Robert, The link that your referred to was helpful. I was finally able to resolve all my installation issue tonight. The following is what I discovered:
1. The tftp negotiation between the net4801 and my server (a PII/250, running FreeBSD 6.2) is very sensitive to latency. In the list archive it sounds like there wasn't enough latency in the connection, so a switch/hub was used to connect the 4801 to the server instead of using a X-over cable. In my case it was the opposite! When I disconnected an uplink cable from my switch (an old 100mb Netgear) thus reducing the traffic going thru the device, I was able to boot using PXE after a couple of tries. Previous to that, I went through almost 2 dozen reboots with no success. 2. I later discovered a posting in the list archive describing the same "non-aligned DMA transfer attempted" error I was recieving which prevented me from installing FreeBSD to my hard drive. According to http://lists.soekris.com/pipermail/soekris-tech/2005-December/009735.html it looked like I needed to disable "ATA-DMA" in FreeBSD. To do this I added the line: ' hw.ata.ata_dma="0" ' to /tftpboot/boot/device.hints on my tftp server. I hope these notes help someone else in the future. Thanks to everyone for your input. -tm ________________________________________ From: Robert Jeppesen [EMAIL PROTECTED] Sent: Thursday, July 26, 2007 9:57 AM To: Tom Mazzotta Subject: Re: [Soekris] Unable to install FreeBSD on 4801 via PXE/TFTP Have you looked in the archieves? Perhaps http://lists.soekris.com/pipermail/soekris-tech/2004-December/007239.html describes the same kind of trouble that you have. /Robert Tom Mazzotta wrote: > Just a followup on the PXE/TFTP boot problems I wrote about yesterday. > After a lot of experimenting I found (by accident) that my 4801 would boot > using only TFTP (w/o NFS). However, it only work occasionally! To get it > to work, I needed to configure the default boot device to "F0" (ethernet > port) using "set bootdrive=F0 80 81" in the combios. (Previously, I was > using the default boot device parameters and I would ^P to get the monitor > screen and type "boot f0"). Anyway, after cycling to power on the unit > anywhere from 6 or 7 to more than a dozen times, it would boot properly. > The way I knew when it would work was by "pxe_open" values reported by the > bootstrap loader. When I saw "pxe_open: gateway ip: 0.0.0.0", I know it > was going to fail! However, on occasion, it would say "pxe_open: gateway > ip: 192.168.168.1", which is the correct gateway address on my LAN, and > in this instance it would boot correctly. > > QUESTION #1 - Is this normal? Could there be a h/w problem w/ my 4801 or > my LAN, etc...? > > Then after going thru a standard FreeBSD install, I get the errors: > ad0: setting up DMA failed > ata0: FAILURE - NON aligned DMA transfer attempted > > Unable to find device node for /dev/ad0s1b in /dev > The creation of the filesystem will be aborted. > > QUESTION #2 - Did I fry my drive by cycling the power so many times, or > might something else be wrong? > > Thanks! > -tm > ________________________________ > From: Tom Mazzotta > Sent: Tuesday, July 24, 2007 11:09 PM > To: [email protected] > Subject: Unable to install FreeBSD on 4801 via PXE/TFTP > > I have been trying to use the procedure on the ultradesic.com website to > install FreeBSD 6.2 on to a net4801-60. I have basically tried 4 > permutations of the install using the boot image file downloaded from > Ultradesic vs. patching the standard bootonly ISO image w/ recompiled > PXEBOOT, LOADER, and KERNEL files (all built according to the Ultradesic > procedure and sample config's). I also tried using FreeBSD vs. Windows XPP > (w/Tftpd32) as my DHCP/TFTP server. No success with any of the scenarios. > Typically, the following is what I see: > > comBIOS ver. 1.31 20070408 Copyright (C) 2000-2007 Soekris Engineering. > net4801 > 0256 Mbyte Memory CPU Geode SC1100 267 Mhz > Pri Mas FUJITSU MHN2100AT LBA Xlt 1024-255-63 9820 Mbyte > Slot Vend Dev ClassRev Cmd Stat CL LT HT Base1 Base2 Int > ------------------------------------------------------------------- > 0:00:0 1078 0001 06000000 0107 0280 00 00 00 00000000 00000000 > 0:06:0 100B 0020 02000000 0107 0290 00 3F 00 0000E101 A0000000 10 > 0:07:0 100B 0020 02000000 0107 0290 00 3F 00 0000E201 A0001000 10 > 0:08:0 100B 0020 02000000 0107 0290 00 3F 00 0000E301 A0002000 10 > 0:10:0 104C AC23 06040002 0107 0210 08 3F 01 00000000 00000000 > 0:18:2 100B 0502 01018001 0005 0280 00 00 00 00000000 00000000 > 0:19:0 0E11 A0F8 0C031008 0117 0280 08 38 00 A0003000 00000000 11 > 1:00:0 100B 0020 02000000 0107 0290 00 3F 00 0000D001 A4000000 05 > 1:01:0 100B 0020 02000000 0107 0290 00 3F 00 0000D101 A4001000 11 > > comBIOS Monitor. Press ? for help. >> boot f0 > NSC DP83815/DP83816 Fast Ethernet UNDI, v1.03 > Copyright (C) 2002, 2003 National Semiconductor Corporation > All rights reserved. > Pre-boot eXecution Environment PXE-2.0 (build 082) > Copyright (C) 1997-2000 Intel Corporation > > CLIENT MAC ADDR: 00 00 24 C8 B4 F8 > CLIENT IP: 192.168.168.11 MASK: 255.255.255.0 DHCP IP: 192.168.168.38 > GATEWAY IP: 192.168.168.1 > PXE Loader 1.00 > Building the boot loader arguments > Relocating the loader and the BTX > Starting the BTX loader > Consoles: internal video/keyboard > BIOS drive C: is disk0 > PXE version 2.1, real mode entry point @9e64:00f6 > BIOS 639kB/261120kB available memory > FreeBSD/i386 bootstrap loader, Revision 1.1 > ([EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>, > Tue Jul 24 17:18:00 EDT 2007) > pxe_open: server addr: 192.168.168.38 > pxe_open: server path: /pxeroot > pxe_open: gateway ip: 0.0.0.0 > NFS MOUNT RPC error: 60 > \ > can't load 'kernel' > Type '?' for a list of commands, 'help' for more detailed help. > OK > ------- > By examining the log from Tftp32, I can tell the PXEBOOT was sent to the > box, but it doesn't get an further: > > Rcvd DHCP Discover Msg for IP 0.0.0.0, Mac 00:00:24:C8:B4:F8 [24/07 > 23:05:12.250] > DHCP: proposed address 192.168.168.12 [24/07 23:05:14.468] > Rcvd DHCP Discover Msg for IP 0.0.0.0, Mac 00:00:24:C8:B4:F8 [24/07 > 23:05:14.468] > DHCP: proposed address 192.168.168.12 [24/07 23:05:15.984] > Rcvd DHCP Rqst Msg for IP 0.0.0.0, Mac 00:00:24:C8:B4:F8 [24/07 > 23:05:18.421] > Previously allocated address 192.168.168.12 acked [24/07 23:05:20.468] > Connection received from 192.168.168.12 on port 2070 [24/07 23:05:20.781] > Read request for file <boot\pxeboot>. Mode octet [24/07 23:05:20.781] > OACK: <tsize=219136,> [24/07 23:05:20.781] > Using local port 1679 [24/07 23:05:20.781] > Peer returns ERROR <TFTP Aborted> -> aborting transfer [24/07 > 23:05:20.781] > Connection received from 192.168.168.12 on port 2071 [24/07 23:05:20.875] > Read request for file <boot\pxeboot>. Mode octet [24/07 23:05:20.875] > Using local port 1680 [24/07 23:05:20.875] > <boot\pxeboot>: sent 429 blks, 219136 bytes in 1 s. 0 blk resent [24/07 > 23:05:21.015] > Rcvd DHCP Discover Msg for IP 0.0.0.0, Mac 00:00:24:C8:B4:F8 [24/07 > 23:05:21.765] > DHCP: proposed address 192.168.168.12 [24/07 23:05:23.265] > Rcvd DHCP Rqst Msg for IP 0.0.0.0, Mac 00:00:24:C8:B4:F8 [24/07 > 23:05:23.265] > Previously allocated address 192.168.168.12 acked [24/07 23:05:24.765] > ------- > > Since I am a newbie at this I'm not sure what to try next. Has anyone > successfully used the Ultradesic procedure to install FreeBSD on a 4801 w/ > comBIOS ver. 1.31? If so, what is the trick to getting it to work? Thanks > in advance! > > -tm > _______________________________________________ > Soekris-tech mailing list > [email protected] > http://lists.soekris.com/mailman/listinfo/soekris-tech > _______________________________________________ Soekris-tech mailing list [email protected] http://lists.soekris.com/mailman/listinfo/soekris-tech
