From: Digestifier To: Linux-Development@senator-bedfellow.mit.edu Reply-To: Linux-Development@senator-bedfellow.mit.edu Date: Sun, 3 Apr 94 01:13:06 EST Subject: Linux-Development Digest #599 Linux-Development Digest #599, Volume #1 Sun, 3 Apr 94 01:13:06 EST Contents: Re: IDE transfer and interrupts, again (Mark Lord) Re: NFS timeouts (Alan Cox) Re: TCP/IP-Bug in 1.0 Kernel? (Klaus Schneider) [Q] Compiling DiamondBase 0.2 (Chris Curtis) WANTED: Sound for pc speaker (randall kevin sharpe) What is GIGO+ (Joseph W. Vigneau) Re: IPX compliancy? (Rob Janssen) Re: TCP/IP-Bug in 1.0 Kernel? (Brandon S. Allbery) Re: BusLogic BT445S driver? (Brandon S. Allbery) Re: telnet in/out problems (Rob Janssen) Re: tcp -> term connection??? (Olaf Titz) Re: Tips on writing serial drivers. (Richard Hodson) Re: mt commands that won't work (Zenon Fortuna) serial port interaction with buffer & swap (JL Gomez) Adaptec 1542C with HD >1GByte (Sam Daniel) Adaptec 152x patch. Please REPOST! (Jonathan Clark) Re: tcp -> term connection??? (Olaf Titz) ---------------------------------------------------------------------------- From: mlord@bnr.ca (Mark Lord) Subject: Re: IDE transfer and interrupts, again Date: 31 Mar 1994 21:11:39 GMT In article <2nenft$p0p@nz12.rz.uni-karlsruhe.de> uknf@rzstud1.rz.uni-karlsruhe.de writes: >In article <2ndj10$8gb@levelland.cs.utexas.edu>, >Daniel Aaron Supernaw-Issen wrote: > >> the following: Whenever I have disk activity, the mouse jumps around under X. >> This has made the system unusable whenever there is any real swapping going >> on. Worse yet, I can't manage to control the mouse enough to be able to > >The problem is that with 32 sectors read at once, you will certainly >lose interrupts. The solution is to enable interrupts during disk >transfer (not to use a 16550 even for the mouse at 1200bps...) > >Someone has claimed the following patch doesn't work for all disk >drives. (No, I don't know which) Careful testing with a r/o partition Equivalent functionality to this is already incorporated into the IDE patches. -- mlord@bnr.ca Mark Lord BNR Ottawa,Canada 613-763-7482 ------------------------------ From: iiitac@uk.ac.swan.pyr (Alan Cox) Subject: Re: NFS timeouts Date: Thu, 31 Mar 1994 20:45:45 GMT In article <1994Mar29.223105.5702@unlv.edu> ftlofaro@unlv.edu (Frank Lofaro) writes: >In article <1994Mar29.013504.25381@cc.gatech.edu> byron@cc.gatech.edu (Byron A Jeff) writes: >It should be easy to get the buffers up to almost 4k, trivial in fact. >After that you'd need to hack kmalloc, use vmalloc, or have the net code >use 2 buffers per large packet or somesuch. The NFS will work up to about 3.5K but for throughput you normally find 1K is better than 2K and 3K is generally bad news (BSD works in 4K disk blocks so 3K falls off boundaries badly) and 2K fragments. > >P.S. Is calling vmalloc from an interrupt bad? Unbelievably so! Alan ------------------------------ From: uk0q@rzstud1.rz.uni-karlsruhe.de (Klaus Schneider) Subject: Re: TCP/IP-Bug in 1.0 Kernel? Date: 2 Apr 1994 14:57:27 GMT Rene COUGNENC (rene@renux.frmug.fr.net) wrote: : Ce brave Rene COUGNENC ecrit: [...] : After 11 hours uptime, 1.0.3 started missing most received characters : on the serial port. Same happens to me! If the problems occurs with SLIP, kermit and Seyon are also infected. That leads me to believe the problem is somewhere within the serial ports drivers or their interrupt handlers. After all it seems to be a wise thing to go back to linux 1.0.2. The only thing that really surprises me is the delay of several hours (or usages of the serial port?) before the problem shows up. [...] Klaus =============================================================================== Klaus Schneider Email: root@ks1i486.dialup.xlink.net Student of Informatics uk0q@rzstud1.rz.uni-karlsruhe.de University of Karlsruhe, Germany klsc@delphi.com =============================================================================== ------------------------------ From: cmcurtis@xmission.com (Chris Curtis) Subject: [Q] Compiling DiamondBase 0.2 Date: 31 Mar 1994 14:11:01 -0700 Hope there are some C++/Linux gurus out there who can help me... I am trying to install DiamondBase 0.2 on my system (Linux 1.0, Slackware 1.1.2) but I keep getting this error: tigger> make gcc -I. -c -g -Wall btree.cc In file included from /usr/g++-include/iostream.h:34 from btree.cc:168: /usr/g++-include/streambuf.h:220: `ios::operator void *(...)' must take `void' /usr/g++-include/streambuf.h:220: confused be earlier errors, bailing out I am far from being a C++ wizard, so I don't even know where to start looking to fix this. Any ideas on this? Also - has anyone managed to compile REQUIEM under Linux? Thanks in advance..... -- ======================================================================== | cmcurtis@xmission.com | This space for rent. | | Absentee expatriate and minor field deity | | ======================================================================== ------------------------------ From: sharpe@sparc60.cs.uiuc.edu (randall kevin sharpe) Subject: WANTED: Sound for pc speaker Date: Sat, 2 Apr 1994 14:24:24 GMT I though I saw somewhere a posting of a sound driver for the standard pc speaker. Where is the post or (better yet) where is the driver itself ? Thanks ------------------------------ From: joev@bigwpi.WPI.EDU (Joseph W. Vigneau) Subject: What is GIGO+ Date: 2 Apr 1994 15:51:14 GMT I don't get it. I just read at least 30 articles in this group from all around the world, all of which had the lines X-Mail-Agent: GIGO+ sn 51 at logo vsn 0.98w32 Organization: Private GIGO testing site in their header... Is there someting I'm missing? -- joev@wpi.edu, joev@hotblack.schunix.dmc.com WPI Computer Science Linux! Click Here! ------------------------------ From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: IPX compliancy? Date: Sat, 2 Apr 1994 11:21:53 GMT Reply-To: pe1chl@rabo.nl In mikenel@netcom.com (Michael Nelson) writes: >Alan Cox (iiitac@uk.ac.swan.pyr) wrote: >: In article <1994Mar14.185508.46244@ucl.ac.uk> zceed04@ucl.ac.uk (Mr Ivan Alastair Beveridge) writes: >: >Basically, I was wondering if anyone has made Linux compliant with Netware >: >at all. As I do not really know much about protocols, I cannot really ask >: >much more than this. >: There is a beta test IPX layer for Linux, but no netware support. Novell >: guards its netware details with lawyers and complex licensing agreements >: involving thousands of dollars. So forget it - Linux does Lan manager and NFS > ^^^^^^^^^^^ >WHERE!? WHERE!? >I am looking for LM support on Linux... please e-mail me anything you can >tell me (mikenel@netcom) :-). Why is it that still so few people read comp.os.linux.announce?? Rob -- ========================================================================= | Rob Janssen | AMPRnet: rob@pe1chl.ampr.org | | e-mail: pe1chl@rabo.nl | AX.25 BBS: PE1CHL@PI8UTR.#UTR.NLD.EU | ========================================================================= ------------------------------ From: bsa@kf8nh.wariat.org (Brandon S. Allbery) Subject: Re: TCP/IP-Bug in 1.0 Kernel? Date: Sat, 2 Apr 1994 15:30:30 GMT In article <2nfbfj$56@renux.frmug.fr.net>, cougnenc@hsc.fr.net (Rene COUGNENC) says: +--------------- | After 11 hours uptime, 1.0.3 started missing most received characters | on the serial port. | | So the problem seems to be in patch3, and related to something about | interrupts... +------------->8 Please post hardware as well as kernel version: 1.0.5 has been running on the Twit for 40 hours straight, with a SLIP connection to kf8nh (still 0.99p15j) and X clients running over that SLIP connection. The connection is exercized at *least* once per second. The Twit is as reliable as it ever is :-) Leading Edge WinPro/486e 486SX25 processor 4MB RAM 16MB swap No external cache 170MB IDE hard drive Two 16450 serial ports SLIP on /dev/cua1 at 19200 baud Slackware 1.1.1 (TransAmeritech) with 1.0.5 kernel XF86_SVGA (WD90c31) running locally olvwm, (ol)clock, xmss, xterm running remotely with local display (Don't try this at home, kids! Running X clients remotely with the above setup is NOT usable; thrashing is faster... It's a configuration intended for "burn-in" testing, not for real work.) So far I've heard several claims relating to network, SCSI (and 1542 in particular), etc. being the cause; I think SCSI has been pretty much ruled out by Rene's system, and mine seems to rule out (non-Ethernet) networking. Since it's working on the Twit I'll try bringing it up on kf8nh sometime this weekend and see what happens. ++Brandon -- Brandon S. Allbery kf8nh@kf8nh.ampr.org bsa@kf8nh.wariat.org The FUDs at Microsoft are shouting "Kill The Wabi!" ------------------------------ From: bsa@kf8nh.wariat.org (Brandon S. Allbery) Subject: Re: BusLogic BT445S driver? Date: Sat, 2 Apr 1994 15:37:06 GMT In article <1994Mar31.113140.3083@pe1chl.ampr.org>, pe1chl@rabo.nl says: +--------------- | It does say ALPHA, doesn't it? +------------->8 Even the (not only ALPHA, but PRIVATE) iBCS2 support code comes with installation instructions... ++Brandon -- Brandon S. Allbery kf8nh@kf8nh.ampr.org bsa@kf8nh.wariat.org The FUDs at Microsoft are shouting "Kill The Wabi!" ------------------------------ From: rob@pe1chl.ampr.org (Rob Janssen) Subject: Re: telnet in/out problems Date: Sat, 2 Apr 1994 21:32:23 GMT Reply-To: pe1chl@rabo.nl In <2nestl$bg6@cmcl2.NYU.EDU> liuyu@acf2.nyu.edu (liuyu) writes: >williams@dewey.nl.nuwc.navy.mil (Dave Williams) writes: >I seem to have a similar problem. Sometimes I can't telnet, ftp, or >rlogin into Linux box. Sometimes there is a huge delay before login >prompt shows up. It just hangs there with connected message. During the >time I can't login, it doesn't seem to get emails either. I suspect that >there are bugs in TCP/IP, inet, or NE2000 driver. I tried a SMC ultra >card. It was worse than NE2000. >Once I get in if I'm lucky, there is no problem getting emails. >BTW, telnet, ftp, and rlogin from Linux to outside machine is no >problem. The problem is there only when I try to get into Linux. >Strange.... >Any ideas? It means you have a problem with your nameserver. Check /etc/resolv.conf and the nameserver configuration files. Rob -- ========================================================================= | Rob Janssen | AMPRnet: rob@pe1chl.ampr.org | | e-mail: pe1chl@rabo.nl | AX.25 BBS: PE1CHL@PI8UTR.#UTR.NLD.EU | ========================================================================= ------------------------------ From: uknf@rzstud1.rz.uni-karlsruhe.de (Olaf Titz) Subject: Re: tcp -> term connection??? Date: 31 Mar 1994 19:54:38 GMT In article <1994Mar22.210615.5628@jupiter.sun.csd.unb.ca>, Neil McIsaac wrote: > I was wondering if ne1 could help me to find out what things I > should look for to change tcp network connections to term network > connections? In general, an active Internet connection is set up with the steps (error checking etc. omitted for brevity): 1. get IP and port #, usually involving gethostbname() 2. s = socket(AF_INET, SOCK_STREAM, 0); 3. connect(s, &adr, sizeof(adr)); After that, s is the fd to read/write to. This translates for term into: 1. nothing (term lets the remote term do that) 2. s = connect_server(term_server); 3. send_command(s, C_PORT, 0, "%s", hostport); send_command(s, C_DUMB, 1, 0); where "adr" is a struct sockaddr_in and "hostport" a string of the format "hostname:portno". For a passive connection, it gets trickier. If you bind to a fixed port, it goes like: 1. s0 = socket(AF_INET, SOCK_STREAM, 0); 2. bind(s0, &adr, sizeof(adr)); listen(s, n); /* n is arbitrary */ 3. s = accept(s0, &adr2, &sadr2); before which adr contains 0 as host and the port number, and after which s is the connected fd and adr2 contains the calling address. Which is termified into: 1. s0 = connect_server(term_server); 2. send_command(s0, C_BIND, 0, "%d", port); 3. read(s0, buff, sizeof(buff)); p = atoi(buff); s = connect_server(term_server); send_command(s, C_ACCEPT, 0, "%d", p); There is no direct mapping between the various stages of operation in this case, especially not between accept() and C_ACCEPT. The point where the waiting for an incoming connection takes place is at the beginning of step 3. If you need an arbitrary port, it is even more complicated: 1. s0 = socket(AF_INET, SOCK_STREAM, 0); 2. bind(s0, &adr, sizeof(adr)); listen(s, n); /* n is arbitrary */ 3. getsockname(s0, &adr, &sadr); /* And communicate the port number in some way */ 4. s = accept(s0, &adr2, &sadr2); translates to: 1. s0 = connect_server(term_server); 2. send_command(s0, C_BINDN, 0, "%d", port); 3. read(s0, buff, sizeof(buff)); /* Now buff contains host IP (important! it's the remote's IP) and port in the format which is required by FTP, i.e. "a,b,c,d,e,f" */ 4. read(s0, buff, sizeof(buff)); p = atoi(buff); s = connect_server(term_server); send_command(s, C_ACCEPT, 0, "%d", p); If you have a WWW client, get my term patch for Lynx from http://rzstud1.rz.uni-karlsruhe.de/~uknf/sw/Lynx-term.txt (sorry, no ftp here) which should make it rather obvious what to look for. Hope this helps. Olaf -- olaf titz o olaf@bigred.ka.sub.org praetorius@irc comp.sc.student _>\ _ s_titz@ira.uka.de LINUX - the choice karlsruhe germany (_)<(_) uknf@dkauni2.bitnet of a GNU generation what good is a photograph of you? everytime i look at it it makes me feel blue ------------------------------ From: richard@radar.demon.co.uk (Richard Hodson) Subject: Re: Tips on writing serial drivers. Date: Thu, 31 Mar 1994 19:53:11 +0000 Cut the legs off with wire cutters, then desolder the pins one by one. This assumes you do not want the chips afterwards. -- Richard Hodson | richard@radar.demon.co.uk And his amazing fetish for dangly earrings... | rhodson@cix.compulink.co.uk ------------------------------ From: zenon@resonex.com (Zenon Fortuna) Subject: Re: mt commands that won't work Date: Sat, 2 Apr 1994 14:34:50 GMT In article ehhchi@epx.cis.umn.edu (Ed H. Chi) writes: >In article <1994Mar30.214944.456@ancient.trillium.se>, >Wolfgang Feldmann wrote: >>I wonder if it's supposed to be so that mt commands others than erase, >>reten, rewind or weof won't work. I'm using linux 1.0, the newest >>ftape and slackware's mt. > >"mt fsf" works, but "mt bsf" doesn't work well for me, (it seems to get >very confused.) "mt status" never really tell anything useful (I wish it >did.) I don't think "fsr" and "bsr" works right, but I have no proof of >this. "mt eod/seod" works. > >I have a Tandberg 3600 SCSI tape drive. > >>I'm longing for beeing able to back up more than one tar file to a tape. > >I can do it by rewinding to the beginning, and then do a "mt eod". And >then I could write my second tar file. > >I have not looked at st.c, so I don't know if that is the cause of the >broken-ish behavior I am experiencing. > The mt(1) from the SLACKWARE 1.2.0 is the GNU mt version 2.3 (type "mt --version"). Interesting, in ftp.cdrom.com:/pub/linux/slackware_source/ap/mt_st you can find a mt_st-0.1.tar.gz file which contains the Berkeley version of the mt(1) - version 5.6 6/6/91 - which compiles find under Linux and works (in some aspects, at least) better than the above GNU version. The "mt eof" works OK, and even "mt rew" works find (so I don't have to type the "mt rewind" :)). Still, this "mt_bsd bsf" has similar problems like "mt bsf", so maybe there is a driver problem with the backspacing, rather than with the mt(1) itself. I will investigate it more closely tonight. zenon@netcom.com ------------------------------ From: gomez@enuxsa.eas.asu.edu (JL Gomez) Subject: serial port interaction with buffer & swap Date: Fri, 1 Apr 1994 14:37:47 GMT I'm running PL 1.0 of the kernal. One thing I'm noticing is that the buffer and swap space is expanding. When I was running PL15 of the kernal, I did not have this problem. Never used swap at all. Is there something going on with serial port code? Anyone has a fix? Here is my 'free' output: total used free shared buffers Mem: 15136 14952 184 3064 10784 Swap: 8188 128 8060 -- gomez@enuxsa.eas.asu.edu ------------------------------ Crossposted-To: comp.os.linux.help,comp.os.linux.misc From: daniel@netcom.com (Sam Daniel) Subject: Adaptec 1542C with HD >1GByte Date: Sat, 2 Apr 1994 20:30:30 GMT I have a friend with no access to the Net, so I'm posting this for him: He's running a 486/66 with 16MB memory, an Adaptec 1542C, and a Quantum 2.4Gbyte SCSI drive. His 1542C came with the old prom, that did not support disks above 1Gbyte. Linux worked fine, and he installed it in the lower 1GB of his disk successfully. When he upgraded the prom to use >1Gb disks, Linux stopped working, in fact, won't even install. He can disable the extended access feature in the 1542C and everything works again. Obviously, he can't afford to give up 1.4Gb of disk space just to use Linux. Can anyone help? E-mail or followups here, I read them all.... -- Sam Daniel * UUCP (Smart): daniel@netcom.com Unisys Government Systems * (Dumb): {...}!uunet!netcom!daniel 1001 W. Maude Ave. * Voice: 1-408-235-2661 Sunnyvale, CA 94086 * Disclaimer: It's only my opinion, but it's a DAMN good one... -- -- Sam Daniel * UUCP (Smart): daniel@netcom.com Unisys Government Systems * (Dumb): {...}!uunet!netcom!daniel 1001 W. Maude Ave. * Voice: 1-408-235-2661 Sunnyvale, CA 94086 * Disclaimer: It's only my opinion, but it's a DAMN good one... ------------------------------ From: jc@sleepy.cc.utexas.edu (Jonathan Clark) Subject: Adaptec 152x patch. Please REPOST! Date: 2 Apr 1994 16:55:11 -0600 I heard there was a patch to the Linux Adaptec 152x driver posted within the last two weeks (on this newsgroup?) I do not see it on my news host anymore, could someone repost it or tell me where I can find it. Thanks, Jonathan Clark ------------------------------ From: uknf@rzstud1.rz.uni-karlsruhe.de (Olaf Titz) Subject: Re: tcp -> term connection??? Date: 31 Mar 1994 21:47:25 GMT I wrote: > send_command(s, C_ACCEPT, 0, "%d", p); After this you can give a C_DUMB like in the first example too, since the channel s will serve for data only and get no commands. Olaf -- olaf titz o olaf@bigred.ka.sub.org praetorius@irc comp.sc.student _>\ _ s_titz@ira.uka.de LINUX - the choice karlsruhe germany (_)<(_) uknf@dkauni2.bitnet of a GNU generation what good is a photograph of you? everytime i look at it it makes me feel blue ------------------------------ ** 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-Request@NEWS-DIGESTS.MIT.EDU You can send mail to the entire list (and comp.os.linux.development) via: Internet: Linux-Development@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 Digest ******************************