From: Dave Chinner <david@fromorbit.com>
To: "Darrick J. Wong" <darrick.wong@oracle.com>
Cc: linux-xfs@vger.kernel.org
Subject: Re: [PATCH 12/25] xfs: scrub free space btrees
Date: Thu, 5 Oct 2017 11:59:43 +1100 [thread overview]
Message-ID: <20171005005943.GI3666@dastard> (raw)
In-Reply-To: <150706332522.19351.3718364611675005450.stgit@magnolia>
On Tue, Oct 03, 2017 at 01:42:05PM -0700, Darrick J. Wong wrote:
> +
> +/*
> + * Set us up to scrub free space btrees.
> + * Push everything out of the log so that the busy extent list is empty.
> + */
> +int
> +xfs_scrub_setup_ag_allocbt(
> + struct xfs_scrub_context *sc,
> + struct xfs_inode *ip)
> +{
> + return xfs_scrub_setup_ag_btree(sc, ip, sc->try_harder);
These setup calls are getting deeply nested and intertwined. And
I really don't know why we pass sc->try_harder as a separate
parameter when we are already passing sc, especically as
xfs_scrub_setup_ag_btree() doesn't use it at all...
> +/* Scrub a bnobt/cntbt record. */
> +STATIC int
> +xfs_scrub_allocbt_helper(
> + struct xfs_scrub_btree *bs,
> + union xfs_btree_rec *rec)
> +{
> + struct xfs_mount *mp = bs->cur->bc_mp;
> + struct xfs_agf *agf;
> + unsigned long long rec_end;
> + xfs_agblock_t bno;
> + xfs_extlen_t len;
> + int error = 0;
> +
> + bno = be32_to_cpu(rec->alloc.ar_startblock);
> + len = be32_to_cpu(rec->alloc.ar_blockcount);
> + agf = XFS_BUF_TO_AGF(bs->sc->sa.agf_bp);
> + rec_end = (unsigned long long)bno + len;
> +
> + if (bno >= mp->m_sb.sb_agblocks ||
Needs to take into account short last AG, so....
> + bno >= be32_to_cpu(agf->agf_length) ||
> + len == 0 ||
> + rec_end > mp->m_sb.sb_agblocks ||
> + rec_end > be32_to_cpu(agf->agf_length))
> + xfs_scrub_btree_set_corrupt(bs->sc, bs->cur, 0);
... it should probably just validate the agbno is good. i.e.
if (!xfs_agbno_verify(bno) || !len ||
!xfs_agbno_verify(rec_end)) {
.....
> +xfs_scrub_allocbt(
> + struct xfs_scrub_context *sc,
> + xfs_btnum_t which)
> +{
> + struct xfs_owner_info oinfo;
> + struct xfs_btree_cur *cur;
> +
> + xfs_rmap_ag_owner(&oinfo, XFS_RMAP_OWN_AG);
> + cur = which == XFS_BTNUM_BNO ? sc->sa.bno_cur : sc->sa.cnt_cur;
> + return xfs_scrub_btree(sc, cur, xfs_scrub_allocbt_helper,
> + &oinfo, NULL);
> +}
I'm assuming the owner info is for later functionality to cross
check btree blocks against the rmap btree?
Cheers,
Dave.
--
Dave Chinner
david@fromorbit.com
next prev parent reply other threads:[~2017-10-05 0:59 UTC|newest]
Thread overview: 91+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-10-03 20:40 [PATCH v11 00/25] xfs: online scrub support Darrick J. Wong
2017-10-03 20:40 ` [PATCH 01/25] xfs: create an ioctl to scrub AG metadata Darrick J. Wong
2017-10-03 20:41 ` [PATCH 02/25] xfs: dispatch metadata scrub subcommands Darrick J. Wong
2017-10-03 20:41 ` [PATCH 03/25] xfs: probe the scrub ioctl Darrick J. Wong
2017-10-03 23:32 ` Dave Chinner
2017-10-04 0:02 ` Darrick J. Wong
2017-10-04 1:56 ` Dave Chinner
2017-10-04 3:14 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 04/25] xfs: create helpers to record and deal with scrub problems Darrick J. Wong
2017-10-03 23:44 ` Dave Chinner
2017-10-04 0:56 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 05/25] xfs: create helpers to scrub a metadata btree Darrick J. Wong
2017-10-03 23:49 ` Dave Chinner
2017-10-04 0:13 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 06/25] xfs: scrub the shape of " Darrick J. Wong
2017-10-04 0:15 ` Dave Chinner
2017-10-04 3:51 ` Darrick J. Wong
2017-10-04 5:48 ` Dave Chinner
2017-10-04 17:48 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 07/25] xfs: scrub btree keys and records Darrick J. Wong
2017-10-04 20:52 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 08/25] xfs: create helpers to scan an allocation group Darrick J. Wong
2017-10-04 0:46 ` Dave Chinner
2017-10-04 3:58 ` Darrick J. Wong
2017-10-04 5:59 ` Dave Chinner
2017-10-04 17:51 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 09/25] xfs: scrub the backup superblocks Darrick J. Wong
2017-10-04 0:57 ` Dave Chinner
2017-10-04 4:06 ` Darrick J. Wong
2017-10-04 6:13 ` Dave Chinner
2017-10-04 17:56 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 10/25] xfs: scrub AGF and AGFL Darrick J. Wong
2017-10-04 1:31 ` Dave Chinner
2017-10-04 4:21 ` Darrick J. Wong
2017-10-04 6:28 ` Dave Chinner
2017-10-04 17:57 ` Darrick J. Wong
2017-10-03 20:41 ` [PATCH 11/25] xfs: scrub the AGI Darrick J. Wong
2017-10-04 1:43 ` Dave Chinner
2017-10-04 4:25 ` Darrick J. Wong
2017-10-04 6:43 ` Dave Chinner
2017-10-04 18:02 ` Darrick J. Wong
2017-10-04 22:16 ` Dave Chinner
2017-10-04 23:12 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 12/25] xfs: scrub free space btrees Darrick J. Wong
2017-10-05 0:59 ` Dave Chinner [this message]
2017-10-05 1:13 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 13/25] xfs: scrub inode btrees Darrick J. Wong
2017-10-05 2:08 ` Dave Chinner
2017-10-05 5:47 ` Darrick J. Wong
2017-10-05 7:22 ` Dave Chinner
2017-10-05 18:26 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 14/25] xfs: scrub rmap btrees Darrick J. Wong
2017-10-05 2:56 ` Dave Chinner
2017-10-05 5:02 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 15/25] xfs: scrub refcount btrees Darrick J. Wong
2017-10-05 2:59 ` Dave Chinner
2017-10-05 5:02 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 16/25] xfs: scrub inodes Darrick J. Wong
2017-10-05 4:04 ` Dave Chinner
2017-10-05 5:22 ` Darrick J. Wong
2017-10-05 7:13 ` Dave Chinner
2017-10-05 19:56 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 17/25] xfs: scrub inode block mappings Darrick J. Wong
2017-10-06 2:51 ` Dave Chinner
2017-10-06 17:00 ` Darrick J. Wong
2017-10-07 23:10 ` Dave Chinner
2017-10-08 3:54 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 18/25] xfs: scrub directory/attribute btrees Darrick J. Wong
2017-10-06 5:07 ` Dave Chinner
2017-10-06 18:30 ` Darrick J. Wong
2017-10-03 20:42 ` [PATCH 19/25] xfs: scrub directory metadata Darrick J. Wong
2017-10-06 7:07 ` Dave Chinner
2017-10-06 19:45 ` Darrick J. Wong
2017-10-06 22:16 ` Dave Chinner
2017-10-03 20:42 ` [PATCH 20/25] xfs: scrub directory freespace Darrick J. Wong
2017-10-09 1:44 ` Dave Chinner
2017-10-09 22:54 ` Darrick J. Wong
2017-10-03 20:43 ` [PATCH 21/25] xfs: scrub extended attributes Darrick J. Wong
2017-10-09 2:13 ` Dave Chinner
2017-10-09 21:14 ` Darrick J. Wong
2017-10-03 20:43 ` [PATCH 22/25] xfs: scrub symbolic links Darrick J. Wong
2017-10-09 2:17 ` Dave Chinner
2017-10-03 20:43 ` [PATCH 23/25] xfs: scrub parent pointers Darrick J. Wong
2017-10-03 20:43 ` [PATCH 24/25] xfs: scrub realtime bitmap/summary Darrick J. Wong
2017-10-09 2:28 ` Dave Chinner
2017-10-09 20:24 ` Darrick J. Wong
2017-10-03 20:43 ` [PATCH 25/25] xfs: scrub quota information Darrick J. Wong
2017-10-09 2:51 ` Dave Chinner
2017-10-09 20:03 ` Darrick J. Wong
2017-10-09 22:17 ` Dave Chinner
2017-10-09 23:08 ` 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=20171005005943.GI3666@dastard \
--to=david@fromorbit.com \
--cc=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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).