[PATCH] add page->mapping handling interface [0/35] intro

[PATCH] add page->mapping handling interface [0/35] intro

[KAMEZAWA Hiroyuki]

Hi, this patch set adds following functions

 - page_inode(page) ... returns inode from page, (page->mapping->host)
 - page_mapping_cache(page) ... returns addrees_space from page
 - page_mapping_anon(page) ... return anon_vma from page
 - page_is_pagecache(page) ... returns 1 if the page is page cache
 - pagecache_consistent(page, mapping) ... returns if page_mapping_cache(page)
   equals to mapping.

By adding aboves, this patch set removes all *direct* references to
page->mapping in usual codes. (compile tested with all mod config.)

I think this can improve VM/FS dependency and make things robust.
In addition,  page->mapping is not a just address_space, now.
(And we can hide page->mapping details from moduled FSs.)

patch set is structured as
[1] ... new interface definition
[2] ... changes in /mm
[3] ... changes in /kernel and /fs
[4...] ... changes in each FSs. (most of patches are very small.)

Any comments are welcome.

Thanks,
-Kame

--
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>

[PATCH] add page->mapping handling interface [1/35] interface definitions

[KAMEZAWA Hiroyuki]

 - changes page->mapping from address_space* to unsigned long
 - add page_mapping_anon() function.
 - add linux/page-cache.h
 - add page_inode() function
 - add page_is_pagecache() function
 - add pagecaceh_consisten() function for pagecache consistency test.
 - expoterd swapper_space. inline function page_mapping() refers this.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 include/linux/fs.h         |    1 +
 include/linux/mm.h         |   20 +++++++++++++++++---
 include/linux/mm_types.h   |    2 +-
 include/linux/page-cache.h |   39 +++++++++++++++++++++++++++++++++++++++
 mm/swap_state.c            |    2 ++
 5 files changed, 60 insertions(+), 4 deletions(-)

Index: test-2.6.23-rc4-mm1/include/linux/page-cache.h
===================================================================
--- /dev/null
+++ test-2.6.23-rc4-mm1/include/linux/page-cache.h
@@ -0,0 +1,39 @@
+/*
+ * For interface definitions between memory management and file systems.
+ * - This file defines small interface functions for handling page cache.
+ */
+
+#ifndef _LINUX_PAGECACHE_H
+#define _LINUX_PAGECACHE_H
+
+#include <linux/mm.h>
+/* page_mapping_xxx() function is defined in mm.h */
+
+static inline int page_is_pagecache(struct page *page)
+{
+	if (!page->mapping || (page->mapping & PAGE_MAPPING_ANON))
+		return 0;
+	return 1;
+}
+
+/*
+ * Return an inode this page belongs to
+ */
+
+static inline struct inode *page_inode(struct page *page)
+{
+	if (!page_is_pagecache(page))
+		return NULL;
+	return page_mapping_cache(page)->host;
+}
+
+/*
+ * Test a page is a page-cache of an address_space.
+ */
+static inline int
+pagecache_consistent(struct page *page, struct address_space *as)
+{
+	return (page_mapping(page) == as);
+}
+
+#endif
Index: test-2.6.23-rc4-mm1/include/linux/fs.h
===================================================================
--- test-2.6.23-rc4-mm1.orig/include/linux/fs.h
+++ test-2.6.23-rc4-mm1/include/linux/fs.h
@@ -582,6 +582,7 @@ static inline int mapping_writably_mappe
 	return mapping->i_mmap_writable != 0;
 }
 
+#include <linux/page-cache.h>
 /*
  * Use sequence counter to get consistent i_size on 32-bit processors.
  */
Index: test-2.6.23-rc4-mm1/include/linux/mm_types.h
===================================================================
--- test-2.6.23-rc4-mm1.orig/include/linux/mm_types.h
+++ test-2.6.23-rc4-mm1/include/linux/mm_types.h
@@ -48,7 +48,7 @@ struct page {
 						 * indicates order in the buddy
 						 * system if PG_buddy is set.
 						 */
-		struct address_space *mapping;	/* If low bit clear, points to
+		unsigned long mapping;		/* If low bit clear, points to
 						 * inode address_space, or NULL.
 						 * If page mapped as anonymous
 						 * memory, low bit is set, and
Index: test-2.6.23-rc4-mm1/include/linux/mm.h
===================================================================
--- test-2.6.23-rc4-mm1.orig/include/linux/mm.h
+++ test-2.6.23-rc4-mm1/include/linux/mm.h
@@ -563,7 +563,7 @@ void page_address_init(void);
 extern struct address_space swapper_space;
 static inline struct address_space *page_mapping(struct page *page)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = (struct address_space *)page->mapping;
 
 	VM_BUG_ON(PageSlab(page));
 	if (unlikely(PageSwapCache(page)))
@@ -579,7 +579,21 @@ static inline struct address_space *page
 
 static inline int PageAnon(struct page *page)
 {
-	return ((unsigned long)page->mapping & PAGE_MAPPING_ANON) != 0;
+	return (page->mapping & PAGE_MAPPING_ANON) != 0;
+}
+
+static inline struct anon_vma *page_mapping_anon(struct page *page)
+{
+	if (PageAnon(page))
+		return (struct anon_vma *)(page->mapping - PAGE_MAPPING_ANON);
+	return NULL;
+}
+
+static inline struct address_space *page_mapping_cache(struct page *page)
+{
+	if (PageAnon(page))
+		return NULL;
+	return (struct address_space *) page->mapping;
 }
 
 /*
@@ -848,7 +862,7 @@ static inline pmd_t *pmd_alloc(struct mm
 #define pte_lock_init(_page)	do {					\
 	spin_lock_init(__pte_lockptr(_page));				\
 } while (0)
-#define pte_lock_deinit(page)	((page)->mapping = NULL)
+#define pte_lock_deinit(page)	((page)->mapping = 0)
 #define pte_lockptr(mm, pmd)	({(void)(mm); __pte_lockptr(pmd_page(*(pmd)));})
 #else
 /*
Index: test-2.6.23-rc4-mm1/mm/swap_state.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/swap_state.c
+++ test-2.6.23-rc4-mm1/mm/swap_state.c
@@ -45,6 +45,8 @@ struct address_space swapper_space = {
 	.backing_dev_info = &swap_backing_dev_info,
 };
 
+EXPORT_SYMBOL(swapper_space);
+
 #define INC_CACHE_INFO(x)	do { swap_cache_info.x++; } while (0)
 
 static struct {

--
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>

[PATCH] add page->mapping handling interface [2/35] changes in /mm

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in /mm directory.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>


---
 mm/filemap.c        |   24 +++++++++++++-----------
 mm/memory.c         |    6 ++++--
 mm/migrate.c        |   17 ++++++-----------
 mm/page-writeback.c |    4 ++--
 mm/rmap.c           |   27 ++++++++++++---------------
 mm/shmem.c          |    4 ++--
 mm/truncate.c       |   15 ++++++++-------
 7 files changed, 47 insertions(+), 50 deletions(-)

Index: test-2.6.23-rc4-mm1/mm/filemap.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/filemap.c
+++ test-2.6.23-rc4-mm1/mm/filemap.c
@@ -115,11 +115,11 @@ generic_file_direct_IO(int rw, struct ki
  */
 void __remove_from_page_cache(struct page *page)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping(page);
 
 	mem_container_uncharge_page(page);
 	radix_tree_delete(&mapping->page_tree, page->index);
-	page->mapping = NULL;
+	page->mapping = 0;
 	mapping->nrpages--;
 	__dec_zone_page_state(page, NR_FILE_PAGES);
 	BUG_ON(page_mapped(page));
@@ -127,7 +127,7 @@ void __remove_from_page_cache(struct pag
 
 void remove_from_page_cache(struct page *page)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping(page);
 
 	BUG_ON(!PageLocked(page));
 
@@ -454,7 +454,7 @@ int add_to_page_cache(struct page *page,
 		if (!error) {
 			page_cache_get(page);
 			SetPageLocked(page);
-			page->mapping = mapping;
+			page->mapping = (unsigned long)mapping;
 			page->index = offset;
 			mapping->nrpages++;
 			__inc_zone_page_state(page, NR_FILE_PAGES);
@@ -642,7 +642,7 @@ repeat:
 			__lock_page(page);
 
 			/* Has the page been truncated while we slept? */
-			if (unlikely(page->mapping != mapping)) {
+			if (unlikely(!pagecache_consistent(page, mapping))) {
 				unlock_page(page);
 				page_cache_release(page);
 				goto repeat;
@@ -751,7 +751,8 @@ unsigned find_get_pages_contig(struct ad
 	ret = radix_tree_gang_lookup(&mapping->page_tree,
 				(void **)pages, index, nr_pages);
 	for (i = 0; i < ret; i++) {
-		if (pages[i]->mapping == NULL || pages[i]->index != index)
+		if (!page_is_pagecache(pages[i]) ||
+		     pages[i]->index != index)
 			break;
 
 		page_cache_get(pages[i]);
@@ -980,7 +981,7 @@ page_not_up_to_date:
 		lock_page(page);
 
 		/* Did it get truncated before we got the lock? */
-		if (!page->mapping) {
+		if (!page_is_pagecache(page)) {
 			unlock_page(page);
 			page_cache_release(page);
 			continue;
@@ -1007,7 +1008,7 @@ readpage:
 		if (!PageUptodate(page)) {
 			lock_page(page);
 			if (!PageUptodate(page)) {
-				if (page->mapping == NULL) {
+				if (!page_is_pagecache(page)) {
 					/*
 					 * invalidate_inode_pages got it
 					 */
@@ -1546,7 +1547,7 @@ retry:
 		goto out;
 
 	lock_page(page);
-	if (!page->mapping) {
+	if (page_is_pagecache(page)) {
 		unlock_page(page);
 		page_cache_release(page);
 		goto retry;
@@ -2113,7 +2114,8 @@ static ssize_t generic_perform_write_2co
 			 * use a non-zeroing copy, but the APIs aren't too
 			 * consistent.
 			 */
-			if (unlikely(!page->mapping || PageUptodate(page))) {
+			if (unlikely(!page_is_pagecache(page) ||
+					PageUptodate(page))) {
 				unlock_page(page);
 				page_cache_release(page);
 				page_cache_release(src_page);
@@ -2556,7 +2558,7 @@ out:
  */
 int try_to_release_page(struct page *page, gfp_t gfp_mask)
 {
-	struct address_space * const mapping = page->mapping;
+	struct address_space * const mapping = page_mapping(page);
 
 	BUG_ON(!PageLocked(page));
 	if (PageWriteback(page))
Index: test-2.6.23-rc4-mm1/mm/memory.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/memory.c
+++ test-2.6.23-rc4-mm1/mm/memory.c
@@ -650,7 +650,8 @@ static unsigned long zap_pte_range(struc
 				 * unmap shared but keep private pages.
 				 */
 				if (details->check_mapping &&
-				    details->check_mapping != page->mapping)
+				    !pagecache_consistent(page,
+						details->check_mapping))
 					continue;
 				/*
 				 * Each page->index must be checked when
@@ -2309,7 +2310,8 @@ static int __do_fault(struct mm_struct *
 				 * reworking page_mkwrite locking API, which
 				 * is better done later.
 				 */
-				if (!page->mapping) {
+				if (!page_is_pagecache(page) &&
+				    !PageAnon(page)) {
 					ret = 0;
 					anon = 1; /* no anon but release vmf.page */
 					goto out;
Index: test-2.6.23-rc4-mm1/mm/migrate.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/migrate.c
+++ test-2.6.23-rc4-mm1/mm/migrate.c
@@ -223,17 +223,12 @@ static void remove_anon_migration_ptes(s
 {
 	struct anon_vma *anon_vma;
 	struct vm_area_struct *vma;
-	unsigned long mapping;
-
-	mapping = (unsigned long)new->mapping;
-
-	if (!mapping || (mapping & PAGE_MAPPING_ANON) == 0)
-		return;
-
 	/*
 	 * We hold the mmap_sem lock. So no need to call page_lock_anon_vma.
 	 */
-	anon_vma = (struct anon_vma *) (mapping - PAGE_MAPPING_ANON);
+	anon_vma = page_mapping_anon(new);
+	if (!anon_vma)
+		return;
 	spin_lock(&anon_vma->lock);
 
 	list_for_each_entry(vma, &anon_vma->head, anon_vma_node)
@@ -388,7 +383,7 @@ static void migrate_page_copy(struct pag
 	ClearPageActive(page);
 	ClearPagePrivate(page);
 	set_page_private(page, 0);
-	page->mapping = NULL;
+	page->mapping = 0;
 
 	/*
 	 * If any waiters have accumulated on the new page then
@@ -601,7 +596,7 @@ static int move_to_new_page(struct page 
 	if (!rc)
 		remove_migration_ptes(page, newpage);
 	else
-		newpage->mapping = NULL;
+		newpage->mapping = 0;
 
 	unlock_page(newpage);
 
@@ -658,7 +653,7 @@ static int unmap_and_move(new_page_t get
 	 * Calling try_to_unmap() against a page->mapping==NULL page is
 	 * BUG. So handle it here.
 	 */
-	if (!page->mapping)
+	if (!page_is_pagecache(page) && !PageAnon(page))
 		goto rcu_unlock;
 	/* Establish migration ptes or remove ptes */
 	try_to_unmap(page, 1);
Index: test-2.6.23-rc4-mm1/mm/page-writeback.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/page-writeback.c
+++ test-2.6.23-rc4-mm1/mm/page-writeback.c
@@ -650,7 +650,7 @@ retry:
 			 */
 			lock_page(page);
 
-			if (unlikely(page->mapping != mapping)) {
+			if (unlikely(!pagecache_consistent(page, mapping))) {
 				unlock_page(page);
 				continue;
 			}
@@ -758,7 +758,7 @@ int do_writepages(struct address_space *
  */
 int write_one_page(struct page *page, int wait)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping(page);
 	int ret = 0;
 	struct writeback_control wbc = {
 		.sync_mode = WB_SYNC_ALL,
Index: test-2.6.23-rc4-mm1/mm/rmap.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/rmap.c
+++ test-2.6.23-rc4-mm1/mm/rmap.c
@@ -160,16 +160,14 @@ void __init anon_vma_init(void)
 static struct anon_vma *page_lock_anon_vma(struct page *page)
 {
 	struct anon_vma *anon_vma;
-	unsigned long anon_mapping;
 
 	rcu_read_lock();
-	anon_mapping = (unsigned long) page->mapping;
-	if (!(anon_mapping & PAGE_MAPPING_ANON))
-		goto out;
+
 	if (!page_mapped(page))
 		goto out;
-
-	anon_vma = (struct anon_vma *) (anon_mapping - PAGE_MAPPING_ANON);
+	anon_vma = page_mapping_anon(page);
+	if (!anon_vma)
+		goto out;
 	spin_lock(&anon_vma->lock);
 	return anon_vma;
 out:
@@ -208,12 +206,11 @@ vma_address(struct page *page, struct vm
 unsigned long page_address_in_vma(struct page *page, struct vm_area_struct *vma)
 {
 	if (PageAnon(page)) {
-		if ((void *)vma->anon_vma !=
-		    (void *)page->mapping - PAGE_MAPPING_ANON)
+		if (vma->anon_vma != page_mapping_anon(page))
 			return -EFAULT;
 	} else if (page->mapping && !(vma->vm_flags & VM_NONLINEAR)) {
 		if (!vma->vm_file ||
-		    vma->vm_file->f_mapping != page->mapping)
+		    vma->vm_file->f_mapping != page_mapping_cache(page))
 			return -EFAULT;
 	} else
 		return -EFAULT;
@@ -344,7 +341,7 @@ static int page_referenced_file(struct p
 				struct mem_container *mem_cont)
 {
 	unsigned int mapcount;
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	pgoff_t pgoff = page->index << (PAGE_CACHE_SHIFT - PAGE_SHIFT);
 	struct vm_area_struct *vma;
 	struct prio_tree_iter iter;
@@ -422,7 +419,7 @@ int page_referenced(struct page *page, i
 		else if (TestSetPageLocked(page))
 			referenced++;
 		else {
-			if (page->mapping)
+			if (page_mapping_cache(page))
 				referenced +=
 					page_referenced_file(page, mem_cont);
 			unlock_page(page);
@@ -514,7 +511,7 @@ static void __page_set_anon_rmap(struct 
 
 	BUG_ON(!anon_vma);
 	anon_vma = (void *) anon_vma + PAGE_MAPPING_ANON;
-	page->mapping = (struct address_space *) anon_vma;
+	page->mapping = (unsigned long) anon_vma;
 
 	page->index = linear_page_index(vma, address);
 
@@ -549,7 +546,7 @@ static void __page_check_anon_rmap(struc
 	 */
 	struct anon_vma *anon_vma = vma->anon_vma;
 	anon_vma = (void *) anon_vma + PAGE_MAPPING_ANON;
-	BUG_ON(page->mapping != (struct address_space *)anon_vma);
+	BUG_ON(page->mapping != (unsigned long)anon_vma);
 	BUG_ON(page->index != linear_page_index(vma, address));
 #endif
 }
@@ -649,7 +646,7 @@ void page_remove_rmap(struct page *page,
 			printk (KERN_EMERG "  page pfn = %lx\n", page_to_pfn(page));
 			printk (KERN_EMERG "  page->flags = %lx\n", page->flags);
 			printk (KERN_EMERG "  page->count = %x\n", page_count(page));
-			printk (KERN_EMERG "  page->mapping = %p\n", page->mapping);
+			printk (KERN_EMERG "  page->mapping = %p\n", (void*)page->mapping);
 			print_symbol (KERN_EMERG "  vma->vm_ops = %s\n", (unsigned long)vma->vm_ops);
 			if (vma->vm_ops) {
 				print_symbol (KERN_EMERG "  vma->vm_ops->nopage = %s\n", (unsigned long)vma->vm_ops->nopage);
@@ -894,7 +891,7 @@ static int try_to_unmap_anon(struct page
  */
 static int try_to_unmap_file(struct page *page, int migration)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	pgoff_t pgoff = page->index << (PAGE_CACHE_SHIFT - PAGE_SHIFT);
 	struct vm_area_struct *vma;
 	struct prio_tree_iter iter;
Index: test-2.6.23-rc4-mm1/mm/shmem.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/shmem.c
+++ test-2.6.23-rc4-mm1/mm/shmem.c
@@ -917,7 +917,7 @@ static int shmem_writepage(struct page *
 	BUG_ON(!PageLocked(page));
 	BUG_ON(page_mapped(page));
 
-	mapping = page->mapping;
+	mapping = page_mapping_cache(page);
 	index = page->index;
 	inode = mapping->host;
 	info = SHMEM_I(inode);
@@ -1454,7 +1454,7 @@ static const struct inode_operations shm
  */
 static int shmem_readpage(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int error = shmem_getpage(inode, page->index, &page, SGP_CACHE, NULL);
 	unlock_page(page);
 	return error;
Index: test-2.6.23-rc4-mm1/mm/truncate.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/mm/truncate.c
+++ test-2.6.23-rc4-mm1/mm/truncate.c
@@ -37,7 +37,7 @@
 void do_invalidatepage(struct page *page, unsigned long offset)
 {
 	void (*invalidatepage)(struct page *, unsigned long);
-	invalidatepage = page->mapping->a_ops->invalidatepage;
+	invalidatepage = page_mapping_cache(page)->a_ops->invalidatepage;
 #ifdef CONFIG_BLOCK
 	if (!invalidatepage)
 		invalidatepage = block_invalidatepage;
@@ -70,7 +70,7 @@ static inline void truncate_partial_page
 void cancel_dirty_page(struct page *page, unsigned int account_size)
 {
 	if (TestClearPageDirty(page)) {
-		struct address_space *mapping = page->mapping;
+		struct address_space *mapping = page_mapping_cache(page);
 		if (mapping && mapping_cap_account_dirty(mapping)) {
 			dec_zone_page_state(page, NR_FILE_DIRTY);
 			if (account_size)
@@ -93,7 +93,7 @@ EXPORT_SYMBOL(cancel_dirty_page);
 static void
 truncate_complete_page(struct address_space *mapping, struct page *page)
 {
-	if (page->mapping != mapping)
+	if (!pagecache_consistent(page, mapping))
 		return;
 
 	cancel_dirty_page(page, PAGE_CACHE_SIZE);
@@ -120,7 +120,7 @@ invalidate_complete_page(struct address_
 {
 	int ret;
 
-	if (page->mapping != mapping)
+	if (!pagecache_consistent(page, mapping))
 		return 0;
 
 	if (PagePrivate(page) && !try_to_release_page(page, 0))
@@ -342,7 +342,7 @@ EXPORT_SYMBOL(invalidate_mapping_pages);
 static int
 invalidate_complete_page2(struct address_space *mapping, struct page *page)
 {
-	if (page->mapping != mapping)
+	if (!pagecache_consistent(page, mapping))
 		return 0;
 
 	if (PagePrivate(page) && !try_to_release_page(page, GFP_KERNEL))
@@ -367,7 +367,8 @@ static int do_launder_page(struct addres
 {
 	if (!PageDirty(page))
 		return 0;
-	if (page->mapping != mapping || mapping->a_ops->launder_page == NULL)
+	if (!pagecache_consistent(page, mapping) ||
+		mapping->a_ops->launder_page == NULL)
 		return 0;
 	return mapping->a_ops->launder_page(page);
 }
@@ -403,7 +404,7 @@ int invalidate_inode_pages2_range(struct
 			pgoff_t page_index;
 
 			lock_page(page);
-			if (page->mapping != mapping) {
+			if (!pagecache_consistent(page, mapping)) {
 				unlock_page(page);
 				continue;
 			}

--
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>

[PATCH] add page->mapping handling interface [3/35] changes in generic parts

[KAMEZAWA Hiroyuki]

Changes page->mapping hanlding of generic fs routine and kexec.
(other than mm layer..)

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>


---
 fs/buffer.c    |   43 ++++++++++++++++++++++---------------------
 fs/libfs.c     |    2 +-
 fs/mpage.c     |   13 +++++++------
 kernel/kexec.c |    2 +-
 4 files changed, 31 insertions(+), 29 deletions(-)

Index: test-2.6.23-rc4-mm1/kernel/kexec.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/kernel/kexec.c
+++ test-2.6.23-rc4-mm1/kernel/kexec.c
@@ -347,7 +347,7 @@ static struct page *kimage_alloc_pages(g
 	pages = alloc_pages(gfp_mask, order);
 	if (pages) {
 		unsigned int count, i;
-		pages->mapping = NULL;
+		pages->mapping = 0;
 		set_page_private(pages, order);
 		count = 1 << order;
 		for (i = 0; i < count; i++)
Index: test-2.6.23-rc4-mm1/fs/buffer.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/buffer.c
+++ test-2.6.23-rc4-mm1/fs/buffer.c
@@ -467,7 +467,7 @@ static void end_buffer_async_write(struc
 					"I/O error on %s\n",
 			       bdevname(bh->b_bdev, b));
 		}
-		set_bit(AS_EIO, &page->mapping->flags);
+		set_bit(AS_EIO, &page_mapping_cache(page)->flags);
 		set_buffer_write_io_error(bh);
 		clear_buffer_uptodate(bh);
 		SetPageError(page);
@@ -678,7 +678,7 @@ void write_boundary_block(struct block_d
 void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode)
 {
 	struct address_space *mapping = inode->i_mapping;
-	struct address_space *buffer_mapping = bh->b_page->mapping;
+	struct address_space *buffer_mapping = page_mapping_cache(bh->b_page);
 
 	mark_buffer_dirty(bh);
 	if (!mapping->assoc_mapping) {
@@ -713,7 +713,7 @@ static int __set_page_dirty(struct page 
 		return 0;
 
 	write_lock_irq(&mapping->tree_lock);
-	if (page->mapping) {	/* Race with truncate? */
+	if (page_is_pagecache(page)) {	/* Race with truncate? */
 		WARN_ON_ONCE(warn && !PageUptodate(page));
 
 		if (mapping_cap_account_dirty(mapping)) {
@@ -1202,7 +1202,8 @@ void __bforget(struct buffer_head *bh)
 {
 	clear_buffer_dirty(bh);
 	if (!list_empty(&bh->b_assoc_buffers)) {
-		struct address_space *buffer_mapping = bh->b_page->mapping;
+		struct address_space *buffer_mapping =
+					page_mapping_cache(bh->b_page);
 
 		spin_lock(&buffer_mapping->private_lock);
 		list_del_init(&bh->b_assoc_buffers);
@@ -1542,7 +1543,7 @@ void create_empty_buffers(struct page *p
 	} while (bh);
 	tail->b_this_page = head;
 
-	spin_lock(&page->mapping->private_lock);
+	spin_lock(&page_mapping_cache(page)->private_lock);
 	if (PageUptodate(page) || PageDirty(page)) {
 		bh = head;
 		do {
@@ -1554,7 +1555,7 @@ void create_empty_buffers(struct page *p
 		} while (bh != head);
 	}
 	attach_page_buffers(page, head);
-	spin_unlock(&page->mapping->private_lock);
+	spin_unlock(&page_mapping_cache(page)->private_lock);
 }
 EXPORT_SYMBOL(create_empty_buffers);
 
@@ -1761,7 +1762,7 @@ recover:
 	} while ((bh = bh->b_this_page) != head);
 	SetPageError(page);
 	BUG_ON(PageWriteback(page));
-	mapping_set_error(page->mapping, err);
+	mapping_set_error(page_mapping_cache(page), err);
 	set_page_writeback(page);
 	do {
 		struct buffer_head *next = bh->b_this_page;
@@ -2075,7 +2076,7 @@ EXPORT_SYMBOL(generic_write_end);
  */
 int block_read_full_page(struct page *page, get_block_t *get_block)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	sector_t iblock, lblock;
 	struct buffer_head *bh, *head, *arr[MAX_BUF_PER_PAGE];
 	unsigned int blocksize;
@@ -2296,7 +2297,7 @@ out:
 int block_prepare_write(struct page *page, unsigned from, unsigned to,
 			get_block_t *get_block)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int err = __block_prepare_write(inode, page, from, to, get_block);
 	if (err)
 		ClearPageUptodate(page);
@@ -2305,7 +2306,7 @@ int block_prepare_write(struct page *pag
 
 int block_commit_write(struct page *page, unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	__block_commit_write(inode,page,from,to);
 	return 0;
 }
@@ -2313,7 +2314,7 @@ int block_commit_write(struct page *page
 int generic_commit_write(struct file *file, struct page *page,
 		unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
 	__block_commit_write(inode,page,from,to);
 	/*
@@ -2353,7 +2354,7 @@ block_page_mkwrite(struct vm_area_struct
 
 	lock_page(page);
 	size = i_size_read(inode);
-	if ((page->mapping != inode->i_mapping) ||
+	if ((!pagecache_consistent(page, inode->i_mapping)) ||
 	    (page_offset(page) > size)) {
 		/* page got truncated out from underneath us */
 		goto out_unlock;
@@ -2391,7 +2392,7 @@ static void end_buffer_read_nobh(struct 
 int nobh_prepare_write(struct page *page, unsigned from, unsigned to,
 			get_block_t *get_block)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	const unsigned blkbits = inode->i_blkbits;
 	const unsigned blocksize = 1 << blkbits;
 	struct buffer_head *head, *bh;
@@ -2505,7 +2506,7 @@ failed:
 	 * the handling of potential IO errors during writeout would be hard
 	 * (could try doing synchronous writeout, but what if that fails too?)
 	 */
-	spin_lock(&page->mapping->private_lock);
+	spin_lock(&page_mapping_cache(page)->private_lock);
 	bh = head;
 	block_start = 0;
 	do {
@@ -2535,7 +2536,7 @@ next:
 		bh = bh->b_this_page;
 	} while (bh != head);
 	attach_page_buffers(page, head);
-	spin_unlock(&page->mapping->private_lock);
+	spin_unlock(&page_mapping_cache(page)->private_lock);
 
 	return ret;
 }
@@ -2548,7 +2549,7 @@ EXPORT_SYMBOL(nobh_prepare_write);
 int nobh_commit_write(struct file *file, struct page *page,
 		unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
 
 	if (page_has_buffers(page))
@@ -2572,7 +2573,7 @@ EXPORT_SYMBOL(nobh_commit_write);
 int nobh_writepage(struct page *page, get_block_t *get_block,
 			struct writeback_control *wbc)
 {
-	struct inode * const inode = page->mapping->host;
+	struct inode * const inode = page_inode(page);
 	loff_t i_size = i_size_read(inode);
 	const pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
 	unsigned offset;
@@ -2737,7 +2738,7 @@ out:
 int block_write_full_page(struct page *page, get_block_t *get_block,
 			struct writeback_control *wbc)
 {
-	struct inode * const inode = page->mapping->host;
+	struct inode * const inode = page_inode(page);
 	loff_t i_size = i_size_read(inode);
 	const pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
 	unsigned offset;
@@ -2966,8 +2967,8 @@ drop_buffers(struct page *page, struct b
 
 	bh = head;
 	do {
-		if (buffer_write_io_error(bh) && page->mapping)
-			set_bit(AS_EIO, &page->mapping->flags);
+		if (buffer_write_io_error(bh) && page_is_pagecache(page))
+			set_bit(AS_EIO, &page_mapping_cache(page)->flags);
 		if (buffer_busy(bh))
 			goto failed;
 		bh = bh->b_this_page;
@@ -2989,7 +2990,7 @@ failed:
 
 int try_to_free_buffers(struct page *page)
 {
-	struct address_space * const mapping = page->mapping;
+	struct address_space * const mapping = page_mapping_cache(page);
 	struct buffer_head *buffers_to_free = NULL;
 	int ret = 0;
 
Index: test-2.6.23-rc4-mm1/fs/libfs.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/libfs.c
+++ test-2.6.23-rc4-mm1/fs/libfs.c
@@ -374,7 +374,7 @@ int simple_write_begin(struct file *file
 static int simple_commit_write(struct file *file, struct page *page,
 			       unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
 
 	if (!PageUptodate(page))
Index: test-2.6.23-rc4-mm1/fs/mpage.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/mpage.c
+++ test-2.6.23-rc4-mm1/fs/mpage.c
@@ -81,8 +81,9 @@ static int mpage_end_io_write(struct bio
 
 		if (!uptodate){
 			SetPageError(page);
-			if (page->mapping)
-				set_bit(AS_EIO, &page->mapping->flags);
+			if (page_is_pagecache(page))
+				set_bit(AS_EIO,
+					&page_mapping_cache(page)->flags);
 		}
 		end_page_writeback(page);
 	} while (bvec >= bio->bi_io_vec);
@@ -133,7 +134,7 @@ mpage_alloc(struct block_device *bdev,
 static void 
 map_buffer_to_page(struct page *page, struct buffer_head *bh, int page_block) 
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct buffer_head *page_bh, *head;
 	int block = 0;
 
@@ -177,7 +178,7 @@ do_mpage_readpage(struct bio *bio, struc
 		sector_t *last_block_in_bio, struct buffer_head *map_bh,
 		unsigned long *first_logical_block, get_block_t get_block)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	const unsigned blkbits = inode->i_blkbits;
 	const unsigned blocks_per_page = PAGE_CACHE_SIZE >> blkbits;
 	const unsigned blocksize = 1 << blkbits;
@@ -460,8 +461,8 @@ static int __mpage_writepage(struct page
 {
 	struct mpage_data *mpd = data;
 	struct bio *bio = mpd->bio;
-	struct address_space *mapping = page->mapping;
-	struct inode *inode = page->mapping->host;
+	struct address_space *mapping = page_mapping_cache(page);
+	struct inode *inode = page_inode(page);
 	const unsigned blkbits = inode->i_blkbits;
 	unsigned long end_index;
 	const unsigned blocks_per_page = PAGE_CACHE_SIZE >> blkbits;

--
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>

[PATCH] add page->mapping handling interface [4/35] changes in AFFS

[KAMEZAWA Hiroyuki]

use page_inode() in AFFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/affs/file.c    |    4 ++--
 fs/affs/symlink.c |    2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/affs/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/affs/file.c
+++ test-2.6.23-rc4-mm1/fs/affs/file.c
@@ -485,7 +485,7 @@ affs_getemptyblk_ino(struct inode *inode
 static int
 affs_do_readpage_ofs(struct file *file, struct page *page, unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct super_block *sb = inode->i_sb;
 	struct buffer_head *bh;
 	char *data;
@@ -593,7 +593,7 @@ out:
 static int
 affs_readpage_ofs(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	u32 to;
 	int err;
 
Index: test-2.6.23-rc4-mm1/fs/affs/symlink.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/affs/symlink.c
+++ test-2.6.23-rc4-mm1/fs/affs/symlink.c
@@ -13,7 +13,7 @@
 static int affs_symlink_readpage(struct file *file, struct page *page)
 {
 	struct buffer_head *bh;
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	char *link = kmap(page);
 	struct slink_front *lf;
 	int err;

--
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>

[PATCH] add page->mapping handling interface [5/35] changes in AFS

[KAMEZAWA Hiroyuki]

Use page->mapping interface in AFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/afs/file.c |    7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/afs/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/afs/file.c
+++ test-2.6.23-rc4-mm1/fs/afs/file.c
@@ -145,7 +145,7 @@ static int afs_readpage(struct file *fil
 	off_t offset;
 	int ret;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 
 	ASSERT(file != NULL);
 	key = file->private_data;
@@ -253,8 +253,7 @@ static void afs_invalidatepage(struct pa
 
 			ret = 0;
 			if (!PageWriteback(page))
-				ret = page->mapping->a_ops->releasepage(page,
-									0);
+				ret = page_mapping_cache(page)->a_ops->releasepage(page, 0);
 			/* possibly should BUG_ON(!ret); - neilb */
 		}
 	}
@@ -277,7 +276,7 @@ static int afs_launder_page(struct page 
  */
 static int afs_releasepage(struct page *page, gfp_t gfp_flags)
 {
-	struct afs_vnode *vnode = AFS_FS_I(page->mapping->host);
+	struct afs_vnode *vnode = AFS_FS_I(page_inode(page));
 	struct afs_writeback *wb;
 
 	_enter("{{%x:%u}[%lu],%lx},%x",

--
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>

[PATCH] add page->mapping handling interface [6/35] changes in CIFS

[KAMEZAWA Hiroyuki]

Change page->mapping handling in CIFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>


---
 fs/cifs/file.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/cifs/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/cifs/file.c
+++ test-2.6.23-rc4-mm1/fs/cifs/file.c
@@ -1056,7 +1056,7 @@ struct cifsFileInfo *find_writable_file(
 
 static int cifs_partialpagewrite(struct page *page, unsigned from, unsigned to)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	loff_t offset = (loff_t)page->index << PAGE_CACHE_SHIFT;
 	char *write_data;
 	int rc = -EFAULT;
@@ -1069,7 +1069,7 @@ static int cifs_partialpagewrite(struct 
 	if (!mapping || !mapping->host)
 		return -EFAULT;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	cifs_sb = CIFS_SB(inode->i_sb);
 	pTcon = cifs_sb->tcon;
 
@@ -1209,7 +1209,7 @@ retry:
 			else if (TestSetPageLocked(page))
 				break;
 
-			if (unlikely(page->mapping != mapping)) {
+			if (unlikely(pagecache_consistent(page, mapping))) {
 				unlock_page(page);
 				break;
 			}
@@ -1371,7 +1371,7 @@ static int cifs_commit_write(struct file
 {
 	int xid;
 	int rc = 0;
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t position = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
 	char *page_data;
 
@@ -1973,7 +1973,7 @@ static int cifs_prepare_write(struct fil
 	}
 
 	offset = (loff_t)page->index << PAGE_CACHE_SHIFT;
-	i_size = i_size_read(page->mapping->host);
+	i_size = i_size_read(page_inode(page));
 
 	if ((offset >= i_size) ||
 	    ((from == 0) && (offset + to) >= i_size)) {

--
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>

[PATCH] add page->mapping handling interface [7/35] changes in CODA

[KAMEZAWA Hiroyuki]

Change page->mapping handling in CODA

Singed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/coda/symlink.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/coda/symlink.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/coda/symlink.c
+++ test-2.6.23-rc4-mm1/fs/coda/symlink.c
@@ -23,7 +23,7 @@
 
 static int coda_symlink_filler(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int error;
 	struct coda_inode_info *cii;
 	unsigned int len = PAGE_SIZE;

--
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>

[PATCH] add page->mapping handling interface [8/35] changes in CRAMFS

[KAMEZAWA Hiroyuki]

patches for handling page->mapping in CRAMFS.

Signed-off-by : KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/cramfs/inode.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/cramfs/inode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/cramfs/inode.c
+++ test-2.6.23-rc4-mm1/fs/cramfs/inode.c
@@ -469,7 +469,7 @@ static struct dentry * cramfs_lookup(str
 
 static int cramfs_readpage(struct file *file, struct page * page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	u32 maxblock, bytes_filled;
 	void *pgdata;
 

--
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>

[PATCH] add page->mapping handling interface [9/35] changes in ECRYPTFS

[KAMEZAWA Hiroyuki]

Change page->mapping handling in ecryptfs

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>


---
 fs/ecryptfs/crypto.c |    9 ++++-----
 fs/ecryptfs/mmap.c   |   14 +++++++-------
 2 files changed, 11 insertions(+), 12 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/ecryptfs/crypto.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ecryptfs/crypto.c
+++ test-2.6.23-rc4-mm1/fs/ecryptfs/crypto.c
@@ -504,8 +504,8 @@ int ecryptfs_encrypt_page(struct ecryptf
 #define ECRYPTFS_PAGE_STATE_WRITTEN   3
 	int page_state;
 
-	lower_inode = ecryptfs_inode_to_lower(ctx->page->mapping->host);
-	inode_info = ecryptfs_inode_to_private(ctx->page->mapping->host);
+	lower_inode = ecryptfs_inode_to_lower(page_inode(ctx->page));
+	inode_info = ecryptfs_inode_to_private(page_inode(ctx->page));
 	crypt_stat = &inode_info->crypt_stat;
 	if (!(crypt_stat->flags & ECRYPTFS_ENCRYPTED)) {
 		rc = ecryptfs_copy_page_to_lower(ctx->page, lower_inode,
@@ -636,9 +636,8 @@ int ecryptfs_decrypt_page(struct file *f
 	int num_extents_per_page;
 	int page_state;
 
-	crypt_stat = &(ecryptfs_inode_to_private(
-			       page->mapping->host)->crypt_stat);
-	lower_inode = ecryptfs_inode_to_lower(page->mapping->host);
+	crypt_stat = &(ecryptfs_inode_to_private(page_inode(page))->crypt_stat);
+	lower_inode = ecryptfs_inode_to_lower(page_inode(page));
 	if (!(crypt_stat->flags & ECRYPTFS_ENCRYPTED)) {
 		rc = ecryptfs_do_readpage(file, page, page->index);
 		if (rc)
Index: test-2.6.23-rc4-mm1/fs/ecryptfs/mmap.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ecryptfs/mmap.c
+++ test-2.6.23-rc4-mm1/fs/ecryptfs/mmap.c
@@ -363,7 +363,7 @@ out:
  */
 static int fill_zeros_to_end_of_page(struct page *page, unsigned int to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int end_byte_in_page;
 
 	if ((i_size_read(inode) / PAGE_CACHE_SIZE) != page->index)
@@ -411,7 +411,7 @@ static int ecryptfs_prepare_write(struct
 	if (page->index != 0) {
 		loff_t end_of_prev_pg_pos = page_offset(page) - 1;
 
-		if (end_of_prev_pg_pos > i_size_read(page->mapping->host)) {
+		if (end_of_prev_pg_pos > i_size_read(page_inode(page))) {
 			rc = ecryptfs_truncate(file->f_path.dentry,
 					       end_of_prev_pg_pos);
 			if (rc) {
@@ -421,7 +421,7 @@ static int ecryptfs_prepare_write(struct
 				goto out;
 			}
 		}
-		if (end_of_prev_pg_pos + 1 > i_size_read(page->mapping->host))
+		if (end_of_prev_pg_pos + 1 > i_size_read(page_inode(page)))
 			zero_user_page(page, 0, PAGE_CACHE_SIZE, KM_USER0);
 	}
 out:
@@ -683,7 +683,7 @@ static int ecryptfs_commit_write(struct 
 	struct ecryptfs_crypt_stat *crypt_stat;
 	int rc;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	lower_inode = ecryptfs_inode_to_lower(inode);
 	lower_file = ecryptfs_file_to_lower(file);
 	mutex_lock(&lower_inode->i_mutex);
@@ -805,7 +805,7 @@ static void ecryptfs_sync_page(struct pa
 	struct inode *lower_inode;
 	struct page *lower_page;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	lower_inode = ecryptfs_inode_to_lower(inode);
 	/* NOTE: Recently swapped with grab_cache_page(), since
 	 * sync_page() just makes sure that pending I/O gets done. */
@@ -814,8 +814,8 @@ static void ecryptfs_sync_page(struct pa
 		ecryptfs_printk(KERN_DEBUG, "find_lock_page failed\n");
 		return;
 	}
-	if (lower_page->mapping->a_ops->sync_page)
-		lower_page->mapping->a_ops->sync_page(lower_page);
+	if (page_mapping_cache(lower_page)->a_ops->sync_page)
+		page_mapping_cache(lower_page)->a_ops->sync_page(lower_page);
 	ecryptfs_printk(KERN_DEBUG, "Unlocking page with index = [0x%.16x]\n",
 			lower_page->index);
 	unlock_page(lower_page);

--
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>

[PATCH] add page->mapping handling interface [10/35] changes in EFS

[KAMEZAWA Hiroyuki]

Change page->mapping handling in EFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>


Index: test-2.6.23-rc4-mm1/fs/efs/symlink.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/efs/symlink.c
+++ test-2.6.23-rc4-mm1/fs/efs/symlink.c
@@ -16,7 +16,7 @@ static int efs_symlink_readpage(struct f
 {
 	char *link = kmap(page);
 	struct buffer_head * bh;
-	struct inode * inode = page->mapping->host;
+	struct inode * inode = page_inode(page);
 	efs_block_t size = inode->i_size;
 	int err;
   

--
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>

[PATCH] add page->mapping handling interface [11/35] changes in EXT2

[KAMEZAWA Hiroyuki]

Change page->mapping handling in ext2

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/ext2/dir.c |   20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/ext2/dir.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ext2/dir.c
+++ test-2.6.23-rc4-mm1/fs/ext2/dir.c
@@ -65,7 +65,7 @@ ext2_last_byte(struct inode *inode, unsi
 
 static int ext2_commit_chunk(struct page *page, loff_t pos, unsigned len)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *dir = mapping->host;
 	int err = 0;
 
@@ -87,7 +87,7 @@ static int ext2_commit_chunk(struct page
 
 static void ext2_check_page(struct page *page)
 {
-	struct inode *dir = page->mapping->host;
+	struct inode *dir = page_inode(page);
 	struct super_block *sb = dir->i_sb;
 	unsigned chunk_size = ext2_chunk_size(dir);
 	char *kaddr = page_address(page);
@@ -429,7 +429,7 @@ void ext2_set_link(struct inode *dir, st
 	int err;
 
 	lock_page(page);
-	err = __ext2_write_begin(NULL, page->mapping, pos, len,
+	err = __ext2_write_begin(NULL, page_mapping_cache(page), pos, len,
 				AOP_FLAG_UNINTERRUPTIBLE, &page, NULL);
 	BUG_ON(err);
 	de->inode = cpu_to_le32(inode->i_ino);
@@ -512,8 +512,8 @@ int ext2_add_link (struct dentry *dentry
 got_it:
 	pos = page_offset(page) +
 		(char*)de - (char*)page_address(page);
-	err = __ext2_write_begin(NULL, page->mapping, pos, rec_len, 0,
-							&page, NULL);
+	err = __ext2_write_begin(NULL, page_mapping_cache(page), pos, rec_len,
+				0, &page, NULL);
 	if (err)
 		goto out_unlock;
 	if (de->inode) {
@@ -546,7 +546,7 @@ out_unlock:
  */
 int ext2_delete_entry (struct ext2_dir_entry_2 * dir, struct page * page )
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *inode = mapping->host;
 	char *kaddr = page_address(page);
 	unsigned from = ((char*)dir - kaddr) & ~(ext2_chunk_size(inode)-1);
@@ -570,8 +570,8 @@ int ext2_delete_entry (struct ext2_dir_e
 		from = (char*)pde - (char*)page_address(page);
 	pos = page_offset(page) + from;
 	lock_page(page);
-	err = __ext2_write_begin(NULL, page->mapping, pos, to - from, 0,
-							&page, NULL);
+	err = __ext2_write_begin(NULL, page_mapping_cache(page), pos,
+				to - from, 0, &page, NULL);
 	BUG_ON(err);
 	if (pde)
 		pde->rec_len = cpu_to_le16(to - from);
@@ -600,8 +600,8 @@ int ext2_make_empty(struct inode *inode,
 	if (!page)
 		return -ENOMEM;
 
-	err = __ext2_write_begin(NULL, page->mapping, 0, chunk_size, 0,
-							&page, NULL);
+	err = __ext2_write_begin(NULL, page_mapping_cache(page), 0, chunk_size,
+					0, &page, NULL);
 	if (err) {
 		unlock_page(page);
 		goto fail;

--
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>

[PATCH] add page->mapping handling interface [12/35] changes in EXT3

[KAMEZAWA Hiroyuki]

Change page->mapping handling in EXT3

Signed-off-by:	KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/ext3/inode.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/ext3/inode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ext3/inode.c
+++ test-2.6.23-rc4-mm1/fs/ext3/inode.c
@@ -1484,7 +1484,7 @@ static int journal_dirty_data_fn(handle_
 static int ext3_ordered_writepage(struct page *page,
 				struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct buffer_head *page_bufs;
 	handle_t *handle = NULL;
 	int ret = 0;
@@ -1550,7 +1550,7 @@ out_fail:
 static int ext3_writeback_writepage(struct page *page,
 				struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	handle_t *handle = NULL;
 	int ret = 0;
 	int err;
@@ -1583,7 +1583,7 @@ out_fail:
 static int ext3_journalled_writepage(struct page *page,
 				struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	handle_t *handle = NULL;
 	int ret = 0;
 	int err;
@@ -1653,7 +1653,7 @@ ext3_readpages(struct file *file, struct
 
 static void ext3_invalidatepage(struct page *page, unsigned long offset)
 {
-	journal_t *journal = EXT3_JOURNAL(page->mapping->host);
+	journal_t *journal = EXT3_JOURNAL(page_inode(page));
 
 	/*
 	 * If it's a full truncate we just forget about the pending dirtying
@@ -1666,7 +1666,7 @@ static void ext3_invalidatepage(struct p
 
 static int ext3_releasepage(struct page *page, gfp_t wait)
 {
-	journal_t *journal = EXT3_JOURNAL(page->mapping->host);
+	journal_t *journal = EXT3_JOURNAL(page_inode(page));
 
 	WARN_ON(PageChecked(page));
 	if (!page_has_buffers(page))

--
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>

[PATCH] add page->mapping handling interface [13/35] changes in EXT4

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in EXT4

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/ext4/inode.c     |   10 +++++-----
 fs/ext4/writeback.c |   24 ++++++++++++------------
 2 files changed, 17 insertions(+), 17 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/ext4/inode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ext4/inode.c
+++ test-2.6.23-rc4-mm1/fs/ext4/inode.c
@@ -1482,7 +1482,7 @@ static int jbd2_journal_dirty_data_fn(ha
 static int ext4_ordered_writepage(struct page *page,
 				struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct buffer_head *page_bufs;
 	handle_t *handle = NULL;
 	int ret = 0;
@@ -1548,7 +1548,7 @@ out_fail:
 static int ext4_writeback_writepage(struct page *page,
 				struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	handle_t *handle = NULL;
 	int ret = 0;
 	int err;
@@ -1581,7 +1581,7 @@ out_fail:
 static int ext4_journalled_writepage(struct page *page,
 				struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	handle_t *handle = NULL;
 	int ret = 0;
 	int err;
@@ -1651,7 +1651,7 @@ ext4_readpages(struct file *file, struct
 
 static void ext4_invalidatepage(struct page *page, unsigned long offset)
 {
-	journal_t *journal = EXT4_JOURNAL(page->mapping->host);
+	journal_t *journal = EXT4_JOURNAL(page_inode(page));
 
 	/*
 	 * If it's a full truncate we just forget about the pending dirtying
@@ -1664,7 +1664,7 @@ static void ext4_invalidatepage(struct p
 
 static int ext4_releasepage(struct page *page, gfp_t wait)
 {
-	journal_t *journal = EXT4_JOURNAL(page->mapping->host);
+	journal_t *journal = EXT4_JOURNAL(page_inode(page));
 
 	WARN_ON(PageChecked(page));
 	if (!page_has_buffers(page))
Index: test-2.6.23-rc4-mm1/fs/ext4/writeback.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ext4/writeback.c
+++ test-2.6.23-rc4-mm1/fs/ext4/writeback.c
@@ -175,7 +175,7 @@ static struct bio *ext4_wb_bio_submit(st
 
 int inline ext4_wb_reserve_space_page(struct page *page, int blocks)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int total, mdb, err;
 
 	wb_debug("reserve %d blocks for page %lu from inode %lu\n",
@@ -263,7 +263,7 @@ static inline int ext4_wb_drop_page_rese
 	 * reserved blocks. we could release reserved blocks right
 	 * now, but I'd prefer to make this once per several blocks */
 	wb_debug("drop reservation from page %lu from inode %lu\n",
-			page->index, page->mapping->host->i_ino);
+			page->index, page_inode(page)->i_ino);
 	BUG_ON(!PageBooked(page));
 	ClearPageBooked(page);
 	return 0;
@@ -711,7 +711,7 @@ int ext4_wb_writepages(struct address_sp
 			if (wbc->sync_mode != WB_SYNC_NONE)
 				wait_on_page_writeback(page);
 
-			if (page->mapping != mapping) {
+			if (pagecache_consistent(page, mapping)) {
 				unlock_page(page);
 				continue;
 			}
@@ -853,12 +853,12 @@ static void ext4_wb_clear_page(struct pa
 int ext4_wb_prepare_write(struct file *file, struct page *page,
 			      unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct buffer_head bh, *bhw = &bh;
 	int err = 0;
 
 	wb_debug("prepare page %lu (%u-%u) for inode %lu\n",
-			page->index, from, to, page->mapping->host->i_ino);
+			page->index, from, to, page_inode(page)->i_ino);
 
 	/* if page is uptodate this means that ->prepare_write() has
 	 * been called on page before and page is mapped to disk or
@@ -912,7 +912,7 @@ int ext4_wb_commit_write(struct file *fi
 			     unsigned from, unsigned to)
 {
 	loff_t pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to;
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int err = 0;
 
 	wb_debug("commit page %lu (%u-%u) for inode %lu\n",
@@ -952,7 +952,7 @@ int ext4_wb_commit_write(struct file *fi
 int ext4_wb_write_single_page(struct page *page,
 					struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct ext4_wb_control wc;
 	int err;
 
@@ -964,7 +964,7 @@ int ext4_wb_write_single_page(struct pag
 		atomic_inc(&EXT4_SB(inode->i_sb)->s_wb_collisions_sp);
 #endif
 
-	ext4_wb_init_control(&wc, page->mapping);
+	ext4_wb_init_control(&wc, page_mapping_cache(page));
 
 	BUG_ON(PageWriteback(page));
 	set_page_writeback(page);
@@ -988,7 +988,7 @@ int ext4_wb_write_single_page(struct pag
 
 int ext4_wb_writepage(struct page *page, struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t i_size = i_size_read(inode);
 	pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT;
 	unsigned offset;
@@ -1002,7 +1002,7 @@ int ext4_wb_writepage(struct page *page,
 	 * hasn't space on a disk yet, leave it for that thread
 	 */
 #if 1
-	if (atomic_read(&EXT4_I(page->mapping->host)->i_wb_writers)
+	if (atomic_read(&EXT4_I(page_inode(page))->i_wb_writers)
 			&& !PageMappedToDisk(page)) {
 		__set_page_dirty_nobuffers(page);
 		unlock_page(page);
@@ -1054,7 +1054,7 @@ int ext4_wb_releasepage(struct page *pag
 	wb_debug("release %sM%sR page %lu from inode %lu (wait %d)\n",
 			PageMappedToDisk(page) ? "" : "!",
 			PageBooked(page) ? "" : "!",
-			page->index, page->mapping->host->i_ino, wait);
+			page->index, page_inode(page)->i_ino, wait);
 
 	if (PageWriteback(page))
 		return 0;
@@ -1066,7 +1066,7 @@ int ext4_wb_releasepage(struct page *pag
 
 void ext4_wb_invalidatepage(struct page *page, unsigned long offset)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int ret = 0;
 
 	/* ->invalidatepage() is called when page is marked Private.

--
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>

[PATCH] add page->mapping handling interface [14/35] changes in freevxfs

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in freevxfs.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/freevxfs/vxfs_immed.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/freevxfs/vxfs_immed.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/freevxfs/vxfs_immed.c
+++ test-2.6.23-rc4-mm1/fs/freevxfs/vxfs_immed.c
@@ -98,7 +98,7 @@ vxfs_immed_follow_link(struct dentry *dp
 static int
 vxfs_immed_readpage(struct file *fp, struct page *pp)
 {
-	struct vxfs_inode_info	*vip = VXFS_INO(pp->mapping->host);
+	struct vxfs_inode_info	*vip = VXFS_INO(page_inode(pp));
 	u_int64_t	offset = (u_int64_t)pp->index << PAGE_CACHE_SHIFT;
 	caddr_t		kaddr;
 

--
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>

[PATCH] add page->mapping handling interface [15/35] changes in FUSE

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in FUSE

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
---
 fs/fuse/file.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/fuse/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/fuse/file.c
+++ test-2.6.23-rc4-mm1/fs/fuse/file.c
@@ -310,7 +310,7 @@ static size_t fuse_send_read(struct fuse
 
 static int fuse_readpage(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct fuse_conn *fc = get_fuse_conn(inode);
 	struct fuse_req *req;
 	int err;
@@ -342,7 +342,7 @@ static void fuse_readpages_end(struct fu
 {
 	int i;
 
-	fuse_invalidate_attr(req->pages[0]->mapping->host); /* atime changed */
+	fuse_invalidate_attr(page_inode(req->pages[0])); /* atime changed */
 
 	for (i = 0; i < req->num_pages; i++) {
 		struct page *page = req->pages[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>

[PATCH] add page->mapping handling interface [16/35] changes in GFS2

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in GFS2

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/gfs2/log.c         |    4 ++--
 fs/gfs2/lops.c        |    2 +-
 fs/gfs2/meta_io.c     |    2 +-
 fs/gfs2/ops_address.c |   16 ++++++++--------
 4 files changed, 12 insertions(+), 12 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/gfs2/log.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/gfs2/log.c
+++ test-2.6.23-rc4-mm1/fs/gfs2/log.c
@@ -229,8 +229,8 @@ static void gfs2_ail2_empty_one(struct g
 		list_del(&bd->bd_ail_st_list);
 		list_del(&bd->bd_ail_gl_list);
 		atomic_dec(&bd->bd_gl->gl_ail_count);
-		if (bd->bd_bh->b_page->mapping) {
-			bh_ip = GFS2_I(bd->bd_bh->b_page->mapping->host);
+		if (page_is_pagecache(bd->bd_bh->b_page)) {
+			bh_ip = GFS2_I(page_inode(bd->bd_bh->b_page));
 			gfs2_meta_cache_flush(bh_ip);
 		}
 		brelse(bd->bd_bh);
Index: test-2.6.23-rc4-mm1/fs/gfs2/lops.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/gfs2/lops.c
+++ test-2.6.23-rc4-mm1/fs/gfs2/lops.c
@@ -473,7 +473,7 @@ static void databuf_lo_add(struct gfs2_s
 {
 	struct gfs2_bufdata *bd = container_of(le, struct gfs2_bufdata, bd_le);
 	struct gfs2_trans *tr = current->journal_info;
-	struct address_space *mapping = bd->bd_bh->b_page->mapping;
+	struct address_space *mapping = page_mapping_cache(bd->bd_bh->b_page);
 	struct gfs2_inode *ip = GFS2_I(mapping->host);
 
 	gfs2_log_lock(sdp);
Index: test-2.6.23-rc4-mm1/fs/gfs2/meta_io.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/gfs2/meta_io.c
+++ test-2.6.23-rc4-mm1/fs/gfs2/meta_io.c
@@ -388,7 +388,7 @@ void gfs2_meta_wipe(struct gfs2_inode *i
 			if (test_clear_buffer_pinned(bh)) {
 				struct gfs2_trans *tr = current->journal_info;
 				struct gfs2_inode *bh_ip =
-					GFS2_I(bh->b_page->mapping->host);
+					GFS2_I(page_inode(bh->b_page));
 
 				gfs2_log_lock(sdp);
 				list_del_init(&bd->bd_le.le_list);
Index: test-2.6.23-rc4-mm1/fs/gfs2/ops_address.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/gfs2/ops_address.c
+++ test-2.6.23-rc4-mm1/fs/gfs2/ops_address.c
@@ -114,7 +114,7 @@ static int gfs2_get_block_direct(struct 
 
 static int gfs2_writepage(struct page *page, struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	loff_t i_size = i_size_read(inode);
@@ -133,7 +133,7 @@ static int gfs2_writepage(struct page *p
 	/* Is the page fully outside i_size? (truncate in progress) */
         offset = i_size & (PAGE_CACHE_SIZE-1);
 	if (page->index > end_index || (page->index == end_index && !offset)) {
-		page->mapping->a_ops->invalidatepage(page, 0);
+		page_mapping_cache(page)->a_ops->invalidatepage(page, 0);
 		unlock_page(page);
 		return 0; /* don't care */
 	}
@@ -241,8 +241,8 @@ static int stuffed_readpage(struct gfs2_
 
 static int gfs2_readpage(struct file *file, struct page *page)
 {
-	struct gfs2_inode *ip = GFS2_I(page->mapping->host);
-	struct gfs2_sbd *sdp = GFS2_SB(page->mapping->host);
+	struct gfs2_inode *ip = GFS2_I(page_inode(page));
+	struct gfs2_sbd *sdp = GFS2_SB(page_inode(page));
 	struct gfs2_file *gf = NULL;
 	struct gfs2_holder gh;
 	int error;
@@ -560,7 +560,7 @@ static int gfs2_write_end(struct file *f
 			  loff_t pos, unsigned len, unsigned copied,
 			  struct page *page, void *fsdata)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	struct buffer_head *dibh;
@@ -624,8 +624,8 @@ failed:
  
 static int gfs2_set_page_dirty(struct page *page)
 {
-	struct gfs2_inode *ip = GFS2_I(page->mapping->host);
-	struct gfs2_sbd *sdp = GFS2_SB(page->mapping->host);
+	struct gfs2_inode *ip = GFS2_I(page_inode(page));
+	struct gfs2_sbd *sdp = GFS2_SB(page_inode(page));
 
 	if (sdp->sd_args.ar_data == GFS2_DATA_ORDERED || gfs2_is_jdata(ip))
 		SetPageChecked(page);
@@ -746,7 +746,7 @@ out:
 
 int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 {
-	struct inode *aspace = page->mapping->host;
+	struct inode *aspace = page_inode(page);
 	struct gfs2_sbd *sdp = aspace->i_sb->s_fs_info;
 	struct buffer_head *bh, *head;
 	struct gfs2_bufdata *bd;

--
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>

[PATCH] add page->mapping handling interface [17/35] changes in HFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in HFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/hfs/inode.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/hfs/inode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/hfs/inode.c
+++ test-2.6.23-rc4-mm1/fs/hfs/inode.c
@@ -52,7 +52,7 @@ static sector_t hfs_bmap(struct address_
 
 static int hfs_releasepage(struct page *page, gfp_t mask)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct super_block *sb = inode->i_sb;
 	struct hfs_btree *tree;
 	struct hfs_bnode *node;

--
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>

[PATCH] add page->mapping handling interface [18/35] changes in HFSPLUS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in HFSPLUS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/hfsplus/inode.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/hfsplus/inode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/hfsplus/inode.c
+++ test-2.6.23-rc4-mm1/fs/hfsplus/inode.c
@@ -44,7 +44,7 @@ static sector_t hfsplus_bmap(struct addr
 
 static int hfsplus_releasepage(struct page *page, gfp_t mask)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct super_block *sb = inode->i_sb;
 	struct hfs_btree *tree;
 	struct hfs_bnode *node;

--
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>

[PATCH] add page->mapping handling interface [19/35] changes in HPFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in HPFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/hpfs/namei.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/hpfs/namei.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/hpfs/namei.c
+++ test-2.6.23-rc4-mm1/fs/hpfs/namei.c
@@ -511,7 +511,7 @@ out:
 static int hpfs_symlink_readpage(struct file *file, struct page *page)
 {
 	char *link = kmap(page);
-	struct inode *i = page->mapping->host;
+	struct inode *i = page_inode(page);
 	struct fnode *fnode;
 	struct buffer_head *bh;
 	int err;

--
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>

[PATCH] add page->mapping handling interface [20/35] changes in ISOFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in ISOFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/isofs/rock.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/isofs/rock.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/isofs/rock.c
+++ test-2.6.23-rc4-mm1/fs/isofs/rock.c
@@ -640,7 +640,7 @@ int parse_rock_ridge_inode(struct iso_di
  */
 static int rock_ridge_symlink_readpage(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct iso_inode_info *ei = ISOFS_I(inode);
 	char *link = kmap(page);
 	unsigned long bufsize = ISOFS_BUFFER_SIZE(inode);

--
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>

[PATCH] add page->mapping handling interface [21/35] changes in JBD

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in JBD

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/jbd/journal.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/jbd/journal.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/jbd/journal.c
+++ test-2.6.23-rc4-mm1/fs/jbd/journal.c
@@ -1822,7 +1822,7 @@ repeat:
 		jh = bh2jh(bh);
 	} else {
 		if (!(atomic_read(&bh->b_count) > 0 ||
-				(bh->b_page && bh->b_page->mapping))) {
+				(bh->b_page && page_mapping_cache(bh->b_page)))) {
 			printk(KERN_EMERG "%s: bh->b_count=%d\n",
 				__FUNCTION__, atomic_read(&bh->b_count));
 			printk(KERN_EMERG "%s: bh->b_page=%p\n",
@@ -1830,7 +1830,7 @@ repeat:
 			if (bh->b_page)
 				printk(KERN_EMERG "%s: "
 						"bh->b_page->mapping=%p\n",
-					__FUNCTION__, bh->b_page->mapping);
+				 __FUNCTION__, page_mapping_cache(bh->b_page));
 		}
 
 		if (!new_jh) {

--
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>

[PATCH] add page->mapping handling interface [22/35] changes in JFFS2

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in JFFS2

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/jffs2/file.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/jffs2/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/jffs2/file.c
+++ test-2.6.23-rc4-mm1/fs/jffs2/file.c
@@ -111,11 +111,11 @@ int jffs2_do_readpage_unlock(struct inod
 
 static int jffs2_readpage (struct file *filp, struct page *pg)
 {
-	struct jffs2_inode_info *f = JFFS2_INODE_INFO(pg->mapping->host);
+	struct jffs2_inode_info *f = JFFS2_INODE_INFO(page_inode(pg));
 	int ret;
 
 	down(&f->sem);
-	ret = jffs2_do_readpage_unlock(pg->mapping->host, pg);
+	ret = jffs2_do_readpage_unlock(page_inode(pg), pg);
 	up(&f->sem);
 	return ret;
 }

--
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>

[PATCH] add page->mapping handling interface [23/35] changes in JFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in JFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/jfs/jfs_metapage.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/jfs/jfs_metapage.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/jfs/jfs_metapage.c
+++ test-2.6.23-rc4-mm1/fs/jfs/jfs_metapage.c
@@ -113,7 +113,7 @@ static inline int insert_metapage(struct
 	}
 
 	if (mp) {
-		l2mp_blocks = L2PSIZE - page->mapping->host->i_blkbits;
+		l2mp_blocks = L2PSIZE - page_inode(page)->i_blkbits;
 		index = (mp->index >> l2mp_blocks) & (MPS_PER_PAGE - 1);
 		a->mp_count++;
 		a->mp[index] = mp;
@@ -125,7 +125,7 @@ static inline int insert_metapage(struct
 static inline void remove_metapage(struct page *page, struct metapage *mp)
 {
 	struct meta_anchor *a = mp_anchor(page);
-	int l2mp_blocks = L2PSIZE - page->mapping->host->i_blkbits;
+	int l2mp_blocks = L2PSIZE - page_inode(page)->i_blkbits;
 	int index;
 
 	index = (mp->index >> l2mp_blocks) & (MPS_PER_PAGE - 1);
@@ -364,7 +364,7 @@ static int metapage_writepage(struct pag
 {
 	struct bio *bio = NULL;
 	unsigned int block_offset;	/* block offset of mp within page */
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	unsigned int blocks_per_mp = JFS_SBI(inode->i_sb)->nbperpage;
 	unsigned int len;
 	unsigned int xlen;
@@ -484,7 +484,7 @@ skip:
 
 static int metapage_readpage(struct file *fp, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct bio *bio = NULL;
 	unsigned int block_offset;
 	unsigned int blocks_per_page = PAGE_CACHE_SIZE >> inode->i_blkbits;

--
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>

[PATCH] add page->mapping handling interface [24/35] changes in MINIX FS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in MINIXFS


Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
---
 fs/minix/dir.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/minix/dir.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/minix/dir.c
+++ test-2.6.23-rc4-mm1/fs/minix/dir.c
@@ -52,7 +52,7 @@ static inline unsigned long dir_pages(st
 
 static int dir_commit_chunk(struct page *page, loff_t pos, unsigned len)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *dir = mapping->host;
 	int err = 0;
 	block_write_end(NULL, mapping, pos, len, len, page, NULL);
@@ -281,7 +281,8 @@ int minix_add_link(struct dentry *dentry
 
 got_it:
 	pos = (page->index >> PAGE_CACHE_SHIFT) + p - (char*)page_address(page);
-	err = __minix_write_begin(NULL, page->mapping, pos, sbi->s_dirsize,
+	err = __minix_write_begin(NULL,
+				page_mapping_cache(page), pos, sbi->s_dirsize,
 					AOP_FLAG_UNINTERRUPTIBLE, &page, NULL);
 	if (err)
 		goto out_unlock;
@@ -307,7 +308,7 @@ out_unlock:
 
 int minix_delete_entry(struct minix_dir_entry *de, struct page *page)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *inode = (struct inode*)mapping->host;
 	char *kaddr = page_address(page);
 	loff_t pos = page_offset(page) + (char*)de - kaddr;
@@ -431,7 +432,7 @@ not_empty:
 void minix_set_link(struct minix_dir_entry *de, struct page *page,
 	struct inode *inode)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *dir = mapping->host;
 	struct minix_sb_info *sbi = minix_sb(dir->i_sb);
 	loff_t pos = page_offset(page) +

--
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>

Re: [PATCH] add page->mapping handling interface [22/35] changes in JFFS2

[David Woodhouse]

On Mon, 2007-09-10 at 19:16 +0900, KAMEZAWA Hiroyuki wrote:
> Changes page->mapping handling in JFFS2
> 
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

Looks reasonable to me; I assume it's not intended for me to take it and
apply it yet, before the core parts are merged? I'll let you shepherd it
upstream...

-- 
dwmw2

--
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>

[PATCH] add page->mapping handling interface [25/35] changes in NCPFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in NCPFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/ncpfs/symlink.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/ncpfs/symlink.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ncpfs/symlink.c
+++ test-2.6.23-rc4-mm1/fs/ncpfs/symlink.c
@@ -42,7 +42,7 @@
 
 static int ncp_symlink_readpage(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int error, length, len;
 	char *link, *rawlink;
 	char *buf = kmap(page);

--
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>

[PATCH] add page->mapping handling interface [26/35] changes in NFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in NFS

Singed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/nfs/file.c     |   11 ++++++-----
 fs/nfs/internal.h |    2 +-
 fs/nfs/pagelist.c |    2 +-
 fs/nfs/read.c     |    4 ++--
 fs/nfs/write.c    |   35 ++++++++++++++++++-----------------
 5 files changed, 28 insertions(+), 26 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/nfs/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/nfs/file.c
+++ test-2.6.23-rc4-mm1/fs/nfs/file.c
@@ -357,7 +357,7 @@ static void nfs_invalidate_page(struct p
 	if (offset != 0)
 		return;
 	/* Cancel any unstarted writes on this page */
-	nfs_wb_page_cancel(page->mapping->host, page);
+	nfs_wb_page_cancel(page_inode(page), page);
 }
 
 static int nfs_release_page(struct page *page, gfp_t gfp)
@@ -368,7 +368,7 @@ static int nfs_release_page(struct page 
 
 static int nfs_launder_page(struct page *page)
 {
-	return nfs_wb_page(page->mapping->host, page);
+	return nfs_wb_page(page_inode(page), page);
 }
 
 const struct address_space_operations nfs_file_aops = {
@@ -395,16 +395,17 @@ static int nfs_vm_page_mkwrite(struct vm
 	void *fsdata;
 
 	lock_page(page);
-	if (page->mapping != vma->vm_file->f_path.dentry->d_inode->i_mapping)
+	if (!pagecache_consistent(page,
+			vma->vm_file->f_path.dentry->d_inode->i_mapping))
 		goto out_unlock;
 	pagelen = nfs_page_length(page);
 	if (pagelen == 0)
 		goto out_unlock;
-	ret = nfs_write_begin(filp, page->mapping,
+	ret = nfs_write_begin(filp, page_mapping_cache(page),
 				(loff_t)page->index << PAGE_CACHE_SHIFT,
 				pagelen, 0, &page, &fsdata);
 	if (!ret)
-		ret = nfs_write_end(filp, page->mapping,
+		ret = nfs_write_end(filp, page_mapping_cache(page),
 				(loff_t)page->index << PAGE_CACHE_SHIFT,
 				pagelen, pagelen, page, fsdata);
 out_unlock:
Index: test-2.6.23-rc4-mm1/fs/nfs/internal.h
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/nfs/internal.h
+++ test-2.6.23-rc4-mm1/fs/nfs/internal.h
@@ -220,7 +220,7 @@ void nfs_super_set_maxbytes(struct super
 static inline
 unsigned int nfs_page_length(struct page *page)
 {
-	loff_t i_size = i_size_read(page->mapping->host);
+	loff_t i_size = i_size_read(page_inode(page));
 
 	if (i_size > 0) {
 		pgoff_t end_index = (i_size - 1) >> PAGE_CACHE_SHIFT;
Index: test-2.6.23-rc4-mm1/fs/nfs/read.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/nfs/read.c
+++ test-2.6.23-rc4-mm1/fs/nfs/read.c
@@ -466,7 +466,7 @@ static const struct rpc_call_ops nfs_rea
 int nfs_readpage(struct file *file, struct page *page)
 {
 	struct nfs_open_context *ctx;
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int		error;
 
 	dprintk("NFS: nfs_readpage (%p %ld@%lu)\n",
@@ -517,7 +517,7 @@ static int
 readpage_async_filler(void *data, struct page *page)
 {
 	struct nfs_readdesc *desc = (struct nfs_readdesc *)data;
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct nfs_page *new;
 	unsigned int len;
 	int error;
Index: test-2.6.23-rc4-mm1/fs/nfs/write.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/nfs/write.c
+++ test-2.6.23-rc4-mm1/fs/nfs/write.c
@@ -131,7 +131,7 @@ static struct nfs_page *nfs_page_find_re
 
 static struct nfs_page *nfs_page_find_request(struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct nfs_page *req = NULL;
 
 	spin_lock(&inode->i_lock);
@@ -143,7 +143,7 @@ static struct nfs_page *nfs_page_find_re
 /* Adjust the file length if we're writing beyond the end */
 static void nfs_grow_file(struct page *page, unsigned int offset, unsigned int count)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t end, i_size = i_size_read(inode);
 	pgoff_t end_index = (i_size - 1) >> PAGE_CACHE_SHIFT;
 
@@ -160,7 +160,7 @@ static void nfs_grow_file(struct page *p
 static void nfs_set_pageerror(struct page *page)
 {
 	SetPageError(page);
-	nfs_zap_mapping(page->mapping->host, page->mapping);
+	nfs_zap_mapping(page_inode(page), page_mapping_cache(page));
 }
 
 /* We can set the PG_uptodate flag if we see that a write request
@@ -192,7 +192,7 @@ static int nfs_writepage_setup(struct nf
 		ret = PTR_ERR(req);
 		if (ret != -EBUSY)
 			return ret;
-		ret = nfs_wb_page(page->mapping->host, page);
+		ret = nfs_wb_page(page_inode(page), page);
 		if (ret != 0)
 			return ret;
 	}
@@ -226,7 +226,7 @@ static int nfs_set_page_writeback(struct
 	int ret = test_set_page_writeback(page);
 
 	if (!ret) {
-		struct inode *inode = page->mapping->host;
+		struct inode *inode = page_inode(page);
 		struct nfs_server *nfss = NFS_SERVER(inode);
 
 		if (atomic_long_inc_return(&nfss->writeback) >
@@ -238,7 +238,7 @@ static int nfs_set_page_writeback(struct
 
 static void nfs_end_page_writeback(struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct nfs_server *nfss = NFS_SERVER(inode);
 
 	end_page_writeback(page);
@@ -255,7 +255,7 @@ static void nfs_end_page_writeback(struc
 static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
 				struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_page *req;
 	int ret;
@@ -301,7 +301,7 @@ static int nfs_page_async_flush(struct n
 
 static int nfs_do_writepage(struct page *page, struct writeback_control *wbc, struct nfs_pageio_descriptor *pgio)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 
 	nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGE);
 	nfs_add_stats(inode, NFSIOS_WRITEPAGES, 1);
@@ -318,7 +318,7 @@ static int nfs_writepage_locked(struct p
 	struct nfs_pageio_descriptor pgio;
 	int err;
 
-	nfs_pageio_init_write(&pgio, page->mapping->host, wb_priority(wbc));
+	nfs_pageio_init_write(&pgio, page_inode(page), wb_priority(wbc));
 	err = nfs_do_writepage(page, wbc, &pgio);
 	nfs_pageio_complete(&pgio);
 	if (err < 0)
@@ -585,7 +585,7 @@ static inline int nfs_scan_commit(struct
 static struct nfs_page * nfs_update_request(struct nfs_open_context* ctx,
 		struct page *page, unsigned int offset, unsigned int bytes)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *inode = mapping->host;
 	struct nfs_page		*req, *new = NULL;
 	pgoff_t		rqend, end;
@@ -687,7 +687,7 @@ int nfs_flush_incompatible(struct file *
 		nfs_release_request(req);
 		if (!do_flush)
 			return 0;
-		status = nfs_wb_page(page->mapping->host, page);
+		status = nfs_wb_page(page_inode(page), page);
 	} while (status == 0);
 	return status;
 }
@@ -702,7 +702,7 @@ int nfs_updatepage(struct file *file, st
 		unsigned int offset, unsigned int count)
 {
 	struct nfs_open_context *ctx = nfs_file_open_context(file);
-	struct inode	*inode = page->mapping->host;
+	struct inode	*inode = page_inode(page);
 	int		status = 0;
 
 	nfs_inc_stats(inode, NFSIOS_VFSUPDATEPAGE);
@@ -958,7 +958,7 @@ static void nfs_writeback_done_partial(s
 	}
 
 	if (nfs_write_need_commit(data)) {
-		struct inode *inode = page->mapping->host;
+		struct inode *inode = page_inode(page);
 
 		spin_lock(&inode->i_lock);
 		if (test_bit(PG_NEED_RESCHED, &req->wb_flags)) {
@@ -1386,7 +1386,7 @@ int nfs_wb_page_cancel(struct inode *ino
 	loff_t range_start = page_offset(page);
 	loff_t range_end = range_start + (loff_t)(PAGE_CACHE_SIZE - 1);
 	struct writeback_control wbc = {
-		.bdi = page->mapping->backing_dev_info,
+		.bdi = page_mapping_cache(page)->backing_dev_info,
 		.sync_mode = WB_SYNC_ALL,
 		.nr_to_write = LONG_MAX,
 		.range_start = range_start,
@@ -1419,7 +1419,8 @@ int nfs_wb_page_cancel(struct inode *ino
 	}
 	if (!PagePrivate(page))
 		return 0;
-	ret = nfs_sync_mapping_wait(page->mapping, &wbc, FLUSH_INVALIDATE);
+	ret = nfs_sync_mapping_wait(page_mapping_cache(page),
+					&wbc, FLUSH_INVALIDATE);
 out:
 	return ret;
 }
@@ -1429,7 +1430,7 @@ int nfs_wb_page_priority(struct inode *i
 	loff_t range_start = page_offset(page);
 	loff_t range_end = range_start + (loff_t)(PAGE_CACHE_SIZE - 1);
 	struct writeback_control wbc = {
-		.bdi = page->mapping->backing_dev_info,
+		.bdi = page_mapping_cache(page)->backing_dev_info,
 		.sync_mode = WB_SYNC_ALL,
 		.nr_to_write = LONG_MAX,
 		.range_start = range_start,
@@ -1445,7 +1446,7 @@ int nfs_wb_page_priority(struct inode *i
 	}
 	if (!PagePrivate(page))
 		return 0;
-	ret = nfs_sync_mapping_wait(page->mapping, &wbc, how);
+	ret = nfs_sync_mapping_wait(page_mapping_cache(page), &wbc, how);
 	if (ret >= 0)
 		return 0;
 out:
Index: test-2.6.23-rc4-mm1/fs/nfs/pagelist.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/nfs/pagelist.c
+++ test-2.6.23-rc4-mm1/fs/nfs/pagelist.c
@@ -81,7 +81,7 @@ nfs_create_request(struct nfs_open_conte
 	page_cache_get(page);
 	BUG_ON(PagePrivate(page));
 	BUG_ON(!PageLocked(page));
-	BUG_ON(page->mapping->host != inode);
+	BUG_ON(page_inode(page) != inode);
 	req->wb_offset  = offset;
 	req->wb_pgbase	= offset;
 	req->wb_bytes   = count;

--
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>

[PATCH] add page->mapping handling interface [27/35] changes in NTFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in NTFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/ntfs/aops.c     |   14 +++++++-------
 fs/ntfs/compress.c |    2 +-
 fs/ntfs/file.c     |    6 +++---
 3 files changed, 11 insertions(+), 11 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/ntfs/aops.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ntfs/aops.c
+++ test-2.6.23-rc4-mm1/fs/ntfs/aops.c
@@ -65,7 +65,7 @@ static void ntfs_end_buffer_async_read(s
 	int page_uptodate = 1;
 
 	page = bh->b_page;
-	vi = page->mapping->host;
+	vi = page_inode(page);
 	ni = NTFS_I(vi);
 
 	if (likely(uptodate)) {
@@ -194,7 +194,7 @@ static int ntfs_read_block(struct page *
 	int i, nr;
 	unsigned char blocksize_bits;
 
-	vi = page->mapping->host;
+	vi = page_inode(page);
 	ni = NTFS_I(vi);
 	vol = ni->vol;
 
@@ -413,7 +413,7 @@ retry_readpage:
 		unlock_page(page);
 		return 0;
 	}
-	vi = page->mapping->host;
+	vi = page_inode(page);
 	ni = NTFS_I(vi);
 	/*
 	 * Only $DATA attributes can be encrypted and only unnamed $DATA
@@ -553,7 +553,7 @@ static int ntfs_write_block(struct page 
 	bool need_end_writeback;
 	unsigned char blocksize_bits;
 
-	vi = page->mapping->host;
+	vi = page_inode(page);
 	ni = NTFS_I(vi);
 	vol = ni->vol;
 
@@ -909,7 +909,7 @@ static int ntfs_write_mst_block(struct p
 		struct writeback_control *wbc)
 {
 	sector_t block, dblock, rec_block;
-	struct inode *vi = page->mapping->host;
+	struct inode *vi = page_inode(page);
 	ntfs_inode *ni = NTFS_I(vi);
 	ntfs_volume *vol = ni->vol;
 	u8 *kaddr;
@@ -1342,7 +1342,7 @@ done:
 static int ntfs_writepage(struct page *page, struct writeback_control *wbc)
 {
 	loff_t i_size;
-	struct inode *vi = page->mapping->host;
+	struct inode *vi = page_inode(page);
 	ntfs_inode *base_ni = NULL, *ni = NTFS_I(vi);
 	char *kaddr;
 	ntfs_attr_search_ctx *ctx = NULL;
@@ -1579,7 +1579,7 @@ const struct address_space_operations nt
  * need the lock since try_to_free_buffers() does not free dirty buffers.
  */
 void mark_ntfs_record_dirty(struct page *page, const unsigned int ofs) {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	ntfs_inode *ni = NTFS_I(mapping->host);
 	struct buffer_head *bh, *head, *buffers_to_free = NULL;
 	unsigned int end, bh_size, bh_ofs;
Index: test-2.6.23-rc4-mm1/fs/ntfs/compress.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ntfs/compress.c
+++ test-2.6.23-rc4-mm1/fs/ntfs/compress.c
@@ -482,7 +482,7 @@ int ntfs_read_compressed_block(struct pa
 {
 	loff_t i_size;
 	s64 initialized_size;
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	ntfs_inode *ni = NTFS_I(mapping->host);
 	ntfs_volume *vol = ni->vol;
 	struct super_block *sb = vol->sb;
Index: test-2.6.23-rc4-mm1/fs/ntfs/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ntfs/file.c
+++ test-2.6.23-rc4-mm1/fs/ntfs/file.c
@@ -520,7 +520,7 @@ static int ntfs_prepare_pages_for_non_re
 	BUG_ON(!nr_pages);
 	BUG_ON(!pages);
 	BUG_ON(!*pages);
-	vi = pages[0]->mapping->host;
+	vi = page_inode(pages[0]);
 	ni = NTFS_I(vi);
 	vol = ni->vol;
 	ntfs_debug("Entering for inode 0x%lx, attribute type 0x%x, start page "
@@ -1494,7 +1494,7 @@ static inline int ntfs_commit_pages_afte
 	unsigned blocksize, u;
 	int err;
 
-	vi = pages[0]->mapping->host;
+	vi = page_inode(pages[0]);
 	ni = NTFS_I(vi);
 	blocksize = vi->i_sb->s_blocksize;
 	end = pos + bytes;
@@ -1654,7 +1654,7 @@ static int ntfs_commit_pages_after_write
 	BUG_ON(!pages);
 	page = pages[0];
 	BUG_ON(!page);
-	vi = page->mapping->host;
+	vi = page_inode(page);
 	ni = NTFS_I(vi);
 	ntfs_debug("Entering for inode 0x%lx, attribute type 0x%x, start page "
 			"index 0x%lx, nr_pages 0x%x, pos 0x%llx, bytes 0x%zx.",

--
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>

[PATCH] add page->mapping handling interface [28/35] changes in OCFS2

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in OCFS2

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/ocfs2/aops.c |    8 ++++----
 fs/ocfs2/mmap.c |    3 ++-
 2 files changed, 6 insertions(+), 5 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/ocfs2/aops.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ocfs2/aops.c
+++ test-2.6.23-rc4-mm1/fs/ocfs2/aops.c
@@ -208,7 +208,7 @@ bail:
 
 static int ocfs2_readpage(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t start = (loff_t)page->index << PAGE_CACHE_SHIFT;
 	int ret, unlock = 1;
 
@@ -540,14 +540,14 @@ static void ocfs2_dio_end_io(struct kioc
  */
 static void ocfs2_invalidatepage(struct page *page, unsigned long offset)
 {
-	journal_t *journal = OCFS2_SB(page->mapping->host->i_sb)->journal->j_journal;
+	journal_t *journal = OCFS2_SB(page_inode(page)->i_sb)->journal->j_journal;
 
 	journal_invalidatepage(journal, page, offset);
 }
 
 static int ocfs2_releasepage(struct page *page, gfp_t wait)
 {
-	journal_t *journal = OCFS2_SB(page->mapping->host->i_sb)->journal->j_journal;
+	journal_t *journal = OCFS2_SB(page_inode(page)->i_sb)->journal->j_journal;
 
 	if (!page_has_buffers(page))
 		return 0;
@@ -1065,7 +1065,7 @@ static int ocfs2_grab_pages_for_write(st
 			 */
 			lock_page(mmap_page);
 
-			if (mmap_page->mapping != mapping) {
+			if (!pagecache_consistent(mmap_page, mapping)) {
 				unlock_page(mmap_page);
 				/*
 				 * Sanity check - the locking in
Index: test-2.6.23-rc4-mm1/fs/ocfs2/mmap.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ocfs2/mmap.c
+++ test-2.6.23-rc4-mm1/fs/ocfs2/mmap.c
@@ -112,7 +112,8 @@ static int __ocfs2_page_mkwrite(struct i
 	 * page mapping after taking the page lock inside of
 	 * ocfs2_write_begin_nolock().
 	 */
-	if (!PageUptodate(page) || page->mapping != inode->i_mapping) {
+	if (!PageUptodate(page) ||
+	    !pagecache_consistent(page, inode->i_mapping)) {
 		ret = -EINVAL;
 		goto out;
 	}

--
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>

[PATCH] add page->mapping handling interface [29/35] changes in REISER4/REISERFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in reiser4
(sorry, changes for /reiserfs is also included.)

Todo: 
Fix this warning caused by this patch(set). does anyone have an adivce ?
fs/reiser4/page_cache.c: In function ‘reiser4_tree_by_page’:
fs/reiser4/page_cache.c:315: warning: passing argument 1 of ‘page_inode’ discards qualifiers from pointer target type

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/reiser4/as_ops.c                      |   34 ++++++++++++++++---------------
 fs/reiser4/entd.c                        |    6 ++---
 fs/reiser4/jnode.c                       |   18 +++++++++-------
 fs/reiser4/page_cache.c                  |   10 ++++-----
 fs/reiser4/plugin/cluster.h              |   14 ++++++------
 fs/reiser4/plugin/file/cryptcompress.c   |   15 +++++++------
 fs/reiser4/plugin/file/file.c            |   22 ++++++++++----------
 fs/reiser4/plugin/file_ops.c             |    8 +++----
 fs/reiser4/plugin/item/ctail.c           |   16 +++++++-------
 fs/reiser4/plugin/item/extent_file_ops.c |    6 ++---
 fs/reiser4/plugin/item/tail.c            |    6 ++---
 fs/reiser4/wander.c                      |    2 -
 fs/reiserfs/inode.c                      |   16 +++++++-------
 fs/reiserfs/journal.c                    |    3 +-
 fs/reiserfs/tail_conversion.c            |    2 -
 15 files changed, 92 insertions(+), 86 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/reiser4/as_ops.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/as_ops.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/as_ops.c
@@ -64,24 +64,25 @@
 int reiser4_set_page_dirty(struct page *page)
 {
 	/* this page can be unformatted only */
-	assert("vs-1734", (page->mapping &&
-			   page->mapping->host &&
-			   reiser4_get_super_fake(page->mapping->host->i_sb) !=
-			   page->mapping->host
-			   && reiser4_get_cc_fake(page->mapping->host->i_sb) !=
-			   page->mapping->host
-			   && reiser4_get_bitmap_fake(page->mapping->host->i_sb) !=
-			   page->mapping->host));
+	assert("vs-1734", (page_mapping_cache(page) &&
+			   page_inode(page) &&
+			   reiser4_get_super_fake(page_inode(page)->i_sb) !=
+			   page_inode(page)
+			   && reiser4_get_cc_fake(page_inode(page)->i_sb) !=
+			   page_inode(page)
+			   && reiser4_get_bitmap_fake(page_inode(page)->i_sb) !=
+			   page_inode(page)));
 
 	if (!TestSetPageDirty(page)) {
-		struct address_space *mapping = page->mapping;
+		struct address_space *mapping = page_mapping_cache(page);
 
 		if (mapping) {
 			write_lock_irq(&mapping->tree_lock);
 
 			/* check for race with truncate */
-			if (page->mapping) {
-				assert("vs-1652", page->mapping == mapping);
+			if (page_is_pagecache(page)) {
+				assert("vs-1652",
+					page_mapping_cache(page) == mapping);
 				if (mapping_cap_account_dirty(mapping))
 					inc_zone_page_state(page,
 							NR_FILE_DIRTY);
@@ -140,7 +141,7 @@ void reiser4_invalidatepage(struct page 
 
 	assert("nikita-3137", PageLocked(page));
 	assert("nikita-3138", !PageWriteback(page));
-	inode = page->mapping->host;
+	inode = page_inode(page);
 
 	/*
 	 * ->invalidatepage() should only be called for the unformatted
@@ -157,7 +158,8 @@ void reiser4_invalidatepage(struct page 
 		return;
 	assert("vs-1426", PagePrivate(page));
 	assert("vs-1427",
-	       page->mapping == jnode_get_mapping(jnode_by_page(page)));
+	       pagecache_consistent(page,
+			jnode_get_mapping(jnode_by_page(page))));
 	assert("", jprivate(page) != NULL);
 	assert("", ergo(inode_file_plugin(inode) !=
 			file_plugin_by_id(CRYPTCOMPRESS_FILE_PLUGIN_ID),
@@ -287,8 +289,8 @@ int reiser4_releasepage(struct page *pag
 
 	node = jnode_by_page(page);
 	assert("nikita-2258", node != NULL);
-	assert("reiser4-4", page->mapping != NULL);
-	assert("reiser4-5", page->mapping->host != NULL);
+	assert("reiser4-4", page_mapping_cache(page) != NULL);
+	assert("reiser4-5", page_inode(page) != NULL);
 
 	if (PageDirty(page))
 		return 0;
@@ -305,7 +307,7 @@ int reiser4_releasepage(struct page *pag
 	if (jnode_is_releasable(node)) {
 		struct address_space *mapping;
 
-		mapping = page->mapping;
+		mapping = page_mapping_cache(page);
 		jref(node);
 		/* there is no need to synchronize against
 		 * jnode_extent_write() here, because pages seen by
Index: test-2.6.23-rc4-mm1/fs/reiser4/entd.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/entd.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/entd.c
@@ -266,9 +266,9 @@ int write_page_by_ent(struct page *page,
 	struct wbq rq;
 
 	assert("", PageLocked(page));
-	assert("", page->mapping != NULL);
+	assert("", page_mapping_cache(page) != NULL);
 
-	sb = page->mapping->host->i_sb;
+	sb = page_inode(page)->i_sb;
 	ent = get_entd_context(sb);
 	assert("", ent && ent->done == 0);
 
@@ -283,7 +283,7 @@ int write_page_by_ent(struct page *page,
 	 * pin inode in memory, unlock page, entd_flush will iput. We can not
 	 * iput here becasue we can not allow delete_inode to be called here
 	 */
-	inode = igrab(page->mapping->host);
+	inode = igrab(page_inode(page));
 	unlock_page(page);
 	if (inode == NULL)
 		/* inode is getting freed */
Index: test-2.6.23-rc4-mm1/fs/reiser4/jnode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/jnode.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/jnode.c
@@ -594,7 +594,7 @@ static jnode *do_jget(reiser4_tree * tre
 	 */
 
 	jnode *result;
-	oid_t oid = get_inode_oid(pg->mapping->host);
+	oid_t oid = get_inode_oid(page_inode(pg));
 
 	assert("umka-176", pg != NULL);
 	assert("nikita-2394", PageLocked(pg));
@@ -606,18 +606,18 @@ static jnode *do_jget(reiser4_tree * tre
 	tree = reiser4_tree_by_page(pg);
 
 	/* check hash-table first */
-	result = jfind(pg->mapping, pg->index);
+	result = jfind(page_mapping_cache(pg), pg->index);
 	if (unlikely(result != NULL)) {
 		spin_lock_jnode(result);
 		jnode_attach_page(result, pg);
 		spin_unlock_jnode(result);
-		result->key.j.mapping = pg->mapping;
+		result->key.j.mapping = page_mapping_cache(pg);
 		return result;
 	}
 
 	/* since page is locked, jnode should be allocated with GFP_NOFS flag */
 	reiser4_ctx_gfp_mask_force(GFP_NOFS);
-	result = find_get_jnode(tree, pg->mapping, oid, pg->index);
+	result = find_get_jnode(tree, page_mapping_cache(pg), oid, pg->index);
 	if (unlikely(IS_ERR(result)))
 		return result;
 	/* attach jnode to page */
@@ -646,13 +646,14 @@ jnode *jnode_of_page(struct page * pg)
 			assert("nikita-2364",
 			       jprivate(pg)->key.j.index == pg->index);
 			assert("nikita-2367",
-			       jprivate(pg)->key.j.mapping == pg->mapping);
+			       jprivate(pg)->key.j.mapping ==
+				page_mapping_cache(pg));
 			assert("nikita-2365",
 			       jprivate(pg)->key.j.objectid ==
-			       get_inode_oid(pg->mapping->host));
+			       get_inode_oid(page_inode(pg)));
 			assert("vs-1200",
 			       jprivate(pg)->key.j.objectid ==
-			       pg->mapping->host->i_ino);
+			       page_inode(pg)->i_ino);
 			assert("nikita-2356",
 			       jnode_is_unformatted(jnode_by_page(pg)));
 		}
@@ -812,7 +813,8 @@ static struct page *jnode_get_page_locke
 		page_cache_get(page);
 		spin_unlock_jnode(node);
 		lock_page(page);
-		assert("nikita-3134", page->mapping == jnode_get_mapping(node));
+		assert("nikita-3134",
+			page_mapping_cache(page) == jnode_get_mapping(node));
 	}
 
 	spin_lock_jnode(node);
Index: test-2.6.23-rc4-mm1/fs/reiser4/page_cache.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/page_cache.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/page_cache.c
@@ -312,7 +312,7 @@ void reiser4_wait_page_writeback(struct 
 reiser4_tree *reiser4_tree_by_page(const struct page *page /* page to query */ )
 {
 	assert("nikita-2461", page != NULL);
-	return &get_super_private(page->mapping->host->i_sb)->tree;
+	return &get_super_private(page_inode(page)->i_sb)->tree;
 }
 
 /* completion handler for single page bio-based read.
@@ -400,7 +400,7 @@ int reiser4_page_io(struct page *page, j
 	assert("nikita-2893", rw == READ || rw == WRITE);
 
 	if (rw) {
-		if (unlikely(page->mapping->host->i_sb->s_flags & MS_RDONLY)) {
+		if (unlikely(page_inode(page)->i_sb->s_flags & MS_RDONLY)) {
 			unlock_page(page);
 			return 0;
 		}
@@ -441,7 +441,7 @@ static struct bio *page_bio(struct page 
 		struct super_block *super;
 		reiser4_block_nr blocknr;
 
-		super = page->mapping->host->i_sb;
+		super = page_inode(page)->i_sb;
 		assert("nikita-2029", super != NULL);
 		blksz = super->s_blocksize;
 		assert("nikita-2028", blksz == (int)PAGE_CACHE_SIZE);
@@ -479,7 +479,7 @@ int reiser4_set_page_dirty_internal(stru
 {
 	struct address_space *mapping;
 
-	mapping = page->mapping;
+	mapping = page_mapping_cache(page);
 	BUG_ON(mapping == NULL);
 
 	if (!TestSetPageDirty(page)) {
@@ -528,7 +528,7 @@ int reiser4_writepage(struct page *page,
 
 	assert("vs-828", PageLocked(page));
 
-	s = page->mapping->host->i_sb;
+	s = page_inode(page)->i_sb;
 	ctx = get_current_context_check();
 
 	//assert("", can_hit_entd(ctx, s));
Index: test-2.6.23-rc4-mm1/fs/reiser4/wander.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/wander.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/wander.c
@@ -765,7 +765,7 @@ static int write_jnodes_to_disk_extent(
 
 			spin_lock_jnode(cur);
 			assert("nikita-3166",
-			       pg->mapping == jnode_get_mapping(cur));
+			      page_mapping_cache(pg) == jnode_get_mapping(cur));
 			assert("zam-912", !JF_ISSET(cur, JNODE_WRITEBACK));
 #if REISER4_DEBUG
 			spin_lock(&cur->load);
Index: test-2.6.23-rc4-mm1/fs/reiserfs/inode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiserfs/inode.c
+++ test-2.6.23-rc4-mm1/fs/reiserfs/inode.c
@@ -2331,7 +2331,7 @@ static int map_block_for_writepage(struc
 static int reiserfs_write_full_page(struct page *page,
 				    struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	unsigned long end_index = inode->i_size >> PAGE_CACHE_SHIFT;
 	int error = 0;
 	unsigned long block;
@@ -2546,7 +2546,7 @@ static int reiserfs_readpage(struct file
 
 static int reiserfs_writepage(struct page *page, struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	reiserfs_wait_on_write_block(inode->i_sb);
 	return reiserfs_write_full_page(page, wbc);
 }
@@ -2624,7 +2624,7 @@ static int reiserfs_write_begin(struct f
 int reiserfs_prepare_write(struct file *f, struct page *page,
 			   unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int ret;
 	int old_ref = 0;
 
@@ -2679,7 +2679,7 @@ static int reiserfs_write_end(struct fil
 			      loff_t pos, unsigned len, unsigned copied,
 			      struct page *page, void *fsdata)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	int ret = 0;
 	int update_sd = 0;
 	struct reiserfs_transaction_handle *th;
@@ -2772,7 +2772,7 @@ static int reiserfs_write_end(struct fil
 int reiserfs_commit_write(struct file *f, struct page *page,
 			  unsigned from, unsigned to)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t pos = ((loff_t) page->index << PAGE_CACHE_SHIFT) + to;
 	int ret = 0;
 	int update_sd = 0;
@@ -2951,7 +2951,7 @@ static int invalidatepage_can_drop(struc
 static void reiserfs_invalidatepage(struct page *page, unsigned long offset)
 {
 	struct buffer_head *head, *bh, *next;
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	unsigned int curr_off = 0;
 	int ret = 1;
 
@@ -2997,7 +2997,7 @@ static void reiserfs_invalidatepage(stru
 
 static int reiserfs_set_page_dirty(struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	if (reiserfs_file_data_log(inode)) {
 		SetPageChecked(page);
 		return __set_page_dirty_nobuffers(page);
@@ -3016,7 +3016,7 @@ static int reiserfs_set_page_dirty(struc
  */
 static int reiserfs_releasepage(struct page *page, gfp_t unused_gfp_flags)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct reiserfs_journal *j = SB_JOURNAL(inode->i_sb);
 	struct buffer_head *head;
 	struct buffer_head *bh;
Index: test-2.6.23-rc4-mm1/fs/reiserfs/journal.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiserfs/journal.c
+++ test-2.6.23-rc4-mm1/fs/reiserfs/journal.c
@@ -889,7 +889,8 @@ static int write_ordered_buffers(spinloc
 		 * the buffer. We're safe if we write the page one last time
 		 * after freeing the journal header.
 		 */
-		if (buffer_dirty(bh) && unlikely(bh->b_page->mapping == NULL)) {
+		if (buffer_dirty(bh) &&
+		    unlikely(!page_mapping_cache(bh->b_page))) {
 			spin_unlock(lock);
 			ll_rw_block(WRITE, 1, &bh);
 			spin_lock(lock);
Index: test-2.6.23-rc4-mm1/fs/reiserfs/tail_conversion.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiserfs/tail_conversion.c
+++ test-2.6.23-rc4-mm1/fs/reiserfs/tail_conversion.c
@@ -151,7 +151,7 @@ void reiserfs_unmap_buffer(struct buffer
 	   interested in removing it from per-sb j_dirty_buffers list, to avoid
 	   BUG() on attempt to write not mapped buffer */
 	if ((!list_empty(&bh->b_assoc_buffers) || bh->b_private) && bh->b_page) {
-		struct inode *inode = bh->b_page->mapping->host;
+		struct inode *inode = page_inode(bh->b_page);
 		struct reiserfs_journal *j = SB_JOURNAL(inode->i_sb);
 		spin_lock(&j->j_dirty_buffers_lock);
 		list_del_init(&bh->b_assoc_buffers);
Index: test-2.6.23-rc4-mm1/fs/reiser4/plugin/cluster.h
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/plugin/cluster.h
+++ test-2.6.23-rc4-mm1/fs/reiser4/plugin/cluster.h
@@ -92,9 +92,9 @@ static inline unsigned off_to_cloff(loff
 static inline  pgoff_t offset_in_clust(struct page * page)
 {
 	assert("edward-1488", page != NULL);
-	assert("edward-1489", page->mapping != NULL);
+	assert("edward-1489", page_mapping_cache(page) != NULL);
 
-	return page_index(page) & ((cluster_nrpages(page->mapping->host)) - 1);
+	return page_index(page) & ((cluster_nrpages(page_inode(page))) - 1);
 }
 
 static inline int first_page_in_cluster(struct page * page)
@@ -105,7 +105,7 @@ static inline int first_page_in_cluster(
 static inline int last_page_in_cluster(struct page * page)
 {
 	return offset_in_clust(page) ==
-		cluster_nrpages(page->mapping->host) - 1;
+		cluster_nrpages(page_inode(page)) - 1;
 }
 
 static inline unsigned
@@ -200,11 +200,11 @@ static inline int same_page_cluster(stru
 {
 	assert("edward-1490", p1 != NULL);
 	assert("edward-1491", p2 != NULL);
-	assert("edward-1492", p1->mapping != NULL);
-	assert("edward-1493", p2->mapping != NULL);
+	assert("edward-1492", page_is_pagecache(p1));
+	assert("edward-1493", page_is_pagecache(p2));
 
-	return (pg_to_clust(page_index(p1), p1->mapping->host) ==
-		pg_to_clust(page_index(p2), p2->mapping->host));
+	return (pg_to_clust(page_index(p1), page_inode(p1)) ==
+		pg_to_clust(page_index(p2), page_inode(p2)));
 }
 
 static inline int cluster_is_complete(struct cluster_handle * clust,
Index: test-2.6.23-rc4-mm1/fs/reiser4/plugin/file/cryptcompress.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/plugin/file/cryptcompress.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/plugin/file/cryptcompress.c
@@ -1233,16 +1233,16 @@ int readpage_cryptcompress(struct file *
 
 	assert("edward-88", PageLocked(page));
 	assert("vs-976", !PageUptodate(page));
-	assert("edward-89", page->mapping && page->mapping->host);
+	assert("edward-89", page_mapping_cache(page) && page_inode(page));
 
-	ctx = reiser4_init_context(page->mapping->host->i_sb);
+	ctx = reiser4_init_context(page_inode(page)->i_sb);
 	if (IS_ERR(ctx)) {
 		unlock_page(page);
 		return PTR_ERR(ctx);
 	}
 	assert("edward-113",
 	       ergo(file != NULL,
-		    page->mapping == file->f_dentry->d_inode->i_mapping));
+		    pagecache_consistent(page, file->f_dentry->d_inode->i_mapping)));
 
 	if (PageUptodate(page)) {
 		warning("edward-1338", "page is already uptodate\n");
@@ -1873,7 +1873,8 @@ static void checkout_page_cluster(struct
 			assert("edward-1480",
 			       i_size_read(inode) <= page_offset(clust->pages[i]));
 			assert("edward-1481",
-			       clust->pages[i]->mapping != inode->i_mapping);
+			       !pagecache_consistent(clust->pages[i],
+						 	inode->i_mapping));
 			unlock_page(clust->pages[i]);
 			break;
 		}
@@ -2651,13 +2652,13 @@ int set_cluster_by_page(struct cluster_h
 
 	assert("edward-1358", clust != NULL);
 	assert("edward-1359", page != NULL);
-	assert("edward-1360", page->mapping != NULL);
-	assert("edward-1361", page->mapping->host != NULL);
+	assert("edward-1360", page_mapping_cache(page) != NULL);
+	assert("edward-1361", page_inode(page) != NULL);
 
 	setting_actor =
 		(clust->pages ? reset_cluster_pgset : alloc_cluster_pgset);
 	result = setting_actor(clust, count);
-	clust->index = pg_to_clust(page->index, page->mapping->host);
+	clust->index = pg_to_clust(page->index, page_inode(page));
 	return result;
 }
 
Index: test-2.6.23-rc4-mm1/fs/reiser4/plugin/file/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/plugin/file/file.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/plugin/file/file.c
@@ -790,8 +790,8 @@ int find_or_create_extent(struct page *p
 
 	jnode *node;
 
-	assert("vs-1065", page->mapping && page->mapping->host);
-	inode = page->mapping->host;
+	assert("vs-1065", page_mapping_cache(page) && page_inode(page));
+	inode = page_inode(page);
 
 	lock_page(page);
 	node = jnode_of_page(page);
@@ -866,8 +866,8 @@ static int capture_page_and_create_exten
 	int result;
 	struct inode *inode;
 
-	assert("vs-1084", page->mapping && page->mapping->host);
-	inode = page->mapping->host;
+	assert("vs-1084", page_mapping_cache(page) && page_inode(page));
+	inode = page_inode(page);
 	assert("vs-1139",
 	       unix_file_inode_data(inode)->container == UF_CONTAINER_EXTENTS);
 	/* page belongs to file */
@@ -905,8 +905,8 @@ commit_write_unix_file(struct file *file
 
 	SetPageUptodate(page);
 
-	inode = page->mapping->host;
-	ctx = reiser4_init_context(page->mapping->host->i_sb);
+	inode = page_inode(page);
+	ctx = reiser4_init_context(page_inode(page)->i_sb);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
 	page_cache_get(page);
@@ -1433,9 +1433,9 @@ int readpage_unix_file(struct file *file
 
 	assert("vs-1062", PageLocked(page));
 	assert("vs-976", !PageUptodate(page));
-	assert("vs-1061", page->mapping && page->mapping->host);
+	assert("vs-1061", page_mapping_cache(page) && page_inode(page));
 
-	if (page->mapping->host->i_size <= page_offset(page)) {
+	if (page_inode(page)->i_size <= page_offset(page)) {
 		/* page is out of file */
 		zero_user_page(page, 0, PAGE_CACHE_SIZE, KM_USER0);
 		SetPageUptodate(page);
@@ -1443,7 +1443,7 @@ int readpage_unix_file(struct file *file
 		return 0;
 	}
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	ctx = reiser4_init_context(inode->i_sb);
 	if (IS_ERR(ctx)) {
 		unlock_page(page);
@@ -1476,7 +1476,7 @@ int readpage_unix_file(struct file *file
 	lock_page(page);
 	page_cache_release(page);
 
-	if (page->mapping == NULL) {
+	if (!page_is_pagecache(page)) {
 		/*
 		 * readpage allows truncate to run concurrently. Page was
 		 * truncated while it was not locked
@@ -1604,7 +1604,7 @@ static int uf_readpages_filler(void * da
 	reiser4_extent *ext;
 	__u64 ext_index;
 	int cbk_done = 0;
-	struct address_space * mapping = page->mapping;
+	struct address_space * mapping = page_mapping_cache(page);
 
 	if (PageUptodate(page)) {
 		unlock_page(page);
Index: test-2.6.23-rc4-mm1/fs/reiser4/plugin/file_ops.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/plugin/file_ops.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/plugin/file_ops.c
@@ -93,7 +93,7 @@ prepare_write_common(struct file *file, 
 	reiser4_context *ctx;
 	int result;
 
-	ctx = reiser4_init_context(page->mapping->host->i_sb);
+	ctx = reiser4_init_context(page_inode(page)->i_sb);
 	result = do_prepare_write(file, page, from, to);
 
 	/* don't commit transaction under inode semaphore */
@@ -120,13 +120,13 @@ do_prepare_write(struct file *file, stru
 	if (to - from == PAGE_CACHE_SIZE || PageUptodate(page))
 		return 0;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	fplug = inode_file_plugin(inode);
 
-	if (page->mapping->a_ops->readpage == NULL)
+	if (page_mapping_cache(page)->a_ops->readpage == NULL)
 		return RETERR(-EINVAL);
 
-	result = page->mapping->a_ops->readpage(file, page);
+	result = page_mapping_cache(page)->a_ops->readpage(file, page);
 	if (result != 0) {
 		SetPageError(page);
 		ClearPageUptodate(page);
Index: test-2.6.23-rc4-mm1/fs/reiser4/plugin/item/ctail.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/plugin/item/ctail.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/plugin/item/ctail.c
@@ -580,7 +580,7 @@ static int ctail_read_disk_cluster(struc
 	 */
 	assert("edward-1528", znode_is_any_locked(clust->hint->lh.node));
 
-	if (page->mapping != inode->i_mapping) {
+	if (pagecache_consistent(page, inode->i_mapping)) {
 		/* page was truncated */
 		reiser4_unset_hint(clust->hint);
 		reset_cluster_params(clust);
@@ -632,7 +632,7 @@ int do_readpage_ctail(struct inode * ino
 
 	assert("edward-212", PageLocked(page));
 
-	if (unlikely(page->mapping != inode->i_mapping))
+	if (unlikely(pagecache_consistent(page, inode->i_mapping)))
 		return AOP_TRUNCATED_PAGE;
 	if (PageUptodate(page))
 		goto exit;
@@ -713,7 +713,7 @@ int readpage_ctail(void *vp, struct page
 	assert("edward-114", clust != NULL);
 	assert("edward-115", PageLocked(page));
 	assert("edward-116", !PageUptodate(page));
-	assert("edward-118", page->mapping && page->mapping->host);
+	assert("edward-118", page_mapping_cache(page) && page_inode(page));
 	assert("edward-867", !tfm_cluster_is_uptodate(&clust->tc));
 
 	hint = kmalloc(sizeof(*hint), reiser4_ctx_gfp_mask_get());
@@ -730,7 +730,7 @@ int readpage_ctail(void *vp, struct page
 	}
 	assert("vs-25", hint->ext_coord.lh == &hint->lh);
 
-	result = do_readpage_ctail(page->mapping->host, clust, page,
+	result = do_readpage_ctail(page_inode(page), clust, page,
 				   ZNODE_READ_LOCK);
 	assert("edward-213", PageLocked(page));
 	assert("edward-1163", ergo(!result, PageUptodate(page)));
@@ -781,7 +781,7 @@ static int ctail_readpages_filler(void *
 	struct cluster_handle * clust = data;
 	struct inode * inode = clust->file->f_dentry->d_inode;
 
-	assert("edward-1525", page->mapping == inode->i_mapping);
+	assert("edward-1525", pagecache_consistent(page, inode->i_mapping));
 
 	if (PageUptodate(page)) {
 		unlock_page(page);
@@ -1110,7 +1110,7 @@ int scan_ctail(flush_scan * scan)
 	assert("edward-639", znode_is_write_locked(scan->parent_lock.node));
 
 	page = jnode_page(node);
-	inode = page->mapping->host;
+	inode = page_inode(page);
 
 	if (!reiser4_scanning_left(scan))
 		return result;
@@ -1516,9 +1516,9 @@ int convert_ctail(flush_pos_t * pos)
 			assert("edward-264", pos->child != NULL);
 			assert("edward-265", jnode_page(pos->child) != NULL);
 			assert("edward-266",
-			       jnode_page(pos->child)->mapping != NULL);
+			    page_mapping_cache(jnode_page(pos->child)) != NULL);
 
-			inode = jnode_page(pos->child)->mapping->host;
+			inode = page_inode(jnode_page(pos->child));
 
 			assert("edward-267", inode != NULL);
 
Index: test-2.6.23-rc4-mm1/fs/reiser4/plugin/item/extent_file_ops.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/plugin/item/extent_file_ops.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/plugin/item/extent_file_ops.c
@@ -1124,7 +1124,7 @@ int reiser4_do_readpage_extent(reiser4_e
 	oid_t oid;
 	reiser4_block_nr block;
 
-	mapping = page->mapping;
+	mapping = page_mapping_cache(page);
 	oid = get_inode_oid(mapping->host);
 	index = page->index;
 
@@ -1324,14 +1324,14 @@ int reiser4_readpage_extent(void *vp, st
 
 	assert("vs-1040", PageLocked(page));
 	assert("vs-1050", !PageUptodate(page));
-	assert("vs-1039", page->mapping && page->mapping->host);
+	assert("vs-1039", page_mapping_cache(page) && page_inode(page));
 
 	assert("vs-1044", znode_is_loaded(coord->node));
 	assert("vs-758", item_is_extent(coord));
 	assert("vs-1046", coord_is_existing_unit(coord));
 	assert("vs-1045", znode_is_rlocked(coord->node));
 	assert("vs-1047",
-	       page->mapping->host->i_ino ==
+	       page_inode(page)->i_ino ==
 	       get_key_objectid(item_key_by_coord(coord, &key)));
 	check_uf_coord(uf_coord, NULL);
 
Index: test-2.6.23-rc4-mm1/fs/reiser4/plugin/item/tail.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/reiser4/plugin/item/tail.c
+++ test-2.6.23-rc4-mm1/fs/reiser4/plugin/item/tail.c
@@ -317,7 +317,7 @@ static int do_readpage_tail(uf_coord_t *
 	/* saving passed coord in order to do not move it by tap. */
 	init_lh(&lh);
 	copy_lh(&lh, uf_coord->lh);
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	coord_dup(&coord, &uf_coord->coord);
 
 	reiser4_tap_init(&tap, &coord, &lh, ZNODE_READ_LOCK);
@@ -421,14 +421,14 @@ int readpage_tail(void *vp, struct page 
 	assert("umka-2515", PageLocked(page));
 	assert("umka-2516", !PageUptodate(page));
 	assert("umka-2517", !jprivate(page) && !PagePrivate(page));
-	assert("umka-2518", page->mapping && page->mapping->host);
+	assert("umka-2518", page_mapping_cache(page) && page_inode(page));
 
 	assert("umka-2519", znode_is_loaded(coord->node));
 	assert("umka-2520", item_is_tail(coord));
 	assert("umka-2521", coord_is_existing_unit(coord));
 	assert("umka-2522", znode_is_rlocked(coord->node));
 	assert("umka-2523",
-	       page->mapping->host->i_ino ==
+	       page_inode(page)->i_ino ==
 	       get_key_objectid(item_key_by_coord(coord, &key)));
 
 	return do_readpage_tail(uf_coord, page);

--
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>

[PATCH] add page->mapping handling interface [30/35] changes ROMFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in ROMFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/romfs/inode.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: test-2.6.23-rc4-mm1/fs/romfs/inode.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/romfs/inode.c
+++ test-2.6.23-rc4-mm1/fs/romfs/inode.c
@@ -417,7 +417,7 @@ out:	unlock_kernel();
 static int
 romfs_readpage(struct file *file, struct page * page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	loff_t offset, avail, readlen;
 	void *buf;
 	int result = -EIO;

--
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>

[PATCH] add page->mapping handling interface [31/35] changes in SYSVFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handlingi in SYSVFS.

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/sysv/dir.c |    9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/sysv/dir.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/sysv/dir.c
+++ test-2.6.23-rc4-mm1/fs/sysv/dir.c
@@ -40,7 +40,7 @@ static inline unsigned long dir_pages(st
 
 static int dir_commit_chunk(struct page *page, loff_t pos, unsigned len)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *dir = mapping->host;
 	int err = 0;
 
@@ -221,7 +221,8 @@ got_it:
 	pos = page_offset(page) +
 			(char*)de - (char*)page_address(page);
 	lock_page(page);
-	err = __sysv_write_begin(NULL, page->mapping, pos, SYSV_DIRSIZE,
+	err = __sysv_write_begin(NULL, page_mapping_cache(page),
+				pos, SYSV_DIRSIZE,
 				AOP_FLAG_UNINTERRUPTIBLE, &page, NULL);
 	if (err)
 		goto out_unlock;
@@ -242,7 +243,7 @@ out_unlock:
 
 int sysv_delete_entry(struct sysv_dir_entry *de, struct page *page)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *inode = (struct inode*)mapping->host;
 	char *kaddr = (char*)page_address(page);
 	loff_t pos = page_offset(page) + (char *)de - kaddr;
@@ -344,7 +345,7 @@ not_empty:
 void sysv_set_link(struct sysv_dir_entry *de, struct page *page,
 	struct inode *inode)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *dir = mapping->host;
 	loff_t pos = page_offset(page) +
 			(char *)de-(char*)page_address(page);


--
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>

[PATCH] add page->mapping handling interface [32/35] changes in UDFFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in UDFFS

Signed-off-by: KAMEZAWA hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
---
 fs/udf/file.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/udf/file.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/udf/file.c
+++ test-2.6.23-rc4-mm1/fs/udf/file.c
@@ -43,7 +43,7 @@
 
 static int udf_adinicb_readpage(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	char *kaddr;
 
 	BUG_ON(!PageLocked(page));
@@ -61,7 +61,7 @@ static int udf_adinicb_readpage(struct f
 
 static int udf_adinicb_writepage(struct page *page, struct writeback_control *wbc)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	char *kaddr;
 
 	BUG_ON(!PageLocked(page));
Index: test-2.6.23-rc4-mm1/fs/udf/symlink.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/udf/symlink.c
+++ test-2.6.23-rc4-mm1/fs/udf/symlink.c
@@ -73,7 +73,7 @@ static void udf_pc_to_char(struct super_
 
 static int udf_symlink_filler(struct file *file, struct page *page)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = page_inode(page);
 	struct buffer_head *bh = NULL;
 	char *symlink;
 	int err = -EIO;

--
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>

[PATCH] add page->mapping handling interface [33/35] changes in UFS

[KAMEZAWA Hiroyuki]

Changes  page->mapping handling in UFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/ufs/dir.c  |   10 +++++-----
 fs/ufs/util.c |    2 +-
 2 files changed, 6 insertions(+), 6 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/ufs/dir.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ufs/dir.c
+++ test-2.6.23-rc4-mm1/fs/ufs/dir.c
@@ -42,7 +42,7 @@ static inline int ufs_match(struct super
 
 static int ufs_commit_chunk(struct page *page, loff_t pos, unsigned len)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	struct inode *dir = mapping->host;
 	int err = 0;
 
@@ -95,7 +95,7 @@ void ufs_set_link(struct inode *dir, str
 	int err;
 
 	lock_page(page);
-	err = __ufs_write_begin(NULL, page->mapping, pos, len,
+	err = __ufs_write_begin(NULL, page_mapping_cache(page), pos, len,
 				AOP_FLAG_UNINTERRUPTIBLE, &page, NULL);
 	BUG_ON(err);
 
@@ -111,7 +111,7 @@ void ufs_set_link(struct inode *dir, str
 
 static void ufs_check_page(struct page *page)
 {
-	struct inode *dir = page->mapping->host;
+	struct inode *dir = page_inode(page);
 	struct super_block *sb = dir->i_sb;
 	char *kaddr = page_address(page);
 	unsigned offs, rec_len;
@@ -381,7 +381,7 @@ int ufs_add_link(struct dentry *dentry, 
 got_it:
 	pos = page_offset(page) +
 			(char*)de - (char*)page_address(page);
-	err = __ufs_write_begin(NULL, page->mapping, pos, rec_len,
+	err = __ufs_write_begin(NULL, page_mapping_cache(page), pos, rec_len,
 				AOP_FLAG_UNINTERRUPTIBLE, &page, NULL);
 	if (err)
 		goto out_unlock;
@@ -518,7 +518,7 @@ int ufs_delete_entry(struct inode *inode
 		     struct page * page)
 {
 	struct super_block *sb = inode->i_sb;
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = page_mapping_cache(page);
 	char *kaddr = page_address(page);
 	unsigned from = ((char*)dir - kaddr) & ~(UFS_SB(sb)->s_uspi->s_dirblksize - 1);
 	unsigned to = ((char*)dir - kaddr) + fs16_to_cpu(sb, dir->d_reclen);
Index: test-2.6.23-rc4-mm1/fs/ufs/util.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/ufs/util.c
+++ test-2.6.23-rc4-mm1/fs/ufs/util.c
@@ -263,7 +263,7 @@ struct page *ufs_get_locked_page(struct 
 
 		lock_page(page);
 
-		if (unlikely(page->mapping == NULL)) {
+		if (unlikely(!page_is_pagecache(page))) {
 			/* Truncate got there first */
 			unlock_page(page);
 			page_cache_release(page);

--
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>

[PATCH] add page->mapping handling interface [34/35] changes in UNIONFS

[KAMEZAWA Hiroyuki]

Changes page->mapping handling in UNIONFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>

---
 fs/unionfs/mmap.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/unionfs/mmap.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/unionfs/mmap.c
+++ test-2.6.23-rc4-mm1/fs/unionfs/mmap.c
@@ -61,7 +61,7 @@ static int unionfs_writepage(struct page
 	char *kaddr, *lower_kaddr;
 	int saved_for_writepages = wbc->for_writepages;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	lower_inode = unionfs_lower_inode(inode);
 
 	/* find lower page (returns a locked page) */
@@ -225,7 +225,7 @@ static int unionfs_commit_write(struct f
 	if ((err = unionfs_file_revalidate(file, 1)))
 		goto out;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	lower_inode = unionfs_lower_inode(inode);
 
 	if (UNIONFS_F(file) != NULL)
@@ -283,7 +283,7 @@ static void unionfs_sync_page(struct pag
 	struct page *lower_page;
 	struct address_space *mapping;
 
-	inode = page->mapping->host;
+	inode = page_inode(page);
 	lower_inode = unionfs_lower_inode(inode);
 
 	/* find lower page (returns a locked page) */
@@ -292,7 +292,7 @@ static void unionfs_sync_page(struct pag
 		goto out;
 
 	/* do the actual sync */
-	mapping = lower_page->mapping;
+	mapping = page_mapping_cache(lower_page);
 	/*
 	 * XXX: can we optimize ala RAIF and set the lower page to be
 	 * discarded after a successful sync_page?

--
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>

[PATCH] add page->mapping handling interface [35/35] changes in XFS

[KAMEZAWA Hiroyuki]

Change page->mapping handling in XFS

Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>


---
 fs/xfs/linux-2.6/xfs_aops.c |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

Index: test-2.6.23-rc4-mm1/fs/xfs/linux-2.6/xfs_aops.c
===================================================================
--- test-2.6.23-rc4-mm1.orig/fs/xfs/linux-2.6/xfs_aops.c
+++ test-2.6.23-rc4-mm1/fs/xfs/linux-2.6/xfs_aops.c
@@ -595,7 +595,7 @@ xfs_probe_page(
 	if (PageWriteback(page))
 		return 0;
 
-	if (page->mapping && PageDirty(page)) {
+	if (page_mapping_cache(page) && PageDirty(page)) {
 		if (page_has_buffers(page)) {
 			struct buffer_head	*bh, *head;
 
@@ -697,7 +697,7 @@ xfs_is_delayed_page(
 	if (PageWriteback(page))
 		return 0;
 
-	if (page->mapping && page_has_buffers(page)) {
+	if (page_mapping_cache(page) && page_has_buffers(page)) {
 		struct buffer_head	*bh, *head;
 		int			acceptable = 0;
 
@@ -752,7 +752,7 @@ xfs_convert_page(
 		goto fail;
 	if (PageWriteback(page))
 		goto fail_unlock_page;
-	if (page->mapping != inode->i_mapping)
+	if (pagecache_consistent(page, inode->i_mapping))
 		goto fail_unlock_page;
 	if (!xfs_is_delayed_page(page, (*ioendp)->io_type))
 		goto fail_unlock_page;
@@ -1178,7 +1178,7 @@ xfs_vm_writepage(
 	int			error;
 	int			need_trans;
 	int			delalloc, unmapped, unwritten;
-	struct inode		*inode = page->mapping->host;
+	struct inode		*inode = page_inode(page);
 
 	xfs_page_trace(XFS_WRITEPAGE_ENTER, inode, page, 0);
 
@@ -1270,7 +1270,7 @@ xfs_vm_releasepage(
 	struct page		*page,
 	gfp_t			gfp_mask)
 {
-	struct inode		*inode = page->mapping->host;
+	struct inode		*inode = page_inode(page);
 	int			dirty, delalloc, unmapped, unwritten;
 	struct writeback_control wbc = {
 		.sync_mode = WB_SYNC_ALL,
@@ -1562,7 +1562,7 @@ xfs_vm_invalidatepage(
 	unsigned long		offset)
 {
 	xfs_page_trace(XFS_INVALIDPAGE_ENTER,
-			page->mapping->host, page, offset);
+			page_inode(page), page, offset);
 	block_invalidatepage(page, offset);
 }
 

--
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>

Re: [PATCH] add page->mapping handling interface [22/35] changes in JFFS2

[KAMEZAWA Hiroyuki]

On Mon, 10 Sep 2007 11:19:51 +0100
David Woodhouse <dwmw2@infradead.org> wrote:

> On Mon, 2007-09-10 at 19:16 +0900, KAMEZAWA Hiroyuki wrote:
> > Changes page->mapping handling in JFFS2
> > 
> > Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> 
> Looks reasonable to me; I assume it's not intended for me to take it and
> apply it yet, before the core parts are merged? I'll let you shepherd it
> upstream...
> 
Ah, no. core patches are not merged. This patch's target is -mm, now.
I just CC:ed to each FS maintainers to show what this patch set does.

Thanks,
-Kame

--
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>

Re: [PATCH] add page->mapping handling interface [1/35] interface definitions

[Richard Knutsson]

KAMEZAWA Hiroyuki wrote:
>  - changes page->mapping from address_space* to unsigned long
>  - add page_mapping_anon() function.
>  - add linux/page-cache.h
>  - add page_inode() function
>  - add page_is_pagecache() function
>  - add pagecaceh_consisten() function for pagecache consistency test.
>  - expoterd swapper_space. inline function page_mapping() refers this.
>
> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
>
> ---
>  include/linux/fs.h         |    1 +
>  include/linux/mm.h         |   20 +++++++++++++++++---
>  include/linux/mm_types.h   |    2 +-
>  include/linux/page-cache.h |   39 +++++++++++++++++++++++++++++++++++++++
>  mm/swap_state.c            |    2 ++
>  5 files changed, 60 insertions(+), 4 deletions(-)
>
> Index: test-2.6.23-rc4-mm1/include/linux/page-cache.h
> ===================================================================
> --- /dev/null
> +++ test-2.6.23-rc4-mm1/include/linux/page-cache.h
> @@ -0,0 +1,39 @@
> +/*
> + * For interface definitions between memory management and file systems.
> + * - This file defines small interface functions for handling page cache.
> + */
> +
> +#ifndef _LINUX_PAGECACHE_H
> +#define _LINUX_PAGECACHE_H
> +
> +#include <linux/mm.h>
> +/* page_mapping_xxx() function is defined in mm.h */
> +
> +static inline int page_is_pagecache(struct page *page)
>   
Why return it as an 'int' instead of 'bool'?
> +{
> +	if (!page->mapping || (page->mapping & PAGE_MAPPING_ANON))
> +		return 0;
> +	return 1;
> +}
>   
Not easier with 'return page->mapping && (page->mapping & 
PAGE_MAPPING_ANON) == 0;'?
> +
> +/*
> + * Return an inode this page belongs to
> + */
> +
> +static inline struct inode *page_inode(struct page *page)
> +{
> +	if (!page_is_pagecache(page))
> +		return NULL;
> +	return page_mapping_cache(page)->host;
> +}
> +
> +/*
> + * Test a page is a page-cache of an address_space.
> + */
> +static inline int
> +pagecache_consistent(struct page *page, struct address_space *as)
> +{
> +	return (page_mapping(page) == as);
> +}
> +
> +#endif
<snip>
> Index: test-2.6.23-rc4-mm1/include/linux/mm.h
> ===================================================================
> --- test-2.6.23-rc4-mm1.orig/include/linux/mm.h
> +++ test-2.6.23-rc4-mm1/include/linux/mm.h
> @@ -563,7 +563,7 @@ void page_address_init(void);
>  extern struct address_space swapper_space;
>  static inline struct address_space *page_mapping(struct page *page)
>  {
> -	struct address_space *mapping = page->mapping;
> +	struct address_space *mapping = (struct address_space *)page->mapping;
>  
>  	VM_BUG_ON(PageSlab(page));
>  	if (unlikely(PageSwapCache(page)))
> @@ -579,7 +579,21 @@ static inline struct address_space *page
>  
>  static inline int PageAnon(struct page *page)
>   
Change to bool? Then "you" can also remove the '!!' from:
mm/memory.c:483:                rss[!!PageAnon(page)]++;
>  {
> -	return ((unsigned long)page->mapping & PAGE_MAPPING_ANON) != 0;
> +	return (page->mapping & PAGE_MAPPING_ANON) != 0;
> +}
> +
>   
<snip>

If you don't mind bool(eans) (for some reason), I can/will check out the 
rest.

Richard Knutsson

--
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>

Re: [PATCH] add page->mapping handling interface [1/35] interface definitions

[KAMEZAWA Hiroyuki]

On Thu, 13 Sep 2007 22:19:20 +0200
Richard Knutsson <ricknu-0@student.ltu.se> wrote:
> > +static inline int page_is_pagecache(struct page *page)
> >   
> Why return it as an 'int' instead of 'bool'?
> > +{
> > +	if (!page->mapping || (page->mapping & PAGE_MAPPING_ANON))
> > +		return 0;
> > +	return 1;
> > +}

Ah, I missed bool type just because I have no experience to use 'bool' in
Linux kernel. ok, will try in the next version. thank you.

> >   
> Not easier with 'return page->mapping && (page->mapping & 
> PAGE_MAPPING_ANON) == 0;'?
> > +

yours seems better.


> 
> >  static inline int PageAnon(struct page *page)
> >   
> Change to bool? Then "you" can also remove the '!!' from:
> mm/memory.c:483:                rss[!!PageAnon(page)]++;

Hmm, will try unless it makes diff big.

> >  {
> > -	return ((unsigned long)page->mapping & PAGE_MAPPING_ANON) != 0;
> > +	return (page->mapping & PAGE_MAPPING_ANON) != 0;
> > +}
> > +
> >   
> <snip>
> 
> If you don't mind bool(eans) (for some reason), I can/will check out the 
> rest.
> 

Thank you. I'll try 'bool' type. 

Regards,
-Kame

--
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>