No dice on changing the USB settings in the BIOS.  Or a bunch of other
settings on/off, that I tried.  The only thing that was able to get an error
message to show was turning off Extended IDE for my SATA HDD :/.  Strange.
Thanks, though :).

On 11/3/07, erik quanstrom <quanstro@quanstro.net> wrote:
>
> > The Nov 3rd image didn't work for me.  But, by turning off extended IDE
> mode
> > for my HDD, I was able to get an error message in assembly rather than
> it
> > just hanging without any error message.  (Not sure if that makes it
> > irrelevant, but anyway.)
> >
> > http://home.earthlink.net/~eldanen/pastes/dmesg
> > http://home.earthlink.net/~eldanen/pastes/plan9output
> >
> > The following was appended to "plan9output":
> >
> > FLAGS=10246 TRAP=0 ENCODE=0 PC=80030089
> > AX 00000000 BX 8004483c CX 00000000 DX 00000000
> > SI 0000020 DI 00000000 BP 80047e6c
> > CS 0010 DS 0008 ES 0008 FS 0008 GS 0008
> > CRO 80000011 CR2 00f0c05c CR3 0000c000
> > panic: exception/interrupt 0
> >
> > Press almost any key to reset...
>
> it didn't print anything else, did it?  the pc indicates that these
> lines should have been executed
>
> ; acid /n/sources/plan9/386/9loaddebug
> /n/sources/plan9/386/9loaddebug:386 plan 9 boot image
> /sys/lib/acid/port
> /sys/lib/acid/386
> acid: src(0x80030089)
> /sys/src/boot/pc/ether.c:119
>
> /n/sources/plan9/sys/src/boot/pc/ether.c
> 108                     print("ether#%d: %s: port 0x%luX irq %lud",
> 109                             ctlr->ctlrno, ctlr->type, ctlr->port,
> ctlr->irq);
> [...]
> 118                     ctlr->rb = ialloc(sizeof(RingBuf)*ctlr->nrb, 0);
> >119                    if(ctlr->ntb == 0
>
> here's a wild stab in the dark which has bitten us for an embedded
> kernel we run.  make sure to turn off legacy emulation for usb devices
> in the bios.  (phoenix bios has this option in the second tab under one
> of the bottom selections.)  legacy emulation can turn on system management
> mode.  system management mode takes over system resources (like memory)
> to run the emulation code and then puts things back.  the os is not
> supposed to notice, but sometimes the emulation is not seemless.
> (or perhaps we don't know all the magic spots in memory.)  either way,
> it's worth a shot.
>
> - erik
>
>