RE: Kernel reboot on loading (port 0xEF)

From: Richard B. Johnson (root@chaos.analogic.com)
Date: Tue Jul 25 2000 - 08:56:52 EST


On Tue, 25 Jul 2000, Agust Karlsson wrote:

> Hi Dick.
> Reading from port 0xEE yields 0xff and nothing seems to happen, but actually
> it is an alternative method to propagate the A20 signal.
> But if you read from port 0xEF (in the bios debugger) the Elan SC400
> reboots.
> That is why I'm wondering if the kernel reads that port somewhere, maybe as
> a part of checking what ports are available ?
>

I am pretty sure it doesn't. There is no PCI on the SC400 so make sure
you compile the kernel without PCI support. This will prevent any
possible probes of PCI ports that may be aliased because of incomplete
decodes. I can boot Linux-2.2.15 on my SC400 from a floppy. It, of course,
fails to mount the root file system because there isn't any. However,
to get up that far seems to show that it should go all the way.

The math coprocessor uses 0xf0 to 0xff for reset/control. The SC400
does not have a math coprocessor. This is why you need to pretend
it's a '386 and enable math emulation. Somebody may have snuck in
some FP math into your kernel version (it sometimes happens). If
the traps in the kernel, with no handler, you have a double-fault
which is fatal. FP math gets in when somebody does something like:

        int foo = (int) (3.141592654 * bar);

.... where bar and 3.14... are constants so the programmer assumes that
the compiler will do the math. Often it doesn't.

Cheers,
Dick Johnson

Penguin : Linux version 2.2.15 on an i686 machine (797.90 BogoMips).

"Memory is like gasoline. You use it up when you are running. Of
course you get it all back when you reboot..."; Actual explanation
obtained from the Micro$oft help desk.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Jul 31 2000 - 21:00:19 EST