From: "Darrick J. Wong" <darrick.wong@oracle.com>
To: darrick.wong@oracle.com
Cc: linux-xfs@vger.kernel.org
Subject: [PATCH 5/8] xfs: inode scrubber shouldn't bother with raw checks
Date: Thu, 22 Feb 2018 18:00:58 -0800 [thread overview]
Message-ID: <151935125879.21842.10380711479854109278.stgit@magnolia> (raw)
In-Reply-To: <151935122813.21842.14996049216693190946.stgit@magnolia>
From: Darrick J. Wong <darrick.wong@oracle.com>
The inode scrubber tries to _iget the inode prior to running checks.
If that _iget call fails with corruption errors that's an automatic
fail, regardless of whether it was the inode buffer read verifier,
the ifork verifier, or the ifork formatter that errored out.
Therefore, get rid of the raw mode scrub code because it's not needed.
Found by trying to fix some test failures in xfs/379 and xfs/415.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
fs/xfs/scrub/inode.c | 98 +++++---------------------------------------------
1 file changed, 9 insertions(+), 89 deletions(-)
diff --git a/fs/xfs/scrub/inode.c b/fs/xfs/scrub/inode.c
index 21297be..0332a01 100644
--- a/fs/xfs/scrub/inode.c
+++ b/fs/xfs/scrub/inode.c
@@ -515,72 +515,6 @@ xfs_scrub_dinode(
}
}
-/* Map and read a raw inode. */
-STATIC int
-xfs_scrub_inode_map_raw(
- struct xfs_scrub_context *sc,
- xfs_ino_t ino,
- struct xfs_buf **bpp,
- struct xfs_dinode **dipp)
-{
- struct xfs_imap imap;
- struct xfs_mount *mp = sc->mp;
- struct xfs_buf *bp = NULL;
- struct xfs_dinode *dip;
- int error;
-
- error = xfs_imap(mp, sc->tp, ino, &imap, XFS_IGET_UNTRUSTED);
- if (error == -EINVAL) {
- /*
- * Inode could have gotten deleted out from under us;
- * just forget about it.
- */
- error = -ENOENT;
- goto out;
- }
- if (!xfs_scrub_process_error(sc, XFS_INO_TO_AGNO(mp, ino),
- XFS_INO_TO_AGBNO(mp, ino), &error))
- goto out;
-
- error = xfs_trans_read_buf(mp, sc->tp, mp->m_ddev_targp,
- imap.im_blkno, imap.im_len, XBF_UNMAPPED, &bp,
- NULL);
- if (!xfs_scrub_process_error(sc, XFS_INO_TO_AGNO(mp, ino),
- XFS_INO_TO_AGBNO(mp, ino), &error))
- goto out;
-
- /*
- * Is this really an inode? We disabled verifiers in the above
- * xfs_trans_read_buf call because the inode buffer verifier
- * fails on /any/ inode record in the inode cluster with a bad
- * magic or version number, not just the one that we're
- * checking. Therefore, grab the buffer unconditionally, attach
- * the inode verifiers by hand, and run the inode verifier only
- * on the one inode we want.
- */
- bp->b_ops = &xfs_inode_buf_ops;
- dip = xfs_buf_offset(bp, imap.im_boffset);
- if (xfs_dinode_verify(mp, ino, dip) != NULL ||
- !xfs_dinode_good_version(mp, dip->di_version)) {
- xfs_scrub_ino_set_corrupt(sc, ino, bp);
- goto out_buf;
- }
-
- /* ...and is it the one we asked for? */
- if (be32_to_cpu(dip->di_gen) != sc->sm->sm_gen) {
- error = -ENOENT;
- goto out_buf;
- }
-
- *dipp = dip;
- *bpp = bp;
-out:
- return error;
-out_buf:
- xfs_trans_brelse(sc->tp, bp);
- return error;
-}
-
/*
* Make sure the finobt doesn't think this inode is free.
* We don't have to check the inobt ourselves because we got the inode via
@@ -727,43 +661,29 @@ xfs_scrub_inode(
struct xfs_scrub_context *sc)
{
struct xfs_dinode di;
- struct xfs_buf *bp = NULL;
- struct xfs_dinode *dip;
- xfs_ino_t ino;
int error = 0;
- /* Did we get the in-core inode, or are we doing this manually? */
- if (sc->ip) {
- ino = sc->ip->i_ino;
- xfs_inode_to_disk(sc->ip, &di, 0);
- dip = &di;
- } else {
- /* Map & read inode. */
- ino = sc->sm->sm_ino;
- error = xfs_scrub_inode_map_raw(sc, ino, &bp, &dip);
- if (error || !bp)
- goto out;
+ /* iget failed means automatic fail. */
+ if (!sc->ip) {
+ xfs_scrub_ino_set_corrupt(sc, sc->sm->sm_ino, NULL);
+ return 0;
}
- xfs_scrub_dinode(sc, bp, dip, ino);
+ /* Scrub the inode core. */
+ xfs_inode_to_disk(sc->ip, &di, 0);
+ xfs_scrub_dinode(sc, NULL, &di, sc->ip->i_ino);
if (sc->sm->sm_flags & XFS_SCRUB_OFLAG_CORRUPT)
goto out;
- /* Now let's do the things that require a live inode. */
- if (!sc->ip)
- goto out;
-
/*
* Look for discrepancies between file's data blocks and the reflink
* iflag. We already checked the iflag against the file mode when
* we scrubbed the dinode.
*/
if (S_ISREG(VFS_I(sc->ip)->i_mode))
- xfs_scrub_inode_check_reflink_iflag(sc, ino, bp);
+ xfs_scrub_inode_check_reflink_iflag(sc, sc->ip->i_ino, NULL);
- xfs_scrub_inode_xref(sc, ino, dip);
+ xfs_scrub_inode_xref(sc, sc->ip->i_ino, &di);
out:
- if (bp)
- xfs_trans_brelse(sc->tp, bp);
return error;
}
next prev parent reply other threads:[~2018-02-23 2:06 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-02-23 2:00 [PATCH 0/8] xfs: online scrub fixes Darrick J. Wong
2018-02-23 2:00 ` [PATCH 1/8] xfs: refactor bmap record valiation Darrick J. Wong
2018-02-23 2:00 ` [PATCH 2/8] xfs: refactor inode verifier error logging Darrick J. Wong
2018-02-23 2:00 ` [PATCH 3/8] xfs: refactor inode buffer " Darrick J. Wong
2018-02-23 2:00 ` [PATCH 4/8] xfs: bmap scrubber should do rmap xref with bmap for sparse files Darrick J. Wong
2018-02-23 2:00 ` Darrick J. Wong [this message]
2018-02-23 2:01 ` [PATCH 6/8] xfs: remove xfs_buf parameter from inode scrub methods Darrick J. Wong
2018-02-23 2:01 ` [PATCH 7/8] xfs: record inode buf errors as a xref error in inode scrubber Darrick J. Wong
2018-02-23 2:01 ` [PATCH 8/8] xfs: move inode extent size hint validation to libxfs Darrick J. Wong
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=151935125879.21842.10380711479854109278.stgit@magnolia \
--to=darrick.wong@oracle.com \
--cc=linux-xfs@vger.kernel.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.