From: Eric Sandeen <sandeen@sandeen.net>
To: Brian Foster <bfoster@redhat.com>, Eric Sandeen <sandeen@redhat.com>
Cc: xfs@oss.sgi.com
Subject: Re: [PATCH 3/5] xfs_repair: fix dir refcount when '.' missing and dir is rebuilt
Date: Mon, 08 Sep 2014 09:33:55 -0500 [thread overview]
Message-ID: <540DBE53.7040600@sandeen.net> (raw)
In-Reply-To: <20140908134524.GC52419@bfoster.bfoster>
On 9/8/14 8:45 AM, Brian Foster wrote:
> On Sun, Sep 07, 2014 at 11:41:03AM -0500, Eric Sandeen wrote:
>> In phase 6's longform_dir2_entry_check, if we never
>> find a '.' entry we never add a reference to that entry;
>> if we subsequently rebuild it, '.' gets added, but
>> no ref to it is ever made. This leads to Phase 7 doing
>> i.e.:
>>
>> Phase 7 - verify and correct link counts...
>> resetting inode 5184 nlinks from 2 to 1
>>
>> and the next run will do:
>>
>> Phase 7 - verify and correct link counts...
>> resetting inode 5184 nlinks from 1 to 2
>>
>> So if '.' was never found, but the directory got
>> rebuilt, manually add the ref for it.
>>
>> Signed-off-by: Eric Sandeen <sandeen@redhat.com>
>> ---
>> repair/phase6.c | 6 ++++++
>> 1 files changed, 6 insertions(+), 0 deletions(-)
>>
>> diff --git a/repair/phase6.c b/repair/phase6.c
>> index f13069f..cc36a9c 100644
>> --- a/repair/phase6.c
>> +++ b/repair/phase6.c
>> @@ -2288,6 +2288,12 @@ out_fix:
>> if (bplist[i])
>> libxfs_putbuf(bplist[i]);
>> longform_dir2_rebuild(mp, ino, ip, irec, ino_offset, hashtab);
>> + /*
>> + * If we didn't find a dot, we never added a ref for it;
>> + * it's there now after the rebuild, so mark it as reached.
>> + */
>> + if (*need_dot)
>> + add_inode_ref(irec, ino_offset);
>
> So if I follow this correctly, we iterate through the dir, add each name
> to the hashtable and handle the inode reference count in the first
> longform_dir2_entry_check() loop. If something is wrong, we call
> longform_dir2_rebuild() to rebuild the dir from the hashtable of
> names/inodes. We may or may not have added a reference for dot at that
> point, and need_dot is set appropriately.
>
> This seems Ok, but where is the dot entry actually added? Hmm, I see
> that we handle dot in the longform_dir2_rebuild() loop by just skipping
> over it...
longform_dir2_rebuild calls this before the loop:
/*
* Initialize a directory with its "." and ".." entries.
*/
int
xfs_dir_init()
But it doesn't actually create .; this creates a shortform dir,
and shortform dirs have no '.' - I guess the comment could use
an update.
In the loop we call xfs_dir_createname() for everything in the hash,
eventually things don't fit in shortform and we do xfs_dir2_sf_to_block,
which creates the dot:
/*
* Create entry for .
*/
dep = xfs_dir3_data_dot_entry_p(mp, hdr);
dep->inumber = cpu_to_be64(dp->i_ino);
dep->namelen = 1;
dep->name[0] = '.';
-Eric
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
next prev parent reply other threads:[~2014-09-08 14:33 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-07 16:41 [PATCH 0/5] xfs_repair fixes, part 1 Eric Sandeen
2014-09-07 16:41 ` [PATCH 1/5] xfs_repair: clear bad flgs in process_dinode_int Eric Sandeen
2014-09-08 13:45 ` Brian Foster
2014-09-09 22:28 ` Christoph Hellwig
2014-09-09 22:33 ` Eric Sandeen
2014-09-09 23:48 ` Dave Chinner
2014-09-07 16:41 ` [PATCH 2/5] xfs_repair: preserve error state in process_shortform_attr Eric Sandeen
2014-09-08 13:45 ` Brian Foster
2014-09-09 22:29 ` Christoph Hellwig
2014-09-07 16:41 ` [PATCH 3/5] xfs_repair: fix dir refcount when '.' missing and dir is rebuilt Eric Sandeen
2014-09-08 13:45 ` Brian Foster
2014-09-08 14:25 ` Brian Foster
2014-09-08 14:44 ` Eric Sandeen
2014-09-08 14:33 ` Eric Sandeen [this message]
2014-09-07 16:41 ` [PATCH 4/5] xfs_repair: don't ASSERT on corrupt ftype Eric Sandeen
2014-09-08 0:10 ` Dave Chinner
2014-09-08 1:02 ` Eric Sandeen
2014-09-08 3:16 ` Dave Chinner
2014-09-08 3:18 ` Eric Sandeen
2014-09-08 6:23 ` Dave Chinner
2014-09-07 16:41 ` [PATCH 5/5] xfs_repair: set proper ftype when moving to lost+found Eric Sandeen
2014-09-07 21:26 ` Eric Sandeen
2014-09-07 17:02 ` [PATCH 6/5] xfs_repair: don't re-add root dotdot if root dir was rebuilt Eric Sandeen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=540DBE53.7040600@sandeen.net \
--to=sandeen@sandeen.net \
--cc=bfoster@redhat.com \
--cc=sandeen@redhat.com \
--cc=xfs@oss.sgi.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox