Xref: utzoo comp.sys.att:12318 comp.unix.questions:31733 comp.unix.sysv386:8475 comp.dcom.modems:10045 Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!crdgw1!uunet!cis.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!mcs.kent.edu!neoucom.edu!wtm From: wtm@uhura.neoucom.EDU (Bill Mayhew) Newsgroups: comp.sys.att,comp.unix.questions,comp.unix.sysv386,comp.dcom.modems Subject: Re: Internal Modem and AT&T SV/386 R3.x Message-ID: <1991May26.105023.17157@uhura.neoucom.EDU> Date: 26 May 91 10:50:23 GMT References: <1991May25.004941.26239@mccc.edu> Organization: Northeastern Ohio Universities College of Medicine Lines: 42 I'm not sure if just changing the address and IRQ of sdevice.d and cf.d entries for the standard asy port driver will work for other than the standard COM1 3f8 and COM2 2f8 addresses and IRQs 4 and 3 respectively. There is also a problem in that COM3 shares IRQ 4 with COM1. You may experience either hardware, software or both conflicts with the interrupts. The problem is that in the standard MSDOS ISA bus architecture interrupts are edge triggered. Almost all cards are stupidly designed so that they hold the IRQ lead low until their hardware is serviced. This means that any subsequent interrupts generated by other boards in the system while the first card is holding the IRQ lead low will not be noticed by the 8259 controller. This is a really p*ss poor design. I know why board designers do that, it allows an IRQ service routine to psuedo poll devices, but really violates the intended way of making things work. The other problem is the &**#@! boards themselves. You could share the interrupt if the boards had open collector outputs, but I've seen quite a few that have regular totempole TTL outputs driving the bus because they were too cheap to use one extra open collector driver part. Anyway, it seems that although the Sys V r 3.2 documentation hints that you can share IRQs, that it doesn't work. The idbuild process will bitch if you have two drivers with the same IRQ. I don't know if that is a "bug" in the idbuild process or if that is a limitation of the asy driver. I'm not an expert at writing drivers. I agree that the FAS 2.06 driver that Peter recommends is a good idea. I've installed it on my system. I also rigged my COM3 port to use IRQ 7 so that I could still have the built-in ports. At least the newer designs fix the quick and dirty design of the original IBM PC. Both the EISA bus and MCA bus have level triggered interrupts. Bill -- Bill Mayhew NEOUCOM Computer Services Department Rootstown, OH 44272-9995 USA phone: 216-325-2511 wtm@uhura.neoucom.edu ....!uunet!aablue!neoucom!wtm via internet: (140.220.001.001)