Date: Mon, 16 Feb 1998 16:40:34 +1030 From: Greg Lehey <grog@lemis.com> To: Mike Smith <mike@smith.net.au> Cc: Wilko Bulte <wilko@yedi.iaf.nl>, FreeBSD hackers list <FreeBSD-hackers@FreeBSD.ORG> Subject: Re: opinions requested on new mainboard purchase Message-ID: <19980216164034.56790@freebie.lemis.com> In-Reply-To: <199802160120.RAA04039@dingo.cdrom.com>; from Mike Smith on Sun, Feb 15, 1998 at 05:20:27PM -0800 References: <19980216113304.18975@freebie.lemis.com> <199802160120.RAA04039@dingo.cdrom.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 15 February 1998 at 17:20:27 -0800, Mike Smith wrote: >>>> A number of TX based motherboards, including the IWill P55XB2 and an >>>> unnamed Totem board, have problems with the serial ports. I can't get >>>> FreeBSD to recognize them at all. >>> >>> I recall asking you to try overriding the probe failure to see if the >>> ports actually worked OK. Have you had a chance to do this yet? >> >> Yup. We seem to be having a communication problem on this subject. > > Yes. Sorry, but I have bogloads on my plate just now. > > I *do* recall that we identified your I/O blob as an ASUS part. Any > luck getting data from them? No, I didn't follow it up. It was in the same message--do you need it again? >>> On Sun, Feb 01, 1998 at 11:11:34PM +1030, Mike Smith wrote: >>>> Hmm, can you try forcing a 'result = IO_COMSIZE;' after the failures[] >>>> loop and see whether the ports actually work? I think we may be seeing >>>> a pattern here, as whilst the interrupt is supposed to go away, your >>>> machine and now the Tecra 750 are failing that test. >>> >>> Hmmm. I see that other people have also worked in this area in the >>> month or two, apparently for a different (mobile) UART that wouldn't >>> probe. I hoped it might help, but no. I've put in delays at every >>> conceivable place, and also your suggestion. Here's the dmesg: >>> >>> Feb 3 16:05:05 freebie /kernel: sio0: probe test 5 failed >>> Feb 3 16:05:05 freebie /kernel: sio0: probe test 8 failed >>> Feb 3 16:05:05 freebie /kernel: sio0 at 0x3f8-0x3ff irq 4 flags 0x80 on isa >>> Feb 3 16:05:05 freebie /kernel: sio0: type 16550A >>> Feb 3 16:05:05 freebie /kernel: sio1: probe test 5 failed >>> Feb 3 16:05:05 freebie /kernel: sio1: probe test 8 failed >>> Feb 3 16:05:05 freebie /kernel: sio1 at 0x2f8-0x2ff irq 3 flags 0x80 on isa >>> Feb 3 16:05:05 freebie /kernel: sio1: type 16550A >>> Feb 3 16:05:05 freebie /kernel: sio2 at 0x3e8-0x3ef irq 9 on isa >>> (etc) >>> Feb 3 16:05:06 freebie /kernel: npx0 on motherboard >>> Feb 3 16:05:06 freebie /kernel: npx0: INT 16 interface >>> Feb 3 16:05:06 freebie /kernel: sio0: 65 events for device with no tp >>> Feb 3 16:05:06 freebie /kernel: sio1: 65 events for device with no tp > > I bet I have this one saved somewhere. Check it out. I can bounce the whole message otherwise. >>> As you can imagine from the last two messages, the ports were dead as >>> dodos. > > Actually, the last two messages *don't* imply that the ports were dead, > rather that there is some nasty interrupt crap going on that is causing > the FIFO depth probe data to be seriously delayed, such that it doesn't > get recognised until after the system comes up. That many delays would suggest to me that there's something seriously wrong. I could imagine they would be dead :-) > The 'events for device with no tp' means that the driver has received > that many events, but the port is not open. > > I think there is a timing or configuration issue here. 8( Sure. I can't think of much else. >> Any ideas? I can spare a little time to try them if needed. > > At the moment, I would perhaps try: > > - scan the BIOS configuration for anything relating to interrupts or > the I/O ports; there was something somwhere I recall about some > new interrupt feature, but I can't put my brainfinger on it. I went through the entire BIOS setup looking for that sort of thing, but I didn't fine anything. > - elaborate on "dead as a dodo", and especially whether you removed > the other I/O board (though you imply that you did). Yes, of course I took out the (not other) I/O board for the tests. Basically, the ports just didn't react at all at a system level. I didn't go and poke at the registers. I wonder if we're looking in the wrong place here. Maybe this is an interrupt controller problem. On the other hand, I've had no trouble at all with the parallel port, which runs just fine with interrupts. Greg To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980216164034.56790>