mm: alloc_large_system_hash check order
authorHugh Dickins <hugh.dickins@tiscali.co.uk>
Tue, 16 Jun 2009 22:31:50 +0000 (15:31 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 17 Jun 2009 02:47:31 +0000 (19:47 -0700)
On an x86_64 with 4GB ram, tcp_init()'s call to alloc_large_system_hash(),
to allocate tcp_hashinfo.ehash, is now triggering an mmotm WARN_ON_ONCE on
order >= MAX_ORDER - it's hoping for order 11.  alloc_large_system_hash()
had better make its own check on the order.

Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Cc: David Miller <davem@davemloft.net>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Eric Dumazet <dada1@cosmosbay.com>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/page_alloc.c

index 7cc3179e359177f93e767c936fb5f71b93271cfb..cbed869fd831ae39c1ffe6200f58cc5682646a13 100644 (file)
@@ -4511,7 +4511,10 @@ void *__init alloc_large_system_hash(const char *tablename,
                        table = __vmalloc(size, GFP_ATOMIC, PAGE_KERNEL);
                else {
                        unsigned long order = get_order(size);
-                       table = (void*) __get_free_pages(GFP_ATOMIC, order);
+
+                       if (order < MAX_ORDER)
+                               table = (void *)__get_free_pages(GFP_ATOMIC,
+                                                               order);
                        /*
                         * If bucketsize is not a power-of-two, we may free
                         * some pages at the end of hash table.