xfs
[Top] [All Lists]

[PATCH 07/16] xfs: define the on-disk rmap btree format

To: xfs@xxxxxxxxxxx
Subject: [PATCH 07/16] xfs: define the on-disk rmap btree format
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Tue, 8 Mar 2016 15:16:09 +1100
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <1457410578-30233-1-git-send-email-david@xxxxxxxxxxxxx>
References: <1457410578-30233-1-git-send-email-david@xxxxxxxxxxxxx>
From: Dave Chinner <dchinner@xxxxxxxxxx>

Now we have all the surrounding call infrastructure in place, we can
start filling out the rmap btree implementation. Start with the
on-disk btree format; add everything needed to read, write and
manipulate rmap btree blocks. This prepares the way for adding the
btree operations implementation.

[darrick.wong@xxxxxxxxxx: record owner and offset info in rmap btree]
[darrick.wong@xxxxxxxxxx: fork, bmbt and unwritten state in rmap btree]

Signed-off-by: Dave Chinner <dchinner@xxxxxxxxxx>
Signed-off-by: Darrick J. Wong <darrick.wong@xxxxxxxxxx>
Reviewed-by: Dave Chinner <dchinner@xxxxxxxxxx>
Signed-off-by: Dave Chinner <david@xxxxxxxxxxxxx>
---
 fs/xfs/Makefile                |   1 +
 fs/xfs/libxfs/xfs_btree.c      |   3 +
 fs/xfs/libxfs/xfs_btree.h      |  18 +++--
 fs/xfs/libxfs/xfs_format.h     |  95 +++++++++++++++++++++++
 fs/xfs/libxfs/xfs_rmap_btree.c | 170 +++++++++++++++++++++++++++++++++++++++++
 fs/xfs/libxfs/xfs_rmap_btree.h |  31 ++++++++
 fs/xfs/libxfs/xfs_sb.c         |   6 ++
 fs/xfs/libxfs/xfs_shared.h     |   2 +
 fs/xfs/xfs_mount.h             |   2 +
 fs/xfs/xfs_ondisk.h            |   3 +
 10 files changed, 323 insertions(+), 8 deletions(-)
 create mode 100644 fs/xfs/libxfs/xfs_rmap_btree.c

diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index c202ce3..9391080 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -52,6 +52,7 @@ xfs-y                         += $(addprefix libxfs/, \
                                   xfs_inode_buf.o \
                                   xfs_log_rlimit.o \
                                   xfs_rmap.o \
+                                  xfs_rmap_btree.o \
                                   xfs_sb.o \
                                   xfs_symlink_remote.o \
                                   xfs_trans_resv.o \
diff --git a/fs/xfs/libxfs/xfs_btree.c b/fs/xfs/libxfs/xfs_btree.c
index 5953764..1240117 100644
--- a/fs/xfs/libxfs/xfs_btree.c
+++ b/fs/xfs/libxfs/xfs_btree.c
@@ -1144,6 +1144,9 @@ xfs_btree_set_refs(
        case XFS_BTNUM_BMAP:
                xfs_buf_set_ref(bp, XFS_BMAP_BTREE_REF);
                break;
+       case XFS_BTNUM_RMAP:
+               xfs_buf_set_ref(bp, XFS_RMAP_BTREE_REF);
+               break;
        default:
                ASSERT(0);
        }
diff --git a/fs/xfs/libxfs/xfs_btree.h b/fs/xfs/libxfs/xfs_btree.h
index 1fcf272..07fa27e 100644
--- a/fs/xfs/libxfs/xfs_btree.h
+++ b/fs/xfs/libxfs/xfs_btree.h
@@ -38,17 +38,19 @@ union xfs_btree_ptr {
 };
 
 union xfs_btree_key {
-       xfs_bmbt_key_t          bmbt;
-       xfs_bmdr_key_t          bmbr;   /* bmbt root block */
-       xfs_alloc_key_t         alloc;
-       xfs_inobt_key_t         inobt;
+       struct xfs_bmbt_key             bmbt;
+       xfs_bmdr_key_t                  bmbr;   /* bmbt root block */
+       xfs_alloc_key_t                 alloc;
+       struct xfs_inobt_key            inobt;
+       struct xfs_rmap_key             rmap;
 };
 
 union xfs_btree_rec {
-       xfs_bmbt_rec_t          bmbt;
-       xfs_bmdr_rec_t          bmbr;   /* bmbt root block */
-       xfs_alloc_rec_t         alloc;
-       xfs_inobt_rec_t         inobt;
+       struct xfs_bmbt_rec             bmbt;
+       xfs_bmdr_rec_t                  bmbr;   /* bmbt root block */
+       struct xfs_alloc_rec            alloc;
+       struct xfs_inobt_rec            inobt;
+       struct xfs_rmap_rec             rmap;
 };
 
 /*
diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h
index 232c145..508b3ea 100644
--- a/fs/xfs/libxfs/xfs_format.h
+++ b/fs/xfs/libxfs/xfs_format.h
@@ -1384,11 +1384,106 @@ xfs_rmap_ino_owner(
 #define XFS_RMAP_OWN_INODES    (-7ULL) /* Inode chunk */
 #define XFS_RMAP_OWN_MIN       (-8ULL) /* guard */
 
+#define XFS_RMAP_NON_INODE_OWNER(owner)        (!!((owner) & (1ULL << 63)))
+
+/*
+ * Data record structure
+ */
+struct xfs_rmap_rec {
+       __be32          rm_startblock;  /* extent start block */
+       __be32          rm_blockcount;  /* extent length */
+       __be64          rm_owner;       /* extent owner */
+       __be64          rm_offset;      /* offset within the owner */
+};
+
+/*
+ * rmap btree record
+ *  rm_blockcount:31 is the unwritten extent flag (same as l0:63 in bmbt)
+ *  rm_blockcount:0-30 are the extent length
+ *  rm_offset:63 is the attribute fork flag
+ *  rm_offset:62 is the bmbt block flag
+ *  rm_offset:0-61 is the block offset within the inode
+ */
+#define XFS_RMAP_OFF_ATTR      ((__uint64_t)1ULL << 63)
+#define XFS_RMAP_OFF_BMBT      ((__uint64_t)1ULL << 62)
+#define XFS_RMAP_LEN_UNWRITTEN ((xfs_extlen_t)1U << 31)
+
+#define XFS_RMAP_OFF_MASK      ~(XFS_RMAP_OFF_ATTR | XFS_RMAP_OFF_BMBT)
+#define XFS_RMAP_LEN_MASK      ~XFS_RMAP_LEN_UNWRITTEN
+
+#define XFS_RMAP_OFF(off)              ((off) & XFS_RMAP_OFF_MASK)
+#define XFS_RMAP_LEN(len)              ((len) & XFS_RMAP_LEN_MASK)
+
+#define XFS_RMAP_IS_BMBT(off)          (!!((off) & XFS_RMAP_OFF_BMBT))
+#define XFS_RMAP_IS_ATTR_FORK(off)     (!!((off) & XFS_RMAP_OFF_ATTR))
+#define XFS_RMAP_IS_UNWRITTEN(len)     (!!((len) & XFS_RMAP_LEN_UNWRITTEN))
+
+#define RMAPBT_STARTBLOCK_BITLEN       32
+#define RMAPBT_EXNTFLAG_BITLEN         1
+#define RMAPBT_BLOCKCOUNT_BITLEN       31
+#define RMAPBT_OWNER_BITLEN            64
+#define RMAPBT_ATTRFLAG_BITLEN         1
+#define RMAPBT_BMBTFLAG_BITLEN         1
+#define RMAPBT_OFFSET_BITLEN           62
+
+struct xfs_rmap_irec {
+       xfs_agblock_t   rm_startblock;  /* extent start block */
+       xfs_extlen_t    rm_blockcount;  /* extent length */
+       __uint64_t      rm_owner;       /* extent owner */
+       __uint64_t      rm_offset;      /* offset within the owner */
+};
+
+/*
+ * Key structure
+ *
+ * We don't use the length for lookups
+ */
+struct xfs_rmap_key {
+       __be32          rm_startblock;  /* extent start block */
+       __be64          rm_owner;       /* extent owner */
+       __be64          rm_offset;      /* offset within the owner */
+} __attribute__((packed));
+
+/* btree pointer type */
+typedef __be32 xfs_rmap_ptr_t;
+
 #define        XFS_RMAP_BLOCK(mp) \
        (xfs_sb_version_hasfinobt(&((mp)->m_sb)) ? \
         XFS_FIBT_BLOCK(mp) + 1 : \
         XFS_IBT_BLOCK(mp) + 1)
 
+static inline void
+xfs_owner_info_unpack(
+       struct xfs_owner_info   *oinfo,
+       uint64_t                *owner,
+       uint64_t                *offset)
+{
+       __uint64_t              r;
+
+       *owner = oinfo->oi_owner;
+       r = oinfo->oi_offset;
+       if (oinfo->oi_flags & XFS_RMAP_INO_ATTR_FORK)
+               r |= XFS_RMAP_OFF_ATTR;
+       if (oinfo->oi_flags & XFS_RMAP_BMBT_BLOCK)
+               r |= XFS_RMAP_OFF_BMBT;
+       *offset = r;
+}
+
+static inline void
+xfs_owner_info_pack(
+       struct xfs_owner_info   *oinfo,
+       uint64_t                owner,
+       uint64_t                offset)
+{
+       oinfo->oi_owner = owner;
+       oinfo->oi_offset = XFS_RMAP_OFF(offset);
+       oinfo->oi_flags = 0;
+       if (XFS_RMAP_IS_ATTR_FORK(offset))
+               oinfo->oi_flags |= XFS_RMAP_INO_ATTR_FORK;
+       if (XFS_RMAP_IS_BMBT(offset))
+               oinfo->oi_flags |= XFS_RMAP_BMBT_BLOCK;
+}
+
 /*
  * BMAP Btree format definitions
  *
diff --git a/fs/xfs/libxfs/xfs_rmap_btree.c b/fs/xfs/libxfs/xfs_rmap_btree.c
new file mode 100644
index 0000000..882e8e2
--- /dev/null
+++ b/fs/xfs/libxfs/xfs_rmap_btree.c
@@ -0,0 +1,170 @@
+/*
+ * Copyright (c) 2014 Red Hat, Inc.
+ * All Rights Reserved.
+ *
+ * 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.
+ *
+ * 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_log_format.h"
+#include "xfs_trans_resv.h"
+#include "xfs_bit.h"
+#include "xfs_sb.h"
+#include "xfs_mount.h"
+#include "xfs_inode.h"
+#include "xfs_trans.h"
+#include "xfs_alloc.h"
+#include "xfs_btree.h"
+#include "xfs_rmap_btree.h"
+#include "xfs_trace.h"
+#include "xfs_cksum.h"
+#include "xfs_error.h"
+#include "xfs_extent_busy.h"
+
+static struct xfs_btree_cur *
+xfs_rmapbt_dup_cursor(
+       struct xfs_btree_cur    *cur)
+{
+       return xfs_rmapbt_init_cursor(cur->bc_mp, cur->bc_tp,
+                       cur->bc_private.a.agbp, cur->bc_private.a.agno);
+}
+
+static bool
+xfs_rmapbt_verify(
+       struct xfs_buf          *bp)
+{
+       struct xfs_mount        *mp = bp->b_target->bt_mount;
+       struct xfs_btree_block  *block = XFS_BUF_TO_BLOCK(bp);
+       struct xfs_perag        *pag = bp->b_pag;
+       unsigned int            level;
+
+       /*
+        * magic number and level verification
+        *
+        * During growfs operations, we can't verify the exact level or owner as
+        * the perag is not fully initialised and hence not attached to the
+        * buffer.  In this case, check against the maximum tree depth.
+        *
+        * Similarly, during log recovery we will have a perag structure
+        * attached, but the agf information will not yet have been initialised
+        * from the on disk AGF. Again, we can only check against maximum limits
+        * in this case.
+        */
+       if (block->bb_magic != cpu_to_be32(XFS_RMAP_CRC_MAGIC))
+               return false;
+
+       if (!xfs_sb_version_hasrmapbt(&mp->m_sb))
+               return false;
+       if (!xfs_btree_sblock_v5hdr_verify(bp))
+               return false;
+
+       level = be16_to_cpu(block->bb_level);
+       if (pag && pag->pagf_init) {
+               if (level >= pag->pagf_levels[XFS_BTNUM_RMAPi])
+                       return false;
+       } else if (level >= mp->m_ag_maxlevels)
+               return false;
+
+       return xfs_btree_sblock_verify(bp, mp->m_rmap_mxr[level != 0]);
+}
+
+static void
+xfs_rmapbt_read_verify(
+       struct xfs_buf  *bp)
+{
+       if (!xfs_btree_sblock_verify_crc(bp))
+               xfs_buf_ioerror(bp, -EFSBADCRC);
+       else if (!xfs_rmapbt_verify(bp))
+               xfs_buf_ioerror(bp, -EFSCORRUPTED);
+
+       if (bp->b_error) {
+               trace_xfs_btree_corrupt(bp, _RET_IP_);
+               xfs_verifier_error(bp);
+       }
+}
+
+static void
+xfs_rmapbt_write_verify(
+       struct xfs_buf  *bp)
+{
+       if (!xfs_rmapbt_verify(bp)) {
+               trace_xfs_btree_corrupt(bp, _RET_IP_);
+               xfs_buf_ioerror(bp, -EFSCORRUPTED);
+               xfs_verifier_error(bp);
+               return;
+       }
+       xfs_btree_sblock_calc_crc(bp);
+
+}
+
+const struct xfs_buf_ops xfs_rmapbt_buf_ops = {
+       .name                   = "xfs_rmapbt",
+       .verify_read            = xfs_rmapbt_read_verify,
+       .verify_write           = xfs_rmapbt_write_verify,
+};
+
+static const struct xfs_btree_ops xfs_rmapbt_ops = {
+       .rec_len                = sizeof(struct xfs_rmap_rec),
+       .key_len                = sizeof(struct xfs_rmap_key),
+
+       .dup_cursor             = xfs_rmapbt_dup_cursor,
+       .buf_ops                = &xfs_rmapbt_buf_ops,
+};
+
+/*
+ * Allocate a new allocation btree cursor.
+ */
+struct xfs_btree_cur *
+xfs_rmapbt_init_cursor(
+       struct xfs_mount        *mp,
+       struct xfs_trans        *tp,
+       struct xfs_buf          *agbp,
+       xfs_agnumber_t          agno)
+{
+       struct xfs_agf          *agf = XFS_BUF_TO_AGF(agbp);
+       struct xfs_btree_cur    *cur;
+
+       cur = kmem_zone_zalloc(xfs_btree_cur_zone, KM_SLEEP);
+       cur->bc_tp = tp;
+       cur->bc_mp = mp;
+       cur->bc_btnum = XFS_BTNUM_RMAP;
+       cur->bc_flags = XFS_BTREE_CRC_BLOCKS;
+       cur->bc_blocklog = mp->m_sb.sb_blocklog;
+       cur->bc_ops = &xfs_rmapbt_ops;
+       cur->bc_nlevels = be32_to_cpu(agf->agf_levels[XFS_BTNUM_RMAP]);
+
+       cur->bc_private.a.agbp = agbp;
+       cur->bc_private.a.agno = agno;
+
+       return cur;
+}
+
+/*
+ * Calculate number of records in an rmap btree block.
+ */
+int
+xfs_rmapbt_maxrecs(
+       struct xfs_mount        *mp,
+       int                     blocklen,
+       int                     leaf)
+{
+       blocklen -= XFS_RMAP_BLOCK_LEN;
+
+       if (leaf)
+               return blocklen / sizeof(struct xfs_rmap_rec);
+       return blocklen /
+               (sizeof(struct xfs_rmap_key) + sizeof(xfs_rmap_ptr_t));
+}
diff --git a/fs/xfs/libxfs/xfs_rmap_btree.h b/fs/xfs/libxfs/xfs_rmap_btree.h
index a3b8f90..2e02362 100644
--- a/fs/xfs/libxfs/xfs_rmap_btree.h
+++ b/fs/xfs/libxfs/xfs_rmap_btree.h
@@ -19,6 +19,37 @@
 #define        __XFS_RMAP_BTREE_H__
 
 struct xfs_buf;
+struct xfs_btree_cur;
+struct xfs_mount;
+
+/* rmaps only exist on crc enabled filesystems */
+#define XFS_RMAP_BLOCK_LEN     XFS_BTREE_SBLOCK_CRC_LEN
+
+/*
+ * Record, key, and pointer address macros for btree blocks.
+ *
+ * (note that some of these may appear unused, but they are used in userspace)
+ */
+#define XFS_RMAP_REC_ADDR(block, index) \
+       ((struct xfs_rmap_rec *) \
+               ((char *)(block) + XFS_RMAP_BLOCK_LEN + \
+                (((index) - 1) * sizeof(struct xfs_rmap_rec))))
+
+#define XFS_RMAP_KEY_ADDR(block, index) \
+       ((struct xfs_rmap_key *) \
+               ((char *)(block) + XFS_RMAP_BLOCK_LEN + \
+                ((index) - 1) * sizeof(struct xfs_rmap_key)))
+
+#define XFS_RMAP_PTR_ADDR(block, index, maxrecs) \
+       ((xfs_rmap_ptr_t *) \
+               ((char *)(block) + XFS_RMAP_BLOCK_LEN + \
+                (maxrecs) * sizeof(struct xfs_rmap_key) + \
+                ((index) - 1) * sizeof(xfs_rmap_ptr_t)))
+
+struct xfs_btree_cur *xfs_rmapbt_init_cursor(struct xfs_mount *mp,
+                               struct xfs_trans *tp, struct xfs_buf *bp,
+                               xfs_agnumber_t agno);
+int xfs_rmapbt_maxrecs(struct xfs_mount *mp, int blocklen, int leaf);
 
 int xfs_rmap_alloc(struct xfs_trans *tp, struct xfs_buf *agbp,
                   xfs_agnumber_t agno, xfs_agblock_t bno, xfs_extlen_t len,
diff --git a/fs/xfs/libxfs/xfs_sb.c b/fs/xfs/libxfs/xfs_sb.c
index 8a53eaa..fadf750 100644
--- a/fs/xfs/libxfs/xfs_sb.c
+++ b/fs/xfs/libxfs/xfs_sb.c
@@ -36,6 +36,7 @@
 #include "xfs_alloc_btree.h"
 #include "xfs_ialloc_btree.h"
 #include "xfs_log.h"
+#include "xfs_rmap_btree.h"
 
 /*
  * Physical superblock buffer manipulations. Shared with libxfs in userspace.
@@ -729,6 +730,11 @@ xfs_sb_mount_common(
        mp->m_bmap_dmnr[0] = mp->m_bmap_dmxr[0] / 2;
        mp->m_bmap_dmnr[1] = mp->m_bmap_dmxr[1] / 2;
 
+       mp->m_rmap_mxr[0] = xfs_rmapbt_maxrecs(mp, sbp->sb_blocksize, 1);
+       mp->m_rmap_mxr[1] = xfs_rmapbt_maxrecs(mp, sbp->sb_blocksize, 0);
+       mp->m_rmap_mnr[0] = mp->m_rmap_mxr[0] / 2;
+       mp->m_rmap_mnr[1] = mp->m_rmap_mxr[1] / 2;
+
        mp->m_bsize = XFS_FSB_TO_BB(mp, 1);
        mp->m_ialloc_inos = (int)MAX((__uint16_t)XFS_INODES_PER_CHUNK,
                                        sbp->sb_inopblock);
diff --git a/fs/xfs/libxfs/xfs_shared.h b/fs/xfs/libxfs/xfs_shared.h
index 81ac870..4912072 100644
--- a/fs/xfs/libxfs/xfs_shared.h
+++ b/fs/xfs/libxfs/xfs_shared.h
@@ -38,6 +38,7 @@ extern const struct xfs_buf_ops xfs_agi_buf_ops;
 extern const struct xfs_buf_ops xfs_agf_buf_ops;
 extern const struct xfs_buf_ops xfs_agfl_buf_ops;
 extern const struct xfs_buf_ops xfs_allocbt_buf_ops;
+extern const struct xfs_buf_ops xfs_rmapbt_buf_ops;
 extern const struct xfs_buf_ops xfs_attr3_leaf_buf_ops;
 extern const struct xfs_buf_ops xfs_attr3_rmt_buf_ops;
 extern const struct xfs_buf_ops xfs_bmbt_buf_ops;
@@ -212,6 +213,7 @@ int xfs_log_calc_minimum_size(struct xfs_mount *);
 #define        XFS_INO_BTREE_REF       3
 #define        XFS_ALLOC_BTREE_REF     2
 #define        XFS_BMAP_BTREE_REF      2
+#define        XFS_RMAP_BTREE_REF      2
 #define        XFS_DIR_BTREE_REF       2
 #define        XFS_INO_REF             2
 #define        XFS_ATTR_BTREE_REF      1
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 9788686..b409a2a 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -90,6 +90,8 @@ typedef struct xfs_mount {
        uint                    m_bmap_dmnr[2]; /* min bmap btree records */
        uint                    m_inobt_mxr[2]; /* max inobt btree records */
        uint                    m_inobt_mnr[2]; /* min inobt btree records */
+       uint                    m_rmap_mxr[2];  /* max rmap btree records */
+       uint                    m_rmap_mnr[2];  /* min rmap btree records */
        uint                    m_ag_maxlevels; /* XFS_AG_MAXLEVELS */
        uint                    m_bm_maxlevels[2]; /* XFS_BM_MAXLEVELS */
        uint                    m_in_maxlevels; /* max inobt btree levels. */
diff --git a/fs/xfs/xfs_ondisk.h b/fs/xfs/xfs_ondisk.h
index 9a78408..07fb5d8 100644
--- a/fs/xfs/xfs_ondisk.h
+++ b/fs/xfs/xfs_ondisk.h
@@ -42,11 +42,14 @@ xfs_check_ondisk_structs(void)
        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);
+       XFS_CHECK_STRUCT_SIZE(struct xfs_rmap_key,              20);
+       XFS_CHECK_STRUCT_SIZE(struct xfs_rmap_rec,              24);
        XFS_CHECK_STRUCT_SIZE(struct xfs_timestamp,             8);
        XFS_CHECK_STRUCT_SIZE(xfs_alloc_key_t,                  8);
        XFS_CHECK_STRUCT_SIZE(xfs_alloc_ptr_t,                  4);
        XFS_CHECK_STRUCT_SIZE(xfs_alloc_rec_t,                  8);
        XFS_CHECK_STRUCT_SIZE(xfs_inobt_ptr_t,                  4);
+       XFS_CHECK_STRUCT_SIZE(xfs_rmap_ptr_t,                   4);
 
        /* dir/attr trees */
        XFS_CHECK_STRUCT_SIZE(struct xfs_attr3_leaf_hdr,        80);
-- 
2.7.0

<Prev in Thread] Current Thread [Next in Thread>