[PATCH] decouple swapper_space treatment from other address_spaces

swapper_space is different.  Its pages are locked during writeout, it
uses PAGE_SIZE rather than PAGE_CACHE_SIZE.  Converting swap to look
more like the other address_spaces is a separate project.  This patch
fully restores the old swap behaviour.

- Don't dirty swapcache page buffers in set_page_dirty().

  Fixes a problem where __free_pte() runs set_page_dirty() and then
  immediately runs ClearPageDirty().  The pages ends up clean, with
  dirty buffers, and is unfreeable.

- Hence, don't mark the page clean if its buffers are clean - swap
  does not have page/buffers dirty state coherency.

fs/buffer.c

@@ -2159,14 +2159,6 @@ static void check_ttfb_buffer(struct page *page, struct buffer_head *bh)
  * total exclusion from __set_page_dirty_buffers().  That is obtained with
  * i_bufferlist_lock.
  *
- * Nobody should be calling try_to_free_buffers against a page which is
- * eligible for set_page_dirty() treatment anyway - the page is clearly
- * not freeable.  So we could just test page_count(page) here and complain
- * then scram if it's wrong.
- *
- * If any buffer is not uptodate then the entire page is set not uptodate,
- * as the partial uptodateness information is about to be lost.
- *
  * try_to_free_buffers() is non-blocking.
  */
 static inline int buffer_busy(struct buffer_head *bh)
@@ -2222,8 +2214,17 @@ int try_to_free_buffers(struct page *page)
 	inode = page->mapping->host;
 	spin_lock(&inode->i_bufferlist_lock);
 	ret = drop_buffers(page);
-	if (ret)
+	if (ret && !PageSwapCache(page)) {
+		/*
+		 * If the filesystem writes its buffers by hand (eg ext3)
+		 * then we can have clean buffers against a dirty page.  We
+		 * clean the page here; otherwise later reattachment of buffers
+		 * could encounter a non-uptodate page, which is unresolvable.
+		 * This only applies in the rare case where try_to_free_buffers
+		 * succeeds but the page is not freed.
+		 */
 		ClearPageDirty(page);
+	}
 	spin_unlock(&inode->i_bufferlist_lock);
 	return ret;
 }

mm/page-writeback.c

@@ -454,6 +454,9 @@ EXPORT_SYMBOL(write_one_page);
  *
  * FIXME: may need to call ->reservepage here as well.  That's rather up to the
  * address_space though.
+ *
+ * For now, we treat swapper_space specially.  It doesn't use the normal
+ * block a_ops.
  */
 int __set_page_dirty_buffers(struct page *page)
 {
@@ -470,7 +473,7 @@ int __set_page_dirty_buffers(struct page *page)
 
 	spin_lock(&inode->i_bufferlist_lock);
 
-	if (page_has_buffers(page)) {
+	if (page_has_buffers(page) && !PageSwapCache(page)) {
 		struct buffer_head *head = page_buffers(page);
 		struct buffer_head *bh = head;