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/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 8628507..5303b3d 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -2040,6 +2040,14 @@
}
bp->b_ops = &xfs_attr3_rmt_buf_ops;
break;
+ case XFS_BLFT_SB_BUF:
+ if (magic32 != XFS_SB_MAGIC) {
+ xfs_warn(mp, "Bad SB block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_sb_buf_ops;
+ break;
default:
xfs_warn(mp, "Unknown buffer type %d!",
xfs_blft_from_flags(buf_f));
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index 3806088..140136c 100644
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -43,6 +43,8 @@
#include "xfs_utils.h"
#include "xfs_trace.h"
#include "xfs_icache.h"
+#include "xfs_cksum.h"
+#include "xfs_buf_item.h"
#ifdef HAVE_PERCPU_SB
@@ -109,6 +111,12 @@
{ 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 }
};
@@ -319,11 +327,23 @@
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_targp == mp->m_ddev_targp)) {
xfs_warn(mp,
@@ -557,6 +577,11 @@
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);
}
/*
@@ -612,13 +637,12 @@
}
}
-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));
@@ -626,16 +650,46 @@
* 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);
+ }
}
/*
@@ -648,11 +702,10 @@
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;
@@ -663,9 +716,27 @@
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 = {
@@ -687,7 +758,8 @@
xfs_readsb(xfs_mount_t *mp, int flags)
{
unsigned int sector_size;
- xfs_buf_t *bp;
+ struct xfs_buf *bp;
+ struct xfs_sb *sbp = &mp->m_sb;
int error;
int loud = !(flags & XFS_MFSI_QUIET);
@@ -726,10 +798,10 @@
/*
* We must be able to do sector-sized and sector-aligned IO.
*/
- if (sector_size > mp->m_sb.sb_sectsize) {
+ if (sector_size > sbp->sb_sectsize) {
if (loud)
xfs_warn(mp, "device supports %u byte sectors (not %u)",
- sector_size, mp->m_sb.sb_sectsize);
+ sector_size, sbp->sb_sectsize);
error = ENOSYS;
goto release_buf;
}
@@ -738,15 +810,18 @@
* If device sector size is smaller than the superblock size,
* re-read the superblock so the buffer is correctly sized.
*/
- if (sector_size < mp->m_sb.sb_sectsize) {
+ if (sector_size < sbp->sb_sectsize) {
xfs_buf_relse(bp);
- sector_size = mp->m_sb.sb_sectsize;
+ sector_size = sbp->sb_sectsize;
goto reread;
}
/* Initialize per-cpu counters */
xfs_icsb_reinit_counters(mp);
+ /* no need to be quiet anymore, so reset the buf ops */
+ bp->b_ops = &xfs_sb_buf_ops;
+
mp->m_sb_bp = bp;
xfs_buf_unlock(bp);
return 0;
@@ -1633,6 +1708,7 @@
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);
}
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 687c171..b004cec 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -391,6 +391,7 @@
#endif /* __KERNEL__ */
+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/fs/xfs/xfs_sb.h b/fs/xfs/xfs_sb.h
index a05b451..457fefa 100644
--- a/fs/xfs/xfs_sb.h
+++ b/fs/xfs/xfs_sb.h
@@ -32,6 +32,7 @@
#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 @@
*/
__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 @@
* 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 @@
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 @@
#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 @@
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 @@
return 1;
}
+ if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5)
+ return 1;
return 0;
}
@@ -365,7 +400,7 @@
{
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 @@
{
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 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 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 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_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_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)
@@ -500,14 +542,14 @@
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 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;
}
/*