gvdg@gvdgpc.UUCP (Gerard J van der Grinten PA0GRI) (01/03/90)
Hello Folks, Happy New Year to all. In response on my inquiry for tape support for Microports SysV AT 2.4 the following is found. I got a response from Bob -obiwan!bob- saying all goes well here but have problems with Sony tapes. Then i got a response from Johannes -sgren@dynas.se- telling he had the same problem I had when he had been thinking. Thinking was a bad idea..... To get to the solution: I was using int2 (hardware line). That is in reality int9 (second PIC). Now it seems Microport did a good job in screwing up interrupt support. The second interrupt controller can only handle the disc controller (int14) and any other service requested fails with no response. Having both an (inteligent) excelan card and a (inteligent) tape controller I was short on interrupts. (int7 for lp, int6 for floppy, int5 for excelan int4 for 1 standard comm1, int3 for ast 4 port serial, and int2 unusable due to braindamage at Uport. (adding a wire to int10 - int15 proved that it does not work)). I had to give up another interrupt line and comm1 was the one getting shot. I removed the interrupt wire (jumper) but left the port in for satisfaction of the BOOTROM. (it plugs vectors needed). Then i moved the interrupt for the ast card to 4 and the tape to int3. That is a working configuration. The int4 MUST be the first comm interrupt otherwise IT DOES NOT WORK. It is surely a pitty that the second PIC is unusable for extensions. Another hint: tar uses /dev/mt/0m as a major/minor for its default device. Normaly it is linked to /dev/rdsk/fd but if you relink to /dev/rmt0 it goes default to the tape drive. I hope You gain some information of this lesson for your situation if you ever get to it. An other question: Are there archives for Microport (messages and / or programs) besides the dead Microport bbs. (was not updated anymore for a year) Regards, gerard.
johnk@opel.uu.net (John Kennedy) (01/05/90)
Yes, the slave PIC is masked out as delivered. However, this table is patchable in the kernel, as I have asy interrupts working on ints 9 ,10, 11, 12 and 15. You probably will have had responses to this effect by the time you read this, but I can email you details if you wish. John -- John Kennedy johnk@opel.uu.uunet Second Source, Inc. Annapolis, MD
mdm@cocktrice.uucp (Mike Mitchell) (01/06/90)
johnk@opel.uu.net (John Kennedy) writes: >Yes, the slave PIC is masked out as delivered. However, this table is patchable >in the kernel, as I have asy interrupts working on ints 9 ,10, 11, 12 and 15. >You probably will have had responses to this effect by the time you read this, >but I can email you details if you wish. Could you post this information (or email me a copy)? I am sure that there are others which would be interested in this. Thank you. -- Mike Mitchell BELL H (505) 471-7639 W (505) 473-4482 2020 Calle Lorca #43 ARPA mdm@cocktrice.UUCP Santa Fe, NM 87505 UUCP ...!uunet!dmk3b1!cocktrice!mdm