The real time clock code in 1.3.89 is interesting, Unfortunatly the
author, Paul Gortmaker, doesn't leave an email address anywhere, so
I'm asking to the list.
Are there any current plans to add asynIO support to the device? If the device
could generate SIGIOs, then it could serve as an itimer type device, but with
10 times the precision of the general purpose itimers (providing the hardware
and system load leave enough time to keep up :-).
While I'm rambling would it be usefull/possible to add a new field to the
sigcontext_struct so that asyncIO events can pass the file descriptor number
responsible for the SIGIO signal. Currently if you have more than one asyncIO
source you need to waste time in a select() call in order to demultiplex the
signal source, it would be nice to have a more direct/faster method. Espcially
when one of the async events arrives at 1KHz...
-- Jon. <jon@gte.esi.us.es, http://www.esi.us.es/~jon>