Alan Cox wrote:
> 
> > stage LILO was loaded, it happily over-wrote part of the SMP config
> > table sending the kernel into nirwana during boot (BTW The fix, although
> > accepted by Werner, never made it into an official LILO release. SuSE
> 
> That shouldnt be possible - its checksummed

Linux won't accept a corrupted SMP config table, that's right, but the
code essentially searched memory for a properly checksummed SMP config
table and if it didn't find one that was appropriate, it didn't come up.

> > Perhaps it was a generic Phoenix BIOS problem? At least Intel's SMP spec
> > (v1.4) specifically allowed this setup. BTW Werner Almesberger also
> 
> Phoenix BIOS is allowed to do this. Totally so. The kernel however assumed
> the EBDA would be <4K. So the combination triggered it

It wasn't the kernel, it was LILO which corrupted the EBDA and thus the
SMP config table. Once I modified LILO (first to load at a slightly
lower address of 0x9900 iirc, later to load in two parts at two
different locations: first stage loader at its original address 0x9a00
and the second stage loader at 0x0b00, thereby repairing the damage that
the first fix did because the kernel was then unable to find the
parameter page), the SMP config table remained untouched and the kernel
was able to find it though the (untouched) MP floating table in BIOS
memory (0xf0000...0xfffff.) 

> > > Free software is good at being reliable - but its a testing/review/fixing
> > > scheme that is nothing like as effective at finding situations where the code
> > > works for everyone currently but isnt quite to specification.
> >
> > Please don't get me wrong. I wasn't complaining about the things that
> > happened in the past. I was merely trying to explain my (and our
> > managers') impatience:
> 
> Im not taking it as a complaint. You asked how it managed to get missed, I
> was trying to explain the quirks of free software testing/reliability

OK.

-- 
Josef M�llers
Fujitsu Siemens Computers
SHV Server DS 1

-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to [EMAIL PROTECTED]

Reply via email to