Skip to contentSkip to conversation

xfs_repair: handling a block with bad crc, bad uuid, and bad magic number needs fixing

PATCH In review [email protected] 3 participants 5 messages Mar 21, 2025 – Mar 24, 2025

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 xfs_repair: handling a block with bad crc, bad uuid, and bad magic number needs fixing

From: Bill O'Donnell <[email protected]>

In certain cases, if a block is so messed up that crc, uuid and magic number are all bad, we need to not only detect in phase3 but fix it properly in phase6. In the current code, the mechanism doesn't work in that it only pays attention to one of the parameters.

Note: in this case, the nlink inode link count drops to 1, but re-running xfs_repair fixes it back to 2. This is a side effect that should probably be handled in update_inode_nlinks() with separate patch. Regardless, running xfs_repair twice, with this patch applied fixes the issue. Recognize that this patch is a fix for xfs v5.

v2: remove superfluous needmagic logic
v3: clarify the description

---
 repair/phase6.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)
repair/phase6.c +1 -4
index 4064a84b2450..9cffbb1f4510 100644
@@ -2364,7 +2364,6 @@ longform_dir2_entry_check(
 	     da_bno = (xfs_dablk_t)next_da_bno) {
 		const struct xfs_buf_ops *ops;
 		int			 error;
-		struct xfs_dir2_data_hdr *d;
 
 		next_da_bno = da_bno + mp->m_dir_geo->fsbcount - 1;
 		if (bmap_next_offset(ip, &next_da_bno)) {
@@ -2404,9 +2403,7 @@ longform_dir2_entry_check(
 		}
 
 		/* check v5 metadata */
-		d = bp->b_addr;
-		if (be32_to_cpu(d->magic) == XFS_DIR3_BLOCK_MAGIC ||
-		    be32_to_cpu(d->magic) == XFS_DIR3_DATA_MAGIC) {
+		if (xfs_has_crc(mp)) {
 			error = check_dir3_header(mp, bp, ino);
 			if (error) {
 				fixit++;
Signature · 1 line
2.48.1
#2 lore raw
In reply to [email protected]
Quoted 7 lines · On 3/21/25 5:05 PM, [email protected] wrote:

From: Bill O'Donnell <[email protected]>

In certain cases, if a block is so messed up that crc, uuid and magic number are all bad, we need to not only detect in phase3 but fix it properly in phase6. In the current code, the mechanism doesn't work in that it only pays attention to one of the parameters.

Quoted 5 lines

Note: in this case, the nlink inode link count drops to 1, but re-running xfs_repair fixes it back to 2. This is a side effect that should probably be handled in update_inode_nlinks() with separate patch. Regardless, running xfs_repair twice, with this patch applied fixes the issue. Recognize that this patch is a fix for xfs v5.

The reason this fix leaves the inode nlinks in an inconsistent state is because the dir is (was) a longform directory (XFS_DIR2_FMT_BLOCK), and we go down this path with your patch in place:

                /* check v5 metadata */
                if (xfs_has_crc(mp)) {
                        error = check_dir3_header(mp, bp, ino);
                        if (error) {
                                fixit++;
                                if (fmt == XFS_DIR2_FMT_BLOCK) { <==== true
                                        goto out_fix;	<=== goto
                                }

                                libxfs_buf_relse(bp);
                                bp = NULL;
                                continue;
                        }
                }

                longform_dir2_entry_check_data(mp, ip, num_illegal, need_dot,
                                irec, ino_offset, bp, hashtab,
                                &freetab, da_bno, fmt == XFS_DIR2_FMT_BLOCK);

...

out_fix:

        if (!no_modify && (fixit || dotdot_update)) {
                longform_dir2_rebuild(mp, ino, ip, irec, ino_offset, hashtab);

longform_dir2_rebuild tries to rebuild the directory from the entries found via longform_dir2_entry_check_data() and placed in hashtab, but because we never called longform_dir2_entry_check_data(), hashtab is empty. This is why all entries in the problematic dir end up in lost+found.

That also means that longform_dir2_rebuild completes without adding any entries at all, and so the directory is now shortform. Because shortform directories have no explicit "." entry, I think it would need an extra ref added at this point.

But I wonder - why not call longform_dir2_entry_check_data() before we check the header? That way it /will/ populate hashtab with any found entries in the block, and when the header is found to be corrupt, it will rebuild it with all entries intact, and leave nothing in lost+found.

Thoughts?
-Eric

#3 lore raw

On 3/23/25 10:51 AM, Eric Sandeen wrote:

...

Quoted 4 lines

But I wonder - why not call longform_dir2_entry_check_data() before we check the header? That way it /will/ populate hashtab with any found entries in the block, and when the header is found to be corrupt, it will rebuild it with all entries intact, and leave nothing in lost+found.

Or alternatively, since checking the header first makes intuitive sense, perhaps change the logic so that we still check the header first, set fixit++, and if we are in XFS_DIR2_FMT_BLOCK, allow longform_dir2_entry_check_data() to be called, and then if errors were found in the header, goto out_fix; at that point, with a populated hashtab.

-Eric

Quoted 7 lines

Thoughts?
-Eric

#4 lore raw
In reply to Eric Sandeen
Quoted 54 lines · Eric Sandeen
Quoted 7 lines · On 3/21/25 5:05 PM, [email protected] wrote:

From: Bill O'Donnell <[email protected]>

In certain cases, if a block is so messed up that crc, uuid and magic number are all bad, we need to not only detect in phase3 but fix it properly in phase6. In the current code, the mechanism doesn't work in that it only pays attention to one of the parameters.

Quoted 5 lines

Note: in this case, the nlink inode link count drops to 1, but re-running xfs_repair fixes it back to 2. This is a side effect that should probably be handled in update_inode_nlinks() with separate patch. Regardless, running xfs_repair twice, with this patch applied fixes the issue. Recognize that this patch is a fix for xfs v5.

The reason this fix leaves the inode nlinks in an inconsistent state is because the dir is (was) a longform directory (XFS_DIR2_FMT_BLOCK), and we go down this path with your patch in place:

                /* check v5 metadata */
                if (xfs_has_crc(mp)) {
                        error = check_dir3_header(mp, bp, ino);
                        if (error) {
                                fixit++;
                                if (fmt == XFS_DIR2_FMT_BLOCK) { <==== true
                                        goto out_fix;	<=== goto
                                }

                                libxfs_buf_relse(bp);
                                bp = NULL;
                                continue;
                        }
                }

                longform_dir2_entry_check_data(mp, ip, num_illegal, need_dot,
                                irec, ino_offset, bp, hashtab,
                                &freetab, da_bno, fmt == XFS_DIR2_FMT_BLOCK);

...

out_fix:

        if (!no_modify && (fixit || dotdot_update)) {
                longform_dir2_rebuild(mp, ino, ip, irec, ino_offset, hashtab);

longform_dir2_rebuild tries to rebuild the directory from the entries found via longform_dir2_entry_check_data() and placed in hashtab, but because we never called longform_dir2_entry_check_data(), hashtab is empty. This is why all entries in the problematic dir end up in lost+found.

That also means that longform_dir2_rebuild completes without adding any entries at all, and so the directory is now shortform. Because shortform directories have no explicit "." entry, I think it would need an extra ref added at this point.

Quoted 4 lines

But I wonder - why not call longform_dir2_entry_check_data() before we check the header? That way it /will/ populate hashtab with any found entries in the block, and when the header is found to be corrupt, it will rebuild it with all entries intact, and leave nothing in lost+found.

Yes, this works as you describe.
Thanks-
Bill