At 11/08/2016 01:05 AM, David Sterba wrote:

On Thu, Nov 03, 2016 at 09:58:21AM +0800, Qu Wenruo wrote:

quoted

At 11/02/2016 11:21 PM, David Sterba wrote:

quoted

On Wed, Sep 21, 2016 at 11:15:52AM +0800, Qu Wenruo wrote:

quoted

From: Lu Fengqi <redacted>

Introduce a new function find_dir_item() to find DIR_ITEM for the given
key, and check it with the specified INODE_REF/INODE_EXTREF match.

Signed-off-by: Lu Fengqi <redacted>
Signed-off-by: Qu Wenruo <redacted>
---
 cmds-check.c | 140 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 140 insertions(+)

diff --git a/cmds-check.c b/cmds-check.c
index 998ba63..4e25804 100644
--- a/cmds-check.c
+++ b/cmds-check.c

@@ -3848,6 +3848,146 @@ out:
 	return err;
 }

+#define ROOT_DIR_ERROR		(1<<1)	/* bad root_dir */
+#define DIR_ITEM_MISSING	(1<<2)	/* DIR_ITEM not found */
+#define DIR_ITEM_MISMATCH	(1<<3)	/* DIR_ITEM found but not match */

What's the reason for another definition of the error codes? They're
mostly copied from te I_ERR_* counterparts. I'd rather have one set of
error codes.

The main reason is, in lowmem fsck mode, we are not checking inodes or
ref/backref in batch.

If using I_ERR and REF_ERR, we can mixing them up as they share the same
bits.

So we introduced the new error bitmap, to make sure all error bits won't
cover each other.

It may be better if we rearrange I_ERR/REF_ERR to avoid conflicts.
E.g, let REF_ERR_ starts from lowest bit and let I_ERR_ starts from high
bits.

Yes please. Third namespace for existing error bits is not a good
option. Move the I_ERR bits to start from 32 and use them in the low-mem
code that's been merged to devel.

Hi David,

I didn't see such fix in devel branch.
In cmds-check.c, there I_ERR_* and REF_ERR_* are still using the old range.

Is the fix in another branch? Or I missed something again?

Thanks,
Qu