From: David Howells <dhowells@redhat.com>

The attached patch cleans up the way the bootmem allocator frees pages.

A new function, __free_pages_bootmem(), is provided in mm/page_alloc.c that is
called from mm/bootmem.c to turn pages over to the main allocator.  All the
bits of code to initialise pages (clearing PG_reserved and setting the page
count) are moved to here.  The checks on page validity are removed, on the
assumption that the struct page arrays will have been prepared correctly.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
---

 mm/bootmem.c    |   20 ++++----------------
 mm/internal.h   |    2 ++
 mm/page_alloc.c |   36 +++++++++++++++++++++++++++++++++++-
 3 files changed, 41 insertions(+), 17 deletions(-)

diff -puN mm/bootmem.c~frv-clean-up-bootmem-allocators-page-freeing-algorithm mm/bootmem.c
--- devel/mm/bootmem.c~frv-clean-up-bootmem-allocators-page-freeing-algorithm	2005-11-26 18:50:54.000000000 -0800
+++ devel-akpm/mm/bootmem.c	2005-11-26 18:50:54.000000000 -0800
@@ -294,20 +294,12 @@ static unsigned long __init free_all_boo
 		unsigned long v = ~map[i / BITS_PER_LONG];
 
 		if (gofast && v == ~0UL) {
-			int j, order;
+			int order;
 
 			page = pfn_to_page(pfn);
 			count += BITS_PER_LONG;
-			__ClearPageReserved(page);
 			order = ffs(BITS_PER_LONG) - 1;
-			set_page_refs(page, order);
-			for (j = 1; j < BITS_PER_LONG; j++) {
-				if (j + 16 < BITS_PER_LONG)
-					prefetchw(page + j + 16);
-				__ClearPageReserved(page + j);
-				set_page_count(page + j, 0);
-			}
-			__free_pages(page, order);
+			__free_pages_bootmem(page, order);
 			i += BITS_PER_LONG;
 			page += BITS_PER_LONG;
 		} else if (v) {
@@ -317,9 +309,7 @@ static unsigned long __init free_all_boo
 			for (m = 1; m && i < idx; m<<=1, page++, i++) {
 				if (v & m) {
 					count++;
-					__ClearPageReserved(page);
-					set_page_refs(page, 0);
-					__free_page(page);
+					__free_pages_bootmem(page, 0);
 				}
 			}
 		} else {
@@ -337,9 +327,7 @@ static unsigned long __init free_all_boo
 	count = 0;
 	for (i = 0; i < ((bdata->node_low_pfn-(bdata->node_boot_start >> PAGE_SHIFT))/8 + PAGE_SIZE-1)/PAGE_SIZE; i++,page++) {
 		count++;
-		__ClearPageReserved(page);
-		set_page_count(page, 1);
-		__free_page(page);
+		__free_pages_bootmem(page, 0);
 	}
 	total += count;
 	bdata->node_bootmem_map = NULL;
diff -puN mm/internal.h~frv-clean-up-bootmem-allocators-page-freeing-algorithm mm/internal.h
--- devel/mm/internal.h~frv-clean-up-bootmem-allocators-page-freeing-algorithm	2005-11-26 18:50:54.000000000 -0800
+++ devel-akpm/mm/internal.h	2005-11-26 18:52:04.000000000 -0800
@@ -26,3 +26,5 @@ static inline void set_page_refs(struct 
 #endif /* CONFIG_MMU */
 }
 
+extern void fastcall __init __free_pages_bootmem(struct page *page,
+						unsigned int order);
diff -puN mm/page_alloc.c~frv-clean-up-bootmem-allocators-page-freeing-algorithm mm/page_alloc.c
--- devel/mm/page_alloc.c~frv-clean-up-bootmem-allocators-page-freeing-algorithm	2005-11-26 18:50:54.000000000 -0800
+++ devel-akpm/mm/page_alloc.c	2005-11-26 18:52:20.000000000 -0800
@@ -53,6 +53,8 @@ unsigned long totalram_pages __read_most
 unsigned long totalhigh_pages __read_mostly;
 long nr_swap_pages;
 
+static void fastcall free_hot_cold_page(struct page *page, int cold);
+
 /*
  * results with 256, 32 in the lowmem_reserve sysctl:
  *	1G machine -> (16M dma, 800M-16M normal, 1G-800M high)
@@ -432,6 +434,39 @@ void __free_pages_ok(struct page *page, 
 	local_irq_restore(flags);
 }
 
+/*
+ * permit the bootmem allocator to evade page validation on high-order frees
+ */
+void fastcall __init __free_pages_bootmem(struct page *page, unsigned int order)
+{
+	if (order == 0) {
+		__ClearPageReserved(page);
+		set_page_count(page, 0);
+
+		free_hot_cold_page(page, 0);
+	} else {
+		LIST_HEAD(list);
+		int loop;
+
+		for (loop = 0; loop < BITS_PER_LONG; loop++) {
+			struct page *p = &page[loop];
+
+			if (loop + 16 < BITS_PER_LONG)
+				prefetchw(p + 16);
+			__ClearPageReserved(p);
+			set_page_count(p, 0);
+		}
+
+		arch_free_page(page, order);
+
+		mod_page_state(pgfree, 1 << order);
+
+		list_add(&page->lru, &list);
+		kernel_map_pages(page, 1 << order, 0);
+		free_pages_bulk(page_zone(page), 1, &list, order);
+	}
+}
+
 
 /*
  * The order of subdivision here is critical for the IO subsystem.
@@ -693,7 +728,6 @@ static void zone_statistics(struct zonel
 /*
  * Free a 0-order page
  */
-static void FASTCALL(free_hot_cold_page(struct page *page, int cold));
 static void fastcall free_hot_cold_page(struct page *page, int cold)
 {
 	struct zone *zone = page_zone(page);
_