Date: Sun, 12 May 1996 12:04:55 +0200 (MET DST) From: J Wunsch <j@uriah.heep.sax.de> To: freebsd-current@FreeBSD.org (FreeBSD-current users) Subject: Re: cvs commit: src/sys/i386/isa syscons.c Message-ID: <199605121004.MAA06566@uriah.heep.sax.de> In-Reply-To: <199605120211.MAA13485@godzilla.zeta.org.au> from Bruce Evans at "May 12, 96 12:11:41 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
(moved to -current) As Bruce Evans wrote: > > Give up waiting for a successful keyboard reset after some > > unreasonable time. I've got a PCI mainboard that simply doesn't grok > > it, so continuing with a warning (and a keyboard that's working > > nevertheless :) seems to be better than spin-looping forever. > > Does it work if you fix the well known `!retries' logic bug (the infinite > loop always gets entered if the first loop fails (retries == -1)). No, retries were at 4 when the probe went on. (Well, damn, i could have fixed the retries bug in the same commit.) I've single-stepped the code in DDB, and this keyboard controller always responded with ``keyboard buffer empty'' (first infinite loop) and the byte 0xfa was fetched (KB_ACK, second infinite loop). It suprised me a bit, since it was just an AMIKEY-2 controller as i've seen it working on many other boards before. Alas, the chip has been soldered in (and the board itself was a loaner to me only to debug this problem -- so i couldn't use a soldering iron :), so replacing it was not an option for me. > I have a keyboard with a partly broken cable or connector which causes > interesting problems for keyboard drivers. The BIOS keyboard driver > seems to handle transient disconnections better than syscons. Perhaps > it issues a reset when the keyboard stops working. Perhaps the BIOS reacts properly to the 0xaa (SELF TEST OKAY) message that is sent when the keyboard plugs in, and re-initializes the keyboard. -- cheers, J"org joerg_wunsch@uriah.heep.sax.de -- http://www.sax.de/~joerg/ -- NIC: JW11-RIPE Never trust an operating system you don't have sources for. ;-)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199605121004.MAA06566>