Date: Fri, 22 Dec 2000 16:55:45 -0600 From: seebs@plethora.net (Peter Seebach) To: freebsd-mobile@freebsd.org Subject: Spectrum24 card Message-ID: <200012222255.QAA24337@guild.plethora.net>
next in thread | raw e-mail | index | archive | help
I found a Linux driver for this on SourceForge.
It may not be a PRISM 2, but it's a card with the same registers, at the
same offsets, the same bits set in those registers to indicate the same
states, and even the same quirk of taking too long to handle an "initialize"
command.
My thinking would be that the "correct" thing would be for one driver
to be picked to handle all of the PRISM2 cards. If I were going to pick
one, I'd pick wi(4). I know Bill Paul hates it, and right now I basically
agree with him, but it is *much* nicer than any of the Linux drivers
I've seen (one of which is over 150k of fairly illegible C. Ow.), and it
is, currently, pretty much working except for that one timeout quirk.
I would be happy to contribute patches towards getting this fixed. In
particular, I'd like to suggest replacing the "init, delay, init, delay,
init, delay" cycle with adding
if (cmd == WI_CMD_INI) {
Delay(10);
}
in wi_cmd's timeout loop; this results in the card being up and initialized
as soon as it's ready; generally 990-1000 iterations, so only a few ms.
Anyway, although I'm mostly interested, for my own purposes, in having the
NetBSD driver work, I'd be glad to share code or information with anyone,
especially since I only know of one outstanding bug in my code. :)
(Or in the chipset. I am starting to wonder.)
-s
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-mobile" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200012222255.QAA24337>