Date: Mon, 19 May 1997 23:50:07 +0200 From: Gerhard Sittig <G.Sittig@abo.freiepresse.de> To: FreeBSD-questions <FreeBSD-Questions@FreeBSD.ORG> Subject: Re: CDROM question, again Message-ID: <3380CB0F.1F01@abo.freiepresse.de>
next in thread | raw e-mail | index | archive | help
Yaning Wang wrote: > > Some of you suggested that the CDROM(Matsushita CDROM/Panasonic Drive > CR-563, connected to a soundcard) is not supported by the GENERIC kernel > therefore I need to rebuild the kernel. IMHO any ATAPI cdrom should work, what seems to be the problem is that the controller has to be a supported one. All the answers to the previous questions of that kind (sorry, but I've been listening to that list for only one week yet) have the same direction: put the CDROM onto a "real" controler and leave those soundcards alone (or at least just make some noices with them and nothing else :) > One mystery during my installation (by DOS partition): > If I bootup the machine using the bootup disk (made from the CD), > when I select DOS as installation medium, the error message says: > DOS partition could not been found (something like that). > If I bootup by using a DOS boot disk then lanuch the installation > from the CD (which I can see from DOS), then everything went on > very smoothly. Reminds me of my situation two years ago running LinuX and having a CMD640 controler (since IDE disks are that cheap :) -- I had to boot DOS and load the apropriate driver to enable the second channel. After a few weeks LinuX itself was able to do that, but referring to the code I read they had to do "some magic" and were not sure whether the next chip's release would react in the same way. Seems to me there's something like that for the SB board -- any kind of "little demon" activating the card's onboard IDE controler. That's why it's seen in DOS and maybe later. I guess once it's activated it's found at the same resources as any IDE controler -- but to get there you need a Doze driver :( And in case it's not an IDE controler (i.e. neither 0x1f0/14 nor 0x3f0/15) but something proprietary then you had to check not just that the driver is in the kernel but also the driver's settings. FreeBSD seems to believe more what the user says than what the hardware is configured like -- my 3c509 is "recognized" at 0x300/10 although it's set to 0x300/12. BTW: Is there a special setting to have the driver do some autoprobing ? Since some hardware like 3c509 cards has config ports to read the settings from, that could be taken over once the hardware is detected (i.e. the config port is touched (?)). This would make it less easy to "misconfigure" a setup by ignoring these settings that can be recognized anyway.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3380CB0F.1F01>