From: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
Scrub an individual inode's block mappings to make sure they make sense.
Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
---
fs/xfs/Makefile | 1
fs/xfs/libxfs/xfs_fs.h | 5 +
fs/xfs/scrub/bmap.c | 371 ++++++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/scrub/common.h | 5 +
fs/xfs/scrub/scrub.c | 12 ++
fs/xfs/scrub/scrub.h | 3
6 files changed, 395 insertions(+), 2 deletions(-)
create mode 100644 fs/xfs/scrub/bmap.c
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index 28e14b7..5a77489 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -148,6 +148,7 @@ xfs-y += $(addprefix scrub/, \
trace.o \
agheader.o \
alloc.o \
+ bmap.o \
btree.o \
common.o \
ialloc.o \
diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h
index f8463e0..02ae58b 100644
--- a/fs/xfs/libxfs/xfs_fs.h
+++ b/fs/xfs/libxfs/xfs_fs.h
@@ -495,9 +495,12 @@ struct xfs_scrub_metadata {
#define XFS_SCRUB_TYPE_RMAPBT 9 /* reverse mapping btree */
#define XFS_SCRUB_TYPE_REFCNTBT 10 /* reference count btree */
#define XFS_SCRUB_TYPE_INODE 11 /* inode record */
+#define XFS_SCRUB_TYPE_BMBTD 12 /* data fork block mapping */
+#define XFS_SCRUB_TYPE_BMBTA 13 /* attr fork block mapping */
+#define XFS_SCRUB_TYPE_BMBTC 14 /* CoW fork block mapping */
/* Number of scrub subcommands. */
-#define XFS_SCRUB_TYPE_NR 12
+#define XFS_SCRUB_TYPE_NR 15
/* i: Repair this metadata. */
#define XFS_SCRUB_IFLAG_REPAIR (1 << 0)
diff --git a/fs/xfs/scrub/bmap.c b/fs/xfs/scrub/bmap.c
new file mode 100644
index 0000000..57f1bbe
--- /dev/null
+++ b/fs/xfs/scrub/bmap.c
@@ -0,0 +1,371 @@
+/*
+ * Copyright (C) 2017 Oracle. All Rights Reserved.
+ *
+ * Author: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; either version 2
+ * of the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA.
+ */
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_shared.h"
+#include "xfs_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_mount.h"
+#include "xfs_defer.h"
+#include "xfs_btree.h"
+#include "xfs_bit.h"
+#include "xfs_log_format.h"
+#include "xfs_trans.h"
+#include "xfs_sb.h"
+#include "xfs_inode.h"
+#include "xfs_inode_fork.h"
+#include "xfs_bmap.h"
+#include "xfs_bmap_util.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_rmap.h"
+#include "scrub/xfs_scrub.h"
+#include "scrub/scrub.h"
+#include "scrub/common.h"
+#include "scrub/btree.h"
+#include "scrub/trace.h"
+
+/* Set us up with an inode's bmap. */
+STATIC int
+__xfs_scrub_setup_inode_bmap(
+ struct xfs_scrub_context *sc,
+ struct xfs_inode *ip,
+ bool flush_data)
+{
+ struct xfs_mount *mp = sc->mp;
+ int error;
+
+ error = xfs_scrub_get_inode(sc, ip);
+ if (error)
+ return error;
+
+ sc->ilock_flags = XFS_IOLOCK_EXCL | XFS_MMAPLOCK_EXCL;
+ xfs_ilock(sc->ip, sc->ilock_flags);
+
+ /*
+ * We don't want any ephemeral data fork updates sitting around
+ * while we inspect block mappings, so wait for directio to finish
+ * and flush dirty data if we have delalloc reservations.
+ */
+ if (S_ISREG(VFS_I(sc->ip)->i_mode) && flush_data) {
+ inode_dio_wait(VFS_I(sc->ip));
+ error = filemap_write_and_wait(VFS_I(sc->ip)->i_mapping);
+ if (error)
+ goto out_unlock;
+ error = invalidate_inode_pages2(VFS_I(sc->ip)->i_mapping);
+ if (error)
+ goto out_unlock;
+ }
+
+ /* Got the inode, lock it and we're ready to go. */
+ error = xfs_scrub_trans_alloc(sc->sm, mp, &sc->tp);
+ if (error)
+ goto out_unlock;
+ sc->ilock_flags |= XFS_ILOCK_EXCL;
+ xfs_ilock(sc->ip, XFS_ILOCK_EXCL);
+
+ return 0;
+out_unlock:
+ xfs_iunlock(sc->ip, sc->ilock_flags);
+ if (sc->ip != ip)
+ iput(VFS_I(sc->ip));
+ sc->ip = NULL;
+ return error;
+}
+
+/* Set us up to scrub the data fork. */
+int
+xfs_scrub_setup_inode_bmap_data(
+ struct xfs_scrub_context *sc,
+ struct xfs_inode *ip)
+{
+ return __xfs_scrub_setup_inode_bmap(sc, ip, true);
+}
+
+/* Set us up to scrub the attr or CoW fork. */
+int
+xfs_scrub_setup_inode_bmap(
+ struct xfs_scrub_context *sc,
+ struct xfs_inode *ip)
+{
+ return __xfs_scrub_setup_inode_bmap(sc, ip, false);
+}
+
+/*
+ * Inode fork block mapping (BMBT) scrubber.
+ * More complex than the others because we have to scrub
+ * all the extents regardless of whether or not the fork
+ * is in btree format.
+ */
+
+struct xfs_scrub_bmap_info {
+ struct xfs_scrub_context *sc;
+ xfs_daddr_t eofs;
+ xfs_fileoff_t lastoff;
+ bool is_rt;
+ bool is_shared;
+ int whichfork;
+};
+
+/* Scrub a single extent record. */
+STATIC int
+xfs_scrub_bmap_extent(
+ struct xfs_inode *ip,
+ struct xfs_btree_cur *cur,
+ struct xfs_scrub_bmap_info *info,
+ struct xfs_bmbt_irec *irec)
+{
+ struct xfs_scrub_ag sa = { 0 };
+ struct xfs_mount *mp = info->sc->mp;
+ struct xfs_buf *bp = NULL;
+ xfs_daddr_t daddr;
+ xfs_daddr_t dlen;
+ xfs_fsblock_t bno;
+ xfs_agnumber_t agno;
+ int error = 0;
+
+ if (cur)
+ xfs_btree_get_block(cur, 0, &bp);
+
+ if (irec->br_startoff < info->lastoff ||
+ irec->br_startblock == HOLESTARTBLOCK ||
+ isnullstartblock(irec->br_startblock))
+ xfs_scrub_fblock_set_corrupt(info->sc, info->whichfork,
+ irec->br_startoff);
+
+ /* Actual mapping, so check the block ranges. */
+ if (info->is_rt) {
+ daddr = XFS_FSB_TO_BB(mp, irec->br_startblock);
+ agno = NULLAGNUMBER;
+ bno = irec->br_startblock;
+ } else {
+ daddr = XFS_FSB_TO_DADDR(mp, irec->br_startblock);
+ agno = XFS_FSB_TO_AGNO(mp, irec->br_startblock);
+ if (agno >= mp->m_sb.sb_agcount) {
+ xfs_scrub_fblock_set_corrupt(info->sc, info->whichfork,
+ irec->br_startoff);
+ goto out;
+ }
+ bno = XFS_FSB_TO_AGBNO(mp, irec->br_startblock);
+ if (bno >= mp->m_sb.sb_agblocks)
+ xfs_scrub_fblock_set_corrupt(info->sc,
+ info->whichfork,
+ irec->br_startoff);
+ }
+ dlen = XFS_FSB_TO_BB(mp, irec->br_blockcount);
+ if (irec->br_blockcount <= 0 ||
+ irec->br_blockcount > MAXEXTLEN ||
+ daddr >= info->eofs ||
+ daddr + dlen > info->eofs)
+ xfs_scrub_fblock_set_corrupt(info->sc, info->whichfork,
+ irec->br_startoff);
+
+ if (irec->br_state == XFS_EXT_UNWRITTEN &&
+ !xfs_sb_version_hasextflgbit(&mp->m_sb))
+ xfs_scrub_fblock_set_corrupt(info->sc, info->whichfork,
+ irec->br_startoff);
+
+ /* Set ourselves up for cross-referencing later. */
+ if (!info->is_rt) {
+ error = xfs_scrub_ag_init(info->sc, agno, &sa);
+ if (!xfs_scrub_fblock_op_ok(info->sc, info->whichfork,
+ irec->br_startoff, &error))
+ goto out;
+ }
+
+ xfs_scrub_ag_free(info->sc, &sa);
+out:
+ info->lastoff = irec->br_startoff + irec->br_blockcount;
+ return error;
+}
+
+/* Scrub a bmbt record. */
+STATIC int
+xfs_scrub_bmapbt_helper(
+ struct xfs_scrub_btree *bs,
+ union xfs_btree_rec *rec)
+{
+ struct xfs_bmbt_rec_host ihost;
+ struct xfs_bmbt_irec irec;
+ struct xfs_scrub_bmap_info *info = bs->private;
+ struct xfs_inode *ip = bs->cur->bc_private.b.ip;
+ struct xfs_buf *bp = NULL;
+ struct xfs_btree_block *block;
+ uint64_t owner;
+ int i;
+
+ /*
+ * Check the owners of the btree blocks up to the level below
+ * the root since the verifiers don't do that.
+ */
+ if (xfs_sb_version_hascrc(&bs->cur->bc_mp->m_sb) &&
+ bs->cur->bc_ptrs[0] == 1) {
+ for (i = 0; i < bs->cur->bc_nlevels - 1; i++) {
+ block = xfs_btree_get_block(bs->cur, i, &bp);
+ owner = be64_to_cpu(block->bb_u.l.bb_owner);
+ if (owner != ip->i_ino)
+ xfs_scrub_fblock_set_corrupt(bs->sc,
+ info->whichfork, 0);
+ }
+ }
+
+ /* Set up the in-core record and scrub it. */
+ ihost.l0 = be64_to_cpu(rec->bmbt.l0);
+ ihost.l1 = be64_to_cpu(rec->bmbt.l1);
+ xfs_bmbt_get_all(&ihost, &irec);
+ return xfs_scrub_bmap_extent(ip, bs->cur, info, &irec);
+}
+
+/* Scrub an inode fork's block mappings. */
+STATIC int
+xfs_scrub_bmap(
+ struct xfs_scrub_context *sc,
+ int whichfork)
+{
+ struct xfs_bmbt_irec irec;
+ struct xfs_scrub_bmap_info info = {0};
+ struct xfs_owner_info oinfo;
+ struct xfs_mount *mp = sc->mp;
+ struct xfs_inode *ip = sc->ip;
+ struct xfs_ifork *ifp;
+ struct xfs_btree_cur *cur;
+ xfs_fileoff_t endoff;
+ xfs_extnum_t idx;
+ bool found;
+ int error = 0;
+
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+
+ info.is_rt = whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip);
+ info.eofs = XFS_FSB_TO_BB(mp, info.is_rt ? mp->m_sb.sb_rblocks :
+ mp->m_sb.sb_dblocks);
+ info.whichfork = whichfork;
+ info.is_shared = whichfork == XFS_DATA_FORK && xfs_is_reflink_inode(ip);
+ info.sc = sc;
+
+ switch (whichfork) {
+ case XFS_COW_FORK:
+ /* Non-existent CoW forks are ignorable. */
+ if (!ifp)
+ goto out_unlock;
+ /* No CoW forks on non-reflink inodes/filesystems. */
+ if (!xfs_is_reflink_inode(ip)) {
+ xfs_scrub_ino_set_corrupt(sc, sc->ip->i_ino, NULL);
+ goto out_unlock;
+ }
+ break;
+ case XFS_ATTR_FORK:
+ if (!ifp)
+ goto out_unlock;
+ if (!xfs_sb_version_hasattr(&mp->m_sb) &&
+ !xfs_sb_version_hasattr2(&mp->m_sb))
+ xfs_scrub_ino_set_corrupt(sc, sc->ip->i_ino, NULL);
+ break;
+ }
+
+ /* Check the fork values */
+ switch (XFS_IFORK_FORMAT(ip, whichfork)) {
+ case XFS_DINODE_FMT_UUID:
+ case XFS_DINODE_FMT_DEV:
+ case XFS_DINODE_FMT_LOCAL:
+ /* No mappings to check. */
+ goto out_unlock;
+ case XFS_DINODE_FMT_EXTENTS:
+ if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+ xfs_scrub_fblock_set_corrupt(sc, whichfork, 0);
+ goto out_unlock;
+ }
+ break;
+ case XFS_DINODE_FMT_BTREE:
+ ASSERT(whichfork != XFS_COW_FORK);
+
+ /* Scan the btree records. */
+ cur = xfs_bmbt_init_cursor(mp, sc->tp, ip, whichfork);
+ xfs_rmap_ino_bmbt_owner(&oinfo, ip->i_ino, whichfork);
+ error = xfs_scrub_btree(sc, cur, xfs_scrub_bmapbt_helper,
+ &oinfo, &info);
+ xfs_btree_del_cursor(cur, error ? XFS_BTREE_ERROR :
+ XFS_BTREE_NOERROR);
+ if (error == -EDEADLOCK)
+ return error;
+ else if (error)
+ goto out_unlock;
+ break;
+ default:
+ xfs_scrub_fblock_set_corrupt(sc, whichfork, 0);
+ goto out_unlock;
+ }
+
+ /* Extent data is in memory, so scrub that. */
+
+ /* Find the offset of the last extent in the mapping. */
+ error = xfs_bmap_last_offset(ip, &endoff, whichfork);
+ if (!xfs_scrub_fblock_op_ok(sc, whichfork, 0, &error))
+ goto out_unlock;
+
+ /* Scrub extent records. */
+ info.lastoff = 0;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ for (found = xfs_iext_lookup_extent(ip, ifp, 0, &idx, &irec);
+ found != 0;
+ found = xfs_iext_get_extent(ifp, ++idx, &irec)) {
+ if (xfs_scrub_should_terminate(sc, &error))
+ break;
+ if (isnullstartblock(irec.br_startblock))
+ continue;
+ if (irec.br_startoff >= endoff) {
+ xfs_scrub_fblock_set_corrupt(sc, whichfork,
+ irec.br_startoff);
+ goto out_unlock;
+ }
+ error = xfs_scrub_bmap_extent(ip, NULL, &info, &irec);
+ if (error == -EDEADLOCK)
+ return error;
+ }
+
+out_unlock:
+ return error;
+}
+
+/* Scrub an inode's data fork. */
+int
+xfs_scrub_bmap_data(
+ struct xfs_scrub_context *sc)
+{
+ return xfs_scrub_bmap(sc, XFS_DATA_FORK);
+}
+
+/* Scrub an inode's attr fork. */
+int
+xfs_scrub_bmap_attr(
+ struct xfs_scrub_context *sc)
+{
+ return xfs_scrub_bmap(sc, XFS_ATTR_FORK);
+}
+
+/* Scrub an inode's CoW fork. */
+int
+xfs_scrub_bmap_cow(
+ struct xfs_scrub_context *sc)
+{
+ if (!xfs_is_reflink_inode(sc->ip))
+ return -ENOENT;
+
+ return xfs_scrub_bmap(sc, XFS_COW_FORK);
+}
diff --git a/fs/xfs/scrub/common.h b/fs/xfs/scrub/common.h
index c8143e8..3e87e3a 100644
--- a/fs/xfs/scrub/common.h
+++ b/fs/xfs/scrub/common.h
@@ -90,7 +90,10 @@ int xfs_scrub_setup_ag_refcountbt(struct xfs_scrub_context *sc,
struct xfs_inode *ip);
int xfs_scrub_setup_inode(struct xfs_scrub_context *sc,
struct xfs_inode *ip);
-
+int xfs_scrub_setup_inode_bmap(struct xfs_scrub_context *sc,
+ struct xfs_inode *ip);
+int xfs_scrub_setup_inode_bmap_data(struct xfs_scrub_context *sc,
+ struct xfs_inode *ip);
void xfs_scrub_ag_free(struct xfs_scrub_context *sc, struct xfs_scrub_ag *sa);
int xfs_scrub_ag_init(struct xfs_scrub_context *sc, xfs_agnumber_t agno,
diff --git a/fs/xfs/scrub/scrub.c b/fs/xfs/scrub/scrub.c
index f014ef0..b10d627 100644
--- a/fs/xfs/scrub/scrub.c
+++ b/fs/xfs/scrub/scrub.c
@@ -209,6 +209,18 @@ static const struct xfs_scrub_meta_ops meta_scrub_ops[] = {
.setup = xfs_scrub_setup_inode,
.scrub = xfs_scrub_inode,
},
+ { /* inode data fork */
+ .setup = xfs_scrub_setup_inode_bmap_data,
+ .scrub = xfs_scrub_bmap_data,
+ },
+ { /* inode attr fork */
+ .setup = xfs_scrub_setup_inode_bmap,
+ .scrub = xfs_scrub_bmap_attr,
+ },
+ { /* inode CoW fork */
+ .setup = xfs_scrub_setup_inode_bmap,
+ .scrub = xfs_scrub_bmap_cow,
+ },
};
/* This isn't a stable feature, warn once per day. */
diff --git a/fs/xfs/scrub/scrub.h b/fs/xfs/scrub/scrub.h
index ec635d4..8920ccf 100644
--- a/fs/xfs/scrub/scrub.h
+++ b/fs/xfs/scrub/scrub.h
@@ -79,5 +79,8 @@ int xfs_scrub_finobt(struct xfs_scrub_context *sc);
int xfs_scrub_rmapbt(struct xfs_scrub_context *sc);
int xfs_scrub_refcountbt(struct xfs_scrub_context *sc);
int xfs_scrub_inode(struct xfs_scrub_context *sc);
+int xfs_scrub_bmap_data(struct xfs_scrub_context *sc);
+int xfs_scrub_bmap_attr(struct xfs_scrub_context *sc);
+int xfs_scrub_bmap_cow(struct xfs_scrub_context *sc);
#endif /* __XFS_SCRUB_SCRUB_H__ */
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
