[PATCH] ext3 set_page_dirty fix

The set_page_dirty() in the ext3_writepage() failure path isn't right.
set_page_dirty() will alter buffer states - it's a "whole page"
dirtying.

__set_page_dirty_buffers() is emitting warnings when it refuses to set
dirty a non-uptodate buffer against a partially-mapped page.

All we want to do in there is to move the page back onto
mapping->dirty_pages, without altering the state of its buffers.

fs/ext3/inode.c

@@ -1327,7 +1327,13 @@ static int ext3_writepage(struct page *page)
 out_fail:
 	
 	unlock_kernel();
-	set_page_dirty(page);
+
+	/*
+	 * We have to fail this writepage to avoid cross-fs transactions.
+	 * Put the page back on mapping->dirty_pages, but leave its buffer's
+	 * dirty state as-is.
+	 */
+	__set_page_dirty_nobuffers(page);
 	unlock_page(page);
 	return ret;
 }