ben@cernvax.UUCP (ben) (08/27/85)
Does anyone out there have experience of making PRACTICAL applications using Douglas Comer's Xinu system? We are interested particularly in any real-time applications for 68000's. Ben M. Segal, CERN-DD, 1211 Geneva 23, Switzerland (ben@cernvax via mcvax)
herndon@umn-cs.UUCP (09/02/85)
Regarding using Doug Comer's XINU system: I TA'd a class here where the project was to port Xinu to an HP-9836U workstation. The project succeeded (barely). If you intend to use Comer's system, be sure to get the magtape. Many pieces of code were left out of the book, or were left for exercises. As an additional comment, it is my (biased) opinion that Doug Comer's book does not describe an operating system so much as it describes a run-time library for multi-tasking -- the 'operating system' described is not capable of supporting itself, since any program run under it must be loaded into memory at boot time. (There is no operation similar to 'exec' anywhere within XINU.) If you seriously wish to run APPLICATIONS under xinu, then I would suggest you look into finding a copy of Mini-Unix. This is a version of UNIX version 6 which runs on PDP-11/40s, etc. and requires about 12Kw for the OS, leaving 16 Kw for user programs. (In fact, this appears to be the basis for Comer's Xinu - most of the code looks remarkably familiar.) It is missing pipes, p-tracing, and profiling (the 3 Ps), and does not support memory management. It works fairly well; another guy and I ported it to a PDP-11/03 with A/D's in 1978 and set it up for real-time data sampling from smooth muscle fibers. I don't know where you can find a copy; you might try the Johns Hopkins Dept. of Biomedical Engineering (whom we did the work for) or John Lions (of the Lions commentary fame) at the University of New South Wales. I seem to recall he managed to squash mini-unix into 6Kw and burn it into a ROM. Robert Herndon ...!ihnp4!umn-cs!herndon