RE: I found a box that explodes with the e820 memory detection!

nathan.zook@amd.com
Wed, 6 Oct 1999 21:06:01 -0500


This is a known issue. Not all biosen that do not support e820 have
predictable behavior when this function is called. If your bios does not
support ACPI, STANDARD_MEMORY_BIOS_CALL needs to be defined for
arch/i386/boot/setup.S Grepping source, I don't see a config option to do
so. Linux/config.h should do, however.

There is an effort undway to make these routines more bullet proof. Part of
the solution may be to turn off ACPI support for any non-MMX cpu.

Nathan

> -----Original Message-----
> From: Aaron Tiensivu [SMTP:mojomofo@ctechnix.com]
> Sent: Wednesday, October 06, 1999 7:30 PM
> To: Linux Kernel Digest
> Subject: I found a box that explodes with the e820 memory detection!
>
> Kernel version 2.3.19
> To get it to boot, I have to tell it 'mem=64M' or else it freaks out.
>
> Relevant boot messages:
>
> Loaded 7878 symbols from /boot/System.map-2.3.19.
> Symbols match kernel version 2.3.19.
> No module symbols loaded.
> Linux version 2.3.19 (root@lucretia.wmis.net) (gcc version 2.7.2.3) #1 Tue
> Oct 5
> 21:03:55 EDT 1999
> Initializing CPU#0
> Detected 60000940 Hz processor.
> Console: colour VGA+ 80x60
> Calibrating delay loop... 23.91 BogoMIPS
> memory region: 640k @ 00000000
> memory region: 1070100k @ 00100000
> CPU: Intel Pentium 60/66 stepping 05
> Checking 386/387 coupling... Hmm, FPU using exception 16 error reporting
> with FD
> IV bug.
> Intel Pentium with F0 0F bug - workaround enabled.
> PCI: PCI BIOS revision 2.00 entry at 0xe8200
> PCI: Probing PCI hardware
> PCI: Ignoring ghost devices on bus 0
> PCI: Address space collision on region 6 of device DEC DC21041
>
> It crashes around there..
>
> If I say 'mem=64M', region 0010000 becomes 64MB like expected.
>
> The machine is a freebie Acer P60 that used to be an expensive server but
> now
> is just a fairly slow DNS server. Its age is demonstrated by the PCI 2.0
> only
> specs. I have to force it to use the BIOS routines so it would find PCI
> devices.
>
>
> -
> 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/

-
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/