Re: [PATCH] Fix boot problem in situations where the boot CPU isrunning on a memoryless node

[Pekka J Enberg]

Hi,

On Wed, 23 Jan 2008, Pekka J Enberg wrote:
> > I still think Christoph's kmem_getpages() patch is correct (to fix 
> > cache_grow() oops) but I overlooked the fact that none the callers of 
> > ____cache_alloc_node() deal with bootstrapping (with the exception of 
> > __cache_alloc_node() that even has a comment about it).
> 
> So something like this (totally untested) patch on top of current git:

Sorry, removed a BUG_ON() from cache_alloc_refill() by mistake, here's a 
better one:

[PATCH] slab: fix allocation on memoryless nodes
From: Pekka Enberg <penberg@xxxxxxxxxxxxxx>

As memoryless nodes do not have a nodelist, change cache_alloc_refill() to bail
out for those and let ____cache_alloc_node() always deal with that by resorting
to fallback_alloc().

Furthermore, don't let kmem_getpages() call alloc_pages_node() if nodeid passed
to it is -1 as the latter will always translate that to numa_node_id() which
might not have ->nodelist that caused the invocation of fallback_alloc() in the
first place (for example, during bootstrap).

Signed-off-by: Pekka Enberg <penberg@xxxxxxxxxxxxxx>
---
 mm/slab.c |   19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

Index: linux-2.6/mm/slab.c
===================================================================
--- linux-2.6.orig/mm/slab.c
+++ linux-2.6/mm/slab.c
@@ -1668,7 +1668,11 @@ static void *kmem_getpages(struct kmem_c
 	if (cachep->flags & SLAB_RECLAIM_ACCOUNT)
 		flags |= __GFP_RECLAIMABLE;
 
-	page = alloc_pages_node(nodeid, flags, cachep->gfporder);
+	if (nodeid == -1)
+		page = alloc_pages(flags, cachep->gfporder);
+	else
+		page = alloc_pages_node(nodeid, flags, cachep->gfporder);
+
 	if (!page)
 		return NULL;
 
@@ -2975,9 +2979,11 @@ retry:
 		 */
 		batchcount = BATCHREFILL_LIMIT;
 	}
+	BUG_ON(ac->avail > 0);
 	l3 = cachep->nodelists[node];
+	if (!l3)
+		return NULL;
 
-	BUG_ON(ac->avail > 0 || !l3);
 	spin_lock(&l3->list_lock);
 
 	/* See if we can refill from the shared array */
@@ -3317,7 +3323,8 @@ static void *____cache_alloc_node(struct
 	int x;
 
 	l3 = cachep->nodelists[nodeid];
-	BUG_ON(!l3);
+	if (!l3)
+		return fallback_alloc(cachep, flags);
 
 retry:
 	check_irq_off();
@@ -3394,12 +3401,6 @@ __cache_alloc_node(struct kmem_cache *ca
 	if (unlikely(nodeid == -1))
 		nodeid = numa_node_id();
 
-	if (unlikely(!cachep->nodelists[nodeid])) {
-		/* Node not bootstrapped yet */
-		ptr = fallback_alloc(cachep, flags);
-		goto out;
-	}
-
 	if (nodeid == numa_node_id()) {
 		/*
 		 * Use the locally cached objects if possible.
--
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/