From: Digestifier To: Linux-Misc@senator-bedfellow.mit.edu Reply-To: Linux-Misc@senator-bedfellow.mit.edu Date: Fri, 24 Sep 93 00:13:13 EDT Subject: Linux-Misc Digest #153 Linux-Misc Digest #153, Volume #1 Fri, 24 Sep 93 00:13:13 EDT Contents: need advice from Linux genii (David E Nichols) Re: linux & 3c501 ? experiences ? (Donald J. Becker) Re: Token Ring/Nameserver (Donald J. Becker) Re: [Q] Can I use a Thomas-Conrad arcnet card with NET-2? (Donald J. Becker) Re: linux & 3c501 ? experiences ? (Donald J. Becker) ---------------------------------------------------------------------------- From: davidn@cs.tamu.edu (David E Nichols) Subject: need advice from Linux genii Date: 24 Sep 1993 03:17:38 GMT Hello fellow Linuxophiles. I hope some kind soul can lend some Linux knowledge to me. I am trying to write an operating system kernel in C++. I have a solid scheduler and I am writing the keyboard and console drivers. I have implemented my scheduler using threads, with tasks sharing their timeslices between all their threads, where a non-multithreaded task has one task description block and one thread description block, and where context swtiching between threads is (I believe) very fast and efficient. In addition I am trying to make it as object-oriented as possible, where the root object is the system, and all kernel objects are registered with (and descendents of) their parents (for example, right now the scheduler is a child of "system", and tasks are children of "scheduler", and individual threads are children of their respective tasks). This way objects can inherit characterists from their parents and also there can be a uniform and consistent way of communicating between objects, as this behaviour can be defined on base objects and overridden on objects that require more specialized code. At any rate, that is the background for the following questions: 1) in my code, a wake_up call is called from the interrupt handler (it looks like Linux does it this way too), is this the correct method (it dodn't seem like it at first but now it does). 2) When I switch to gcc, I will be doing my development under Linux again, and I need to know where my DS starts and stuff like that so I can write my own loader. How are Linux object files structured? Are they standard Unix object files? Is there a standard Unix object file format? 3) Where can I find information on COFF or ELF binaries or whatever they are? (as you can see I'm sadly and pitifully informed, but I would like to use a standard object file format) 4) I would like to be able to calculate load values and CPU usage figures for my tasks and threads, but since the code I am using to program the timer was taken from CTASK and I don't know exactly what its doing (i.e. I don't really know how many interrupts per second I'm getting), can anyone point me in the right direction for information on programming the PC timer? 5) I am using the source code for Linux to get lots of good information. However, sometimes it is difficult to read. Are the sources to Minix available? It would be nice to have a second source to cross-reference things. 6) Can anyone suggest a good source for information on the pc's hardware? The only source I have right now is Linux source code, and I'm having trouble writing the keyboard and console drivers. I could take a lot of the code wholesale but I don't want to do that, because then I might not really understand what's going on and be unable to fix bugs later or add features to code that I did not write. 7) can anyone point me to info on POSIX, like what constitutes the newest POSIX standard? Thank you very much for any help and or info you may provide. By the way, I am just writing this OS for my own curiosity, but it has been working out well so far and I would like to keep going with the project. I would like to put a Unix-like API on top of the kernel, but I don't want it to be Unix. It would be great if it could run multiple APIs but I will be happy if it runs one. My 386 book won't arrive for another few weeks so I won't be getting in to protected mode until then. thanks a lot David Nichols davidn@cs.tamu.edu -- ============================================================ David E. Nichols Center for Advanced Moose Technology davidn@cs.tamu.edu Director - Antler Division Texas A&M University "A moose is not a toy" ------------------------------ From: becker@super.org (Donald J. Becker) Subject: Re: linux & 3c501 ? experiences ? Date: Thu, 23 Sep 1993 04:36:29 GMT In article , Jussi Savola wrote: >I know I know, I am masochistic type, but my braindamage is much worse >than that of those 501's... :-) Hmmm, that a lot of braindamage. A really really, well, billleions and billeions ... a monstrous amount of braindamage. >PS. Anyone wanna buy a cheap ethernet card ? :-) I have a standing offer: I'll pay $2 for each 3c501 shipped to me postpaid, but only if you include the 'T' connector and the jumpers. $2.50 if you just send the 'T', jumpers, and address PROM and promise to destroy the board. The 3c501 is an ancient product with documentation that's no longer being printed. It has a single packet buffer that is shared between transmit and receive. By this I mean that when you have a packet it cannot receive another packet. While you are offloading that packet it cannot receive a packet. While you are loading a new packet to transmit it cannot receive a packet. While you are waiting to transmit it cannot receive a packet. While... well, you get the idea -- you miss most packets. And I haven't even started describing the bugs. -- Donald Becker becker@super.org IDA Supercomputing Research Center 17100 Science Drive, Bowie MD 20715 301-805-7482 ------------------------------ From: becker@super.org (Donald J. Becker) Subject: Re: Token Ring/Nameserver Date: Fri, 24 Sep 1993 01:04:42 GMT In article , Glenn T. Wasserman wrote: >I've looked at the FAQ and can't seem to find this out... will >Linux function as a domain name server? Yes. SLS distributions even include all of the files you'll need. >Also, since I lost the compatibility list, can Linux work on a token >ring network with IBM 16/4 adapters? No, there isn't a Linux token-ring driver. I looked at writing a token ring device driver, and concluded that the while the hardware interface wasn't too difficult to do, the writing the support for source routing would take significantly longer than I was willing to spend on an expensive and dying technology. -- Donald Becker becker@super.org IDA Supercomputing Research Center 17100 Science Drive, Bowie MD 20715 301-805-7482 ------------------------------ From: becker@super.org (Donald J. Becker) Subject: Re: [Q] Can I use a Thomas-Conrad arcnet card with NET-2? Date: Fri, 24 Sep 1993 02:07:58 GMT In article <27pt6q$ne0@tamsun.tamu.edu>, Owen P. Crow wrote: >I have access to a couple of Thomas-Conrad 16-bit arcnet cards, so I'd like >to know if they will work with my linux box. The only other thing I know >about them is they use a coax not twisted pair cable. An Arcnet driver hasn't been written yet. -- Donald Becker becker@super.org IDA Supercomputing Research Center 17100 Science Drive, Bowie MD 20715 301-805-7482 ------------------------------ Crossposted-To: comp.os.linux.development From: becker@super.org (Donald J. Becker) Subject: Re: linux & 3c501 ? experiences ? Date: Fri, 24 Sep 1993 02:33:18 GMT OK, there seems to be a number of 3c501 fans out there, and a few of them insist on telling my why it isn't so bad. They are wrong, just because it works fine on your 8088 doesn't mean it's OK hardware. I know what I'm talking about: "seen it, been there, done it". But in a fit of madness I wasted a whole day updating my 3c501 driver and then trying to track down a few more of the 3c501 glitches. It now works well enough to NFS mount filesystems, but the receiver still occasionally hangs. I'm mostly certain that this is a hardware bug. When it hangs, the next set of outgoing packets will reset the board, but that's only useful if you have something occasionally generating outgoing packets. I'll let this out for "pre-alpha" testing, under the following conditions: This is unsupported code. I know my usual copyright says all the code is unsupported, but this is _really_ unsupported. I DON'T want to see bug reports, and I'll accept bug fixes only if I'm in a good mood that day. I don't want to see a fest of "Linux ethercards for sale" postings. A bunch of people have picked dozens of "dumpster special" 3c501s, and they hope to sell them at rip-off prices. A 3c501 is barely worth the shipping cost, and if I see people trying to sell them here by claiming "supported by Linux" I _will_ flame them. They are _not_ supported by Linux. I don't want to flamed later for putting out bad software. I don't know all all of the 3c501 bugs, and I know this driver only handles a few that I've been able to figure out. It has taken a long, intense effort just to get the driver working this well. That said, ftp.super.org:/pub/linux/pre-alpha/3c501. Jumper your card to 0x280, using pl13, add the 3c501.o to the OBJS line in net/inet/Makefile, uncomment the 3c501 line in linux/config.in, 'make config; make' as usual. AutoIRQ works, DMA isn't used, the autoprobe only looks at 0x280, the debug level is set with the third boot-time argument. You'll probably want to change the default EL_DEBUG to '2'. [[ I'm probably going to regret this. I'm going on a much-needed vacation next week, so don't phone in your thanks, mail with cash and goodies will do just fine. ]] -- Donald Becker becker@super.org IDA Supercomputing Research Center 17100 Science Drive, Bowie MD 20715 301-805-7482 ------------------------------ ** 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 ******************************