From owner-freebsd-questions Thu Aug 9 4:59:13 2001 Delivered-To: freebsd-questions@freebsd.org Received: from klima.physik.uni-mainz.de (klima.Physik.Uni-Mainz.DE [134.93.180.162]) by hub.freebsd.org (Postfix) with ESMTP id 9ADA837B406; Thu, 9 Aug 2001 04:59:02 -0700 (PDT) (envelope-from ohartman@klima.physik.uni-mainz.de) Received: from klima.Physik.Uni-Mainz.DE (Sturm@klima.Physik.Uni-Mainz.DE [134.93.180.162]) by klima.physik.uni-mainz.de (8.11.5/8.11.4) with ESMTP id f79Bwsp02475; Thu, 9 Aug 2001 13:58:54 +0200 (CEST) (envelope-from ohartman@klima.physik.uni-mainz.de) Date: Thu, 9 Aug 2001 13:58:54 +0200 (CEST) From: "Hartmann, O." To: Cc: Subject: PXE boot image defective! Message-ID: <20010809131921.G1839-100000@klima.physik.uni-mainz.de> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Dear Sirs. I'm sorry bothering you again with this problem, but it is very serious to me and our department. Description of the fault: As with the cvsupdate from today (09th August 2001 9 o'clock CEST from a German cvsupdate server) the pxeboot image created on a AMD based machine crashes immediately after being loaded via TFTP. All diskless stations show the same symptome: blinking, coloured block graphics with several truncated status rows at the end (I will report this row later). Configuration: FreeBSD 4.4-PRERELEASE after 01. August 2001. Our diskless stations are all the same: DFI AK74-EC main PCB, 700 MHz AMD Duron, 64MB PC133-CL3 RAM, Intel Ether Express Pro Management NIC, Gigabyte GA660+ TNT2 based graphics adapter. The BIOS of the AK74-EC mainboard is up to date. The Flash ROM of the NICs are all brought up to the newest revision as found on Intels server except two machines. Actual revision is R17 of the image, two machines have R12 and R14 (I mention this due to the reports of problems with Intels PXE bootPROM). The server for the diskless stations is a AMD K7 TBird with 850 MHz on which the stuff for the diskless stations is compiled. We use ISC-DHCP2 as DHCP server on this machine. We always compile our sources after a weekly cvsupdate with the specific CPU-type option for each architecture within our server farm and for all peripherial systems running FreeBSD without crisscrossing Intel- and AMD compiled images. All AMD stuff is compiled on AMD machines with AMD CPU type. As CFLAGS and COPTFLAGS we use only "-O -pipe", nothing else. The mentioned fault occurs as well as when ommiting the CPU type FLAG or using a Intel compiled pxeboot image. I personally tried several combination without success. This is the image size of pxeboot: on AMD: 149504 9 Aug 10:53 pxeboot on Intel: 153600 Aug 9 10:52 pxeboot As you can see, they differs ... To make our systems running we use a backuped pxeboot image created on August, 1st 2001. This image has been compiled the same way as the faulty ones, the same optimization flags, the same CPU type! And pxeboot worked for us for the last 10 months pretty well (as long as we boot our X11 terminal pool by this mechnism). Several people from this group reported not having any problem with the new pxeboot code and simply tell me what tehy did: compiling and booting, but they do not tell more about configuration and so on. That's worthless. I do not tell you a heap of bullshit. I try to understand what's going on and I suspected our configuaration here (especially the PXE boot PROM of the Intel NIC). But the fact that all things worked before the 2nd August with several releases and now especially the older ones do not makes it senseless to track down the problem in this direction. It seems to me that immediately after loading the pxeboot image into memory and trying to initialize it a kind of disordering happens and the image crashes. And as said: using with the same hardware the older pxeboot image all things are pretty clear! So, the last thing to do is telling you about that what I see on a crashed diskless station's screen (don't feel fooled, but I think it could be a worthy detail for those who are firm with PXE and the pxeboot image). The only row which makes sense to me is the last row on the screen that has the following truncated, somehow human readable message: (Underscore means: blank or unicoloured 8x16 block): BX_lae_.0_BXvrini__Err_letfra_o_upre Well, hope this gets not a foolish story. If someone is willing or able to test it with a similar configuartion and has no problems, I would appreciate any help on this. If you are interested, I will send you a copy of the crashing image. It would be nice if you can tell me what kind of parameters involves the creation of pxeboot image and whether there are boot parameters who influences the functionality of pxeboot (as I understand this mechanism, pxeboot is the first step of the boot process and therefore nothing could be influenced since this image has to be loaded first to make config parameters active. But in this case, loading fails as the image crashes immediately after it gets loaded by TFTP). Thanks in advance for your tips and hints and help. Oliver -- MfG O. Hartmann ohartman@klima.physik.uni-mainz.de ---------------------------------------------------------------- IT-Administration des Institut fuer Physik der Atmosphaere (IPA) ---------------------------------------------------------------- Johannes Gutenberg Universitaet Mainz Becherweg 21 55099 Mainz Tel: +496131/3924662 (Maschinenraum) Tel: +496131/3924144 FAX: +496131/3923532 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message