[IPSEC]: Don't warn if high-order hash resize fails

Multi-page allocations are always likely to fail.  Since such failures
are expected and non-critical in xfrm_hash_alloc, we shouldn't warn about
them.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/xfrm/xfrm_hash.c

@@ -22,7 +22,8 @@ struct hlist_head *xfrm_hash_alloc(unsigned int sz)
 		n = __vmalloc(sz, GFP_KERNEL, PAGE_KERNEL);
 	else
 		n = (struct hlist_head *)
-			__get_free_pages(GFP_KERNEL, get_order(sz));
+			__get_free_pages(GFP_KERNEL | __GFP_NOWARN,
+					 get_order(sz));
 
 	if (n)
 		memset(n, 0, sz);