linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
* [RFC][PATCH] unify build_zonelists()
@ 2005-09-08 16:13 Dave Hansen
  0 siblings, 0 replies; only message in thread
From: Dave Hansen @ 2005-09-08 16:13 UTC (permalink / raw)
  To: linux-kernel; +Cc: linux-mm, colpatch, Dave Hansen

page_alloc.c currently has two implementations of build_zonelists()
based on a CONFIG_NUMA #ifdef.  Does anybody see a reason to keep
the !NUMA one?  The NUMA version uses one more global __init
variable than the flat one.  But, all of its extra loops should
optimize down to nothing when NUMA is off.

Using all of the skill of a blind butcher, I hacked the non-NUMA
version out and booted it.  Seems to work just fine.  Any reason
we shouldn't do this?

P.S. The guys doing all of the work on the NUMA and CPU bitmaps
deserve a pat on the back.  All of those abstractions pay off
in cases like this.

---

 memhotplug-dave/mm/page_alloc.c |   46 ----------------------------------------
 1 files changed, 46 deletions(-)

diff -puN mm/page_alloc.c~build_zonelists_fix mm/page_alloc.c
--- memhotplug/mm/page_alloc.c~build_zonelists_fix	2005-09-08 09:04:48.000000000 -0700
+++ memhotplug-dave/mm/page_alloc.c	2005-09-08 09:05:27.000000000 -0700
@@ -1451,7 +1451,6 @@ static int __init build_zonelists_node(p
 	return j;
 }
 
-#ifdef CONFIG_NUMA
 #define MAX_NODE_LOAD (num_online_nodes())
 static int __initdata node_load[MAX_NUMNODES];
 /**
@@ -1559,51 +1558,6 @@ static void __init build_zonelists(pg_da
 	}
 }
 
-#else	/* CONFIG_NUMA */
-
-static void __init build_zonelists(pg_data_t *pgdat)
-{
-	int i, j, k, node, local_node;
-
-	local_node = pgdat->node_id;
-	for (i = 0; i < GFP_ZONETYPES; i++) {
-		struct zonelist *zonelist;
-
-		zonelist = pgdat->node_zonelists + i;
-
-		j = 0;
-		k = ZONE_NORMAL;
-		if (i & __GFP_HIGHMEM)
-			k = ZONE_HIGHMEM;
-		if (i & __GFP_DMA)
-			k = ZONE_DMA;
-
- 		j = build_zonelists_node(pgdat, zonelist, j, k);
- 		/*
- 		 * Now we build the zonelist so that it contains the zones
- 		 * of all the other nodes.
- 		 * We don't want to pressure a particular node, so when
- 		 * building the zones for node N, we make sure that the
- 		 * zones coming right after the local ones are those from
- 		 * node N+1 (modulo N)
- 		 */
-		for (node = local_node + 1; node < MAX_NUMNODES; node++) {
-			if (!node_online(node))
-				continue;
-			j = build_zonelists_node(NODE_DATA(node), zonelist, j, k);
-		}
-		for (node = 0; node < local_node; node++) {
-			if (!node_online(node))
-				continue;
-			j = build_zonelists_node(NODE_DATA(node), zonelist, j, k);
-		}
-
-		zonelist->zones[j] = NULL;
-	}
-}
-
-#endif	/* CONFIG_NUMA */
-
 void __init build_all_zonelists(void)
 {
 	int i;
_
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2005-09-08 16:15 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2005-09-08 16:13 [RFC][PATCH] unify build_zonelists() Dave Hansen

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox