Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!apple!sun-barr!texsun!letni!merch!cpe!hal6000!trsvax!uhclem
From: uhclem@trsvax.UUCP
Newsgroups: comp.unix.xenix
Subject: Re: VP/ix problems
Message-ID: <196500029@trsvax>
Date: 23 Jun 89 13:46:00 GMT
References: <192@chiron.UUCP>
Lines: 22
Nf-ID: #R:chiron.UUCP:192:trsvax:196500029:000:922
Nf-From: trsvax.UUCP!uhclem    Jun 23 08:46:00 1989


<>
Make sure the serial adapters do not have a 8250_ or 8250B UART
part installed.  If so, replace with a 8250A, 82C50 or 16450.  If your bus
speed is above 4 MHZ, you really should have a 16450 or 82C50 since all
the others have too slow a recovery time, resulting in double reads of status
and dubious success in programming the chip.   The 8250_ / 8250B have numerous
race conditions that are unimportant if you aren't using interrupts (MS-DOS)
but fatal if you are (XENIX/UNIX).  All pre-AT IBM adapter cards and many
current clones used/use the 8250_ so the BIOS would be happy (in pre-AT, it
was coded to expect the bugs.)

Finally, make sure the modems are not set for LOCAL COPY or anything weird
like that.

						
					"Thank you, Uh Clem."
					Frank Durda IV @ <trsvax!uhclem>
				...decvax!microsoft!trsvax!uhclem
				...sys1!hal6000!trsvax!uhclem

"If we took the bugs out, it wouldn't be MS-DOS, would it?"