Replacement async driver -- experience with SV/ATa

[Charles Hedrick]

My great thanks to Jim Murray for this driver.  I have it working on
SV/AT with a standard AT comm port configuration.  It wasn't clear
from the posting that this driver might make sense even for those of
us without exotic async hardware.  One of the biggest curses of SV/AT
has been hangs in the async driver.  Preliminary tests suggest that
this driver does not have those problems.  In order to use it with
SV/AT, you'll need to fix the declarations of open_device and
close_device in asy.c.  They are first declared in forward
declarations as "static void".  When the actual declaration is given,
the "static" is omitted.  This causes the compiler to complain.  You
want to change

   void
   open_device(aip, opening_getty_version)

to

   static void
   open_device(aip, opening_getty_version)

and similarly for close_device.  Other than that the installation went
smoothly with my 2.4 system.  I'm using it at 2400 baud.  The README
suggests that performance may not be as good as the Microport driver
if you have the standard async chip.  But of course I wouldn't notice
that at 2400.  I'm more interested in reliability than speed.