Date: Mon, 28 Oct 1996 20:41:10 -0500 From: "Drew Derbyshire" <kew-freebsd@kew.com> To: dwhite@resnet.uoregon.edu Cc: questions@freebsd.org Subject: Re: can't find modem on 03e8, irq 5 Message-ID: <327560be.kendra@pandora.kew.com>
next in thread | raw e-mail | index | archive | help
On Mon, 28 Oct 1996 12:25:25 -0800 (PST), "Doug White" <dwhite@gdi.uoregon.edu> wrote: > On Mon, 28 Oct 1996, Drew Derbyshire wrote: > > After cleaning my disk up (hello, McAfee), I've now regenned my kernel > > 2.1.5R kernel for my 386/25, removing the chaff and leaving: > > > > sio2 not found at 0x3e8 <=== *** PROBLEM *** > > > As noted above, there is supposed to be a 16550 based port at x03e8 with > > IRQ 5, and the system is not finding it. A boot into DOS shows > > Kermit-MS finding it with no problem, and casual inspection of other > > IRQ's (lest I set the jumpers wrong) makes me think I've got the right > > one. Any suggestions on how to debug what FreeBSD is finding? > > 1. Check the line in your kernel config; make sure it does line up. Yes. IRQ 5, address 3e8. > 2. IRQ 5 must not be used by another device. Make sure it is free. I checked, and also tried COM4 IRQ 9, regenning the kernel, of course. > 3. Some video adapters stomp on the upper comports. Try changing the > base address. I used to own a S-3 chip based F-1280+, which is one of the guilty parties (great board though), so I know what you mean. Normally, COM4 gets stomped, not COM3. I checked the card in question, a Diamond SpeedStar Plus, and it uses a ET4000 series chip, which is not one of the famous port stompers. > 4. Make sure you are using the kernel you just built. I've done this on > many occaisions. :-) Yea, I've done it too. Not these times though. (Well, once tonight, but I caught it. :-)) On Mon, 28 Oct 1996 15:46:27 -0500 (EST), "Charlie ROOT" <root@k7.datadesign.com> wrote: > On Mon, 28 Oct 1996, Doug White wrote: > > > 2. IRQ 5 must not be used by another device. Make sure it is free. If another device is live on the system (for example 14, used by primary IDE disk controller), then the system will not enable the second using the IRQ. Actually, IRQ 14 would be a REALLY bad one to try to share, if the kernel assigns it to the serial I/O port first, you're gonna crash when the IDE disk is not accessible. In response to another message from a third party and not copied to the list, the device _is_ enabled, you only need to enable it if the config file includes "disable" -- the message listed above, BTW, would not be issued if the device was disabled. Any other ideas? My boot adventures have delayed this project enough that the board needs to go back tomorrow (10 day return policy at CompUSA, it used to be 30) if I don't solve this. -ahd-
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?327560be.kendra>