[PATCH] retries in ext4_prepare_write() violate ordering requirements
authorAndrey Savochkin <saw@sw.ru>
Thu, 7 Dec 2006 04:37:36 +0000 (20:37 -0800)
committerLinus Torvalds <torvalds@woody.osdl.org>
Thu, 7 Dec 2006 16:39:37 +0000 (08:39 -0800)
In journal=ordered or journal=data mode retry in ext4_prepare_write()
breaks the requirements of journaling of data with respect to metadata.
The fix is to call commit_write to commit allocated zero blocks before
retry.

Signed-off-by: Kirill Korotaev <dev@openvz.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Ken Chen <kenneth.w.chen@intel.com>
Cc: <linux-ext4@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
fs/ext4/inode.c

index 0a60ec5a16dbc4bb92645dfafcd644e895d05161..1d85d4ec9598e063968894e1cc9aa592e65dc849 100644 (file)
@@ -1147,37 +1147,102 @@ static int do_journal_get_write_access(handle_t *handle,
        return ext4_journal_get_write_access(handle, bh);
 }
 
+/*
+ * The idea of this helper function is following:
+ * if prepare_write has allocated some blocks, but not all of them, the
+ * transaction must include the content of the newly allocated blocks.
+ * This content is expected to be set to zeroes by block_prepare_write().
+ * 2006/10/14  SAW
+ */
+static int ext4_prepare_failure(struct file *file, struct page *page,
+                               unsigned from, unsigned to)
+{
+       struct address_space *mapping;
+       struct buffer_head *bh, *head, *next;
+       unsigned block_start, block_end;
+       unsigned blocksize;
+       int ret;
+       handle_t *handle = ext4_journal_current_handle();
+
+       mapping = page->mapping;
+       if (ext4_should_writeback_data(mapping->host)) {
+               /* optimization: no constraints about data */
+skip:
+               return ext4_journal_stop(handle);
+       }
+
+       head = page_buffers(page);
+       blocksize = head->b_size;
+       for (   bh = head, block_start = 0;
+               bh != head || !block_start;
+               block_start = block_end, bh = next)
+       {
+               next = bh->b_this_page;
+               block_end = block_start + blocksize;
+               if (block_end <= from)
+                       continue;
+               if (block_start >= to) {
+                       block_start = to;
+                       break;
+               }
+               if (!buffer_mapped(bh))
+               /* prepare_write failed on this bh */
+                       break;
+               if (ext4_should_journal_data(mapping->host)) {
+                       ret = do_journal_get_write_access(handle, bh);
+                       if (ret) {
+                               ext4_journal_stop(handle);
+                               return ret;
+                       }
+               }
+       /*
+        * block_start here becomes the first block where the current iteration
+        * of prepare_write failed.
+        */
+       }
+       if (block_start <= from)
+               goto skip;
+
+       /* commit allocated and zeroed buffers */
+       return mapping->a_ops->commit_write(file, page, from, block_start);
+}
+
 static int ext4_prepare_write(struct file *file, struct page *page,
                              unsigned from, unsigned to)
 {
        struct inode *inode = page->mapping->host;
-       int ret, needed_blocks = ext4_writepage_trans_blocks(inode);
+       int ret, ret2;
+       int needed_blocks = ext4_writepage_trans_blocks(inode);
        handle_t *handle;
        int retries = 0;
 
 retry:
        handle = ext4_journal_start(inode, needed_blocks);
-       if (IS_ERR(handle)) {
-               ret = PTR_ERR(handle);
-               goto out;
-       }
+       if (IS_ERR(handle))
+               return PTR_ERR(handle);
        if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode))
                ret = nobh_prepare_write(page, from, to, ext4_get_block);
        else
                ret = block_prepare_write(page, from, to, ext4_get_block);
        if (ret)
-               goto prepare_write_failed;
+               goto failure;
 
        if (ext4_should_journal_data(inode)) {
                ret = walk_page_buffers(handle, page_buffers(page),
                                from, to, NULL, do_journal_get_write_access);
+               if (ret)
+                       /* fatal error, just put the handle and return */
+                       journal_stop(handle);
        }
-prepare_write_failed:
-       if (ret)
-               ext4_journal_stop(handle);
+       return ret;
+
+failure:
+       ret2 = ext4_prepare_failure(file, page, from, to);
+       if (ret2 < 0)
+               return ret2;
        if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
                goto retry;
-out:
+       /* retry number exceeded, or other error like -EDQUOT */
        return ret;
 }