Date: Sat, 18 Jul 1998 00:32:59 -0500 From: Paul Saab <paul@mu.org> To: Bill Paul <wpaul@skynet.ctr.columbia.edu>, hackers@FreeBSD.ORG Subject: Re: Obtaining 3Com programming documentation Message-ID: <19980718003259.A26528@mu.org> In-Reply-To: <199807171254.IAA06263@skynet.ctr.columbia.edu>; from Bill Paul on Fri, Jul 17, 1998 at 08:54:14AM -0400 References: <199807171254.IAA06263@skynet.ctr.columbia.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
I think there is a driver being written but we never hear anything about it. From what I understand about the 905 and the 905B is that the 905 has a programmed IO interface and the 905B had it removed. Paul Bill Paul (wpaul@skynet.ctr.columbia.edu) wrote: > Anybody know the right place/person from which to obtain programming > documentation for 3Com ethernet chips? We recently received two Dell > GX1 systems as donations from Intel as part of a testbed system that > some people in the EE department will be using in cooperation with a > researcher from Intel. The systems actually have the Dell markings > replaced with big 'Donated by Intel' signs but you can still tell they're > Dells. > > Anyway, the machines have integrated 3Com 3c905B 'Cyclone' chips in > them which, of course, don't work right in FreeBSD. My experiments so > far indicate that the problem is that the first time a packet is received, > the chip generates an interrupt flood. After the interrupt routine is > invoked, there's a function that's supposed to remove the received frame > from the chip: this function is supposed to read a status register from > the window 1 register set so that it can learn the size of the packet, > but the value read is always 0. This means the packet is never taken > off the chip and the chip keeps interrupting forever, which wedges the > system pretty tightly. > > I have the source for a Linux driver which claims to support this chip, > but an actual programming manual would be better. Of course, 3Com doesn't > seem to have any such info on their web servers. Any other driver > programmers know the right trees to shake at 3Com? I really need to get > this stupid thing working to make the Intel guy happy (to say nothing > of certain Columbia EE faculty). > > -Bill > > PS: These systems also came with Adaptec 2940 U/UW controllers with > the 7895 chip. This forced me to download the CAM snapshot in order > to even get the machines running. The CAM stuff seems to work fine > so far, though at the moment we only have the one SCSI disk attached. > Please tell me there will be driver support for these bastiches in > FreeBSD 3.0-RELEASE. > > PPS: Amusingly, whoever assembled the systems at Dell used the wrong > pre-loaded LoseNT 4.0 disk image which did not have the correct driver > loaded to support this controller: the machines blue-screened immediately > after we first turned the systems on, with the Intel researcher looking > on. I got a good laugh. Well, for a little while. Then I had to re-load > LoseNT. > > PPPS: This is the second time I've had to do FreeBSD development work as > part of my job. (The first time was writing the ThunderLAN driver.) Maybe > the world isn't so screwed up after all. > > -- > ============================================================================= > -Bill Paul (212) 854-6020 | System Manager, Master of Unix-Fu > Work: wpaul@ctr.columbia.edu | Center for Telecommunications Research > Home: wpaul@skynet.ctr.columbia.edu | Columbia University, New York City > ============================================================================= > "It is not I who am crazy; it is I who am mad!" - Ren Hoek, "Space Madness" > ============================================================================= > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-hackers" in the body of the message 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?19980718003259.A26528>