xfs: add CRC checks to the superblock

With the addition of CRCs, there is such a wide and varied change to
the on disk format that it makes sense to bump the superblock
version number rather than try to use feature bits for all the new
functionality.

This commit introduces all the new superblock fields needed for all
the new functionality: feature masks similar to ext4, separate
project quota inodes, a LSN field for recovery and the CRC field.

This commit does not bump the superblock version number, however.
That will be done as a separate commit at the end of the series
after all the new functionality is present so we switch it all on in
one commit. This means that we can slowly introduce the changes
without them being active and hence maintain bisectability of the
tree.

This patch is based on a patch originally written by myself back
from SGI days, which was subsequently modified by Christoph Hellwig.
There is relatively little of that patch remaining, but the history
of the patch still should be acknowledged here.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
Signed-off-by: Ben Myers <bpm@sgi.com>

diff --git a/include/xfs_mount.h b/include/xfs_mount.h
index 28bbf46a0249754d7a8fd852efa30882e05975fb..68c02a9c827c6b0c4b071d5e62eed8839c0a2391 100644 (file)
--- a/include/xfs_mount.h
+++ b/include/xfs_mount.h
@@ -391,6 +391,7 @@ struct xfs_perag *xfs_perag_get_tag(struct xfs_mount *mp, xfs_agnumber_t agno,
                                        int tag);
 void   xfs_perag_put(struct xfs_perag *pag);
 
+extern void    xfs_sb_calc_crc(struct xfs_buf  *);
 extern void    xfs_mod_sb(struct xfs_trans *, __int64_t);
 extern int     xfs_initialize_perag(struct xfs_mount *, xfs_agnumber_t,
                                        xfs_agnumber_t *);

diff --git a/include/xfs_sb.h b/include/xfs_sb.h
index 6a7f8b0e49244c56c947817cde59eef6cf006acd..d6709db1931ed4f854a55fa807bc9e2a01eb6f28 100644 (file)
--- a/include/xfs_sb.h
+++ b/include/xfs_sb.h
@@ -32,6 +32,7 @@ struct xfs_mount;
 #define        XFS_SB_VERSION_2        2               /* 6.2 - attributes */
 #define        XFS_SB_VERSION_3        3               /* 6.2 - new inode version */
 #define        XFS_SB_VERSION_4        4               /* 6.2+ - bitmask version */
+#define        XFS_SB_VERSION_5        5               /* CRC enabled filesystem */
 #define        XFS_SB_VERSION_NUMBITS          0x000f
 #define        XFS_SB_VERSION_ALLFBITS         0xfff0
 #define        XFS_SB_VERSION_SASHFBITS        0xf000
@@ -161,6 +162,18 @@ typedef struct xfs_sb {
         */
        __uint32_t      sb_bad_features2;
 
+       /* version 5 superblock fields start here */
+
+       /* feature masks */
+       __uint32_t      sb_features_compat;
+       __uint32_t      sb_features_ro_compat;
+       __uint32_t      sb_features_incompat;
+
+       __uint32_t      sb_crc;         /* superblock crc */
+
+       xfs_ino_t       sb_pquotino;    /* project quota inode */
+       xfs_lsn_t       sb_lsn;         /* last write sequence */
+
        /* must be padded to 64 bit alignment */
 } xfs_sb_t;
 
@@ -229,7 +242,19 @@ typedef struct xfs_dsb {
         * for features2 bits. Easiest just to mark it bad and not use
         * it for anything else.
         */
-       __be32  sb_bad_features2;
+       __be32          sb_bad_features2;
+
+       /* version 5 superblock fields start here */
+
+       /* feature masks */
+       __be32          sb_features_compat;
+       __be32          sb_features_ro_compat;
+       __be32          sb_features_incompat;
+
+       __le32          sb_crc;         /* superblock crc */
+
+       __be64          sb_pquotino;    /* project quota inode */
+       __be64          sb_lsn;         /* last write sequence */
 
        /* must be padded to 64 bit alignment */
 } xfs_dsb_t;
@@ -250,7 +275,9 @@ typedef enum {
        XFS_SBS_GQUOTINO, XFS_SBS_QFLAGS, XFS_SBS_FLAGS, XFS_SBS_SHARED_VN,
        XFS_SBS_INOALIGNMT, XFS_SBS_UNIT, XFS_SBS_WIDTH, XFS_SBS_DIRBLKLOG,
        XFS_SBS_LOGSECTLOG, XFS_SBS_LOGSECTSIZE, XFS_SBS_LOGSUNIT,
-       XFS_SBS_FEATURES2, XFS_SBS_BAD_FEATURES2,
+       XFS_SBS_FEATURES2, XFS_SBS_BAD_FEATURES2, XFS_SBS_FEATURES_COMPAT,
+       XFS_SBS_FEATURES_RO_COMPAT, XFS_SBS_FEATURES_INCOMPAT, XFS_SBS_CRC,
+       XFS_SBS_PQUOTINO, XFS_SBS_LSN,
        XFS_SBS_FIELDCOUNT
 } xfs_sb_field_t;
 
@@ -276,6 +303,11 @@ typedef enum {
 #define XFS_SB_FDBLOCKS                XFS_SB_MVAL(FDBLOCKS)
 #define XFS_SB_FEATURES2       XFS_SB_MVAL(FEATURES2)
 #define XFS_SB_BAD_FEATURES2   XFS_SB_MVAL(BAD_FEATURES2)
+#define XFS_SB_FEATURES_COMPAT XFS_SB_MVAL(FEATURES_COMPAT)
+#define XFS_SB_FEATURES_RO_COMPAT XFS_SB_MVAL(FEATURES_RO_COMPAT)
+#define XFS_SB_FEATURES_INCOMPAT XFS_SB_MVAL(FEATURES_INCOMPAT)
+#define XFS_SB_CRC             XFS_SB_MVAL(CRC)
+#define XFS_SB_PQUOTINO                XFS_SB_MVAL(PQUOTINO)
 #define        XFS_SB_NUM_BITS         ((int)XFS_SBS_FIELDCOUNT)
 #define        XFS_SB_ALL_BITS         ((1LL << XFS_SB_NUM_BITS) - 1)
 #define        XFS_SB_MOD_BITS         \
@@ -283,7 +315,8 @@ typedef enum {
         XFS_SB_VERSIONNUM | XFS_SB_UQUOTINO | XFS_SB_GQUOTINO | \
         XFS_SB_QFLAGS | XFS_SB_SHARED_VN | XFS_SB_UNIT | XFS_SB_WIDTH | \
         XFS_SB_ICOUNT | XFS_SB_IFREE | XFS_SB_FDBLOCKS | XFS_SB_FEATURES2 | \
-        XFS_SB_BAD_FEATURES2)
+        XFS_SB_BAD_FEATURES2 | XFS_SB_FEATURES_COMPAT | \
+        XFS_SB_FEATURES_RO_COMPAT | XFS_SB_FEATURES_INCOMPAT | XFS_SB_PQUOTINO)
 
 
 /*
@@ -325,6 +358,8 @@ static inline int xfs_sb_good_version(xfs_sb_t *sbp)
 
                return 1;
        }
+       if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5)
+               return 1;
 
        return 0;
 }
@@ -365,7 +400,7 @@ static inline int xfs_sb_version_hasattr(xfs_sb_t *sbp)
 {
        return sbp->sb_versionnum == XFS_SB_VERSION_2 ||
                sbp->sb_versionnum == XFS_SB_VERSION_3 ||
-               (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+               (XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
                 (sbp->sb_versionnum & XFS_SB_VERSION_ATTRBIT));
 }
 
@@ -373,7 +408,7 @@ static inline void xfs_sb_version_addattr(xfs_sb_t *sbp)
 {
        if (sbp->sb_versionnum == XFS_SB_VERSION_1)
                sbp->sb_versionnum = XFS_SB_VERSION_2;
-       else if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4)
+       else if (XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4)
                sbp->sb_versionnum |= XFS_SB_VERSION_ATTRBIT;
        else
                sbp->sb_versionnum = XFS_SB_VERSION_4 | XFS_SB_VERSION_ATTRBIT;
@@ -382,7 +417,7 @@ static inline void xfs_sb_version_addattr(xfs_sb_t *sbp)
 static inline int xfs_sb_version_hasnlink(xfs_sb_t *sbp)
 {
        return sbp->sb_versionnum == XFS_SB_VERSION_3 ||
-                (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+                (XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
                  (sbp->sb_versionnum & XFS_SB_VERSION_NLINKBIT));
 }
 
@@ -396,13 +431,13 @@ static inline void xfs_sb_version_addnlink(xfs_sb_t *sbp)
 
 static inline int xfs_sb_version_hasquota(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+       return XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
                (sbp->sb_versionnum & XFS_SB_VERSION_QUOTABIT);
 }
 
 static inline void xfs_sb_version_addquota(xfs_sb_t *sbp)
 {
-       if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4)
+       if (XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4)
                sbp->sb_versionnum |= XFS_SB_VERSION_QUOTABIT;
        else
                sbp->sb_versionnum = xfs_sb_version_tonew(sbp->sb_versionnum) |
@@ -411,13 +446,14 @@ static inline void xfs_sb_version_addquota(xfs_sb_t *sbp)
 
 static inline int xfs_sb_version_hasalign(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
-               (sbp->sb_versionnum & XFS_SB_VERSION_ALIGNBIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
+               (sbp->sb_versionnum & XFS_SB_VERSION_ALIGNBIT));
 }
 
 static inline int xfs_sb_version_hasdalign(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+       return XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
                (sbp->sb_versionnum & XFS_SB_VERSION_DALIGNBIT);
 }
 
@@ -429,38 +465,42 @@ static inline int xfs_sb_version_hasshared(xfs_sb_t *sbp)
 
 static inline int xfs_sb_version_hasdirv2(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
-               (sbp->sb_versionnum & XFS_SB_VERSION_DIRV2BIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+               (sbp->sb_versionnum & XFS_SB_VERSION_DIRV2BIT));
 }
 
 static inline int xfs_sb_version_haslogv2(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
-               (sbp->sb_versionnum & XFS_SB_VERSION_LOGV2BIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
+               (sbp->sb_versionnum & XFS_SB_VERSION_LOGV2BIT));
 }
 
 static inline int xfs_sb_version_hasextflgbit(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
-               (sbp->sb_versionnum & XFS_SB_VERSION_EXTFLGBIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+               (sbp->sb_versionnum & XFS_SB_VERSION_EXTFLGBIT));
 }
 
 static inline int xfs_sb_version_hassector(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+       return XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
                (sbp->sb_versionnum & XFS_SB_VERSION_SECTORBIT);
 }
 
 static inline int xfs_sb_version_hasasciici(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+       return XFS_SB_VERSION_NUM(sbp) >= XFS_SB_VERSION_4 &&
                (sbp->sb_versionnum & XFS_SB_VERSION_BORGBIT);
 }
 
 static inline int xfs_sb_version_hasmorebits(xfs_sb_t *sbp)
 {
-       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
-               (sbp->sb_versionnum & XFS_SB_VERSION_MOREBITSBIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_4 &&
+               (sbp->sb_versionnum & XFS_SB_VERSION_MOREBITSBIT));
 }
 
 /*
@@ -475,14 +515,16 @@ static inline int xfs_sb_version_hasmorebits(xfs_sb_t *sbp)
 
 static inline int xfs_sb_version_haslazysbcount(xfs_sb_t *sbp)
 {
-       return xfs_sb_version_hasmorebits(sbp) &&
-               (sbp->sb_features2 & XFS_SB_VERSION2_LAZYSBCOUNTBIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (xfs_sb_version_hasmorebits(sbp) &&
+               (sbp->sb_features2 & XFS_SB_VERSION2_LAZYSBCOUNTBIT));
 }
 
 static inline int xfs_sb_version_hasattr2(xfs_sb_t *sbp)
 {
-       return xfs_sb_version_hasmorebits(sbp) &&
-               (sbp->sb_features2 & XFS_SB_VERSION2_ATTR2BIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (xfs_sb_version_hasmorebits(sbp) &&
+               (sbp->sb_features2 & XFS_SB_VERSION2_ATTR2BIT));
 }
 
 static inline void xfs_sb_version_addattr2(xfs_sb_t *sbp)
@@ -500,8 +542,9 @@ static inline void xfs_sb_version_removeattr2(xfs_sb_t *sbp)
 
 static inline int xfs_sb_version_hasprojid32bit(xfs_sb_t *sbp)
 {
-       return xfs_sb_version_hasmorebits(sbp) &&
-               (sbp->sb_features2 & XFS_SB_VERSION2_PROJID32BIT);
+       return (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) ||
+              (xfs_sb_version_hasmorebits(sbp) &&
+               (sbp->sb_features2 & XFS_SB_VERSION2_PROJID32BIT));
 }
 
 static inline void xfs_sb_version_addprojid32bit(xfs_sb_t *sbp)
@@ -513,8 +556,7 @@ static inline void xfs_sb_version_addprojid32bit(xfs_sb_t *sbp)
 
 static inline int xfs_sb_version_hascrc(xfs_sb_t *sbp)
 {
-       return (xfs_sb_version_hasmorebits(sbp) &&
-               (sbp->sb_features2 & XFS_SB_VERSION2_CRCBIT));
+       return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5;
 }
 
 /*

diff --git a/libxfs/xfs_mount.c b/libxfs/xfs_mount.c
index 7ab3519b05335947e73f28c5d1280e7e8240d91a..07b892b10700397f15db0ac111d9fd60811da975 100644 (file)
--- a/libxfs/xfs_mount.c
+++ b/libxfs/xfs_mount.c
@@ -70,6 +70,12 @@ static const struct {
     { offsetof(xfs_sb_t, sb_logsunit),  0 },
     { offsetof(xfs_sb_t, sb_features2),         0 },
     { offsetof(xfs_sb_t, sb_bad_features2), 0 },
+    { offsetof(xfs_sb_t, sb_features_compat), 0 },
+    { offsetof(xfs_sb_t, sb_features_ro_compat), 0 },
+    { offsetof(xfs_sb_t, sb_features_incompat), 0 },
+    { offsetof(xfs_sb_t, sb_crc),       0 },
+    { offsetof(xfs_sb_t, sb_pquotino),  0 },
+    { offsetof(xfs_sb_t, sb_lsn),       0 },
     { sizeof(xfs_sb_t),                         0 }
 };
 
@@ -127,11 +133,23 @@ xfs_mount_validate_sb(
                return XFS_ERROR(EWRONGFS);
        }
 
+
        if (!xfs_sb_good_version(sbp)) {
                xfs_warn(mp, "bad version");
                return XFS_ERROR(EWRONGFS);
        }
 
+       /*
+        * Do not allow Version 5 superblocks to mount right now, even though
+        * support is in place. We need to implement the proper feature masks
+        * first.
+        */
+       if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) {
+               xfs_alert(mp,
+       "Version 5 superblock detected. Experimental support not yet enabled!");
+               return XFS_ERROR(EINVAL);
+       }
+
        if (unlikely(
            sbp->sb_logstart == 0 && mp->m_logdev == mp->m_dev)) {
                xfs_warn(mp,
@@ -264,6 +282,11 @@ xfs_sb_from_disk(
        to->sb_logsunit = be32_to_cpu(from->sb_logsunit);
        to->sb_features2 = be32_to_cpu(from->sb_features2);
        to->sb_bad_features2 = be32_to_cpu(from->sb_bad_features2);
+       to->sb_features_compat = be32_to_cpu(from->sb_features_compat);
+       to->sb_features_ro_compat = be32_to_cpu(from->sb_features_ro_compat);
+       to->sb_features_incompat = be32_to_cpu(from->sb_features_incompat);
+       to->sb_pquotino = be64_to_cpu(from->sb_pquotino);
+       to->sb_lsn = be64_to_cpu(from->sb_lsn);
 }
 
 /*
@@ -319,13 +342,12 @@ xfs_sb_to_disk(
        }
 }
 
-static void
+static int
 xfs_sb_verify(
        struct xfs_buf  *bp)
 {
        struct xfs_mount *mp = bp->b_target->bt_mount;
        struct xfs_sb   sb;
-       int             error;
 
        xfs_sb_from_disk(&sb, XFS_BUF_TO_SBP(bp));
 
@@ -333,16 +355,46 @@ xfs_sb_verify(
         * Only check the in progress field for the primary superblock as
         * mkfs.xfs doesn't clear it from secondary superblocks.
         */
-       error = xfs_mount_validate_sb(mp, &sb, bp->b_bn == XFS_SB_DADDR);
-       if (error)
-               xfs_buf_ioerror(bp, error);
+       return xfs_mount_validate_sb(mp, &sb, bp->b_bn == XFS_SB_DADDR);
 }
 
+/*
+ * If the superblock has the CRC feature bit set or the CRC field is non-null,
+ * check that the CRC is valid.  We check the CRC field is non-null because a
+ * single bit error could clear the feature bit and unused parts of the
+ * superblock are supposed to be zero. Hence a non-null crc field indicates that
+ * we've potentially lost a feature bit and we should check it anyway.
+ */
 static void
 xfs_sb_read_verify(
        struct xfs_buf  *bp)
 {
-       xfs_sb_verify(bp);
+       struct xfs_mount *mp = bp->b_target->bt_mount;
+       struct xfs_dsb  *dsb = XFS_BUF_TO_SBP(bp);
+       int             error;
+
+       /*
+        * open code the version check to avoid needing to convert the entire
+        * superblock from disk order just to check the version number
+        */
+       if (dsb->sb_magicnum == cpu_to_be32(XFS_SB_MAGIC) &&
+           (((be16_to_cpu(dsb->sb_versionnum) & XFS_SB_VERSION_NUMBITS) ==
+                                               XFS_SB_VERSION_5) ||
+            dsb->sb_crc != 0)) {
+
+               if (!xfs_verify_cksum(bp->b_addr, be16_to_cpu(dsb->sb_sectsize),
+                                     offsetof(struct xfs_sb, sb_crc))) {
+                       error = EFSCORRUPTED;
+                       goto out_error;
+               }
+       }
+       error = xfs_sb_verify(bp);
+
+out_error:
+       if (error) {
+               XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+               xfs_buf_ioerror(bp, error);
+       }
 }
 
 /*
@@ -355,11 +407,10 @@ static void
 xfs_sb_quiet_read_verify(
        struct xfs_buf  *bp)
 {
-       struct xfs_sb   sb;
+       struct xfs_dsb  *dsb = XFS_BUF_TO_SBP(bp);
 
-       xfs_sb_from_disk(&sb, XFS_BUF_TO_SBP(bp));
 
-       if (sb.sb_magicnum == XFS_SB_MAGIC) {
+       if (dsb->sb_magicnum == cpu_to_be32(XFS_SB_MAGIC)) {
                /* XFS filesystem, verify noisily! */
                xfs_sb_read_verify(bp);
                return;
@@ -370,9 +421,27 @@ xfs_sb_quiet_read_verify(
 
 static void
 xfs_sb_write_verify(
-       struct xfs_buf  *bp)
+       struct xfs_buf          *bp)
 {
-       xfs_sb_verify(bp);
+       struct xfs_mount        *mp = bp->b_target->bt_mount;
+       struct xfs_buf_log_item *bip = bp->b_fspriv;
+       int                     error;
+
+       error = xfs_sb_verify(bp);
+       if (error) {
+               XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+               xfs_buf_ioerror(bp, error);
+               return;
+       }
+
+       if (!xfs_sb_version_hascrc(&mp->m_sb))
+               return;
+
+       if (bip)
+               XFS_BUF_TO_SBP(bp)->sb_lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+       xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+                        offsetof(struct xfs_sb, sb_crc));
 }
 
 const struct xfs_buf_ops xfs_sb_buf_ops = {
@@ -525,5 +594,6 @@ xfs_mod_sb(xfs_trans_t *tp, __int64_t fields)
        ASSERT((1LL << f) & XFS_SB_MOD_BITS);
        first = xfs_sb_info[f].offset;
 
+       xfs_trans_buf_set_type(tp, bp, XFS_BLFT_SB_BUF);
        xfs_trans_log_buf(tp, bp, first, last);
 }