Re: [patch 00/12] Slab defragmentation V3

[Christoph Lameter <clameter@sgi.com>]

On Fri, 8 Jun 2007, Michal Piotrowski wrote:

> bash shared mapping + your script in a loop
> while true;  do sudo ./run.sh; done > res3.txt

Hmmmm... Seems to be triggered from the reclaim path kmem_cache_defrag 
rather than the manual triggered one from the script. Taking the slub_lock 
on the reclaim path is an issue it seems.

Maybe we need to do a trylock in kmem_cache_defrag to defuse the 
situation? This is after all an optimization so we can bug out.

Does this fix it?

---
 mm/slub.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

Index: slub/mm/slub.c
===================================================================
--- slub.orig/mm/slub.c	2007-06-08 11:12:40.000000000 -0700
+++ slub/mm/slub.c	2007-06-08 11:14:34.000000000 -0700
@@ -2738,7 +2738,9 @@ int kmem_cache_defrag(int percent, int n
 	unsigned long pages = 0;
 	void *scratch;
 
-	down_read(&slub_lock);
+	if (!down_read_trylock(&slub_lock))
+		return 0;
+
 	list_for_each_entry(s, &slab_caches, list) {
 
 		/*


-
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/