Looking for FreeBSD kernel debugging help

From: Soren Kristensen (soren_at_soekris.com)
Date: 06/10/03

  • Next message: Raja Saleru: "GDB - Query Section Offset"
    Date: Tue, 10 Jun 2003 12:54:08 -0700
    To: freebsd-hackers@freebsd.org, freebsd-small@freebsd.org
    
    

    Hi Everybody,

    I've been working like a madman on bringing up a new Geode SC1100 based
    embedded board, but are stalled right now, and need help for debugging
    with FreeBSD....

    The hardware basically seems to be working just fine, and I can boot
    both MS-DOS and OpenBSD 2.9 from a CompactFlash. But I can't boot
    FreeBSD, and need to find out why, and I'm looking for somebody that can
    help me, as I'm just the hardware guy....

    When booting freeBSD I get:

    Type '?' for a list of commands, 'help' for more detailed help.
    ok boot -v
    Copyright (c) 1992-2001 The FreeBSD Project.
    Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
             The Regents of the University of California. All rights reserved.
    FreeBSD 4.4-RELEASE #1: Sun Jun 8 19:08:24 PDT 2003
         soren@Server600.soekris.com:/usr/src/sys/compile/KERN_NSC
    Calibrating clock(s) ... TSC clock: 266667486 Hz, i8254 clock: 1193267 Hz
    CLK_USE_I8254_CALIBRATION not specified - using default frequency
    Timecounter "i8254" frequency 1193182 Hz
    CLK_USE_TSC_CALIBRATION not specified - using old calibration method
    CPU: Cyrix GXm (266.65-MHz 586-class CPU)
       Origin = "CyrixInstead" Id = 0x540 DIR=0x81b7 Stepping=8 Revision=1
    real memory = 134217728 (131072K bytes)
    Physical memory chunk(s):
    0x00001000 - 0x0009efff, 647168 bytes (158 pages)
    0x003c1000 - 0x07ff7fff, 130248704 bytes (31799 pages)
    avail memory = 126984192 (124008K bytes)
    bios32: Found BIOS32 Service Directory header at 0xc00f0080
    bios32: Entry = 0xf00c0 (c00f00c0) Rev = 0 Len = 1
    pcibios: PCI BIOS entry at 0xe1
    Other BIOS signatures found:
    ACPI: 00000000
    Preloaded elf kernel "kernel" at 0xc039b000.
    md0: Malloc disk
    Creating DISK md0
    Math emulator present
    pci_open(1): mode 1 addr port (0x0cf8) is 0x800090d0
    pci_open(1a): mode1res=0x80000000 (0x80000000)
    pci_cfgcheck: device 0 [class=060000] [hdr=00] is there (id=00011078)

    And there it just hangs forever.... Normally it would continue with:

    npx0: <math processor> on motherboard
    npx0: INT 16 interface
    pcib0: <Host to PCI bridge> on motherboard
    .
    .
    .

    So it looks like it hangs between first PCI check probe and printing the
    NPX info. The question is, what hardware specific could there be that
    could hang ? It's probably PCI config or floating point related, I tried
    to follow the source, but couldn't really....

    Some additional info:

    The PCI bus seems fine, I can netboot, but it then hangs the same place.
    Same with a 4.2-release and 4.8-release. Same with and without NSC VSA
    extension. Same with GENERIC kernel. I did some search on the net, and
    all info indicated that a stock FreeBSD should work, although a patch
    for the TSC counter is needed for optimal performance and a patch if you
    use the Video (which the SC1100 don't have). I have tried every single
    register in the SC1100 chip (And there are tons of them...) I also tried
    changed the chip ID from "Geode by NSC" to "CyrixInstead", but that just
    resulted in FreeBSD recognizing the CPU and providing a little better
    CPU information.

    More info on the hardware at: http://www.soekris.com/net4801.htm

    So I need:

    All qualified help pinpointing the problem. Or a kernel hacker that can
    come by Morgan Hill, CA and debug the FreeBSD kernel telling me exactly
    where and why it hangs. Once I know that, it's usually easy to fix.

    I'm prepared to pay for solving the problem, cash and/or net4801
    hardware....

    Regards,

    Soren

    _______________________________________________
    freebsd-hackers@freebsd.org mailing list
    http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
    To unsubscribe, send any mail to "freebsd-hackers-unsubscribe@freebsd.org"


  • Next message: Raja Saleru: "GDB - Query Section Offset"

    Relevant Pages

    • FreeBSD Status report for Oct-Dec 2003
      ... Bluetooth stack for FreeBSD ... Not much to report. ... Bluetooth kernel modules appear to be stable. ... concerns and some src committers are willing to commit the patches. ...
      (freebsd-current)
    • FreeBSD Status Report for Oct-Dec 2003
      ... Bluetooth stack for FreeBSD ... Not much to report. ... Bluetooth kernel modules appear to be stable. ... concerns and some src committers are willing to commit the patches. ...
      (freebsd-hackers)
    • FreeBSD Status Report for Oct-Dec 2003
      ... Bluetooth stack for FreeBSD ... Not much to report. ... Bluetooth kernel modules appear to be stable. ... concerns and some src committers are willing to commit the patches. ...
      (freebsd-stable)
    • RE: FreeBSD 4.11 P13 Crash
      ... I do not want to jinx myself, but after back revving to FreeBSD 4.9 + ... think it is related to IPFilter in conjunction with 4 Intel nics and/or ... page fault while in kernel mode ... Okay this time my kernel was recompiled so there are no modules to ...
      (freebsd-hackers)
    • Re: 5.3 interrupt storm (atapicam) and I/O error
      ... attached you will find a copy of my custom kernel ... config which did work without any issue under FreeBSD ... The following hardware configuration *won't* cause ... * harddrive attached to IDE channel #1 ...
      (freebsd-questions)