nvdimm.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
From: Shiyang Ruan <ruansy.fnst@fujitsu.com>
To: <linux-kernel@vger.kernel.org>, <linux-xfs@vger.kernel.org>,
	<nvdimm@lists.linux.dev>, <linux-mm@kvack.org>,
	<linux-fsdevel@vger.kernel.org>, <dm-devel@redhat.com>
Cc: <darrick.wong@oracle.com>, <dan.j.williams@intel.com>,
	<david@fromorbit.com>, <hch@lst.de>, <agk@redhat.com>,
	<snitzer@redhat.com>, <rgoldwyn@suse.de>
Subject: [PATCH v5 6/9] xfs: Implement ->corrupted_range() for XFS
Date: Mon, 28 Jun 2021 08:02:15 +0800	[thread overview]
Message-ID: <20210628000218.387833-7-ruansy.fnst@fujitsu.com> (raw)
In-Reply-To: <20210628000218.387833-1-ruansy.fnst@fujitsu.com>

This function is used to handle errors which may cause data lost in
filesystem.  Such as memory failure in fsdax mode.

If the rmap feature of XFS enabled, we can query it to find files and
metadata which are associated with the corrupt data.  For now all we do
is kill processes with that file mapped into their address spaces, but
future patches could actually do something about corrupt metadata.

After that, the memory failure needs to notify the processes who are
using those files.

Only support data device.  Realtime device is not supported for now.

Signed-off-by: Shiyang Ruan <ruansy.fnst@fujitsu.com>
---
 fs/xfs/xfs_fsops.c |   5 ++
 fs/xfs/xfs_mount.h |   1 +
 fs/xfs/xfs_super.c | 140 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 146 insertions(+)

diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index be9cf88d2ad7..e89ada33d8fc 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -551,6 +551,11 @@ xfs_do_force_shutdown(
 "Corruption of in-memory data detected.  Shutting down filesystem");
 		if (XFS_ERRLEVEL_HIGH <= xfs_error_level)
 			xfs_stack_trace();
+	} else if (flags & SHUTDOWN_CORRUPT_META) {
+		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_CORRUPT,
+"Corruption of on-disk metadata detected.  Shutting down filesystem");
+		if (XFS_ERRLEVEL_HIGH <= xfs_error_level)
+			xfs_stack_trace();
 	} else if (logerror) {
 		xfs_alert_tag(mp, XFS_PTAG_SHUTDOWN_LOGERROR,
 			"Log I/O Error Detected. Shutting down filesystem");
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index bb67274ee23f..c62ccf3e07d0 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -276,6 +276,7 @@ void xfs_do_force_shutdown(struct xfs_mount *mp, int flags, char *fname,
 #define SHUTDOWN_LOG_IO_ERROR	0x0002	/* write attempt to the log failed */
 #define SHUTDOWN_FORCE_UMOUNT	0x0004	/* shutdown from a forced unmount */
 #define SHUTDOWN_CORRUPT_INCORE	0x0008	/* corrupt in-memory data structures */
+#define SHUTDOWN_CORRUPT_META	0x0010  /* corrupt metadata on device */

 /*
  * Flags for xfs_mountfs
diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
index a2dab05332ac..a986ddf3b248 100644
--- a/fs/xfs/xfs_super.c
+++ b/fs/xfs/xfs_super.c
@@ -36,11 +36,18 @@
 #include "xfs_bmap_item.h"
 #include "xfs_reflink.h"
 #include "xfs_pwork.h"
+#include "xfs_alloc.h"
+#include "xfs_rmap.h"
+#include "xfs_rmap_btree.h"
+#include "xfs_rtalloc.h"
+#include "xfs_bit.h"

 #include <linux/magic.h>
 #include <linux/fs_context.h>
 #include <linux/fs_parser.h>
+#include <linux/mm.h>

+static const struct dax_holder_operations xfs_dax_holder_operations;
 static const struct super_operations xfs_super_operations;

 static struct kset *xfs_kset;		/* top-level xfs sysfs dir */
@@ -358,6 +365,7 @@ xfs_close_devices(

 		xfs_free_buftarg(mp->m_logdev_targp);
 		xfs_blkdev_put(logdev);
+		dax_set_holder(dax_logdev, NULL, NULL);
 		fs_put_dax(dax_logdev);
 	}
 	if (mp->m_rtdev_targp) {
@@ -366,9 +374,11 @@ xfs_close_devices(

 		xfs_free_buftarg(mp->m_rtdev_targp);
 		xfs_blkdev_put(rtdev);
+		dax_set_holder(dax_rtdev, NULL, NULL);
 		fs_put_dax(dax_rtdev);
 	}
 	xfs_free_buftarg(mp->m_ddev_targp);
+	dax_set_holder(dax_ddev, NULL, NULL);
 	fs_put_dax(dax_ddev);
 }

@@ -392,6 +402,7 @@ xfs_open_devices(
 	struct block_device	*logdev = NULL, *rtdev = NULL;
 	int			error;

+	dax_set_holder(dax_ddev, mp, &xfs_dax_holder_operations);
 	/*
 	 * Open real time and log devices - order is important.
 	 */
@@ -400,6 +411,9 @@ xfs_open_devices(
 		if (error)
 			goto out;
 		dax_logdev = fs_dax_get_by_bdev(logdev);
+		if (dax_logdev != dax_ddev)
+			dax_set_holder(dax_logdev, mp,
+				       &xfs_dax_holder_operations);
 	}

 	if (mp->m_rtname) {
@@ -414,6 +428,7 @@ xfs_open_devices(
 			goto out_close_rtdev;
 		}
 		dax_rtdev = fs_dax_get_by_bdev(rtdev);
+		dax_set_holder(dax_rtdev, mp, &xfs_dax_holder_operations);
 	}

 	/*
@@ -1076,6 +1091,131 @@ xfs_fs_free_cached_objects(
 	return xfs_reclaim_inodes_nr(XFS_M(sb), sc->nr_to_scan);
 }

+static int
+xfs_corrupt_helper(
+	struct xfs_btree_cur		*cur,
+	struct xfs_rmap_irec		*rec,
+	void				*data)
+{
+	struct xfs_inode		*ip;
+	struct address_space		*mapping;
+	int				error = 0;
+	int				*flags = data;
+
+	if (XFS_RMAP_NON_INODE_OWNER(rec->rm_owner) ||
+	    (rec->rm_flags & (XFS_RMAP_ATTR_FORK | XFS_RMAP_BMBT_BLOCK))) {
+		// TODO check and try to fix metadata
+		xfs_force_shutdown(cur->bc_mp, SHUTDOWN_CORRUPT_META);
+		return -EFSCORRUPTED;
+	}
+
+	/*
+	 * Get files that incore, filter out others that are not in use.
+	 */
+	error = xfs_iget(cur->bc_mp, cur->bc_tp, rec->rm_owner, XFS_IGET_INCORE,
+			0, &ip);
+	if (error)
+		return error;
+
+	mapping = VFS_I(ip)->i_mapping;
+
+#ifdef CONFIG_MEMORY_FAILURE
+	error = mf_dax_kill_procs(mapping, rec->rm_offset, *flags);
+#endif
+
+	// TODO try to fix data
+	xfs_irele(ip);
+
+	return error;
+}
+
+static loff_t
+xfs_dax_bdev_offset(
+	struct xfs_mount *mp,
+	struct dax_device *dax_dev,
+	loff_t disk_offset)
+{
+	struct block_device *bdev;
+
+	if (mp->m_ddev_targp->bt_daxdev == dax_dev)
+		bdev = mp->m_ddev_targp->bt_bdev;
+	else if (mp->m_logdev_targp->bt_daxdev == dax_dev)
+		bdev = mp->m_logdev_targp->bt_bdev;
+	else
+		bdev = mp->m_rtdev_targp->bt_bdev;
+
+	return disk_offset - (get_start_sect(bdev) << SECTOR_SHIFT);
+}
+
+static int
+xfs_dax_notify_failure(
+	struct dax_device	*dax_dev,
+	loff_t			offset,
+	size_t			len,
+	void			*data)
+{
+	struct xfs_mount	*mp = dax_get_holder(dax_dev);
+	struct xfs_trans	*tp = NULL;
+	struct xfs_btree_cur	*cur = NULL;
+	struct xfs_buf		*agf_bp = NULL;
+	struct xfs_rmap_irec	rmap_low, rmap_high;
+	loff_t 			bdev_offset = xfs_dax_bdev_offset(mp, dax_dev,
+								  offset);
+	xfs_fsblock_t		fsbno = XFS_B_TO_FSB(mp, bdev_offset);
+	xfs_filblks_t		bcnt = XFS_B_TO_FSB(mp, len);
+	xfs_agnumber_t		agno = XFS_FSB_TO_AGNO(mp, fsbno);
+	xfs_agblock_t		agbno = XFS_FSB_TO_AGBNO(mp, fsbno);
+	int			error = 0;
+
+	pr_warn("%s bno: %lld\n", __func__, bdev_offset >> SECTOR_SHIFT);
+
+	if (mp->m_rtdev_targp && mp->m_rtdev_targp->bt_daxdev == dax_dev) {
+		xfs_warn(mp, "corrupted_range support not available for realtime device!");
+		return -EOPNOTSUPP;
+	}
+	if (mp->m_logdev_targp && mp->m_logdev_targp->bt_daxdev == dax_dev &&
+	    mp->m_logdev_targp != mp->m_ddev_targp) {
+		xfs_err(mp, "ondisk log corrupt, shutting down fs!");
+		xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_META);
+		return -EFSCORRUPTED;
+	}
+
+	if (!xfs_sb_version_hasrmapbt(&mp->m_sb)) {
+		xfs_warn(mp, "corrupted_range needs rmapbt enabled!");
+		return -EOPNOTSUPP;
+	}
+
+	error = xfs_trans_alloc_empty(mp, &tp);
+	if (error)
+		return error;
+
+	error = xfs_alloc_read_agf(mp, tp, agno, 0, &agf_bp);
+	if (error)
+		goto out_cancel_tp;
+
+	cur = xfs_rmapbt_init_cursor(mp, tp, agf_bp, agno);
+
+	/* Construct a range for rmap query */
+	memset(&rmap_low, 0, sizeof(rmap_low));
+	memset(&rmap_high, 0xFF, sizeof(rmap_high));
+	rmap_low.rm_startblock = rmap_high.rm_startblock = agbno;
+	rmap_low.rm_blockcount = rmap_high.rm_blockcount = bcnt;
+
+	error = xfs_rmap_query_range(cur, &rmap_low, &rmap_high,
+				     xfs_corrupt_helper, data);
+
+	xfs_btree_del_cursor(cur, error);
+	xfs_trans_brelse(tp, agf_bp);
+
+out_cancel_tp:
+	xfs_trans_cancel(tp);
+	return error;
+}
+
+static const struct dax_holder_operations xfs_dax_holder_operations = {
+	.notify_failure = xfs_dax_notify_failure,
+};
+
 static const struct super_operations xfs_super_operations = {
 	.alloc_inode		= xfs_fs_alloc_inode,
 	.destroy_inode		= xfs_fs_destroy_inode,
--
2.32.0




  parent reply	other threads:[~2021-06-28  0:03 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-28  0:02 [PATCH v5 0/9] fsdax: introduce fs query to support reflink Shiyang Ruan
2021-06-28  0:02 ` [PATCH v5 1/9] pagemap: Introduce ->memory_failure() Shiyang Ruan
2021-06-28  0:02 ` [PATCH v5 2/9] dax: Introduce holder for dax_device Shiyang Ruan
2021-07-19 15:17   ` Christoph Hellwig
2021-07-20 10:37     ` ruansy.fnst
2021-06-28  0:02 ` [PATCH v5 3/9] mm: factor helpers for memory_failure_dev_pagemap Shiyang Ruan
2021-06-28  0:02 ` [PATCH v5 4/9] pmem,mm: Implement ->memory_failure in pmem driver Shiyang Ruan
2021-06-28  0:02 ` [PATCH v5 5/9] mm: Introduce mf_dax_kill_procs() for fsdax case Shiyang Ruan
2021-06-28  3:23   ` kernel test robot
2021-06-28 11:49   ` Matthew Wilcox
2021-06-29  7:49     ` ruansy.fnst
2021-06-29 11:46       ` Matthew Wilcox
2021-07-16  6:32         ` ruansy.fnst
2021-06-28  0:02 ` Shiyang Ruan [this message]
2021-06-28  2:02   ` [PATCH v5 6/9] xfs: Implement ->corrupted_range() for XFS kernel test robot
2021-06-28  3:05   ` kernel test robot
2021-06-28  0:02 ` [PATCH v5 7/9] dm: Introduce ->rmap() to find bdev offset Shiyang Ruan
2021-06-28  0:02 ` [PATCH v5 8/9] md: Implement dax_holder_operations Shiyang Ruan
2021-06-28  0:02 ` [PATCH v5 9/9] fs/dax: Remove useless functions Shiyang Ruan
2021-07-19 15:18   ` Christoph Hellwig

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=20210628000218.387833-7-ruansy.fnst@fujitsu.com \
    --to=ruansy.fnst@fujitsu.com \
    --cc=agk@redhat.com \
    --cc=dan.j.williams@intel.com \
    --cc=darrick.wong@oracle.com \
    --cc=david@fromorbit.com \
    --cc=dm-devel@redhat.com \
    --cc=hch@lst.de \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=linux-xfs@vger.kernel.org \
    --cc=nvdimm@lists.linux.dev \
    --cc=rgoldwyn@suse.de \
    --cc=snitzer@redhat.com \
    --subject='Re: [PATCH v5 6/9] xfs: Implement ->corrupted_range() for XFS' \
    /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

This is a public inbox, see mirroring instructions
on how to clone and mirror all data and code used for this inbox