[PATCH] fix a bogus OOM condition for __GFP_NOFS allocations

If a GFP_NOFS allocation is made when the ZONE_NORMAL inactive list is
full of dirty or under-writeback pages, there is nothing the caller can
do to force some page reclaim.  The caller ends up getting oom-killed.

- In mempool_alloc(), don't try to perform page reclaim again.  Just
  go to sleep and wait for some elements to be returned to the pool.

- In try_to_free_pages(): perform a single, short scan of the LRU and
  if that doesn't work, fail the allocation.  GFP_NOFS allocators know
  how to handle that.

mm/mempool.c

@@ -196,10 +196,11 @@ void * mempool_alloc(mempool_t *pool, int gfp_mask)
 		return element;
 
 	/*
-	 * If the pool is less than 50% full then try harder
-	 * to allocate an element:
+	 * If the pool is less than 50% full and we can perform effective
+	 * page reclaim then try harder to allocate an element.
 	 */
-	if ((gfp_mask != gfp_nowait) && (pool->curr_nr <= pool->min_nr/2)) {
+	if ((gfp_mask & __GFP_FS) && (gfp_mask != gfp_nowait) &&
+				(pool->curr_nr <= pool->min_nr/2)) {
 		element = pool->alloc(gfp_mask, pool->pool_data);
 		if (likely(element != NULL))
 			return element;

mm/vmscan.c

@@ -536,6 +536,20 @@ shrink_caches(struct zone *classzone, int priority,
 
 /*
  * This is the main entry point to page reclaim.
+ *
+ * If a full scan of the inactive list fails to free enough memory then we
+ * are "out of memory" and something needs to be killed.
+ *
+ * If the caller is !__GFP_FS then the probability of a failure is reasonably
+ * high - the zone may be full of dirty or under-writeback pages, which this
+ * caller can't do much about.  So for !__GFP_FS callers, we just perform a
+ * small LRU walk and if that didn't work out, fail the allocation back to the
+ * caller.  GFP_NOFS allocators need to know how to deal with it.  Kicking
+ * bdflush, waiting and retrying will work.
+ *
+ * This is a fairly lame algorithm - it can result in excessive CPU burning and
+ * excessive rotation of the inactive list, which is _supposed_ to be an LRU,
+ * yes?
  */
 int
 try_to_free_pages(struct zone *classzone,
@@ -546,12 +560,15 @@ try_to_free_pages(struct zone *classzone,
 
 	KERNEL_STAT_INC(pageoutrun);
 
-	do {
+	for (priority = DEF_PRIORITY; priority; priority--) {
 		nr_pages = shrink_caches(classzone, priority,
 					gfp_mask, nr_pages);
 		if (nr_pages <= 0)
 			return 1;
-	} while (--priority);
+		if (!(gfp_mask & __GFP_FS))
+			break;
+	}
+	if (gfp_mask & __GFP_FS)
 		out_of_memory();
 	return 0;
 }