Flint, MI 48504 ellis@nova.gmi.edu / / / / / / Gopher,News and sendmail maintainer, all around hack /________/ / / / / ------------------------------ ** 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 ******************************