BTX halted trying to boot X4100 from disk

John Baldwin jhb at freebsd.org
Wed Nov 29 11:29:35 PST 2006


On Saturday 25 November 2006 18:10, Michiel Boland wrote:
> 
> On Mon, 13 Nov 2006, John Baldwin wrote:
> 
> > On Monday 13 November 2006 15:12, Michiel Boland wrote:
> >>> Yeah, the X4100 has an mpt(4) controller. Mine is running the original
> > BIOS
> >>> release (6464), and runs FreeBSD fine. You can try fiddling with the 
BIOS
> >>> revision by downloading replacements from Sun's website.
> >>
> >> As far as I can see there are no other BIOSes on the Sun website other
> >> than what the box is already running (ILOM 1.0.7 - I understand ILOM and
> >> BIOS are linked for 4100s)
> >>
> >> I am a bit puzzled why solaris and fedora do not have these problems on
> >> this hardware. Oh well...
> >
> > The real mode breakpoint handler may just ignore breakpoints, or the BIOS
> > takes a different codepath when it runs in vm86 mode.
> 
> The code path is the same. The bios calls INT3 if someone tries to load a 
> block with LBA < 16, or LBA 64 or 65. (I think - my asm skills are a bit 
> rusty.)
> 
> Is there not a way for BTX to ignore these INT3 interrupts? Obviously they 
> are ignored in real mode too.

Is it ignored in real mode?  I had figured the real mode handler would have 
rebooted the box.  If it triggers in real mode then I guess we can just 
ignore them, at least coming from vm86 mode.  I think there is magic to dump 
the registers but keep going when a breakpoint is triggered in protected 
mode.

-- 
John Baldwin


More information about the freebsd-amd64 mailing list