kernel BUGs when removing largish files with the SLOB allocator

From: Aubrey
Date: Mon Sep 04 2006 - 02:53:42 EST


Hi all,

I'm working on the nommu blackfin uclinux(2.6.16) platform and get a
kernel BUG! at mm/nommu.c:124 when removing largish file with the SLOB
allocator.

root:~> df -k
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/mtdblock3 3008 1264 1744 42% /
root:~> cp /bin/busybox /busy
root:~> df -k
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/mtdblock3 3008 1532 1476 51% /
root:~> ls -l /bin/busybox /busy
-rwxr-xr-x 1 0 0 423904 /bin/busybox
-rwxr-xr-x 1 0 0 423904 /busy
root:~> md5sum /bin/busybox
7db253a2259ab71bc854c9e5dac544d6 /bin/busybox
root:~> md5sum /busy
7db253a2259ab71bc854c9e5dac544d6 /busy
root:~> rm /busy
kernel BUG at mm/nommu.c:124!
Kernel panic - not syncing: BUG!

Bug comes from mm/nommu.c:
=======================================
if (!objp || !((page = virt_to_page(objp))) ||
(unsigned long)objp >= memory_end)
return 0;

if (PageSlab(page))
return ksize(objp);

BUG_ON(page->index < 0);
124: BUG_ON(page->index >= MAX_ORDER);
=======================================

This seems that the SLOB allocator doesn't set the SLAB page flag
while nommu.c seem to be written for SLAB only.

On my side the following patch seems to work around the issue
============================================================
--- nommu.c 2006-06-26 14:49:28.000000000 +0800
+++ nommu.c.new 2006-06-26 14:47:20.000000000 +0800
@@ -18,7 +18,9 @@
#include <linux/file.h>
#include <linux/highmem.h>
#include <linux/pagemap.h>
-#include <linux/slab.h>
+#ifdef CONFIG_SLAB
+# include <linux/slab.h>
+#endif
#include <linux/vmalloc.h>
#include <linux/ptrace.h>
#include <linux/blkdev.h>
@@ -117,7 +119,9 @@
if (!objp || !((page = virt_to_page(objp))) || (unsigned
long)objp >= memory_end)
return 0;

+#ifdef CONFIG_SLAB
if (PageSlab(page))
+#endif
return ksize(objp);

BUG_ON(page->index < 0);
============================================================

Is there any solution/patch to fix the issue?

Any suggestions are really appreciated.

Thanks,
-Aubrey

--
VGER BF report: U 0.498985
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/