lkml.org 
[lkml]   [2007]   [May]   [14]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: slab hang on boot
On Mon, 14 May 2007, David Miller wrote:

> From: Meelis Roos <mroos@linux.ee>
> Date: Mon, 14 May 2007 21:36:42 +0300 (EEST)
>
> > Sorry, I could not test the parport patch yet - I did git upgrade as of
> > yesterday and the kernel hangs on boot. boot -p reveals the following
> > slab panic:
>
> Please enable SLUB to work around this for now, SLAB wants
> LARGE_ALLOCS enabled when it really shouldn't require that
> for a 512K or 1MB SLAB when PAGE_SIZE==8192 :-(

Ok.

> > kmem_cache_create: Early error in slab tsb_512KB
> > kernel BUG at mm/slab.c:2165!
>
> Christoph, this is why I had to have LARGE_ALLOCS enabled on sparc64
> all this time, for that MAX_OBJ_ORDER logic in mm/slab.c in order
> to get 512K and 1MB slab caches usable.

Hmmm... The #ifdef crap there has worried me for a long time. Lets try to
clear that mess up too.

> This inconsistency is painful for platforms. If I enable LARGE_ALLOCS
> on sparc64 then SLAB works but it's a total waste for SLUB.
>
> Right now if I enable LARGE_ALLOCS, SLUB will break again due to
> that KMALLOC_SHIFT_HIGH bug we fixed the other week, Christophe
> can you at least push that fix to Linus if you haven't already?

No I thought you had it under control so I held it in my tree.

Andrew please push this patch upstream.



SLUB: CONFIG_LARGE_ALLOCS must consider MAX_ORDER limit

Take MAX_ORDER into consideration when determing KMALLOC_SHIFT_HIGH.
Otherwise we may run into a situation where we attempt to create
general slabs larger than MAX_ORDER.

Signed-off-by: Christoph Lameter <clameter@sgi.com>

---
include/linux/slub_def.h | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)

Index: slub/include/linux/slub_def.h
===================================================================
--- slub.orig/include/linux/slub_def.h 2007-05-10 15:19:39.000000000 -0700
+++ slub/include/linux/slub_def.h 2007-05-10 16:06:15.000000000 -0700
@@ -59,7 +59,8 @@ struct kmem_cache {
#define KMALLOC_SHIFT_LOW 3

#ifdef CONFIG_LARGE_ALLOCS
-#define KMALLOC_SHIFT_HIGH 25
+#define KMALLOC_SHIFT_HIGH ((MAX_ORDER + PAGE_SHIFT) =< 25 ? \
+ MAX_ORDER + PAGE_SHIFT - 1 : 25)
#else
#if !defined(CONFIG_MMU) || NR_CPUS > 512 || MAX_NUMNODES > 256
#define KMALLOC_SHIFT_HIGH 20
@@ -86,6 +87,9 @@ static inline int kmalloc_index(int size
*/
WARN_ON_ONCE(size == 0);

+ if (size >= (1UL << KMALLOC_SHIFT_HIGH))
+ return -1;
+
if (size > 64 && size <= 96)
return 1;
if (size > 128 && size <= 192)

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

\
 
 \ /
  Last update: 2007-05-15 01:49    [W:0.059 / U:1.016 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site