Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!bcm!dimacs.rutgers.edu!rutgers!faatcrl!jprad From: jprad@faatcrl.UUCP (Jack Radigan) Newsgroups: comp.sys.amiga.datacomm Subject: Re: 19200 baud amiga Message-ID: <1032@faatcrl.UUCP> Date: 27 Feb 91 02:45:34 GMT References: <2587@tmiuv0.uucp> <978@faatcrl.UUCP> <19116@cbmvax.commodore.com> <1022@faatcrl.UUCP> <19303@cbmvax.commodore.com> Distribution: comp Organization: FAA Technical Center, Atlantic City NJ Lines: 24 daveh@cbmvax.commodore.com (Dave Haynie) writes: >Certainly a FIFO would be a good idea; they worked fine in the system level >chips like the DMAC. Either a decent FIFO or a dedicated DMA channel like >the floppy gets would solve the problem. Hope so, an A1000 can send at 38.4kbps very nicely, faster than my 12MHz 286 PeeCee, in fact. But, it can't receive at that rate. Hope you can squeeze a 4 byte FIFO in there, sure would be nice. >You sure that's one of the old Supra DMA controllers, or is it their new one, >the non-DMA "WordSync". If their DMA controller is causing problems, they're >doing something evil. No DMA device should tie up the bus unless it has >data to transfer immediately, and no single hard disk can keep a DMA device >device's FIFO/buffer full for long. Polled I/O devices, on the other hand, >are something you would expect to see problems with. Not sure which one it was, this was reported to me by a few users, all were losing data with HST modems and Supra HD controllers. I didn't know the "WordSync" was non-DMA, this may account for one who told me that the device driver for the controller was running at a priority of something like 50... -jack-