Index: 2.6.13-joel2/include/linux/mmzone.h
===================================================================
--- 2.6.13-joel2.orig/include/linux/mmzone.h	2005-09-26 13:58:59.%N -0500
+++ 2.6.13-joel2/include/linux/mmzone.h	2005-09-26 13:59:38.%N -0500
@@ -57,13 +57,28 @@ struct zone_padding {
 #else
 #define ZONE_PADDING(name)
 #endif
+/*
+ * The pcpu_list is to keep kernel and userrclm allocations
+ * apart while still allowing all allocation types to have
+ * per-cpu lists
+ */
+struct pcpu_list {
+	int count;
+	struct list_head list;
+} ____cacheline_aligned_in_smp;
+
+
+/* Indices into pcpu_list */
+#define PCPU_KERN 0
+#define PCPU_USER 1
+#define PCPU_LIST_SIZE 2
 
 struct per_cpu_pages {
-	int count;		/* number of pages in the list */
-	int low;		/* low watermark, refill needed */
-	int high;		/* high watermark, emptying needed */
-	int batch;		/* chunk size for buddy add/remove */
-	struct list_head list;	/* the list of pages */
+	int count;			/* number of pages in the list */
+	struct pcpu_list pcpu_list[PCPU_LIST_SIZE];
+	int low;			/* low watermark, refill needed */
+	int high;			/* high watermark, emptying needed */
+	int batch;			/* chunk size for buddy add/remove */
 };
 
 struct per_cpu_pageset {
Index: 2.6.13-joel2/mm/page_alloc.c
===================================================================
--- 2.6.13-joel2.orig/mm/page_alloc.c	2005-09-26 13:59:27.%N -0500
+++ 2.6.13-joel2/mm/page_alloc.c	2005-09-26 13:59:38.%N -0500
@@ -775,9 +775,18 @@ void drain_remote_pages(void)
 			struct per_cpu_pages *pcp;
 
 			pcp = &pset->pcp[i];
-			if (pcp->count)
-				pcp->count -= free_pages_bulk(zone, pcp->count,
-						&pcp->list, 0);
+			if (pcp->pcpu_list[PCPU_KERN].count)
+				pcp->pcpu_list[PCPU_KERN].count -=
+					free_pages_bulk(zone,
+							pcp->pcpu_list[PCPU_KERN].count,
+							&pcp->pcpu_list[PCPU_KERN].list,
+							0);
+			if (pcp->pcpu_list[PCPU_USER].count)
+				pcp->pcpu_list[PCPU_USER].count -=
+					free_pages_bulk(zone,
+							pcp->pcpu_list[PCPU_USER].count,
+							&pcp->pcpu_list[PCPU_USER].list,
+							0);
 		}
 	}
 	local_irq_restore(flags);
@@ -798,8 +807,18 @@ static void __drain_pages(unsigned int c
 			struct per_cpu_pages *pcp;
 
 			pcp = &pset->pcp[i];
-			pcp->count -= free_pages_bulk(zone, pcp->count,
-						&pcp->list, 0);
+			pcp->pcpu_list[PCPU_KERN].count -=
+				free_pages_bulk(zone,
+						pcp->pcpu_list[PCPU_KERN].count,
+						&pcp->pcpu_list[PCPU_KERN].list,
+						0);
+
+			pcp->pcpu_list[PCPU_USER].count -=
+				free_pages_bulk(zone,
+						pcp->pcpu_list[PCPU_USER].count,
+						&pcp->pcpu_list[PCPU_USER].list,
+						0);
+
 		}
 	}
 }
@@ -881,6 +900,7 @@ static void fastcall free_hot_cold_page(
 	struct zone *zone = page_zone(page);
 	struct per_cpu_pages *pcp;
 	unsigned long flags;
+	struct pcpu_list *plist;
 
 	arch_free_page(page, 0);
 
@@ -890,11 +910,24 @@ static void fastcall free_hot_cold_page(
 		page->mapping = NULL;
 	free_pages_check(__FUNCTION__, page);
 	pcp = &zone_pcp(zone, get_cpu())->pcp[cold];
+
+	/*
+	 * Strictly speaking, we should not be accessing the zone information
+	 * here wihtout the zone lock. In this case, it does not matter if 
+	 * the read is incorrect.
+	 */
+	if (get_pageblock_type(zone, page) == RCLM_USER)
+		plist = &pcp->pcpu_list[PCPU_USER];
+	else
+		plist = &pcp->pcpu_list[PCPU_KERN];
+
+	if (plist->count >= pcp->high)
+		plist->count -= free_pages_bulk(zone, pcp->batch,
+						&plist->list, 0);
+
 	local_irq_save(flags);
-	list_add(&page->lru, &pcp->list);
-	pcp->count++;
-	if (pcp->count >= pcp->high)
-		pcp->count -= free_pages_bulk(zone, pcp->batch, &pcp->list, 0);
+	list_add(&page->lru, &plist->list);
+	plist->count++;
 	local_irq_restore(flags);
 	put_cpu();
 }
@@ -930,19 +963,28 @@ buffered_rmqueue(struct zone *zone, int 
 	unsigned long flags;
 	struct page *page = NULL;
 	int cold = !!(gfp_flags & __GFP_COLD);
+	struct pcpu_list *plist;
 
 	if (order == 0) {
 		struct per_cpu_pages *pcp;
 
 		pcp = &zone_pcp(zone, get_cpu())->pcp[cold];
 		local_irq_save(flags);
-		if (pcp->count <= pcp->low)
-			pcp->count += rmqueue_bulk(zone, pcp->batch,
-						   &pcp->list, alloctype);
-		if (pcp->count) {
-			page = list_entry(pcp->list.next, struct page, lru);
+
+		if (alloctype == __GFP_USER)
+			plist = &pcp->pcpu_list[PCPU_USER];
+		else
+			plist = &pcp->pcpu_list[PCPU_KERN];
+
+		if (plist->count <= pcp->low)
+			plist->count += rmqueue_bulk(zone,
+						     pcp->batch,
+						     &plist->list,
+						     alloctype);
+		if (plist->count) {
+			page = list_entry(plist->list.next, struct page, lru);
 			list_del(&page->lru);
-			pcp->count--;
+			plist->count--;
 		}
 		local_irq_restore(flags);
 		put_cpu();
@@ -2001,18 +2043,23 @@ inline void setup_pageset(struct per_cpu
 	struct per_cpu_pages *pcp;
 
 	pcp = &p->pcp[0];		/* hot */
-	pcp->count = 0;
+	pcp->pcpu_list[PCPU_KERN].count = 0;
+	pcp->pcpu_list[PCPU_USER].count = 0;
 	pcp->low = 2 * batch;
 	pcp->high = 6 * batch;
 	pcp->batch = max(1UL, 1 * batch);
-	INIT_LIST_HEAD(&pcp->list);
+	INIT_LIST_HEAD(&pcp->pcpu_list[PCPU_KERN].list);
+	INIT_LIST_HEAD(&pcp->pcpu_list[PCPU_USER].list);
 
 	pcp = &p->pcp[1];		/* cold*/
-	pcp->count = 0;
+	pcp->pcpu_list[PCPU_KERN].count = 0;
+	pcp->pcpu_list[PCPU_USER].count = 0;
 	pcp->low = 0;
 	pcp->high = 2 * batch;
 	pcp->batch = max(1UL, 1 * batch);
-	INIT_LIST_HEAD(&pcp->list);
+	INIT_LIST_HEAD(&pcp->pcpu_list[PCPU_KERN].list);
+	INIT_LIST_HEAD(&pcp->pcpu_list[PCPU_USER].list);
+
 }
 
 #ifdef CONFIG_NUMA