From: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
Add a field to the superblock to record the rt rmapbt inode and load
it at mount time. The rtrmapbt inode will have a unique extent format
code, which means that we also have to update the inode validation and
flush routines to look for it.
Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
---
fs/xfs/libxfs/xfs_format.h | 10 +++++++++-
fs/xfs/libxfs/xfs_inode_buf.c | 6 ++++++
fs/xfs/libxfs/xfs_inode_fork.c | 11 +++++++++++
fs/xfs/libxfs/xfs_rtrmap_btree.c | 2 +-
fs/xfs/libxfs/xfs_sb.c | 2 ++
fs/xfs/libxfs/xfs_types.c | 1 +
fs/xfs/xfs_inode.c | 9 ++++++++-
fs/xfs/xfs_inode_item.c | 2 ++
fs/xfs/xfs_log_recover.c | 14 +++++++++++++-
fs/xfs/xfs_mount.h | 1 +
fs/xfs/xfs_ondisk.h | 2 +-
fs/xfs/xfs_rtalloc.c | 21 +++++++++++++++++++++
12 files changed, 76 insertions(+), 5 deletions(-)
diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h
index ad767b27dbb3..bb5ea9d46c07 100644
--- a/fs/xfs/libxfs/xfs_format.h
+++ b/fs/xfs/libxfs/xfs_format.h
@@ -174,6 +174,7 @@ typedef struct xfs_sb {
xfs_ino_t sb_pquotino; /* project quota inode */
xfs_lsn_t sb_lsn; /* last write sequence */
uuid_t sb_meta_uuid; /* metadata file system unique id */
+ xfs_ino_t sb_rrmapino; /* realtime reverse map inode */
/* must be padded to 64 bit alignment */
} xfs_sb_t;
@@ -261,6 +262,7 @@ typedef struct xfs_dsb {
__be64 sb_pquotino; /* project quota inode */
__be64 sb_lsn; /* last write sequence */
uuid_t sb_meta_uuid; /* metadata file system unique id */
+ __be64 sb_rrmapino; /* realtime reverse map inode */
/* must be padded to 64 bit alignment */
} xfs_dsb_t;
@@ -544,6 +546,11 @@ static inline bool xfs_sb_version_hasrmapbt(struct xfs_sb *sbp)
(sbp->sb_features_ro_compat & XFS_SB_FEAT_RO_COMPAT_RMAPBT);
}
+static inline bool xfs_sb_version_hasrtrmapbt(struct xfs_sb *sbp)
+{
+ return sbp->sb_rblocks > 0 && xfs_sb_version_hasrmapbt(sbp);
+}
+
static inline bool xfs_sb_version_hasreflink(struct xfs_sb *sbp)
{
return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5 &&
@@ -926,7 +933,8 @@ typedef enum xfs_dinode_fmt {
XFS_DINODE_FMT_LOCAL, /* bulk data */
XFS_DINODE_FMT_EXTENTS, /* struct xfs_bmbt_rec */
XFS_DINODE_FMT_BTREE, /* struct xfs_bmdr_block */
- XFS_DINODE_FMT_UUID /* added long ago, but never used */
+ XFS_DINODE_FMT_UUID, /* added long ago, but never used */
+ XFS_DINODE_FMT_RMAP, /* reverse mapping btree */
} xfs_dinode_fmt_t;
/*
diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c
index 30d1d60f1d46..e33e3fa22b14 100644
--- a/fs/xfs/libxfs/xfs_inode_buf.c
+++ b/fs/xfs/libxfs/xfs_inode_buf.c
@@ -480,6 +480,12 @@ xfs_dinode_verify(
return __this_address;
break;
case S_IFREG:
+ if (dip->di_format == XFS_DINODE_FMT_RMAP) {
+ if (ino != mp->m_sb.sb_rrmapino)
+ return __this_address;
+ break;
+ }
+ /* fall through */
case S_IFLNK:
case S_IFDIR:
fa = xfs_dinode_verify_fork(dip, mp, XFS_DATA_FORK);
diff --git a/fs/xfs/libxfs/xfs_inode_fork.c b/fs/xfs/libxfs/xfs_inode_fork.c
index cfb68d4befc4..d1401cbdd266 100644
--- a/fs/xfs/libxfs/xfs_inode_fork.c
+++ b/fs/xfs/libxfs/xfs_inode_fork.c
@@ -76,6 +76,13 @@ xfs_iformat_fork(
case XFS_DINODE_FMT_BTREE:
error = xfs_iformat_btree(ip, dip, XFS_DATA_FORK);
break;
+ case XFS_DINODE_FMT_RMAP:
+ if (!xfs_sb_version_hasrtrmapbt(&ip->i_mount->m_sb))
+ return -EFSCORRUPTED;
+ if (ip->i_ino != ip->i_mount->m_sb.sb_rrmapino)
+ return -EFSCORRUPTED;
+ /* to be implemented later */
+ break;
default:
return -EFSCORRUPTED;
}
@@ -525,6 +532,10 @@ xfs_iflush_fork(
}
break;
+ case XFS_DINODE_FMT_RMAP:
+ /* to be implemented later */
+ break;
+
default:
ASSERT(0);
break;
diff --git a/fs/xfs/libxfs/xfs_rtrmap_btree.c b/fs/xfs/libxfs/xfs_rtrmap_btree.c
index 4529dbc6545c..ce9932397aa1 100644
--- a/fs/xfs/libxfs/xfs_rtrmap_btree.c
+++ b/fs/xfs/libxfs/xfs_rtrmap_btree.c
@@ -331,7 +331,7 @@ xfs_rtrmapbt_verify(
if (!xfs_sb_version_hasrmapbt(&mp->m_sb))
return __this_address;
- fa = xfs_btree_lblock_v5hdr_verify(bp, XFS_RMAP_OWN_UNKNOWN);
+ fa = xfs_btree_lblock_v5hdr_verify(bp, mp->m_sb.sb_rrmapino);
if (fa)
return fa;
level = be16_to_cpu(block->bb_level);
diff --git a/fs/xfs/libxfs/xfs_sb.c b/fs/xfs/libxfs/xfs_sb.c
index 733bb7e5409c..e21fc6e0d2aa 100644
--- a/fs/xfs/libxfs/xfs_sb.c
+++ b/fs/xfs/libxfs/xfs_sb.c
@@ -522,6 +522,7 @@ __xfs_sb_from_disk(
uuid_copy(&to->sb_meta_uuid, &from->sb_meta_uuid);
else
uuid_copy(&to->sb_meta_uuid, &from->sb_uuid);
+ to->sb_rrmapino = be64_to_cpu(from->sb_rrmapino);
/* Convert on-disk flags to in-memory flags? */
if (convert_xquota)
xfs_sb_quota_from_disk(to);
@@ -665,6 +666,7 @@ xfs_sb_to_disk(
to->sb_lsn = cpu_to_be64(from->sb_lsn);
if (xfs_sb_version_hasmetauuid(from))
uuid_copy(&to->sb_meta_uuid, &from->sb_meta_uuid);
+ to->sb_rrmapino = cpu_to_be64(from->sb_rrmapino);
}
}
diff --git a/fs/xfs/libxfs/xfs_types.c b/fs/xfs/libxfs/xfs_types.c
index 33a5ca346baf..3679ca947954 100644
--- a/fs/xfs/libxfs/xfs_types.c
+++ b/fs/xfs/libxfs/xfs_types.c
@@ -142,6 +142,7 @@ xfs_internal_inum(
xfs_ino_t ino)
{
return ino == mp->m_sb.sb_rbmino || ino == mp->m_sb.sb_rsumino ||
+ ino == mp->m_sb.sb_rrmapino ||
(xfs_sb_version_hasquota(&mp->m_sb) &&
xfs_is_quota_inode(&mp->m_sb, ino));
}
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 0e142787138c..f53e9488e796 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -3524,7 +3524,14 @@ xfs_iflush_int(
__func__, ip->i_ino, be16_to_cpu(dip->di_magic), dip);
goto corrupt_out;
}
- if (S_ISREG(VFS_I(ip)->i_mode)) {
+ if (ip->i_ino == mp->m_sb.sb_rrmapino) {
+ if (ip->i_d.di_format != XFS_DINODE_FMT_RMAP) {
+ xfs_alert_tag(mp, XFS_PTAG_IFLUSH,
+ "%s: Bad rtrmap inode %llu, ptr 0x%p",
+ __func__, ip->i_ino, ip);
+ goto corrupt_out;
+ }
+ } else if (S_ISREG(VFS_I(ip)->i_mode)) {
if (XFS_TEST_ERROR(
(ip->i_d.di_format != XFS_DINODE_FMT_EXTENTS) &&
(ip->i_d.di_format != XFS_DINODE_FMT_BTREE),
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index fa1c4fe2ffbf..1d2eb7bf9e04 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -46,6 +46,7 @@ xfs_inode_item_data_fork_size(
}
break;
case XFS_DINODE_FMT_BTREE:
+ case XFS_DINODE_FMT_RMAP:
if ((iip->ili_fields & XFS_ILOG_DBROOT) &&
ip->i_df.if_broot_bytes > 0) {
*nbytes += ip->i_df.if_broot_bytes;
@@ -166,6 +167,7 @@ xfs_inode_item_format_data_fork(
}
break;
case XFS_DINODE_FMT_BTREE:
+ case XFS_DINODE_FMT_RMAP:
iip->ili_fields &=
~(XFS_ILOG_DDATA | XFS_ILOG_DEXT | XFS_ILOG_DEV);
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index f269c96cb0ab..95af44487b6c 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -3123,7 +3123,19 @@ xlog_recover_inode_pass2(
/* Take the opportunity to reset the flush iteration count */
ldip->di_flushiter = 0;
- if (unlikely(S_ISREG(ldip->di_mode))) {
+ if (in_f->ilf_ino == mp->m_sb.sb_rrmapino) {
+ if (ldip->di_format != XFS_DINODE_FMT_RMAP) {
+ XFS_CORRUPTION_ERROR("xlog_recover_inode_pass2(3)",
+ XFS_ERRLEVEL_LOW, mp, ldip,
+ sizeof(*ldip));
+ xfs_alert(mp,
+ "%s: Bad rtrmapbt inode log record, rec ptr 0x%p, "
+ "ino ptr = 0x%p, ino bp = 0x%p, ino %Ld",
+ __func__, item, dip, bp, in_f->ilf_ino);
+ error = -EFSCORRUPTED;
+ goto out_release;
+ }
+ } else if (unlikely(S_ISREG(ldip->di_mode))) {
if ((ldip->di_format != XFS_DINODE_FMT_EXTENTS) &&
(ldip->di_format != XFS_DINODE_FMT_BTREE)) {
XFS_CORRUPTION_ERROR("xlog_recover_inode_pass2(3)",
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 5a4df6b860e2..712bec475c5d 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -91,6 +91,7 @@ typedef struct xfs_mount {
uint m_rsumsize; /* size of rt summary, bytes */
struct xfs_inode *m_rbmip; /* pointer to bitmap inode */
struct xfs_inode *m_rsumip; /* pointer to summary inode */
+ struct xfs_inode *m_rrmapip; /* pointer to rmap inode */
struct xfs_inode *m_rootip; /* pointer to root directory */
struct xfs_quotainfo *m_quotainfo; /* disk quota information */
xfs_buftarg_t *m_ddev_targp; /* saves taking the address */
diff --git a/fs/xfs/xfs_ondisk.h b/fs/xfs/xfs_ondisk.h
index 23c9b6f047c5..0840ec7e58aa 100644
--- a/fs/xfs/xfs_ondisk.h
+++ b/fs/xfs/xfs_ondisk.h
@@ -33,7 +33,7 @@ xfs_check_ondisk_structs(void)
XFS_CHECK_STRUCT_SIZE(struct xfs_dinode, 176);
XFS_CHECK_STRUCT_SIZE(struct xfs_disk_dquot, 104);
XFS_CHECK_STRUCT_SIZE(struct xfs_dqblk, 136);
- XFS_CHECK_STRUCT_SIZE(struct xfs_dsb, 264);
+ XFS_CHECK_STRUCT_SIZE(struct xfs_dsb, 272);
XFS_CHECK_STRUCT_SIZE(struct xfs_dsymlink_hdr, 56);
XFS_CHECK_STRUCT_SIZE(struct xfs_inobt_key, 4);
XFS_CHECK_STRUCT_SIZE(struct xfs_inobt_rec, 16);
diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c
index 926ed314ffba..1ac69ff79a22 100644
--- a/fs/xfs/xfs_rtalloc.c
+++ b/fs/xfs/xfs_rtalloc.c
@@ -1211,6 +1211,25 @@ xfs_rtmount_inodes(
return error;
}
ASSERT(mp->m_rsumip != NULL);
+
+ /* If we have rmap and a realtime device, look for the rtrmapbt. */
+ if (xfs_sb_version_hasrmapbt(&mp->m_sb) && mp->m_sb.sb_rblocks > 0) {
+ ASSERT(sbp->sb_rrmapino != NULLFSINO);
+ error = xfs_iget(mp, NULL, sbp->sb_rrmapino, 0, 0,
+ &mp->m_rrmapip);
+ if (error) {
+ if (mp->m_rrmapip)
+ xfs_irele(mp->m_rrmapip);
+ return error;
+ }
+ ASSERT(mp->m_rrmapip != NULL);
+ if (mp->m_rrmapip->i_d.di_format != XFS_DINODE_FMT_RMAP) {
+ xfs_irele(mp->m_rrmapip);
+ mp->m_rrmapip = NULL;
+ return -EFSCORRUPTED;
+ }
+ }
+
return 0;
}
@@ -1218,6 +1237,8 @@ void
xfs_rtunmount_inodes(
struct xfs_mount *mp)
{
+ if (mp->m_rrmapip)
+ xfs_irele(mp->m_rrmapip);
if (mp->m_rbmip)
xfs_irele(mp->m_rbmip);
if (mp->m_rsumip)
