Re: "My" crash when reading partition table

david parsons (orc@pell.portland.or.us)
12 Oct 1999 15:22:46 -0700


In article <linux.kernel.19991012100608.A253@bug.ucw.cz>,
Pavel Machek <pavel@suse.cz> wrote:
>Hi!
>
>I reported crash with reading partition table. Now I've found out:
>something is wrong with memory detection. Passing mem=8M makes machine
>boot again. That's probably why it was even in ac-10.

Well, barring something horrible like e820 returning bogus values
that the kernel takes as gospel (what do the memory region: kernel
messages say?), it may be the problem the fellow from fujitsu
reported (boundary conditions -- the one horrible flaw I've got as a
programmer :-( )

Could you try this patch and see if it makes any difference (I don't
think that mm/init.c has changed since then):

--- linux-2.3.19/arch/i386/mm/init.c.orig Tue Oct 5 14:21:20 1999
+++ linux-2.3.19/arch/i386/mm/init.c Tue Oct 5 14:21:49 1999
@@ -444,9 +444,9 @@
* in any case, we don't want to hack mem_map
* entries above end_mem.
*/
- if ( (addr < start_low_mem)
- || (addr >= (HIGH_MEMORY + PAGE_OFFSET)&& addr <= start_mem)
- || (addr > end_mem) )
+ if ( (addr <= start_low_mem)
+ || (addr >= (HIGH_MEMORY + PAGE_OFFSET) && addr < start_mem)
+ || (addr >= end_mem) )
continue;

clear_bit(PG_reserved, &mem_map[MAP_NR(addr)].flags);

____
david parsons \bi/ a boundary condition here, a boundary condition there.
\/ Ahh, the joys of hacking the kernel.

-
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/