Path: utzoo!utgpu!water!watmath!clyde!rutgers!rochester!bbn!uwmcsd1!ig!agate!ucbvax!hplabs!hp-pcd!uoregon!omepd!pcm From: pcm@iwarpo3.intel.com (Phil C. Miller) Newsgroups: comp.os.minix Subject: Interrupt Problems (long) Message-ID: <2815@omepd> Date: 13 Feb 88 00:12:15 GMT Sender: news@omepd Reply-To: pcm@iwarpo3.UUCP (Phil C. Miller) Organization: Intel Corp., Hillsboro Lines: 47 In the seemingly never-ending process of trying to whip my AT-clone into shape long enough to rebuild the Minix kernel, I have run into a fairly troublesome problem. When I attempt a 'make', minix crunches away for 10-15 seconds, then starts giving me an endless stream of unexpected interrupt messages. Occasionally, these problems go away long enough to get something done, but there seems to be no reliable way to get rid of them. As a stop-gap measure, I can power down my PC, rip out the extended memory cards and my bus mouse, and try again. If any of those are causing the problem, the interrupt errors will at least go away long enough to finish building the v1.2 kernel. However, as soon as I plug my toys back in, the problem will come back (again assuming it's a hardware problem induced by the toys). Is there some general mechanism for dealing with such problems? My suggestion would be a program which printed warning messages for the first one or two interrupts, then "ignored" subsequent messages. By "ignoring" I basically mean anything which suppresses 100's of error messages flying by at 600kbaud. Perhaps a program which (a) disabled the offending interrupts after a few token warnings, or (b) tacitly handled the required Minix interrupt protocol and returned control to the interrupted process without action (the degradation in performance would be much better than no performance at all). Any suggestions, oh netlanders? I will post interesting results of a non-pornographic nature. Incidentally, I am running on an Intelligent Data Systems PC-286 Turbo AT with a Western Digital controller, a CDC 42mB/28mS winchester, and it is not inherently friendly toward Minix. Have had hard disk problems of various kinds, hercules display problems, and an occasional floppy problem. My printer, a Panasonic 1080i, also gives me fits with Minix. It is obviously imperative that I get to version 1.2; can't do much with v1.1. This brings up another interesting point: is it possible (legal) for someone to e-mail, snail-mail, or post a copy of the kernel? I really need a copy of v1.2 and I'm just not getting anywhere after 8 months of grief with my PC choking on EVERYTHING Minix tries to do. Thanks for WHATEVER help I get. Phil Miller