From owner-freebsd-mobile Mon Dec 9 03:06:35 1996 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.4/8.8.4) id DAA28645 for mobile-outgoing; Mon, 9 Dec 1996 03:06:35 -0800 (PST) Received: from diablo.ppp.de (diablo.ppp.de [193.141.101.34]) by freefall.freebsd.org (8.8.4/8.8.4) with SMTP id DAA28640 for ; Mon, 9 Dec 1996 03:06:31 -0800 (PST) From: Greg Lehey Received: from freebie.lemis.de by diablo.ppp.de with smtp (Smail3.1.28.1 #1) id m0vX3X9-000QXSC; Mon, 9 Dec 96 12:06 MET Received: (grog@localhost) by freebie.lemis.de (8.8.4/8.6.12) id MAA05173; Mon, 9 Dec 1996 12:05:54 +0100 (MET) Organisation: LEMIS, Schellnhausen 2, 36325 Feldatal, Germany Phone: +49-6637-919123 Fax: +49-6637-919122 Message-Id: <199612091105.MAA05173@freebie.lemis.de> Subject: Re: Installation: still not perfect In-Reply-To: <199612090617.QAA07837@genesis.atrad.adelaide.edu.au> from Michael Smith at "Dec 9, 96 04:47:18 pm" To: msmith@atrad.adelaide.edu.au (Michael Smith) Date: Mon, 9 Dec 1996 12:05:53 +0100 (MET) Cc: freebsd-mobile@FreeBSD.org (FreeBSD mobile Mailing List) X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-mobile@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk Michael Smith writes: > Greg Lehey stands accused of saying: >> >>> These aren't available on the normal boot floppy for space reasons. >> >> Then there should be another boot floppy. I know, there are good >> reasons to oppose this, but it shouldn't be at the expense of not >> being able to install on certain configurations. I think that an >> ethernet-based install on a laptop should be relatively common. > > This is the PAO boot floppy. Those few of us? them? that have the > gear to test this sort of thing often don't have the time 8( I don't really even care if they test them, but it should be possible to put the floppy on the CD. >> I have two cards: one is the 3C589C, which I can't configure because >> the diagnostic doesn't work. It seems that there might have been card > > You should be able to get the diagnostic off 3com's FTP site to work. No. From my (angered) message to 3Com: > After installing the target operating system (UNIX) on the notebook, > the diagnostic does not work at all. It stops immediately with the > message > > Can't Run Install, files missing. Make sure you have INSTALL.EXE, INST1.SAC > and STRINGS.BIN in your directory (-1, 2) > > These files are, of course, all there, on the original, > write-protected diskette. I re-created a second diskette from the > backup, and I downloaded the file 3C589.EXE from your web site, but > the results were the same. > > I suspect that the problem lies in another bug I have observed: the > diagnostic program is unbelievably slow. To get as far as starting > a test takes about two minutes, while the program tries to emulate > Windows 95, redrawing the screen endlessly at a snail's pace. While > doing so, it frequently accesses the hard disk, something that it > should have no reason to do. The message mentioned above first > occurred after installing UNIX on the hard disk, after which point > it presumably didn't find something it thought it should look for. > > This kind of software is useless. I can't imagine that you use it > yourselves. If your customer base is of interest to you, please > reply and tell me where I can find a real diagnostic program. I've since had a reply and a toll-free number from 3Com. I called the toll-free number, and they took lots of information and promised to ring back. I'll be interested to see what they have to say. >> services installed on the machine before I wiped out the Windoze 95% >> partition, but they didn't give me a diskette, and it seems that card >> services relate to the machine and not the board, so I don't have any >> card services at all any more. Without card services, the diagnostic >> just doesn't run. If somebody out there can send me a copy of >> CardWizard or whatever for the AcerNote Light, I'd be grateful. > > I seem to recall that the 3c589 diagnostics have to be run _without_ > any Card Services drivers in place. No, this isn't correct. They say they should be in place (*stupid*). The diagnostic prints an error message and attempts to continue if they are not, but it will fail unless you first write the configuration before running the test. > If you really need that sort of thing though, we have a pile of > "generic" Card Services disks here that came with some PCC modems we > bought that I could squirt your way. I'm looking for something for an AcerNote Light. According to the documentation, it's called CardWizard. Anybody have it? >> There's also a very minimal "manual" (a folded sheet of paper) which >> tells me that it has drivers for every machine under the Sun, is >> Ethernet compliant, has a 16 kB data buffer. I've tried booting with >> the PAO boot floppy, and it recognizes the card, but claims it can't >> find card type (from memory) " ()" in the card database. This also >> happens, including the same name, with the 3C589, so I assume this >> doesn't relate to the board. > > If you build a kernel with PCCARD support (pcic0 and crd0 per LINT), > make the /dev/card0 and /dev/card1 devices (50/0, 50/1), and then say > 'pccardc dumpcis', you should be able to find the manufacturer and > version strings. Good information. Is it written down anywhere? I've just checked the FAQ, and it looks like I'm missing some important information which the FAQ assumes. I'll add the information to "The (now even more) Complete FreeBSD". Anyway, I didn't have the /dev/card? devices, so I created them. I also created an /etc/pccard.conf, with an entry for the 3C589C from the laptop survey, and started pccardd, then: === root@papillon (/dev/ttyp0) /etc 139 -> pccardc dumpcis Configuration data for card in slot 0 Tuple #1, code = 0xff (Terminator), length = 0 Configuration data for card in slot 1 Tuple #1, code = 0xff (Terminator), length = 0 2 slots found === root@papillon (/dev/ttyp0) /etc 145 -> pccardc enabler 0 ep0 -a 0x300 -i 10 drv ep0, mem 0x0, size 0, io 768, irq 0x400, flags 0x0 set driver: Device not configured === root@papillon (/dev/ttyp0) /etc 146 -> pccardc pccardmem PCCARD Memory address set to 0xd0000 === root@papillon (/dev/ttyp0) /etc 147 -> pccardc rdmap Mem 0: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 1: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 2: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 3: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 4: flags 0x000 host 0x0 card 0000 size 0 bytes I/O 0: flags 0x014 port 0x300 size 32 bytes I/O 1: flags 0x000 port 0x 0 size 0 bytes Mem 0: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 1: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 2: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 3: flags 0x000 host 0x0 card 0000 size 0 bytes Mem 4: flags 0x000 host 0x0 card 0000 size 0 bytes I/O 0: flags 0x000 port 0x 0 size 0 bytes I/O 1: flags 0x000 port 0x 0 size 0 bytes === root@papillon (/dev/ttyp0) /etc 148 -> pccardc rdreg Registers for slot 0 00: 82 ff d0 60 00 3f 60 02 00 03 1f 03 00 00 00 00 10: 00 00 00 00 00 00 01 80 00 00 00 00 00 00 00 fc 20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 30: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 Registers for slot 1 00: 82 ff d0 60 00 3f 20 00 00 00 00 00 00 00 00 00 10: 00 00 00 00 00 00 01 80 00 00 00 00 00 00 00 3c 20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 30: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 I get the feeling that there is some memory conflict in the notebook, but before I wiped Windog 95% it ran the diagnostic correctly modulo the problems with the config. Does anybody have any ideas? Here's the pccard.conf entry: # 3Com Etherlink III 3C589B, 3C589C card "3Com Corporation" "3C589" config 0x1 "ep0" 10 insert echo 3Com Etherlink III inserted insert /etc/pccard_ether $device remove echo 3Com Etherlink III removed remove /etc/pccard_ether_remove $device BTW, two problems about this entry: first, I don't find any file called pccard_ether_remove, and secondly the original 'config' line contained the keyword 'default', which pccardd definitely did not like. I'm running 3.0-current as of Friday, so I wouldn't think my pccardd is too old. In general, I think my strategy should be: 1. Find out whether the information from pccard makes sense. You guys can help me on that one. 2. Find out what the manufacturer string is for the second card (for the benefit of those who haven't seen the beginning of the thread, the sum total of manufacturer information on this board is "Made in Taiwan R.O.C." and "P/N: 18-0A040BC S/N: QK007360". The diagnostic looks like it's for an NE2000 clone. 3. See if I can enable either card. If anybody has any other suggestions, I'd be glad to hear of them. I'd also be interested to hear from anybody running an AcerNote Light. Greg