Re: PROBLEM: page_alloc 2.4.1 kernel BUG running java 1.3.0

[Manfred Spraul]

This is a multi-part message in MIME format.
--------------339055CB58C727B46844E66C
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Rob Leathley wrote:
> 
> [X] I have been suffering a lot of memory paging related Oops' on the
> above PC since upgrading to the 2.2.16 kernel.  Most of these problems
> are fixed in 2.4.1 appart from the above.  These problems don't appear
> on a faster machine (e.g. P3 733MHz) so could be related to race
> conditions?  I appreciate that there is probably a bug in java 1.3.0 but
> it would be nice if it didn't kill the whole machine!
>

It's not a bug in java 1.3.0, that certain.
It's either a kernel bug or bad memory. Usually I'd say bad memory, but
your report is the third or forth one with __free_pages_ok(), so I
suspect a kernel bug.

Could you apply the attached patch to your kernel and run it until it
crashes again?

--
	Manfred
--------------339055CB58C727B46844E66C
Content-Type: text/plain; charset=us-ascii;
 name="patch-tst"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="patch-tst"

--- linux.old/mm/page_alloc.c	Sun Feb 18 20:06:11 2001
+++ linux/mm/page_alloc.c	Sun Feb 18 20:05:59 2001
@@ -70,8 +70,15 @@
 
 	if (page->buffers)
 		BUG();
-	if (page->mapping)
+	if (page->mapping) {
+	    	printk(KERN_EMERG "found bad mapping %lxh.\n",
+			(unsigned long)page->mapping);
+		printk(KERN_EMERG "page->mapping->nrpages: %d.\n",
+				(int)page->mapping->nrpages);
+		printk(KERN_EMERG "page->mapping->a_ops: [<%lxh>]\n",
+				(unsigned long)page->mapping->a_ops);
 		BUG();
+	}
 	if (!VALID_PAGE(page))
 		BUG();
 	if (PageSwapCache(page))

--------------339055CB58C727B46844E66C--