* [PATCH 01/16] xfs: rearrange some code in xfs_bmap for better locality
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 14:44 ` Mark Tinguely
2013-03-07 18:48 ` Ben Myers
2013-02-25 1:31 ` [PATCH 02/16] xfs: take inode version into account in XFS_LITINO Dave Chinner
` (14 subsequent siblings)
15 siblings, 2 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
xfs_bmap.c is a big file, and some of the related code is spread all
throughout the file requiring function prototypes for static
function and jumping all through the file to follow a single call
path. Rearrange the code so that:
a) related functionality is grouped together; and
b) functions are grouped in call dependency order
While the diffstat is large, there are no code changes in the patch;
it is just moving the functionality around and removing the function
prototypes at the top of the file. The resulting layout of the code
is as follows (top of file to bottom):
- miscellaneous helper functions
- extent tree block counting routines
- debug/sanity checking code
- bmap free list manipulation functions
- inode fork format manipulation functions
- internal/external extent tree seach functions
- extent tree manipulation functions used during allocation
- functions used during extent read/allocate/removal
operations (i.e. xfs_bmapi_write, xfs_bmapi_read,
xfs_bunmapi and xfs_getbmap)
This means that following logic paths through the bmapi code is much
simpler - most of the code relevant to a specific operation is now
clustered together rather than spread all over the file....
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_bmap.c |10659 ++++++++++++++++++++++++++---------------------------
1 file changed, 5261 insertions(+), 5398 deletions(-)
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index b44af92..d490fe8 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -52,175 +52,72 @@
kmem_zone_t *xfs_bmap_free_item_zone;
/*
- * Prototypes for internal bmap routines.
+ * Miscellaneous helper functions
*/
-#ifdef DEBUG
-STATIC void
-xfs_bmap_check_leaf_extents(
- struct xfs_btree_cur *cur,
- struct xfs_inode *ip,
- int whichfork);
-#else
-#define xfs_bmap_check_leaf_extents(cur, ip, whichfork) do { } while (0)
-#endif
-
-
-/*
- * Called from xfs_bmap_add_attrfork to handle extents format files.
- */
-STATIC int /* error */
-xfs_bmap_add_attrfork_extents(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first block allocated */
- xfs_bmap_free_t *flist, /* blocks to free at commit */
- int *flags); /* inode logging flags */
-
-/*
- * Called from xfs_bmap_add_attrfork to handle local format files.
- */
-STATIC int /* error */
-xfs_bmap_add_attrfork_local(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first block allocated */
- xfs_bmap_free_t *flist, /* blocks to free at commit */
- int *flags); /* inode logging flags */
-
-/*
- * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file.
- * It figures out where to ask the underlying allocator to put the new extent.
- */
-STATIC int /* error */
-xfs_bmap_alloc(
- xfs_bmalloca_t *ap); /* bmap alloc argument struct */
-
-/*
- * Transform a btree format file with only one leaf node, where the
- * extents list will fit in the inode, into an extents format file.
- * Since the file extents are already in-core, all we have to do is
- * give up the space for the btree root and pitch the leaf block.
- */
-STATIC int /* error */
-xfs_bmap_btree_to_extents(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_btree_cur_t *cur, /* btree cursor */
- int *logflagsp, /* inode logging flags */
- int whichfork); /* data or attr fork */
-
-/*
- * Remove the entry "free" from the free item list. Prev points to the
- * previous entry, unless "free" is the head of the list.
- */
-STATIC void
-xfs_bmap_del_free(
- xfs_bmap_free_t *flist, /* free item list header */
- xfs_bmap_free_item_t *prev, /* previous item on list, if any */
- xfs_bmap_free_item_t *free); /* list item to be freed */
-
-/*
- * Convert an extents-format file into a btree-format file.
- * The new file will have a root block (in the inode) and a single child block.
- */
-STATIC int /* error */
-xfs_bmap_extents_to_btree(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first-block-allocated */
- xfs_bmap_free_t *flist, /* blocks freed in xaction */
- xfs_btree_cur_t **curp, /* cursor returned to caller */
- int wasdel, /* converting a delayed alloc */
- int *logflagsp, /* inode logging flags */
- int whichfork); /* data or attr fork */
-
-/*
- * Convert a local file to an extents file.
- * This code is sort of bogus, since the file data needs to get
- * logged so it won't be lost. The bmap-level manipulations are ok, though.
- */
-STATIC int /* error */
-xfs_bmap_local_to_extents(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first block allocated in xaction */
- xfs_extlen_t total, /* total blocks needed by transaction */
- int *logflagsp, /* inode logging flags */
- int whichfork, /* data or attr fork */
- void (*init_fn)(struct xfs_buf *bp,
- struct xfs_inode *ip,
- struct xfs_ifork *ifp));
-
-/*
- * Search the extents list for the inode, for the extent containing bno.
- * If bno lies in a hole, point to the next entry. If bno lies past eof,
- * *eofp will be set, and *prevp will contain the last entry (null if none).
- * Else, *lastxp will be set to the index of the found
- * entry; *gotp will contain the entry.
- */
-STATIC xfs_bmbt_rec_host_t * /* pointer to found extent entry */
-xfs_bmap_search_extents(
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fileoff_t bno, /* block number searched for */
- int whichfork, /* data or attr fork */
- int *eofp, /* out: end of file found */
- xfs_extnum_t *lastxp, /* out: last extent index */
- xfs_bmbt_irec_t *gotp, /* out: extent entry found */
- xfs_bmbt_irec_t *prevp); /* out: previous extent entry found */
-
-/*
- * Compute the worst-case number of indirect blocks that will be used
- * for ip's delayed extent of length "len".
- */
-STATIC xfs_filblks_t
-xfs_bmap_worst_indlen(
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_filblks_t len); /* delayed extent length */
-
-#ifdef DEBUG
/*
- * Perform various validation checks on the values being returned
- * from xfs_bmapi().
+ * Compute and fill in the value of the maximum depth of a bmap btree
+ * in this filesystem. Done once, during mount.
*/
-STATIC void
-xfs_bmap_validate_ret(
- xfs_fileoff_t bno,
- xfs_filblks_t len,
- int flags,
- xfs_bmbt_irec_t *mval,
- int nmap,
- int ret_nmap);
-#else
-#define xfs_bmap_validate_ret(bno,len,flags,mval,onmap,nmap)
-#endif /* DEBUG */
-
-STATIC int
-xfs_bmap_count_tree(
- xfs_mount_t *mp,
- xfs_trans_t *tp,
- xfs_ifork_t *ifp,
- xfs_fsblock_t blockno,
- int levelin,
- int *count);
-
-STATIC void
-xfs_bmap_count_leaves(
- xfs_ifork_t *ifp,
- xfs_extnum_t idx,
- int numrecs,
- int *count);
+void
+xfs_bmap_compute_maxlevels(
+ xfs_mount_t *mp, /* file system mount structure */
+ int whichfork) /* data or attr fork */
+{
+ int level; /* btree level */
+ uint maxblocks; /* max blocks at this level */
+ uint maxleafents; /* max leaf entries possible */
+ int maxrootrecs; /* max records in root block */
+ int minleafrecs; /* min records in leaf block */
+ int minnoderecs; /* min records in node block */
+ int sz; /* root block size */
-STATIC void
-xfs_bmap_disk_count_leaves(
- struct xfs_mount *mp,
- struct xfs_btree_block *block,
- int numrecs,
- int *count);
+ /*
+ * The maximum number of extents in a file, hence the maximum
+ * number of leaf entries, is controlled by the type of di_nextents
+ * (a signed 32-bit number, xfs_extnum_t), or by di_anextents
+ * (a signed 16-bit number, xfs_aextnum_t).
+ *
+ * Note that we can no longer assume that if we are in ATTR1 that
+ * the fork offset of all the inodes will be
+ * (xfs_default_attroffset(ip) >> 3) because we could have mounted
+ * with ATTR2 and then mounted back with ATTR1, keeping the
+ * di_forkoff's fixed but probably at various positions. Therefore,
+ * for both ATTR1 and ATTR2 we have to assume the worst case scenario
+ * of a minimum size available.
+ */
+ if (whichfork == XFS_DATA_FORK) {
+ maxleafents = MAXEXTNUM;
+ sz = XFS_BMDR_SPACE_CALC(MINDBTPTRS);
+ } else {
+ maxleafents = MAXAEXTNUM;
+ sz = XFS_BMDR_SPACE_CALC(MINABTPTRS);
+ }
+ maxrootrecs = xfs_bmdr_maxrecs(mp, sz, 0);
+ minleafrecs = mp->m_bmap_dmnr[0];
+ minnoderecs = mp->m_bmap_dmnr[1];
+ maxblocks = (maxleafents + minleafrecs - 1) / minleafrecs;
+ for (level = 1; maxblocks > 1; level++) {
+ if (maxblocks <= maxrootrecs)
+ maxblocks = 1;
+ else
+ maxblocks = (maxblocks + minnoderecs - 1) / minnoderecs;
+ }
+ mp->m_bm_maxlevels[whichfork] = level;
+}
/*
- * Bmap internal routines.
+ * Convert the given file system block to a disk block. We have to treat it
+ * differently based on whether the file is a real time file or not, because the
+ * bmap code does.
*/
+xfs_daddr_t
+xfs_fsb_to_db(struct xfs_inode *ip, xfs_fsblock_t fsb)
+{
+ return (XFS_IS_REALTIME_INODE(ip) ? \
+ (xfs_daddr_t)XFS_FSB_TO_BB((ip)->i_mount, (fsb)) : \
+ XFS_FSB_TO_DADDR((ip)->i_mount, (fsb)));
+}
STATIC int /* error */
xfs_bmbt_lookup_eq(
@@ -290,5935 +187,5914 @@ xfs_bmbt_update(
}
/*
- * Called from xfs_bmap_add_attrfork to handle btree format files.
+ * Compute the worst-case number of indirect blocks that will be used
+ * for ip's delayed extent of length "len".
*/
-STATIC int /* error */
-xfs_bmap_add_attrfork_btree(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first block allocated */
- xfs_bmap_free_t *flist, /* blocks to free at commit */
- int *flags) /* inode logging flags */
+STATIC xfs_filblks_t
+xfs_bmap_worst_indlen(
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_filblks_t len) /* delayed extent length */
{
- xfs_btree_cur_t *cur; /* btree cursor */
- int error; /* error return value */
- xfs_mount_t *mp; /* file system mount struct */
- int stat; /* newroot status */
+ int level; /* btree level number */
+ int maxrecs; /* maximum record count at this level */
+ xfs_mount_t *mp; /* mount structure */
+ xfs_filblks_t rval; /* return value */
mp = ip->i_mount;
- if (ip->i_df.if_broot_bytes <= XFS_IFORK_DSIZE(ip))
- *flags |= XFS_ILOG_DBROOT;
- else {
- cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK);
- cur->bc_private.b.flist = flist;
- cur->bc_private.b.firstblock = *firstblock;
- if ((error = xfs_bmbt_lookup_ge(cur, 0, 0, 0, &stat)))
- goto error0;
- /* must be at least one entry */
- XFS_WANT_CORRUPTED_GOTO(stat == 1, error0);
- if ((error = xfs_btree_new_iroot(cur, flags, &stat)))
- goto error0;
- if (stat == 0) {
- xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
- return XFS_ERROR(ENOSPC);
- }
- *firstblock = cur->bc_private.b.firstblock;
- cur->bc_private.b.allocated = 0;
- xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
+ maxrecs = mp->m_bmap_dmxr[0];
+ for (level = 0, rval = 0;
+ level < XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK);
+ level++) {
+ len += maxrecs - 1;
+ do_div(len, maxrecs);
+ rval += len;
+ if (len == 1)
+ return rval + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) -
+ level - 1;
+ if (level == 0)
+ maxrecs = mp->m_bmap_dmxr[1];
}
- return 0;
-error0:
- xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
- return error;
+ return rval;
}
/*
- * Called from xfs_bmap_add_attrfork to handle extents format files.
+ * Calculate the default attribute fork offset for newly created inodes.
*/
-STATIC int /* error */
-xfs_bmap_add_attrfork_extents(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first block allocated */
- xfs_bmap_free_t *flist, /* blocks to free at commit */
- int *flags) /* inode logging flags */
+uint
+xfs_default_attroffset(
+ struct xfs_inode *ip)
{
- xfs_btree_cur_t *cur; /* bmap btree cursor */
- int error; /* error return value */
+ struct xfs_mount *mp = ip->i_mount;
+ uint offset;
- if (ip->i_d.di_nextents * sizeof(xfs_bmbt_rec_t) <= XFS_IFORK_DSIZE(ip))
- return 0;
- cur = NULL;
- error = xfs_bmap_extents_to_btree(tp, ip, firstblock, flist, &cur, 0,
- flags, XFS_DATA_FORK);
- if (cur) {
- cur->bc_private.b.allocated = 0;
- xfs_btree_del_cursor(cur,
- error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+ if (mp->m_sb.sb_inodesize == 256) {
+ offset = XFS_LITINO(mp) -
+ XFS_BMDR_SPACE_CALC(MINABTPTRS);
+ } else {
+ offset = XFS_BMDR_SPACE_CALC(6 * MINABTPTRS);
}
- return error;
+
+ ASSERT(offset < XFS_LITINO(mp));
+ return offset;
}
/*
- * Block initialisation functions for local to extent format conversion.
- * As these get more complex, they will be moved to the relevant files,
- * but for now they are too simple to worry about.
+ * Helper routine to reset inode di_forkoff field when switching
+ * attribute fork from local to extent format - we reset it where
+ * possible to make space available for inline data fork extents.
*/
STATIC void
-xfs_bmap_local_to_extents_init_fn(
- struct xfs_buf *bp,
- struct xfs_inode *ip,
- struct xfs_ifork *ifp)
+xfs_bmap_forkoff_reset(
+ xfs_mount_t *mp,
+ xfs_inode_t *ip,
+ int whichfork)
{
- bp->b_ops = &xfs_bmbt_buf_ops;
- memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+ if (whichfork == XFS_ATTR_FORK &&
+ ip->i_d.di_format != XFS_DINODE_FMT_DEV &&
+ ip->i_d.di_format != XFS_DINODE_FMT_UUID &&
+ ip->i_d.di_format != XFS_DINODE_FMT_BTREE) {
+ uint dfl_forkoff = xfs_default_attroffset(ip) >> 3;
+
+ if (dfl_forkoff > ip->i_d.di_forkoff)
+ ip->i_d.di_forkoff = dfl_forkoff;
+ }
}
+/*
+ * Extent tree block counting routines.
+ */
+
+/*
+ * Count leaf blocks given a range of extent records.
+ */
STATIC void
-xfs_symlink_local_to_remote(
- struct xfs_buf *bp,
- struct xfs_inode *ip,
- struct xfs_ifork *ifp)
+xfs_bmap_count_leaves(
+ xfs_ifork_t *ifp,
+ xfs_extnum_t idx,
+ int numrecs,
+ int *count)
{
- /* remote symlink blocks are not verifiable until CRCs come along */
- bp->b_ops = NULL;
- memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+ int b;
+
+ for (b = 0; b < numrecs; b++) {
+ xfs_bmbt_rec_host_t *frp = xfs_iext_get_ext(ifp, idx + b);
+ *count += xfs_bmbt_get_blockcount(frp);
+ }
}
/*
- * Called from xfs_bmap_add_attrfork to handle local format files. Each
- * different data fork content type needs a different callout to do the
- * conversion. Some are basic and only require special block initialisation
- * callouts for the data formating, others (directories) are so specialised they
- * handle everything themselves.
- *
- * XXX (dgc): investigate whether directory conversion can use the generic
- * formatting callout. It should be possible - it's just a very complex
- * formatter. it would also require passing the transaction through to the init
- * function.
+ * Count leaf blocks given a range of extent records originally
+ * in btree format.
*/
-STATIC int /* error */
-xfs_bmap_add_attrfork_local(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first block allocated */
- xfs_bmap_free_t *flist, /* blocks to free at commit */
- int *flags) /* inode logging flags */
+STATIC void
+xfs_bmap_disk_count_leaves(
+ struct xfs_mount *mp,
+ struct xfs_btree_block *block,
+ int numrecs,
+ int *count)
{
- xfs_da_args_t dargs; /* args for dir/attr code */
-
- if (ip->i_df.if_bytes <= XFS_IFORK_DSIZE(ip))
- return 0;
+ int b;
+ xfs_bmbt_rec_t *frp;
- if (S_ISDIR(ip->i_d.di_mode)) {
- memset(&dargs, 0, sizeof(dargs));
- dargs.dp = ip;
- dargs.firstblock = firstblock;
- dargs.flist = flist;
- dargs.total = ip->i_mount->m_dirblkfsbs;
- dargs.whichfork = XFS_DATA_FORK;
- dargs.trans = tp;
- return xfs_dir2_sf_to_block(&dargs);
+ for (b = 1; b <= numrecs; b++) {
+ frp = XFS_BMBT_REC_ADDR(mp, block, b);
+ *count += xfs_bmbt_disk_get_blockcount(frp);
}
-
- if (S_ISLNK(ip->i_d.di_mode))
- return xfs_bmap_local_to_extents(tp, ip, firstblock, 1,
- flags, XFS_DATA_FORK,
- xfs_symlink_local_to_remote);
-
- return xfs_bmap_local_to_extents(tp, ip, firstblock, 1, flags,
- XFS_DATA_FORK,
- xfs_bmap_local_to_extents_init_fn);
}
/*
- * Convert a delayed allocation to a real allocation.
+ * Recursively walks each level of a btree
+ * to count total fsblocks is use.
*/
-STATIC int /* error */
-xfs_bmap_add_extent_delay_real(
- struct xfs_bmalloca *bma)
+STATIC int /* error */
+xfs_bmap_count_tree(
+ xfs_mount_t *mp, /* file system mount point */
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_ifork_t *ifp, /* inode fork pointer */
+ xfs_fsblock_t blockno, /* file system block number */
+ int levelin, /* level in btree */
+ int *count) /* Count of blocks */
{
- struct xfs_bmbt_irec *new = &bma->got;
- int diff; /* temp value */
- xfs_bmbt_rec_host_t *ep; /* extent entry for idx */
- int error; /* error return value */
- int i; /* temp state */
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_fileoff_t new_endoff; /* end offset of new entry */
- xfs_bmbt_irec_t r[3]; /* neighbor extent entries */
- /* left is 0, right is 1, prev is 2 */
- int rval=0; /* return value (logging flags) */
- int state = 0;/* state bits, accessed thru macros */
- xfs_filblks_t da_new; /* new count del alloc blocks used */
- xfs_filblks_t da_old; /* old count del alloc blocks used */
- xfs_filblks_t temp=0; /* value for da_new calculations */
- xfs_filblks_t temp2=0;/* value for da_new calculations */
- int tmp_rval; /* partial logging flags */
+ int error;
+ xfs_buf_t *bp, *nbp;
+ int level = levelin;
+ __be64 *pp;
+ xfs_fsblock_t bno = blockno;
+ xfs_fsblock_t nextbno;
+ struct xfs_btree_block *block, *nextblock;
+ int numrecs;
- ifp = XFS_IFORK_PTR(bma->ip, XFS_DATA_FORK);
+ error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, XFS_BMAP_BTREE_REF,
+ &xfs_bmbt_buf_ops);
+ if (error)
+ return error;
+ *count += 1;
+ block = XFS_BUF_TO_BLOCK(bp);
- ASSERT(bma->idx >= 0);
- ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
- ASSERT(!isnullstartblock(new->br_startblock));
- ASSERT(!bma->cur ||
- (bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
+ if (--level) {
+ /* Not at node above leaves, count this level of nodes */
+ nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
+ while (nextbno != NULLFSBLOCK) {
+ error = xfs_btree_read_bufl(mp, tp, nextbno, 0, &nbp,
+ XFS_BMAP_BTREE_REF,
+ &xfs_bmbt_buf_ops);
+ if (error)
+ return error;
+ *count += 1;
+ nextblock = XFS_BUF_TO_BLOCK(nbp);
+ nextbno = be64_to_cpu(nextblock->bb_u.l.bb_rightsib);
+ xfs_trans_brelse(tp, nbp);
+ }
- XFS_STATS_INC(xs_add_exlist);
+ /* Dive to the next level */
+ pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
+ bno = be64_to_cpu(*pp);
+ if (unlikely((error =
+ xfs_bmap_count_tree(mp, tp, ifp, bno, level, count)) < 0)) {
+ xfs_trans_brelse(tp, bp);
+ XFS_ERROR_REPORT("xfs_bmap_count_tree(1)",
+ XFS_ERRLEVEL_LOW, mp);
+ return XFS_ERROR(EFSCORRUPTED);
+ }
+ xfs_trans_brelse(tp, bp);
+ } else {
+ /* count all level 1 nodes and their leaves */
+ for (;;) {
+ nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
+ numrecs = be16_to_cpu(block->bb_numrecs);
+ xfs_bmap_disk_count_leaves(mp, block, numrecs, count);
+ xfs_trans_brelse(tp, bp);
+ if (nextbno == NULLFSBLOCK)
+ break;
+ bno = nextbno;
+ error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
+ XFS_BMAP_BTREE_REF,
+ &xfs_bmbt_buf_ops);
+ if (error)
+ return error;
+ *count += 1;
+ block = XFS_BUF_TO_BLOCK(bp);
+ }
+ }
+ return 0;
+}
-#define LEFT r[0]
-#define RIGHT r[1]
-#define PREV r[2]
+/*
+ * Count fsblocks of the given fork.
+ */
+int /* error */
+xfs_bmap_count_blocks(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode */
+ int whichfork, /* data or attr fork */
+ int *count) /* out: count of blocks */
+{
+ struct xfs_btree_block *block; /* current btree block */
+ xfs_fsblock_t bno; /* block # of "block" */
+ xfs_ifork_t *ifp; /* fork structure */
+ int level; /* btree level, for checking */
+ xfs_mount_t *mp; /* file system mount structure */
+ __be64 *pp; /* pointer to block address */
+
+ bno = NULLFSBLOCK;
+ mp = ip->i_mount;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ if ( XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS ) {
+ xfs_bmap_count_leaves(ifp, 0,
+ ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t),
+ count);
+ return 0;
+ }
/*
- * Set up a bunch of variables to make the tests simpler.
+ * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
*/
- ep = xfs_iext_get_ext(ifp, bma->idx);
- xfs_bmbt_get_all(ep, &PREV);
- new_endoff = new->br_startoff + new->br_blockcount;
- ASSERT(PREV.br_startoff <= new->br_startoff);
- ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
+ block = ifp->if_broot;
+ level = be16_to_cpu(block->bb_level);
+ ASSERT(level > 0);
+ pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
+ bno = be64_to_cpu(*pp);
+ ASSERT(bno != NULLDFSBNO);
+ ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
+ ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
- da_old = startblockval(PREV.br_startblock);
- da_new = 0;
+ if (unlikely(xfs_bmap_count_tree(mp, tp, ifp, bno, level, count) < 0)) {
+ XFS_ERROR_REPORT("xfs_bmap_count_blocks(2)", XFS_ERRLEVEL_LOW,
+ mp);
+ return XFS_ERROR(EFSCORRUPTED);
+ }
- /*
- * Set flags determining what part of the previous delayed allocation
- * extent is being replaced by a real allocation.
- */
- if (PREV.br_startoff == new->br_startoff)
- state |= BMAP_LEFT_FILLING;
- if (PREV.br_startoff + PREV.br_blockcount == new_endoff)
- state |= BMAP_RIGHT_FILLING;
+ return 0;
+}
- /*
- * Check and set flags if this segment has a left neighbor.
- * Don't set contiguous if the combined extent would be too large.
- */
- if (bma->idx > 0) {
- state |= BMAP_LEFT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &LEFT);
+/*
+ * Debug/sanity checking code
+ */
- if (isnullstartblock(LEFT.br_startblock))
- state |= BMAP_LEFT_DELAY;
- }
+STATIC int
+xfs_bmap_sanity_check(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp,
+ int level)
+{
+ struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
- if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
- LEFT.br_startoff + LEFT.br_blockcount == new->br_startoff &&
- LEFT.br_startblock + LEFT.br_blockcount == new->br_startblock &&
- LEFT.br_state == new->br_state &&
- LEFT.br_blockcount + new->br_blockcount <= MAXEXTLEN)
- state |= BMAP_LEFT_CONTIG;
+ if (block->bb_magic != cpu_to_be32(XFS_BMAP_MAGIC) ||
+ be16_to_cpu(block->bb_level) != level ||
+ be16_to_cpu(block->bb_numrecs) == 0 ||
+ be16_to_cpu(block->bb_numrecs) > mp->m_bmap_dmxr[level != 0])
+ return 0;
+ return 1;
+}
- /*
- * Check and set flags if this segment has a right neighbor.
- * Don't set contiguous if the combined extent would be too large.
- * Also check for all-three-contiguous being too large.
- */
- if (bma->idx < bma->ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
- state |= BMAP_RIGHT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx + 1), &RIGHT);
+#ifdef DEBUG
+STATIC struct xfs_buf *
+xfs_bmap_get_bp(
+ struct xfs_btree_cur *cur,
+ xfs_fsblock_t bno)
+{
+ struct xfs_log_item_desc *lidp;
+ int i;
- if (isnullstartblock(RIGHT.br_startblock))
- state |= BMAP_RIGHT_DELAY;
+ if (!cur)
+ return NULL;
+
+ for (i = 0; i < XFS_BTREE_MAXLEVELS; i++) {
+ if (!cur->bc_bufs[i])
+ break;
+ if (XFS_BUF_ADDR(cur->bc_bufs[i]) == bno)
+ return cur->bc_bufs[i];
}
- if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
- new_endoff == RIGHT.br_startoff &&
- new->br_startblock + new->br_blockcount == RIGHT.br_startblock &&
- new->br_state == RIGHT.br_state &&
- new->br_blockcount + RIGHT.br_blockcount <= MAXEXTLEN &&
- ((state & (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
- BMAP_RIGHT_FILLING)) !=
- (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
- BMAP_RIGHT_FILLING) ||
- LEFT.br_blockcount + new->br_blockcount + RIGHT.br_blockcount
- <= MAXEXTLEN))
- state |= BMAP_RIGHT_CONTIG;
+ /* Chase down all the log items to see if the bp is there */
+ list_for_each_entry(lidp, &cur->bc_tp->t_items, lid_trans) {
+ struct xfs_buf_log_item *bip;
+ bip = (struct xfs_buf_log_item *)lidp->lid_item;
+ if (bip->bli_item.li_type == XFS_LI_BUF &&
+ XFS_BUF_ADDR(bip->bli_buf) == bno)
+ return bip->bli_buf;
+ }
- error = 0;
- /*
- * Switch out based on the FILLING and CONTIG state bits.
- */
- switch (state & (BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
- BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG)) {
- case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
- BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
- /*
- * Filling in all of a previously delayed allocation extent.
- * The left and right neighbors are both contiguous with new.
- */
- bma->idx--;
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
- LEFT.br_blockcount + PREV.br_blockcount +
- RIGHT.br_blockcount);
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+ return NULL;
+}
- xfs_iext_remove(bma->ip, bma->idx + 1, 2, state);
- bma->ip->i_d.di_nextents--;
- if (bma->cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
- RIGHT.br_startblock,
- RIGHT.br_blockcount, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_btree_delete(bma->cur, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_btree_decrement(bma->cur, 0, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
- LEFT.br_startblock,
- LEFT.br_blockcount +
- PREV.br_blockcount +
- RIGHT.br_blockcount, LEFT.br_state);
- if (error)
- goto done;
- }
- break;
+STATIC void
+xfs_check_block(
+ struct xfs_btree_block *block,
+ xfs_mount_t *mp,
+ int root,
+ short sz)
+{
+ int i, j, dmxr;
+ __be64 *pp, *thispa; /* pointer to block address */
+ xfs_bmbt_key_t *prevp, *keyp;
- case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
- /*
- * Filling in all of a previously delayed allocation extent.
- * The left neighbor is contiguous, the right is not.
- */
- bma->idx--;
+ ASSERT(be16_to_cpu(block->bb_level) > 0);
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
- LEFT.br_blockcount + PREV.br_blockcount);
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+ prevp = NULL;
+ for( i = 1; i <= xfs_btree_get_numrecs(block); i++) {
+ dmxr = mp->m_bmap_dmxr[0];
+ keyp = XFS_BMBT_KEY_ADDR(mp, block, i);
- xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
- if (bma->cur == NULL)
- rval = XFS_ILOG_DEXT;
- else {
- rval = 0;
- error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
- LEFT.br_startblock, LEFT.br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
- LEFT.br_startblock,
- LEFT.br_blockcount +
- PREV.br_blockcount, LEFT.br_state);
- if (error)
- goto done;
+ if (prevp) {
+ ASSERT(be64_to_cpu(prevp->br_startoff) <
+ be64_to_cpu(keyp->br_startoff));
}
- break;
+ prevp = keyp;
- case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
/*
- * Filling in all of a previously delayed allocation extent.
- * The right neighbor is contiguous, the left is not.
+ * Compare the block numbers to see if there are dups.
*/
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_startblock(ep, new->br_startblock);
- xfs_bmbt_set_blockcount(ep,
- PREV.br_blockcount + RIGHT.br_blockcount);
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+ if (root)
+ pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, i, sz);
+ else
+ pp = XFS_BMBT_PTR_ADDR(mp, block, i, dmxr);
- xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
- if (bma->cur == NULL)
- rval = XFS_ILOG_DEXT;
- else {
- rval = 0;
- error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
- RIGHT.br_startblock,
- RIGHT.br_blockcount, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, PREV.br_startoff,
- new->br_startblock,
- PREV.br_blockcount +
- RIGHT.br_blockcount, PREV.br_state);
- if (error)
- goto done;
+ for (j = i+1; j <= be16_to_cpu(block->bb_numrecs); j++) {
+ if (root)
+ thispa = XFS_BMAP_BROOT_PTR_ADDR(mp, block, j, sz);
+ else
+ thispa = XFS_BMBT_PTR_ADDR(mp, block, j, dmxr);
+ if (*thispa == *pp) {
+ xfs_warn(mp, "%s: thispa(%d) == pp(%d) %Ld",
+ __func__, j, i,
+ (unsigned long long)be64_to_cpu(*thispa));
+ panic("%s: ptrs are equal in node\n",
+ __func__);
+ }
}
- break;
+ }
+}
- case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
- /*
- * Filling in all of a previously delayed allocation extent.
- * Neither the left nor right neighbors are contiguous with
- * the new one.
- */
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_startblock(ep, new->br_startblock);
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+/*
+ * Check that the extents for the inode ip are in the right order in all
+ * btree leaves.
+ */
- bma->ip->i_d.di_nextents++;
- if (bma->cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 0, done);
- bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
- error = xfs_btree_insert(bma->cur, &i);
+STATIC void
+xfs_bmap_check_leaf_extents(
+ xfs_btree_cur_t *cur, /* btree cursor or null */
+ xfs_inode_t *ip, /* incore inode pointer */
+ int whichfork) /* data or attr fork */
+{
+ struct xfs_btree_block *block; /* current btree block */
+ xfs_fsblock_t bno; /* block # of "block" */
+ xfs_buf_t *bp; /* buffer for "block" */
+ int error; /* error return value */
+ xfs_extnum_t i=0, j; /* index into the extents list */
+ xfs_ifork_t *ifp; /* fork structure */
+ int level; /* btree level, for checking */
+ xfs_mount_t *mp; /* file system mount structure */
+ __be64 *pp; /* pointer to block address */
+ xfs_bmbt_rec_t *ep; /* pointer to current extent */
+ xfs_bmbt_rec_t last = {0, 0}; /* last extent in prev block */
+ xfs_bmbt_rec_t *nextp; /* pointer to next extent */
+ int bp_release = 0;
+
+ if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE) {
+ return;
+ }
+
+ bno = NULLFSBLOCK;
+ mp = ip->i_mount;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ block = ifp->if_broot;
+ /*
+ * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
+ */
+ level = be16_to_cpu(block->bb_level);
+ ASSERT(level > 0);
+ xfs_check_block(block, mp, 1, ifp->if_broot_bytes);
+ pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
+ bno = be64_to_cpu(*pp);
+
+ ASSERT(bno != NULLDFSBNO);
+ ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
+ ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
+
+ /*
+ * Go down the tree until leaf level is reached, following the first
+ * pointer (leftmost) at each level.
+ */
+ while (level-- > 0) {
+ /* See if buf is in cur first */
+ bp_release = 0;
+ bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno));
+ if (!bp) {
+ bp_release = 1;
+ error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp,
+ XFS_BMAP_BTREE_REF,
+ &xfs_bmbt_buf_ops);
if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ goto error_norelse;
}
- break;
+ block = XFS_BUF_TO_BLOCK(bp);
+ XFS_WANT_CORRUPTED_GOTO(
+ xfs_bmap_sanity_check(mp, bp, level),
+ error0);
+ if (level == 0)
+ break;
- case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG:
/*
- * Filling in the first part of a previous delayed allocation.
- * The left neighbor is contiguous.
+ * Check this block for basic sanity (increasing keys and
+ * no duplicate blocks).
*/
- trace_xfs_bmap_pre_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx - 1),
- LEFT.br_blockcount + new->br_blockcount);
- xfs_bmbt_set_startoff(ep,
- PREV.br_startoff + new->br_blockcount);
- trace_xfs_bmap_post_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
- temp = PREV.br_blockcount - new->br_blockcount;
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep, temp);
- if (bma->cur == NULL)
- rval = XFS_ILOG_DEXT;
- else {
- rval = 0;
- error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
- LEFT.br_startblock, LEFT.br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
- LEFT.br_startblock,
- LEFT.br_blockcount +
- new->br_blockcount,
- LEFT.br_state);
- if (error)
- goto done;
+ xfs_check_block(block, mp, 0, 0);
+ pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
+ bno = be64_to_cpu(*pp);
+ XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0);
+ if (bp_release) {
+ bp_release = 0;
+ xfs_trans_brelse(NULL, bp);
}
- da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
- startblockval(PREV.br_startblock));
- xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+ }
- bma->idx--;
- break;
+ /*
+ * Here with bp and block set to the leftmost leaf node in the tree.
+ */
+ i = 0;
+
+ /*
+ * Loop over all leaf nodes checking that all extents are in the right order.
+ */
+ for (;;) {
+ xfs_fsblock_t nextbno;
+ xfs_extnum_t num_recs;
+
+
+ num_recs = xfs_btree_get_numrecs(block);
- case BMAP_LEFT_FILLING:
/*
- * Filling in the first part of a previous delayed allocation.
- * The left neighbor is not contiguous.
+ * Read-ahead the next leaf block, if any.
*/
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_startoff(ep, new_endoff);
- temp = PREV.br_blockcount - new->br_blockcount;
- xfs_bmbt_set_blockcount(ep, temp);
- xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
- bma->ip->i_d.di_nextents++;
- if (bma->cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 0, done);
- bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
- error = xfs_btree_insert(bma->cur, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- }
- if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
- error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
- bma->firstblock, bma->flist,
- &bma->cur, 1, &tmp_rval, XFS_DATA_FORK);
- rval |= tmp_rval;
- if (error)
- goto done;
- }
- da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
- startblockval(PREV.br_startblock) -
- (bma->cur ? bma->cur->bc_private.b.allocated : 0));
- ep = xfs_iext_get_ext(ifp, bma->idx + 1);
- xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
- trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
- break;
+ nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
- case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
/*
- * Filling in the last part of a previous delayed allocation.
- * The right neighbor is contiguous with the new allocation.
+ * Check all the extents to make sure they are OK.
+ * If we had a previous block, the last entry should
+ * conform with the first entry in this one.
*/
- temp = PREV.br_blockcount - new->br_blockcount;
- trace_xfs_bmap_pre_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep, temp);
- xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->idx + 1),
- new->br_startoff, new->br_startblock,
- new->br_blockcount + RIGHT.br_blockcount,
- RIGHT.br_state);
- trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
- if (bma->cur == NULL)
- rval = XFS_ILOG_DEXT;
- else {
- rval = 0;
- error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
- RIGHT.br_startblock,
- RIGHT.br_blockcount, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, new->br_startoff,
- new->br_startblock,
- new->br_blockcount +
- RIGHT.br_blockcount,
- RIGHT.br_state);
- if (error)
- goto done;
- }
-
- da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
- startblockval(PREV.br_startblock));
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
- bma->idx++;
- break;
+ ep = XFS_BMBT_REC_ADDR(mp, block, 1);
+ if (i) {
+ ASSERT(xfs_bmbt_disk_get_startoff(&last) +
+ xfs_bmbt_disk_get_blockcount(&last) <=
+ xfs_bmbt_disk_get_startoff(ep));
+ }
+ for (j = 1; j < num_recs; j++) {
+ nextp = XFS_BMBT_REC_ADDR(mp, block, j + 1);
+ ASSERT(xfs_bmbt_disk_get_startoff(ep) +
+ xfs_bmbt_disk_get_blockcount(ep) <=
+ xfs_bmbt_disk_get_startoff(nextp));
+ ep = nextp;
+ }
- case BMAP_RIGHT_FILLING:
+ last = *ep;
+ i += num_recs;
+ if (bp_release) {
+ bp_release = 0;
+ xfs_trans_brelse(NULL, bp);
+ }
+ bno = nextbno;
/*
- * Filling in the last part of a previous delayed allocation.
- * The right neighbor is not contiguous.
+ * If we've reached the end, stop.
*/
- temp = PREV.br_blockcount - new->br_blockcount;
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep, temp);
- xfs_iext_insert(bma->ip, bma->idx + 1, 1, new, state);
- bma->ip->i_d.di_nextents++;
- if (bma->cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 0, done);
- bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
- error = xfs_btree_insert(bma->cur, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- }
-
- if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
- error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
- bma->firstblock, bma->flist, &bma->cur, 1,
- &tmp_rval, XFS_DATA_FORK);
- rval |= tmp_rval;
- if (error)
- goto done;
- }
- da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
- startblockval(PREV.br_startblock) -
- (bma->cur ? bma->cur->bc_private.b.allocated : 0));
- ep = xfs_iext_get_ext(ifp, bma->idx);
- xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-
- bma->idx++;
- break;
-
- case 0:
- /*
- * Filling in the middle part of a previous delayed allocation.
- * Contiguity is impossible here.
- * This case is avoided almost all the time.
- *
- * We start with a delayed allocation:
- *
- * +ddddddddddddddddddddddddddddddddddddddddddddddddddddddd+
- * PREV @ idx
- *
- * and we are allocating:
- * +rrrrrrrrrrrrrrrrr+
- * new
- *
- * and we set it up for insertion as:
- * +ddddddddddddddddddd+rrrrrrrrrrrrrrrrr+ddddddddddddddddd+
- * new
- * PREV @ idx LEFT RIGHT
- * inserted at idx + 1
- */
- temp = new->br_startoff - PREV.br_startoff;
- temp2 = PREV.br_startoff + PREV.br_blockcount - new_endoff;
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, 0, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep, temp); /* truncate PREV */
- LEFT = *new;
- RIGHT.br_state = PREV.br_state;
- RIGHT.br_startblock = nullstartblock(
- (int)xfs_bmap_worst_indlen(bma->ip, temp2));
- RIGHT.br_startoff = new_endoff;
- RIGHT.br_blockcount = temp2;
- /* insert LEFT (r[0]) and RIGHT (r[1]) at the same time */
- xfs_iext_insert(bma->ip, bma->idx + 1, 2, &LEFT, state);
- bma->ip->i_d.di_nextents++;
- if (bma->cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 0, done);
- bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
- error = xfs_btree_insert(bma->cur, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- }
+ if (bno == NULLFSBLOCK)
+ break;
- if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
- error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
- bma->firstblock, bma->flist, &bma->cur,
- 1, &tmp_rval, XFS_DATA_FORK);
- rval |= tmp_rval;
- if (error)
- goto done;
- }
- temp = xfs_bmap_worst_indlen(bma->ip, temp);
- temp2 = xfs_bmap_worst_indlen(bma->ip, temp2);
- diff = (int)(temp + temp2 - startblockval(PREV.br_startblock) -
- (bma->cur ? bma->cur->bc_private.b.allocated : 0));
- if (diff > 0) {
- error = xfs_icsb_modify_counters(bma->ip->i_mount,
- XFS_SBS_FDBLOCKS,
- -((int64_t)diff), 0);
- ASSERT(!error);
+ bp_release = 0;
+ bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno));
+ if (!bp) {
+ bp_release = 1;
+ error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp,
+ XFS_BMAP_BTREE_REF,
+ &xfs_bmbt_buf_ops);
if (error)
- goto done;
+ goto error_norelse;
}
-
- ep = xfs_iext_get_ext(ifp, bma->idx);
- xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
- trace_xfs_bmap_pre_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
- xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, bma->idx + 2),
- nullstartblock((int)temp2));
- trace_xfs_bmap_post_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
-
- bma->idx++;
- da_new = temp + temp2;
- break;
-
- case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- case BMAP_LEFT_FILLING | BMAP_RIGHT_CONTIG:
- case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
- case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- case BMAP_LEFT_CONTIG:
- case BMAP_RIGHT_CONTIG:
- /*
- * These cases are all impossible.
- */
- ASSERT(0);
- }
-
- /* convert to a btree if necessary */
- if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
- int tmp_logflags; /* partial log flag return val */
-
- ASSERT(bma->cur == NULL);
- error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
- bma->firstblock, bma->flist, &bma->cur,
- da_old > 0, &tmp_logflags, XFS_DATA_FORK);
- bma->logflags |= tmp_logflags;
- if (error)
- goto done;
+ block = XFS_BUF_TO_BLOCK(bp);
}
-
- /* adjust for changes in reserved delayed indirect blocks */
- if (da_old || da_new) {
- temp = da_new;
- if (bma->cur)
- temp += bma->cur->bc_private.b.allocated;
- ASSERT(temp <= da_old);
- if (temp < da_old)
- xfs_icsb_modify_counters(bma->ip->i_mount,
- XFS_SBS_FDBLOCKS,
- (int64_t)(da_old - temp), 0);
+ if (bp_release) {
+ bp_release = 0;
+ xfs_trans_brelse(NULL, bp);
}
+ return;
- /* clear out the allocated field, done with it now in any case. */
- if (bma->cur)
- bma->cur->bc_private.b.allocated = 0;
-
- xfs_bmap_check_leaf_extents(bma->cur, bma->ip, XFS_DATA_FORK);
-done:
- bma->logflags |= rval;
- return error;
-#undef LEFT
-#undef RIGHT
-#undef PREV
+error0:
+ xfs_warn(mp, "%s: at error0", __func__);
+ if (bp_release)
+ xfs_trans_brelse(NULL, bp);
+error_norelse:
+ xfs_warn(mp, "%s: BAD after btree leaves for %d extents",
+ __func__, i);
+ panic("%s: CORRUPTED BTREE OR SOMETHING", __func__);
+ return;
}
/*
- * Convert an unwritten allocation to a real allocation or vice versa.
+ * Add bmap trace insert entries for all the contents of the extent records.
*/
-STATIC int /* error */
-xfs_bmap_add_extent_unwritten_real(
- struct xfs_trans *tp,
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_extnum_t *idx, /* extent number to update/insert */
- xfs_btree_cur_t **curp, /* if *curp is null, not a btree */
- xfs_bmbt_irec_t *new, /* new data to add to file extents */
- xfs_fsblock_t *first, /* pointer to firstblock variable */
- xfs_bmap_free_t *flist, /* list of extents to be freed */
- int *logflagsp) /* inode logging flags */
+void
+xfs_bmap_trace_exlist(
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_extnum_t cnt, /* count of entries in the list */
+ int whichfork, /* data or attr fork */
+ unsigned long caller_ip)
{
- xfs_btree_cur_t *cur; /* btree cursor */
- xfs_bmbt_rec_host_t *ep; /* extent entry for idx */
- int error; /* error return value */
- int i; /* temp state */
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_fileoff_t new_endoff; /* end offset of new entry */
- xfs_exntst_t newext; /* new extent state */
- xfs_exntst_t oldext; /* old extent state */
- xfs_bmbt_irec_t r[3]; /* neighbor extent entries */
- /* left is 0, right is 1, prev is 2 */
- int rval=0; /* return value (logging flags) */
- int state = 0;/* state bits, accessed thru macros */
-
- *logflagsp = 0;
+ xfs_extnum_t idx; /* extent record index */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ int state = 0;
- cur = *curp;
- ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+ if (whichfork == XFS_ATTR_FORK)
+ state |= BMAP_ATTRFORK;
- ASSERT(*idx >= 0);
- ASSERT(*idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
- ASSERT(!isnullstartblock(new->br_startblock));
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ ASSERT(cnt == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)));
+ for (idx = 0; idx < cnt; idx++)
+ trace_xfs_extlist(ip, idx, whichfork, caller_ip);
+}
- XFS_STATS_INC(xs_add_exlist);
+/*
+ * Validate that the bmbt_irecs being returned from bmapi are valid
+ * given the callers original parameters. Specifically check the
+ * ranges of the returned irecs to ensure that they only extent beyond
+ * the given parameters if the XFS_BMAPI_ENTIRE flag was set.
+ */
+STATIC void
+xfs_bmap_validate_ret(
+ xfs_fileoff_t bno,
+ xfs_filblks_t len,
+ int flags,
+ xfs_bmbt_irec_t *mval,
+ int nmap,
+ int ret_nmap)
+{
+ int i; /* index to map values */
-#define LEFT r[0]
-#define RIGHT r[1]
-#define PREV r[2]
+ ASSERT(ret_nmap <= nmap);
- /*
- * Set up a bunch of variables to make the tests simpler.
- */
- error = 0;
- ep = xfs_iext_get_ext(ifp, *idx);
- xfs_bmbt_get_all(ep, &PREV);
- newext = new->br_state;
- oldext = (newext == XFS_EXT_UNWRITTEN) ?
- XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
- ASSERT(PREV.br_state == oldext);
- new_endoff = new->br_startoff + new->br_blockcount;
- ASSERT(PREV.br_startoff <= new->br_startoff);
- ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
+ for (i = 0; i < ret_nmap; i++) {
+ ASSERT(mval[i].br_blockcount > 0);
+ if (!(flags & XFS_BMAPI_ENTIRE)) {
+ ASSERT(mval[i].br_startoff >= bno);
+ ASSERT(mval[i].br_blockcount <= len);
+ ASSERT(mval[i].br_startoff + mval[i].br_blockcount <=
+ bno + len);
+ } else {
+ ASSERT(mval[i].br_startoff < bno + len);
+ ASSERT(mval[i].br_startoff + mval[i].br_blockcount >
+ bno);
+ }
+ ASSERT(i == 0 ||
+ mval[i - 1].br_startoff + mval[i - 1].br_blockcount ==
+ mval[i].br_startoff);
+ ASSERT(mval[i].br_startblock != DELAYSTARTBLOCK &&
+ mval[i].br_startblock != HOLESTARTBLOCK);
+ ASSERT(mval[i].br_state == XFS_EXT_NORM ||
+ mval[i].br_state == XFS_EXT_UNWRITTEN);
+ }
+}
- /*
- * Set flags determining what part of the previous oldext allocation
- * extent is being replaced by a newext allocation.
- */
- if (PREV.br_startoff == new->br_startoff)
- state |= BMAP_LEFT_FILLING;
- if (PREV.br_startoff + PREV.br_blockcount == new_endoff)
- state |= BMAP_RIGHT_FILLING;
+#else
+#define xfs_bmap_check_leaf_extents(cur, ip, whichfork) do { } while (0)
+#define xfs_bmap_validate_ret(bno,len,flags,mval,onmap,nmap)
+#endif /* DEBUG */
- /*
- * Check and set flags if this segment has a left neighbor.
- * Don't set contiguous if the combined extent would be too large.
- */
- if (*idx > 0) {
- state |= BMAP_LEFT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &LEFT);
+/*
+ * bmap free list manipulation functions
+ */
- if (isnullstartblock(LEFT.br_startblock))
- state |= BMAP_LEFT_DELAY;
+/*
+ * Add the extent to the list of extents to be free at transaction end.
+ * The list is maintained sorted (by block number).
+ */
+void
+xfs_bmap_add_free(
+ xfs_fsblock_t bno, /* fs block number of extent */
+ xfs_filblks_t len, /* length of extent */
+ xfs_bmap_free_t *flist, /* list of extents */
+ xfs_mount_t *mp) /* mount point structure */
+{
+ xfs_bmap_free_item_t *cur; /* current (next) element */
+ xfs_bmap_free_item_t *new; /* new element */
+ xfs_bmap_free_item_t *prev; /* previous element */
+#ifdef DEBUG
+ xfs_agnumber_t agno;
+ xfs_agblock_t agbno;
+
+ ASSERT(bno != NULLFSBLOCK);
+ ASSERT(len > 0);
+ ASSERT(len <= MAXEXTLEN);
+ ASSERT(!isnullstartblock(bno));
+ agno = XFS_FSB_TO_AGNO(mp, bno);
+ agbno = XFS_FSB_TO_AGBNO(mp, bno);
+ ASSERT(agno < mp->m_sb.sb_agcount);
+ ASSERT(agbno < mp->m_sb.sb_agblocks);
+ ASSERT(len < mp->m_sb.sb_agblocks);
+ ASSERT(agbno + len <= mp->m_sb.sb_agblocks);
+#endif
+ ASSERT(xfs_bmap_free_item_zone != NULL);
+ new = kmem_zone_alloc(xfs_bmap_free_item_zone, KM_SLEEP);
+ new->xbfi_startblock = bno;
+ new->xbfi_blockcount = (xfs_extlen_t)len;
+ for (prev = NULL, cur = flist->xbf_first;
+ cur != NULL;
+ prev = cur, cur = cur->xbfi_next) {
+ if (cur->xbfi_startblock >= bno)
+ break;
}
+ if (prev)
+ prev->xbfi_next = new;
+ else
+ flist->xbf_first = new;
+ new->xbfi_next = cur;
+ flist->xbf_count++;
+}
- if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
- LEFT.br_startoff + LEFT.br_blockcount == new->br_startoff &&
- LEFT.br_startblock + LEFT.br_blockcount == new->br_startblock &&
- LEFT.br_state == newext &&
- LEFT.br_blockcount + new->br_blockcount <= MAXEXTLEN)
- state |= BMAP_LEFT_CONTIG;
+/*
+ * Remove the entry "free" from the free item list. Prev points to the
+ * previous entry, unless "free" is the head of the list.
+ */
+STATIC void
+xfs_bmap_del_free(
+ xfs_bmap_free_t *flist, /* free item list header */
+ xfs_bmap_free_item_t *prev, /* previous item on list, if any */
+ xfs_bmap_free_item_t *free) /* list item to be freed */
+{
+ if (prev)
+ prev->xbfi_next = free->xbfi_next;
+ else
+ flist->xbf_first = free->xbfi_next;
+ flist->xbf_count--;
+ kmem_zone_free(xfs_bmap_free_item_zone, free);
+}
+
+
+/*
+ * Routine to be called at transaction's end by xfs_bmapi, xfs_bunmapi
+ * caller. Frees all the extents that need freeing, which must be done
+ * last due to locking considerations. We never free any extents in
+ * the first transaction.
+ *
+ * Return 1 if the given transaction was committed and a new one
+ * started, and 0 otherwise in the committed parameter.
+ */
+int /* error */
+xfs_bmap_finish(
+ xfs_trans_t **tp, /* transaction pointer addr */
+ xfs_bmap_free_t *flist, /* i/o: list extents to free */
+ int *committed) /* xact committed or not */
+{
+ xfs_efd_log_item_t *efd; /* extent free data */
+ xfs_efi_log_item_t *efi; /* extent free intention */
+ int error; /* error return value */
+ xfs_bmap_free_item_t *free; /* free extent item */
+ unsigned int logres; /* new log reservation */
+ unsigned int logcount; /* new log count */
+ xfs_mount_t *mp; /* filesystem mount structure */
+ xfs_bmap_free_item_t *next; /* next item on free list */
+ xfs_trans_t *ntp; /* new transaction pointer */
+ ASSERT((*tp)->t_flags & XFS_TRANS_PERM_LOG_RES);
+ if (flist->xbf_count == 0) {
+ *committed = 0;
+ return 0;
+ }
+ ntp = *tp;
+ efi = xfs_trans_get_efi(ntp, flist->xbf_count);
+ for (free = flist->xbf_first; free; free = free->xbfi_next)
+ xfs_trans_log_efi_extent(ntp, efi, free->xbfi_startblock,
+ free->xbfi_blockcount);
+ logres = ntp->t_log_res;
+ logcount = ntp->t_log_count;
+ ntp = xfs_trans_dup(*tp);
+ error = xfs_trans_commit(*tp, 0);
+ *tp = ntp;
+ *committed = 1;
/*
- * Check and set flags if this segment has a right neighbor.
- * Don't set contiguous if the combined extent would be too large.
- * Also check for all-three-contiguous being too large.
+ * We have a new transaction, so we should return committed=1,
+ * even though we're returning an error.
*/
- if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
- state |= BMAP_RIGHT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT);
- if (isnullstartblock(RIGHT.br_startblock))
- state |= BMAP_RIGHT_DELAY;
- }
-
- if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
- new_endoff == RIGHT.br_startoff &&
- new->br_startblock + new->br_blockcount == RIGHT.br_startblock &&
- newext == RIGHT.br_state &&
- new->br_blockcount + RIGHT.br_blockcount <= MAXEXTLEN &&
- ((state & (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
- BMAP_RIGHT_FILLING)) !=
- (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
- BMAP_RIGHT_FILLING) ||
- LEFT.br_blockcount + new->br_blockcount + RIGHT.br_blockcount
- <= MAXEXTLEN))
- state |= BMAP_RIGHT_CONTIG;
+ if (error)
+ return error;
/*
- * Switch out based on the FILLING and CONTIG state bits.
+ * transaction commit worked ok so we can drop the extra ticket
+ * reference that we gained in xfs_trans_dup()
*/
- switch (state & (BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
- BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG)) {
- case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
- BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
- /*
- * Setting all of a previous oldext extent to newext.
- * The left and right neighbors are both contiguous with new.
- */
- --*idx;
+ xfs_log_ticket_put(ntp->t_ticket);
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
- LEFT.br_blockcount + PREV.br_blockcount +
- RIGHT.br_blockcount);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ if ((error = xfs_trans_reserve(ntp, 0, logres, 0, XFS_TRANS_PERM_LOG_RES,
+ logcount)))
+ return error;
+ efd = xfs_trans_get_efd(ntp, efi, flist->xbf_count);
+ for (free = flist->xbf_first; free != NULL; free = next) {
+ next = free->xbfi_next;
+ if ((error = xfs_free_extent(ntp, free->xbfi_startblock,
+ free->xbfi_blockcount))) {
+ /*
+ * The bmap free list will be cleaned up at a
+ * higher level. The EFI will be canceled when
+ * this transaction is aborted.
+ * Need to force shutdown here to make sure it
+ * happens, since this transaction may not be
+ * dirty yet.
+ */
+ mp = ntp->t_mountp;
+ if (!XFS_FORCED_SHUTDOWN(mp))
+ xfs_force_shutdown(mp,
+ (error == EFSCORRUPTED) ?
+ SHUTDOWN_CORRUPT_INCORE :
+ SHUTDOWN_META_IO_ERROR);
+ return error;
+ }
+ xfs_trans_log_efd_extent(ntp, efd, free->xbfi_startblock,
+ free->xbfi_blockcount);
+ xfs_bmap_del_free(flist, NULL, free);
+ }
+ return 0;
+}
- xfs_iext_remove(ip, *idx + 1, 2, state);
- ip->i_d.di_nextents -= 2;
- if (cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
- RIGHT.br_startblock,
- RIGHT.br_blockcount, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_delete(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_decrement(cur, 0, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_delete(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_decrement(cur, 0, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
- LEFT.br_startblock,
- LEFT.br_blockcount + PREV.br_blockcount +
- RIGHT.br_blockcount, LEFT.br_state)))
- goto done;
- }
- break;
-
- case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
- /*
- * Setting all of a previous oldext extent to newext.
- * The left neighbor is contiguous, the right is not.
- */
- --*idx;
+/*
+ * Free up any items left in the list.
+ */
+void
+xfs_bmap_cancel(
+ xfs_bmap_free_t *flist) /* list of bmap_free_items */
+{
+ xfs_bmap_free_item_t *free; /* free list item */
+ xfs_bmap_free_item_t *next;
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
- LEFT.br_blockcount + PREV.br_blockcount);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ if (flist->xbf_count == 0)
+ return;
+ ASSERT(flist->xbf_first != NULL);
+ for (free = flist->xbf_first; free; free = next) {
+ next = free->xbfi_next;
+ xfs_bmap_del_free(flist, NULL, free);
+ }
+ ASSERT(flist->xbf_count == 0);
+}
- xfs_iext_remove(ip, *idx + 1, 1, state);
- ip->i_d.di_nextents--;
- if (cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
- PREV.br_startblock, PREV.br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_delete(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_decrement(cur, 0, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
- LEFT.br_startblock,
- LEFT.br_blockcount + PREV.br_blockcount,
- LEFT.br_state)))
- goto done;
- }
- break;
+/*
+ * Inode fork format manipulation functions
+ */
- case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
- /*
- * Setting all of a previous oldext extent to newext.
- * The right neighbor is contiguous, the left is not.
- */
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep,
- PREV.br_blockcount + RIGHT.br_blockcount);
- xfs_bmbt_set_state(ep, newext);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- xfs_iext_remove(ip, *idx + 1, 1, state);
- ip->i_d.di_nextents--;
- if (cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
- RIGHT.br_startblock,
- RIGHT.br_blockcount, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_delete(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_btree_decrement(cur, 0, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur, new->br_startoff,
- new->br_startblock,
- new->br_blockcount + RIGHT.br_blockcount,
- newext)))
- goto done;
- }
- break;
+/*
+ * Transform a btree format file with only one leaf node, where the
+ * extents list will fit in the inode, into an extents format file.
+ * Since the file extents are already in-core, all we have to do is
+ * give up the space for the btree root and pitch the leaf block.
+ */
+STATIC int /* error */
+xfs_bmap_btree_to_extents(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_btree_cur_t *cur, /* btree cursor */
+ int *logflagsp, /* inode logging flags */
+ int whichfork) /* data or attr fork */
+{
+ /* REFERENCED */
+ struct xfs_btree_block *cblock;/* child btree block */
+ xfs_fsblock_t cbno; /* child block number */
+ xfs_buf_t *cbp; /* child block's buffer */
+ int error; /* error return value */
+ xfs_ifork_t *ifp; /* inode fork data */
+ xfs_mount_t *mp; /* mount point structure */
+ __be64 *pp; /* ptr to block address */
+ struct xfs_btree_block *rblock;/* root btree block */
- case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
- /*
- * Setting all of a previous oldext extent to newext.
- * Neither the left nor right neighbors are contiguous with
- * the new one.
- */
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_state(ep, newext);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ mp = ip->i_mount;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ ASSERT(ifp->if_flags & XFS_IFEXTENTS);
+ ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE);
+ rblock = ifp->if_broot;
+ ASSERT(be16_to_cpu(rblock->bb_level) == 1);
+ ASSERT(be16_to_cpu(rblock->bb_numrecs) == 1);
+ ASSERT(xfs_bmbt_maxrecs(mp, ifp->if_broot_bytes, 0) == 1);
+ pp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, ifp->if_broot_bytes);
+ cbno = be64_to_cpu(*pp);
+ *logflagsp = 0;
+#ifdef DEBUG
+ if ((error = xfs_btree_check_lptr(cur, cbno, 1)))
+ return error;
+#endif
+ error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, XFS_BMAP_BTREE_REF,
+ &xfs_bmbt_buf_ops);
+ if (error)
+ return error;
+ cblock = XFS_BUF_TO_BLOCK(cbp);
+ if ((error = xfs_btree_check_block(cur, cblock, 0, cbp)))
+ return error;
+ xfs_bmap_add_free(cbno, 1, cur->bc_private.b.flist, mp);
+ ip->i_d.di_nblocks--;
+ xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, -1L);
+ xfs_trans_binval(tp, cbp);
+ if (cur->bc_bufs[0] == cbp)
+ cur->bc_bufs[0] = NULL;
+ xfs_iroot_realloc(ip, -1, whichfork);
+ ASSERT(ifp->if_broot == NULL);
+ ASSERT((ifp->if_flags & XFS_IFBROOT) == 0);
+ XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
+ *logflagsp = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
+ return 0;
+}
- if (cur == NULL)
- rval = XFS_ILOG_DEXT;
- else {
- rval = 0;
- if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- newext)))
- goto done;
- }
- break;
+/*
+ * Convert an extents-format file into a btree-format file.
+ * The new file will have a root block (in the inode) and a single child block.
+ */
+STATIC int /* error */
+xfs_bmap_extents_to_btree(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_fsblock_t *firstblock, /* first-block-allocated */
+ xfs_bmap_free_t *flist, /* blocks freed in xaction */
+ xfs_btree_cur_t **curp, /* cursor returned to caller */
+ int wasdel, /* converting a delayed alloc */
+ int *logflagsp, /* inode logging flags */
+ int whichfork) /* data or attr fork */
+{
+ struct xfs_btree_block *ablock; /* allocated (child) bt block */
+ xfs_buf_t *abp; /* buffer for ablock */
+ xfs_alloc_arg_t args; /* allocation arguments */
+ xfs_bmbt_rec_t *arp; /* child record pointer */
+ struct xfs_btree_block *block; /* btree root block */
+ xfs_btree_cur_t *cur; /* bmap btree cursor */
+ xfs_bmbt_rec_host_t *ep; /* extent record pointer */
+ int error; /* error return value */
+ xfs_extnum_t i, cnt; /* extent record index */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_bmbt_key_t *kp; /* root block key pointer */
+ xfs_mount_t *mp; /* mount structure */
+ xfs_extnum_t nextents; /* number of file extents */
+ xfs_bmbt_ptr_t *pp; /* root block address pointer */
- case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG:
- /*
- * Setting the first part of a previous oldext extent to newext.
- * The left neighbor is contiguous.
- */
- trace_xfs_bmap_pre_update(ip, *idx - 1, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx - 1),
- LEFT.br_blockcount + new->br_blockcount);
- xfs_bmbt_set_startoff(ep,
- PREV.br_startoff + new->br_blockcount);
- trace_xfs_bmap_post_update(ip, *idx - 1, state, _THIS_IP_);
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS);
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_startblock(ep,
- new->br_startblock + new->br_blockcount);
- xfs_bmbt_set_blockcount(ep,
- PREV.br_blockcount - new->br_blockcount);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ /*
+ * Make space in the inode incore.
+ */
+ xfs_iroot_realloc(ip, 1, whichfork);
+ ifp->if_flags |= XFS_IFBROOT;
- --*idx;
+ /*
+ * Fill in the root.
+ */
+ block = ifp->if_broot;
+ block->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC);
+ block->bb_level = cpu_to_be16(1);
+ block->bb_numrecs = cpu_to_be16(1);
+ block->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
+ block->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
- if (cur == NULL)
- rval = XFS_ILOG_DEXT;
- else {
- rval = 0;
- if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
- PREV.br_startblock, PREV.br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur,
- PREV.br_startoff + new->br_blockcount,
- PREV.br_startblock + new->br_blockcount,
- PREV.br_blockcount - new->br_blockcount,
- oldext)))
- goto done;
- if ((error = xfs_btree_decrement(cur, 0, &i)))
- goto done;
- error = xfs_bmbt_update(cur, LEFT.br_startoff,
- LEFT.br_startblock,
- LEFT.br_blockcount + new->br_blockcount,
- LEFT.br_state);
- if (error)
- goto done;
- }
- break;
-
- case BMAP_LEFT_FILLING:
- /*
- * Setting the first part of a previous oldext extent to newext.
- * The left neighbor is not contiguous.
- */
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- ASSERT(ep && xfs_bmbt_get_state(ep) == oldext);
- xfs_bmbt_set_startoff(ep, new_endoff);
- xfs_bmbt_set_blockcount(ep,
- PREV.br_blockcount - new->br_blockcount);
- xfs_bmbt_set_startblock(ep,
- new->br_startblock + new->br_blockcount);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-
- xfs_iext_insert(ip, *idx, 1, new, state);
- ip->i_d.di_nextents++;
- if (cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
- PREV.br_startblock, PREV.br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur,
- PREV.br_startoff + new->br_blockcount,
- PREV.br_startblock + new->br_blockcount,
- PREV.br_blockcount - new->br_blockcount,
- oldext)))
- goto done;
- cur->bc_rec.b = *new;
- if ((error = xfs_btree_insert(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ /*
+ * Need a cursor. Can't allocate until bb_level is filled in.
+ */
+ mp = ip->i_mount;
+ cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
+ cur->bc_private.b.firstblock = *firstblock;
+ cur->bc_private.b.flist = flist;
+ cur->bc_private.b.flags = wasdel ? XFS_BTCUR_BPRV_WASDEL : 0;
+ /*
+ * Convert to a btree with two levels, one record in root.
+ */
+ XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_BTREE);
+ memset(&args, 0, sizeof(args));
+ args.tp = tp;
+ args.mp = mp;
+ args.firstblock = *firstblock;
+ if (*firstblock == NULLFSBLOCK) {
+ args.type = XFS_ALLOCTYPE_START_BNO;
+ args.fsbno = XFS_INO_TO_FSB(mp, ip->i_ino);
+ } else if (flist->xbf_low) {
+ args.type = XFS_ALLOCTYPE_START_BNO;
+ args.fsbno = *firstblock;
+ } else {
+ args.type = XFS_ALLOCTYPE_NEAR_BNO;
+ args.fsbno = *firstblock;
+ }
+ args.minlen = args.maxlen = args.prod = 1;
+ args.wasdel = wasdel;
+ *logflagsp = 0;
+ if ((error = xfs_alloc_vextent(&args))) {
+ xfs_iroot_realloc(ip, -1, whichfork);
+ xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
+ return error;
+ }
+ /*
+ * Allocation can't fail, the space was reserved.
+ */
+ ASSERT(args.fsbno != NULLFSBLOCK);
+ ASSERT(*firstblock == NULLFSBLOCK ||
+ args.agno == XFS_FSB_TO_AGNO(mp, *firstblock) ||
+ (flist->xbf_low &&
+ args.agno > XFS_FSB_TO_AGNO(mp, *firstblock)));
+ *firstblock = cur->bc_private.b.firstblock = args.fsbno;
+ cur->bc_private.b.allocated++;
+ ip->i_d.di_nblocks++;
+ xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, 1L);
+ abp = xfs_btree_get_bufl(mp, tp, args.fsbno, 0);
+ /*
+ * Fill in the child block.
+ */
+ abp->b_ops = &xfs_bmbt_buf_ops;
+ ablock = XFS_BUF_TO_BLOCK(abp);
+ ablock->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC);
+ ablock->bb_level = 0;
+ ablock->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
+ ablock->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
+ arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
+ nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
+ for (cnt = i = 0; i < nextents; i++) {
+ ep = xfs_iext_get_ext(ifp, i);
+ if (!isnullstartblock(xfs_bmbt_get_startblock(ep))) {
+ arp->l0 = cpu_to_be64(ep->l0);
+ arp->l1 = cpu_to_be64(ep->l1);
+ arp++; cnt++;
}
- break;
+ }
+ ASSERT(cnt == XFS_IFORK_NEXTENTS(ip, whichfork));
+ xfs_btree_set_numrecs(ablock, cnt);
- case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
- /*
- * Setting the last part of a previous oldext extent to newext.
- * The right neighbor is contiguous with the new allocation.
- */
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep,
- PREV.br_blockcount - new->br_blockcount);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ /*
+ * Fill in the root key and pointer.
+ */
+ kp = XFS_BMBT_KEY_ADDR(mp, block, 1);
+ arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
+ kp->br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(arp));
+ pp = XFS_BMBT_PTR_ADDR(mp, block, 1, xfs_bmbt_get_maxrecs(cur,
+ be16_to_cpu(block->bb_level)));
+ *pp = cpu_to_be64(args.fsbno);
- ++*idx;
+ /*
+ * Do all this logging at the end so that
+ * the root is at the right level.
+ */
+ xfs_btree_log_block(cur, abp, XFS_BB_ALL_BITS);
+ xfs_btree_log_recs(cur, abp, 1, be16_to_cpu(ablock->bb_numrecs));
+ ASSERT(*curp == NULL);
+ *curp = cur;
+ *logflagsp = XFS_ILOG_CORE | xfs_ilog_fbroot(whichfork);
+ return 0;
+}
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
- new->br_startoff, new->br_startblock,
- new->br_blockcount + RIGHT.br_blockcount, newext);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+/*
+ * Convert a local file to an extents file.
+ * This code is out of bounds for data forks of regular files,
+ * since the file data needs to get logged so things will stay consistent.
+ * (The bmap-level manipulations are ok, though).
+ */
+STATIC int /* error */
+xfs_bmap_local_to_extents(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_fsblock_t *firstblock, /* first block allocated in xaction */
+ xfs_extlen_t total, /* total blocks needed by transaction */
+ int *logflagsp, /* inode logging flags */
+ int whichfork,
+ void (*init_fn)(struct xfs_buf *bp,
+ struct xfs_inode *ip,
+ struct xfs_ifork *ifp))
+{
+ int error; /* error return value */
+ int flags; /* logging flags returned */
+ xfs_ifork_t *ifp; /* inode fork pointer */
- if (cur == NULL)
- rval = XFS_ILOG_DEXT;
- else {
- rval = 0;
- if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
- PREV.br_startblock,
- PREV.br_blockcount, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
- PREV.br_startblock,
- PREV.br_blockcount - new->br_blockcount,
- oldext)))
- goto done;
- if ((error = xfs_btree_increment(cur, 0, &i)))
- goto done;
- if ((error = xfs_bmbt_update(cur, new->br_startoff,
- new->br_startblock,
- new->br_blockcount + RIGHT.br_blockcount,
- newext)))
- goto done;
- }
- break;
+ /*
+ * We don't want to deal with the case of keeping inode data inline yet.
+ * So sending the data fork of a regular inode is invalid.
+ */
+ ASSERT(!(S_ISREG(ip->i_d.di_mode) && whichfork == XFS_DATA_FORK));
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
+ flags = 0;
+ error = 0;
+ if (ifp->if_bytes) {
+ xfs_alloc_arg_t args; /* allocation arguments */
+ xfs_buf_t *bp; /* buffer for extent block */
+ xfs_bmbt_rec_host_t *ep;/* extent record pointer */
- case BMAP_RIGHT_FILLING:
+ ASSERT((ifp->if_flags &
+ (XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) == XFS_IFINLINE);
+ memset(&args, 0, sizeof(args));
+ args.tp = tp;
+ args.mp = ip->i_mount;
+ args.firstblock = *firstblock;
/*
- * Setting the last part of a previous oldext extent to newext.
- * The right neighbor is not contiguous.
+ * Allocate a block. We know we need only one, since the
+ * file currently fits in an inode.
*/
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep,
- PREV.br_blockcount - new->br_blockcount);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ if (*firstblock == NULLFSBLOCK) {
+ args.fsbno = XFS_INO_TO_FSB(args.mp, ip->i_ino);
+ args.type = XFS_ALLOCTYPE_START_BNO;
+ } else {
+ args.fsbno = *firstblock;
+ args.type = XFS_ALLOCTYPE_NEAR_BNO;
+ }
+ args.total = total;
+ args.minlen = args.maxlen = args.prod = 1;
+ error = xfs_alloc_vextent(&args);
+ if (error)
+ goto done;
- ++*idx;
- xfs_iext_insert(ip, *idx, 1, new, state);
+ /* Can't fail, the space was reserved. */
+ ASSERT(args.fsbno != NULLFSBLOCK);
+ ASSERT(args.len == 1);
+ *firstblock = args.fsbno;
+ bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0);
- ip->i_d.di_nextents++;
- if (cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
- PREV.br_startblock, PREV.br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
- PREV.br_startblock,
- PREV.br_blockcount - new->br_blockcount,
- oldext)))
- goto done;
- if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 0, done);
- cur->bc_rec.b.br_state = XFS_EXT_NORM;
- if ((error = xfs_btree_insert(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- }
- break;
-
- case 0:
- /*
- * Setting the middle part of a previous oldext extent to
- * newext. Contiguity is impossible here.
- * One extent becomes three extents.
- */
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep,
- new->br_startoff - PREV.br_startoff);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ /* initialise the block and copy the data */
+ init_fn(bp, ip, ifp);
- r[0] = *new;
- r[1].br_startoff = new_endoff;
- r[1].br_blockcount =
- PREV.br_startoff + PREV.br_blockcount - new_endoff;
- r[1].br_startblock = new->br_startblock + new->br_blockcount;
- r[1].br_state = oldext;
+ /* account for the change in fork size and log everything */
+ xfs_trans_log_buf(tp, bp, 0, ifp->if_bytes - 1);
+ xfs_bmap_forkoff_reset(args.mp, ip, whichfork);
+ xfs_idata_realloc(ip, -ifp->if_bytes, whichfork);
+ xfs_iext_add(ifp, 0, 1);
+ ep = xfs_iext_get_ext(ifp, 0);
+ xfs_bmbt_set_allf(ep, 0, args.fsbno, 1, XFS_EXT_NORM);
+ trace_xfs_bmap_post_update(ip, 0,
+ whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0,
+ _THIS_IP_);
+ XFS_IFORK_NEXT_SET(ip, whichfork, 1);
+ ip->i_d.di_nblocks = 1;
+ xfs_trans_mod_dquot_byino(tp, ip,
+ XFS_TRANS_DQ_BCOUNT, 1L);
+ flags |= xfs_ilog_fext(whichfork);
+ } else {
+ ASSERT(XFS_IFORK_NEXTENTS(ip, whichfork) == 0);
+ xfs_bmap_forkoff_reset(ip->i_mount, ip, whichfork);
+ }
+ ifp->if_flags &= ~XFS_IFINLINE;
+ ifp->if_flags |= XFS_IFEXTENTS;
+ XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
+ flags |= XFS_ILOG_CORE;
+done:
+ *logflagsp = flags;
+ return error;
+}
- ++*idx;
- xfs_iext_insert(ip, *idx, 2, &r[0], state);
+/*
+ * Called from xfs_bmap_add_attrfork to handle btree format files.
+ */
+STATIC int /* error */
+xfs_bmap_add_attrfork_btree(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_fsblock_t *firstblock, /* first block allocated */
+ xfs_bmap_free_t *flist, /* blocks to free at commit */
+ int *flags) /* inode logging flags */
+{
+ xfs_btree_cur_t *cur; /* btree cursor */
+ int error; /* error return value */
+ xfs_mount_t *mp; /* file system mount struct */
+ int stat; /* newroot status */
- ip->i_d.di_nextents += 2;
- if (cur == NULL)
- rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
- else {
- rval = XFS_ILOG_CORE;
- if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
- PREV.br_startblock, PREV.br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- /* new right extent - oldext */
- if ((error = xfs_bmbt_update(cur, r[1].br_startoff,
- r[1].br_startblock, r[1].br_blockcount,
- r[1].br_state)))
- goto done;
- /* new left extent - oldext */
- cur->bc_rec.b = PREV;
- cur->bc_rec.b.br_blockcount =
- new->br_startoff - PREV.br_startoff;
- if ((error = xfs_btree_insert(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- /*
- * Reset the cursor to the position of the new extent
- * we are about to insert as we can't trust it after
- * the previous insert.
- */
- if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
- new->br_startblock, new->br_blockcount,
- &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 0, done);
- /* new middle extent - newext */
- cur->bc_rec.b.br_state = new->br_state;
- if ((error = xfs_btree_insert(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ mp = ip->i_mount;
+ if (ip->i_df.if_broot_bytes <= XFS_IFORK_DSIZE(ip))
+ *flags |= XFS_ILOG_DBROOT;
+ else {
+ cur = xfs_bmbt_init_cursor(mp, tp, ip, XFS_DATA_FORK);
+ cur->bc_private.b.flist = flist;
+ cur->bc_private.b.firstblock = *firstblock;
+ if ((error = xfs_bmbt_lookup_ge(cur, 0, 0, 0, &stat)))
+ goto error0;
+ /* must be at least one entry */
+ XFS_WANT_CORRUPTED_GOTO(stat == 1, error0);
+ if ((error = xfs_btree_new_iroot(cur, flags, &stat)))
+ goto error0;
+ if (stat == 0) {
+ xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
+ return XFS_ERROR(ENOSPC);
}
- break;
-
- case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- case BMAP_LEFT_FILLING | BMAP_RIGHT_CONTIG:
- case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
- case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- case BMAP_LEFT_CONTIG:
- case BMAP_RIGHT_CONTIG:
- /*
- * These cases are all impossible.
- */
- ASSERT(0);
+ *firstblock = cur->bc_private.b.firstblock;
+ cur->bc_private.b.allocated = 0;
+ xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR);
}
+ return 0;
+error0:
+ xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
+ return error;
+}
- /* convert to a btree if necessary */
- if (xfs_bmap_needs_btree(ip, XFS_DATA_FORK)) {
- int tmp_logflags; /* partial log flag return val */
-
- ASSERT(cur == NULL);
- error = xfs_bmap_extents_to_btree(tp, ip, first, flist, &cur,
- 0, &tmp_logflags, XFS_DATA_FORK);
- *logflagsp |= tmp_logflags;
- if (error)
- goto done;
- }
+/*
+ * Called from xfs_bmap_add_attrfork to handle extents format files.
+ */
+STATIC int /* error */
+xfs_bmap_add_attrfork_extents(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_fsblock_t *firstblock, /* first block allocated */
+ xfs_bmap_free_t *flist, /* blocks to free at commit */
+ int *flags) /* inode logging flags */
+{
+ xfs_btree_cur_t *cur; /* bmap btree cursor */
+ int error; /* error return value */
- /* clear out the allocated field, done with it now in any case. */
+ if (ip->i_d.di_nextents * sizeof(xfs_bmbt_rec_t) <= XFS_IFORK_DSIZE(ip))
+ return 0;
+ cur = NULL;
+ error = xfs_bmap_extents_to_btree(tp, ip, firstblock, flist, &cur, 0,
+ flags, XFS_DATA_FORK);
if (cur) {
cur->bc_private.b.allocated = 0;
- *curp = cur;
+ xfs_btree_del_cursor(cur,
+ error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
}
-
- xfs_bmap_check_leaf_extents(*curp, ip, XFS_DATA_FORK);
-done:
- *logflagsp |= rval;
return error;
-#undef LEFT
-#undef RIGHT
-#undef PREV
}
/*
- * Convert a hole to a delayed allocation.
+ * Block initialisation functions for local to extent format conversion.
+ * As these get more complex, they will be moved to the relevant files,
+ * but for now they are too simple to worry about.
*/
STATIC void
-xfs_bmap_add_extent_hole_delay(
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_extnum_t *idx, /* extent number to update/insert */
- xfs_bmbt_irec_t *new) /* new data to add to file extents */
+xfs_bmap_local_to_extents_init_fn(
+ struct xfs_buf *bp,
+ struct xfs_inode *ip,
+ struct xfs_ifork *ifp)
{
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_bmbt_irec_t left; /* left neighbor extent entry */
- xfs_filblks_t newlen=0; /* new indirect size */
- xfs_filblks_t oldlen=0; /* old indirect size */
- xfs_bmbt_irec_t right; /* right neighbor extent entry */
- int state; /* state bits, accessed thru macros */
- xfs_filblks_t temp=0; /* temp for indirect calculations */
+ bp->b_ops = &xfs_bmbt_buf_ops;
+ memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+}
- ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
- state = 0;
- ASSERT(isnullstartblock(new->br_startblock));
+STATIC void
+xfs_symlink_local_to_remote(
+ struct xfs_buf *bp,
+ struct xfs_inode *ip,
+ struct xfs_ifork *ifp)
+{
+ /* remote symlink blocks are not verifiable until CRCs come along */
+ bp->b_ops = NULL;
+ memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+}
- /*
- * Check and set flags if this segment has a left neighbor
- */
- if (*idx > 0) {
- state |= BMAP_LEFT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left);
+/*
+ * Called from xfs_bmap_add_attrfork to handle local format files. Each
+ * different data fork content type needs a different callout to do the
+ * conversion. Some are basic and only require special block initialisation
+ * callouts for the data formating, others (directories) are so specialised they
+ * handle everything themselves.
+ *
+ * XXX (dgc): investigate whether directory conversion can use the generic
+ * formatting callout. It should be possible - it's just a very complex
+ * formatter. it would also require passing the transaction through to the init
+ * function.
+ */
+STATIC int /* error */
+xfs_bmap_add_attrfork_local(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_fsblock_t *firstblock, /* first block allocated */
+ xfs_bmap_free_t *flist, /* blocks to free at commit */
+ int *flags) /* inode logging flags */
+{
+ xfs_da_args_t dargs; /* args for dir/attr code */
- if (isnullstartblock(left.br_startblock))
- state |= BMAP_LEFT_DELAY;
- }
-
- /*
- * Check and set flags if the current (right) segment exists.
- * If it doesn't exist, we're converting the hole at end-of-file.
- */
- if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
- state |= BMAP_RIGHT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right);
+ if (ip->i_df.if_bytes <= XFS_IFORK_DSIZE(ip))
+ return 0;
- if (isnullstartblock(right.br_startblock))
- state |= BMAP_RIGHT_DELAY;
+ if (S_ISDIR(ip->i_d.di_mode)) {
+ memset(&dargs, 0, sizeof(dargs));
+ dargs.dp = ip;
+ dargs.firstblock = firstblock;
+ dargs.flist = flist;
+ dargs.total = ip->i_mount->m_dirblkfsbs;
+ dargs.whichfork = XFS_DATA_FORK;
+ dargs.trans = tp;
+ return xfs_dir2_sf_to_block(&dargs);
}
- /*
- * Set contiguity flags on the left and right neighbors.
- * Don't let extents get too large, even if the pieces are contiguous.
- */
- if ((state & BMAP_LEFT_VALID) && (state & BMAP_LEFT_DELAY) &&
- left.br_startoff + left.br_blockcount == new->br_startoff &&
- left.br_blockcount + new->br_blockcount <= MAXEXTLEN)
- state |= BMAP_LEFT_CONTIG;
-
- if ((state & BMAP_RIGHT_VALID) && (state & BMAP_RIGHT_DELAY) &&
- new->br_startoff + new->br_blockcount == right.br_startoff &&
- new->br_blockcount + right.br_blockcount <= MAXEXTLEN &&
- (!(state & BMAP_LEFT_CONTIG) ||
- (left.br_blockcount + new->br_blockcount +
- right.br_blockcount <= MAXEXTLEN)))
- state |= BMAP_RIGHT_CONTIG;
+ if (S_ISLNK(ip->i_d.di_mode))
+ return xfs_bmap_local_to_extents(tp, ip, firstblock, 1,
+ flags, XFS_DATA_FORK,
+ xfs_symlink_local_to_remote);
- /*
- * Switch out based on the contiguity flags.
- */
- switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
- case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- /*
- * New allocation is contiguous with delayed allocations
- * on the left and on the right.
- * Merge all three into a single extent record.
- */
- --*idx;
- temp = left.br_blockcount + new->br_blockcount +
- right.br_blockcount;
+ return xfs_bmap_local_to_extents(tp, ip, firstblock, 1, flags,
+ XFS_DATA_FORK,
+ xfs_bmap_local_to_extents_init_fn);
+}
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
- oldlen = startblockval(left.br_startblock) +
- startblockval(new->br_startblock) +
- startblockval(right.br_startblock);
- newlen = xfs_bmap_worst_indlen(ip, temp);
- xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
- nullstartblock((int)newlen));
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+/*
+ * Convert inode from non-attributed to attributed.
+ * Must not be in a transaction, ip must not be locked.
+ */
+int /* error code */
+xfs_bmap_add_attrfork(
+ xfs_inode_t *ip, /* incore inode pointer */
+ int size, /* space new attribute needs */
+ int rsvd) /* xact may use reserved blks */
+{
+ xfs_fsblock_t firstblock; /* 1st block/ag allocated */
+ xfs_bmap_free_t flist; /* freed extent records */
+ xfs_mount_t *mp; /* mount structure */
+ xfs_trans_t *tp; /* transaction pointer */
+ int blks; /* space reservation */
+ int version = 1; /* superblock attr version */
+ int committed; /* xaction was committed */
+ int logflags; /* logging flags */
+ int error; /* error return value */
- xfs_iext_remove(ip, *idx + 1, 1, state);
- break;
+ ASSERT(XFS_IFORK_Q(ip) == 0);
- case BMAP_LEFT_CONTIG:
+ mp = ip->i_mount;
+ ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
+ tp = xfs_trans_alloc(mp, XFS_TRANS_ADDAFORK);
+ blks = XFS_ADDAFORK_SPACE_RES(mp);
+ if (rsvd)
+ tp->t_flags |= XFS_TRANS_RESERVE;
+ if ((error = xfs_trans_reserve(tp, blks, XFS_ADDAFORK_LOG_RES(mp), 0,
+ XFS_TRANS_PERM_LOG_RES, XFS_ADDAFORK_LOG_COUNT)))
+ goto error0;
+ xfs_ilock(ip, XFS_ILOCK_EXCL);
+ error = xfs_trans_reserve_quota_nblks(tp, ip, blks, 0, rsvd ?
+ XFS_QMOPT_RES_REGBLKS | XFS_QMOPT_FORCE_RES :
+ XFS_QMOPT_RES_REGBLKS);
+ if (error) {
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+ xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES);
+ return error;
+ }
+ if (XFS_IFORK_Q(ip))
+ goto error1;
+ if (ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS) {
/*
- * New allocation is contiguous with a delayed allocation
- * on the left.
- * Merge the new allocation with the left neighbor.
+ * For inodes coming from pre-6.2 filesystems.
*/
- --*idx;
- temp = left.br_blockcount + new->br_blockcount;
+ ASSERT(ip->i_d.di_aformat == 0);
+ ip->i_d.di_aformat = XFS_DINODE_FMT_EXTENTS;
+ }
+ ASSERT(ip->i_d.di_anextents == 0);
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
- oldlen = startblockval(left.br_startblock) +
- startblockval(new->br_startblock);
- newlen = xfs_bmap_worst_indlen(ip, temp);
- xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
- nullstartblock((int)newlen));
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- break;
+ xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
+ xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
- case BMAP_RIGHT_CONTIG:
- /*
- * New allocation is contiguous with a delayed allocation
- * on the right.
- * Merge the new allocation with the right neighbor.
- */
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- temp = new->br_blockcount + right.br_blockcount;
- oldlen = startblockval(new->br_startblock) +
- startblockval(right.br_startblock);
- newlen = xfs_bmap_worst_indlen(ip, temp);
- xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
- new->br_startoff,
- nullstartblock((int)newlen), temp, right.br_state);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ switch (ip->i_d.di_format) {
+ case XFS_DINODE_FMT_DEV:
+ ip->i_d.di_forkoff = roundup(sizeof(xfs_dev_t), 8) >> 3;
+ break;
+ case XFS_DINODE_FMT_UUID:
+ ip->i_d.di_forkoff = roundup(sizeof(uuid_t), 8) >> 3;
+ break;
+ case XFS_DINODE_FMT_LOCAL:
+ case XFS_DINODE_FMT_EXTENTS:
+ case XFS_DINODE_FMT_BTREE:
+ ip->i_d.di_forkoff = xfs_attr_shortform_bytesfit(ip, size);
+ if (!ip->i_d.di_forkoff)
+ ip->i_d.di_forkoff = xfs_default_attroffset(ip) >> 3;
+ else if (mp->m_flags & XFS_MOUNT_ATTR2)
+ version = 2;
break;
+ default:
+ ASSERT(0);
+ error = XFS_ERROR(EINVAL);
+ goto error1;
+ }
- case 0:
- /*
- * New allocation is not contiguous with another
- * delayed allocation.
- * Insert a new entry.
- */
- oldlen = newlen = 0;
- xfs_iext_insert(ip, *idx, 1, new, state);
+ ASSERT(ip->i_afp == NULL);
+ ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP);
+ ip->i_afp->if_flags = XFS_IFEXTENTS;
+ logflags = 0;
+ xfs_bmap_init(&flist, &firstblock);
+ switch (ip->i_d.di_format) {
+ case XFS_DINODE_FMT_LOCAL:
+ error = xfs_bmap_add_attrfork_local(tp, ip, &firstblock, &flist,
+ &logflags);
+ break;
+ case XFS_DINODE_FMT_EXTENTS:
+ error = xfs_bmap_add_attrfork_extents(tp, ip, &firstblock,
+ &flist, &logflags);
+ break;
+ case XFS_DINODE_FMT_BTREE:
+ error = xfs_bmap_add_attrfork_btree(tp, ip, &firstblock, &flist,
+ &logflags);
+ break;
+ default:
+ error = 0;
break;
}
- if (oldlen != newlen) {
- ASSERT(oldlen > newlen);
- xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS,
- (int64_t)(oldlen - newlen), 0);
- /*
- * Nothing to do for disk quota accounting here.
- */
+ if (logflags)
+ xfs_trans_log_inode(tp, ip, logflags);
+ if (error)
+ goto error2;
+ if (!xfs_sb_version_hasattr(&mp->m_sb) ||
+ (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2)) {
+ __int64_t sbfields = 0;
+
+ spin_lock(&mp->m_sb_lock);
+ if (!xfs_sb_version_hasattr(&mp->m_sb)) {
+ xfs_sb_version_addattr(&mp->m_sb);
+ sbfields |= XFS_SB_VERSIONNUM;
+ }
+ if (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2) {
+ xfs_sb_version_addattr2(&mp->m_sb);
+ sbfields |= (XFS_SB_VERSIONNUM | XFS_SB_FEATURES2);
+ }
+ if (sbfields) {
+ spin_unlock(&mp->m_sb_lock);
+ xfs_mod_sb(tp, sbfields);
+ } else
+ spin_unlock(&mp->m_sb_lock);
}
+
+ error = xfs_bmap_finish(&tp, &flist, &committed);
+ if (error)
+ goto error2;
+ return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
+error2:
+ xfs_bmap_cancel(&flist);
+error1:
+ xfs_iunlock(ip, XFS_ILOCK_EXCL);
+error0:
+ xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT);
+ return error;
}
/*
- * Convert a hole to a real allocation.
+ * Internal and external extent tree search functions.
*/
-STATIC int /* error */
-xfs_bmap_add_extent_hole_real(
- struct xfs_bmalloca *bma,
- int whichfork)
-{
- struct xfs_bmbt_irec *new = &bma->got;
- int error; /* error return value */
- int i; /* temp state */
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_bmbt_irec_t left; /* left neighbor extent entry */
- xfs_bmbt_irec_t right; /* right neighbor extent entry */
- int rval=0; /* return value (logging flags) */
- int state; /* state bits, accessed thru macros */
-
- ifp = XFS_IFORK_PTR(bma->ip, whichfork);
- ASSERT(bma->idx >= 0);
- ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
- ASSERT(!isnullstartblock(new->br_startblock));
- ASSERT(!bma->cur ||
- !(bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
-
- XFS_STATS_INC(xs_add_exlist);
-
- state = 0;
- if (whichfork == XFS_ATTR_FORK)
- state |= BMAP_ATTRFORK;
+/*
+ * Read in the extents to if_extents.
+ * All inode fields are set up by caller, we just traverse the btree
+ * and copy the records in. If the file system cannot contain unwritten
+ * extents, the records are checked for no "state" flags.
+ */
+int /* error */
+xfs_bmap_read_extents(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode */
+ int whichfork) /* data or attr fork */
+{
+ struct xfs_btree_block *block; /* current btree block */
+ xfs_fsblock_t bno; /* block # of "block" */
+ xfs_buf_t *bp; /* buffer for "block" */
+ int error; /* error return value */
+ xfs_exntfmt_t exntf; /* XFS_EXTFMT_NOSTATE, if checking */
+ xfs_extnum_t i, j; /* index into the extents list */
+ xfs_ifork_t *ifp; /* fork structure */
+ int level; /* btree level, for checking */
+ xfs_mount_t *mp; /* file system mount structure */
+ __be64 *pp; /* pointer to block address */
+ /* REFERENCED */
+ xfs_extnum_t room; /* number of entries there's room for */
+ bno = NULLFSBLOCK;
+ mp = ip->i_mount;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ exntf = (whichfork != XFS_DATA_FORK) ? XFS_EXTFMT_NOSTATE :
+ XFS_EXTFMT_INODE(ip);
+ block = ifp->if_broot;
/*
- * Check and set flags if this segment has a left neighbor.
+ * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
*/
- if (bma->idx > 0) {
- state |= BMAP_LEFT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &left);
- if (isnullstartblock(left.br_startblock))
- state |= BMAP_LEFT_DELAY;
- }
-
+ level = be16_to_cpu(block->bb_level);
+ ASSERT(level > 0);
+ pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
+ bno = be64_to_cpu(*pp);
+ ASSERT(bno != NULLDFSBNO);
+ ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
+ ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
/*
- * Check and set flags if this segment has a current value.
- * Not true if we're inserting into the "hole" at eof.
+ * Go down the tree until leaf level is reached, following the first
+ * pointer (leftmost) at each level.
*/
- if (bma->idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
- state |= BMAP_RIGHT_VALID;
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &right);
- if (isnullstartblock(right.br_startblock))
- state |= BMAP_RIGHT_DELAY;
+ while (level-- > 0) {
+ error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
+ XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
+ if (error)
+ return error;
+ block = XFS_BUF_TO_BLOCK(bp);
+ XFS_WANT_CORRUPTED_GOTO(
+ xfs_bmap_sanity_check(mp, bp, level),
+ error0);
+ if (level == 0)
+ break;
+ pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
+ bno = be64_to_cpu(*pp);
+ XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0);
+ xfs_trans_brelse(tp, bp);
}
-
/*
- * We're inserting a real allocation between "left" and "right".
- * Set the contiguity flags. Don't let extents get too large.
+ * Here with bp and block set to the leftmost leaf node in the tree.
*/
- if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
- left.br_startoff + left.br_blockcount == new->br_startoff &&
- left.br_startblock + left.br_blockcount == new->br_startblock &&
- left.br_state == new->br_state &&
- left.br_blockcount + new->br_blockcount <= MAXEXTLEN)
- state |= BMAP_LEFT_CONTIG;
-
- if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
- new->br_startoff + new->br_blockcount == right.br_startoff &&
- new->br_startblock + new->br_blockcount == right.br_startblock &&
- new->br_state == right.br_state &&
- new->br_blockcount + right.br_blockcount <= MAXEXTLEN &&
- (!(state & BMAP_LEFT_CONTIG) ||
- left.br_blockcount + new->br_blockcount +
- right.br_blockcount <= MAXEXTLEN))
- state |= BMAP_RIGHT_CONTIG;
-
- error = 0;
+ room = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
+ i = 0;
/*
- * Select which case we're in here, and implement it.
+ * Loop over all leaf nodes. Copy information to the extent records.
*/
- switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
- case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
- /*
- * New allocation is contiguous with real allocations on the
- * left and on the right.
- * Merge all three into a single extent record.
- */
- --bma->idx;
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
- left.br_blockcount + new->br_blockcount +
- right.br_blockcount);
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-
- xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
+ for (;;) {
+ xfs_bmbt_rec_t *frp;
+ xfs_fsblock_t nextbno;
+ xfs_extnum_t num_recs;
+ xfs_extnum_t start;
- XFS_IFORK_NEXT_SET(bma->ip, whichfork,
- XFS_IFORK_NEXTENTS(bma->ip, whichfork) - 1);
- if (bma->cur == NULL) {
- rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
- } else {
- rval = XFS_ILOG_CORE;
- error = xfs_bmbt_lookup_eq(bma->cur, right.br_startoff,
- right.br_startblock, right.br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_btree_delete(bma->cur, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_btree_decrement(bma->cur, 0, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, left.br_startoff,
- left.br_startblock,
- left.br_blockcount +
- new->br_blockcount +
- right.br_blockcount,
- left.br_state);
- if (error)
- goto done;
+ num_recs = xfs_btree_get_numrecs(block);
+ if (unlikely(i + num_recs > room)) {
+ ASSERT(i + num_recs <= room);
+ xfs_warn(ip->i_mount,
+ "corrupt dinode %Lu, (btree extents).",
+ (unsigned long long) ip->i_ino);
+ XFS_CORRUPTION_ERROR("xfs_bmap_read_extents(1)",
+ XFS_ERRLEVEL_LOW, ip->i_mount, block);
+ goto error0;
}
- break;
-
- case BMAP_LEFT_CONTIG:
+ XFS_WANT_CORRUPTED_GOTO(
+ xfs_bmap_sanity_check(mp, bp, 0),
+ error0);
/*
- * New allocation is contiguous with a real allocation
- * on the left.
- * Merge the new allocation with the left neighbor.
+ * Read-ahead the next leaf block, if any.
*/
- --bma->idx;
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
- left.br_blockcount + new->br_blockcount);
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-
- if (bma->cur == NULL) {
- rval = xfs_ilog_fext(whichfork);
- } else {
- rval = 0;
- error = xfs_bmbt_lookup_eq(bma->cur, left.br_startoff,
- left.br_startblock, left.br_blockcount,
- &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, left.br_startoff,
- left.br_startblock,
- left.br_blockcount +
- new->br_blockcount,
- left.br_state);
- if (error)
- goto done;
- }
- break;
-
- case BMAP_RIGHT_CONTIG:
+ nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
+ if (nextbno != NULLFSBLOCK)
+ xfs_btree_reada_bufl(mp, nextbno, 1,
+ &xfs_bmbt_buf_ops);
/*
- * New allocation is contiguous with a real allocation
- * on the right.
- * Merge the new allocation with the right neighbor.
+ * Copy records into the extent records.
*/
- trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
- xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->idx),
- new->br_startoff, new->br_startblock,
- new->br_blockcount + right.br_blockcount,
- right.br_state);
- trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-
- if (bma->cur == NULL) {
- rval = xfs_ilog_fext(whichfork);
- } else {
- rval = 0;
- error = xfs_bmbt_lookup_eq(bma->cur,
- right.br_startoff,
- right.br_startblock,
- right.br_blockcount, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- error = xfs_bmbt_update(bma->cur, new->br_startoff,
- new->br_startblock,
- new->br_blockcount +
- right.br_blockcount,
- right.br_state);
- if (error)
- goto done;
+ frp = XFS_BMBT_REC_ADDR(mp, block, 1);
+ start = i;
+ for (j = 0; j < num_recs; j++, i++, frp++) {
+ xfs_bmbt_rec_host_t *trp = xfs_iext_get_ext(ifp, i);
+ trp->l0 = be64_to_cpu(frp->l0);
+ trp->l1 = be64_to_cpu(frp->l1);
}
- break;
-
- case 0:
+ if (exntf == XFS_EXTFMT_NOSTATE) {
+ /*
+ * Check all attribute bmap btree records and
+ * any "older" data bmap btree records for a
+ * set bit in the "extent flag" position.
+ */
+ if (unlikely(xfs_check_nostate_extents(ifp,
+ start, num_recs))) {
+ XFS_ERROR_REPORT("xfs_bmap_read_extents(2)",
+ XFS_ERRLEVEL_LOW,
+ ip->i_mount);
+ goto error0;
+ }
+ }
+ xfs_trans_brelse(tp, bp);
+ bno = nextbno;
/*
- * New allocation is not contiguous with another
- * real allocation.
- * Insert a new entry.
+ * If we've reached the end, stop.
*/
- xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
- XFS_IFORK_NEXT_SET(bma->ip, whichfork,
- XFS_IFORK_NEXTENTS(bma->ip, whichfork) + 1);
- if (bma->cur == NULL) {
- rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
- } else {
- rval = XFS_ILOG_CORE;
- error = xfs_bmbt_lookup_eq(bma->cur,
- new->br_startoff,
- new->br_startblock,
- new->br_blockcount, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 0, done);
- bma->cur->bc_rec.b.br_state = new->br_state;
- error = xfs_btree_insert(bma->cur, &i);
- if (error)
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- }
- break;
+ if (bno == NULLFSBLOCK)
+ break;
+ error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
+ XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
+ if (error)
+ return error;
+ block = XFS_BUF_TO_BLOCK(bp);
}
+ ASSERT(i == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)));
+ ASSERT(i == XFS_IFORK_NEXTENTS(ip, whichfork));
+ XFS_BMAP_TRACE_EXLIST(ip, i, whichfork);
+ return 0;
+error0:
+ xfs_trans_brelse(tp, bp);
+ return XFS_ERROR(EFSCORRUPTED);
+}
- /* convert to a btree if necessary */
- if (xfs_bmap_needs_btree(bma->ip, whichfork)) {
- int tmp_logflags; /* partial log flag return val */
- ASSERT(bma->cur == NULL);
- error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
- bma->firstblock, bma->flist, &bma->cur,
- 0, &tmp_logflags, whichfork);
- bma->logflags |= tmp_logflags;
- if (error)
- goto done;
- }
+/*
+ * Search the extent records for the entry containing block bno.
+ * If bno lies in a hole, point to the next entry. If bno lies
+ * past eof, *eofp will be set, and *prevp will contain the last
+ * entry (null if none). Else, *lastxp will be set to the index
+ * of the found entry; *gotp will contain the entry.
+ */
+STATIC xfs_bmbt_rec_host_t * /* pointer to found extent entry */
+xfs_bmap_search_multi_extents(
+ xfs_ifork_t *ifp, /* inode fork pointer */
+ xfs_fileoff_t bno, /* block number searched for */
+ int *eofp, /* out: end of file found */
+ xfs_extnum_t *lastxp, /* out: last extent index */
+ xfs_bmbt_irec_t *gotp, /* out: extent entry found */
+ xfs_bmbt_irec_t *prevp) /* out: previous extent entry found */
+{
+ xfs_bmbt_rec_host_t *ep; /* extent record pointer */
+ xfs_extnum_t lastx; /* last extent index */
- /* clear out the allocated field, done with it now in any case. */
- if (bma->cur)
- bma->cur->bc_private.b.allocated = 0;
+ /*
+ * Initialize the extent entry structure to catch access to
+ * uninitialized br_startblock field.
+ */
+ gotp->br_startoff = 0xffa5a5a5a5a5a5a5LL;
+ gotp->br_blockcount = 0xa55a5a5a5a5a5a5aLL;
+ gotp->br_state = XFS_EXT_INVALID;
+#if XFS_BIG_BLKNOS
+ gotp->br_startblock = 0xffffa5a5a5a5a5a5LL;
+#else
+ gotp->br_startblock = 0xffffa5a5;
+#endif
+ prevp->br_startoff = NULLFILEOFF;
- xfs_bmap_check_leaf_extents(bma->cur, bma->ip, whichfork);
-done:
- bma->logflags |= rval;
- return error;
+ ep = xfs_iext_bno_to_ext(ifp, bno, &lastx);
+ if (lastx > 0) {
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx - 1), prevp);
+ }
+ if (lastx < (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))) {
+ xfs_bmbt_get_all(ep, gotp);
+ *eofp = 0;
+ } else {
+ if (lastx > 0) {
+ *gotp = *prevp;
+ }
+ *eofp = 1;
+ ep = NULL;
+ }
+ *lastxp = lastx;
+ return ep;
}
/*
- * Adjust the size of the new extent based on di_extsize and rt extsize.
+ * Search the extents list for the inode, for the extent containing bno.
+ * If bno lies in a hole, point to the next entry. If bno lies past eof,
+ * *eofp will be set, and *prevp will contain the last entry (null if none).
+ * Else, *lastxp will be set to the index of the found
+ * entry; *gotp will contain the entry.
*/
-STATIC int
-xfs_bmap_extsize_align(
- xfs_mount_t *mp,
- xfs_bmbt_irec_t *gotp, /* next extent pointer */
- xfs_bmbt_irec_t *prevp, /* previous extent pointer */
- xfs_extlen_t extsz, /* align to this extent size */
- int rt, /* is this a realtime inode? */
- int eof, /* is extent at end-of-file? */
- int delay, /* creating delalloc extent? */
- int convert, /* overwriting unwritten extent? */
- xfs_fileoff_t *offp, /* in/out: aligned offset */
- xfs_extlen_t *lenp) /* in/out: aligned length */
+STATIC xfs_bmbt_rec_host_t * /* pointer to found extent entry */
+xfs_bmap_search_extents(
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_fileoff_t bno, /* block number searched for */
+ int fork, /* data or attr fork */
+ int *eofp, /* out: end of file found */
+ xfs_extnum_t *lastxp, /* out: last extent index */
+ xfs_bmbt_irec_t *gotp, /* out: extent entry found */
+ xfs_bmbt_irec_t *prevp) /* out: previous extent entry found */
{
- xfs_fileoff_t orig_off; /* original offset */
- xfs_extlen_t orig_alen; /* original length */
- xfs_fileoff_t orig_end; /* original off+len */
- xfs_fileoff_t nexto; /* next file offset */
- xfs_fileoff_t prevo; /* previous file offset */
- xfs_fileoff_t align_off; /* temp for offset */
- xfs_extlen_t align_alen; /* temp for length */
- xfs_extlen_t temp; /* temp for calculations */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_bmbt_rec_host_t *ep; /* extent record pointer */
- if (convert)
- return 0;
+ XFS_STATS_INC(xs_look_exlist);
+ ifp = XFS_IFORK_PTR(ip, fork);
- orig_off = align_off = *offp;
- orig_alen = align_alen = *lenp;
- orig_end = orig_off + orig_alen;
+ ep = xfs_bmap_search_multi_extents(ifp, bno, eofp, lastxp, gotp, prevp);
- /*
- * If this request overlaps an existing extent, then don't
- * attempt to perform any additional alignment.
- */
- if (!delay && !eof &&
- (orig_off >= gotp->br_startoff) &&
- (orig_end <= gotp->br_startoff + gotp->br_blockcount)) {
- return 0;
+ if (unlikely(!(gotp->br_startblock) && (*lastxp != NULLEXTNUM) &&
+ !(XFS_IS_REALTIME_INODE(ip) && fork == XFS_DATA_FORK))) {
+ xfs_alert_tag(ip->i_mount, XFS_PTAG_FSBLOCK_ZERO,
+ "Access to block zero in inode %llu "
+ "start_block: %llx start_off: %llx "
+ "blkcnt: %llx extent-state: %x lastx: %x\n",
+ (unsigned long long)ip->i_ino,
+ (unsigned long long)gotp->br_startblock,
+ (unsigned long long)gotp->br_startoff,
+ (unsigned long long)gotp->br_blockcount,
+ gotp->br_state, *lastxp);
+ *lastxp = NULLEXTNUM;
+ *eofp = 1;
+ return NULL;
}
+ return ep;
+}
- /*
- * If the file offset is unaligned vs. the extent size
- * we need to align it. This will be possible unless
- * the file was previously written with a kernel that didn't
- * perform this alignment, or if a truncate shot us in the
- * foot.
- */
- temp = do_mod(orig_off, extsz);
- if (temp) {
- align_alen += temp;
- align_off -= temp;
+/*
+ * Returns the file-relative block number of the first unused block(s)
+ * in the file with at least "len" logically contiguous blocks free.
+ * This is the lowest-address hole if the file has holes, else the first block
+ * past the end of file.
+ * Return 0 if the file is currently local (in-inode).
+ */
+int /* error */
+xfs_bmap_first_unused(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode */
+ xfs_extlen_t len, /* size of hole to find */
+ xfs_fileoff_t *first_unused, /* unused block */
+ int whichfork) /* data or attr fork */
+{
+ int error; /* error return value */
+ int idx; /* extent record index */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_fileoff_t lastaddr; /* last block number seen */
+ xfs_fileoff_t lowest; /* lowest useful block */
+ xfs_fileoff_t max; /* starting useful block */
+ xfs_fileoff_t off; /* offset for this block */
+ xfs_extnum_t nextents; /* number of extent entries */
+
+ ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE ||
+ XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS ||
+ XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
+ if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
+ *first_unused = 0;
+ return 0;
}
- /*
- * Same adjustment for the end of the requested area.
- */
- if ((temp = (align_alen % extsz))) {
- align_alen += extsz - temp;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ if (!(ifp->if_flags & XFS_IFEXTENTS) &&
+ (error = xfs_iread_extents(tp, ip, whichfork)))
+ return error;
+ lowest = *first_unused;
+ nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
+ for (idx = 0, lastaddr = 0, max = lowest; idx < nextents; idx++) {
+ xfs_bmbt_rec_host_t *ep = xfs_iext_get_ext(ifp, idx);
+ off = xfs_bmbt_get_startoff(ep);
+ /*
+ * See if the hole before this extent will work.
+ */
+ if (off >= lowest + len && off - max >= len) {
+ *first_unused = max;
+ return 0;
+ }
+ lastaddr = off + xfs_bmbt_get_blockcount(ep);
+ max = XFS_FILEOFF_MAX(lastaddr, lowest);
}
- /*
- * If the previous block overlaps with this proposed allocation
- * then move the start forward without adjusting the length.
- */
- if (prevp->br_startoff != NULLFILEOFF) {
- if (prevp->br_startblock == HOLESTARTBLOCK)
- prevo = prevp->br_startoff;
- else
- prevo = prevp->br_startoff + prevp->br_blockcount;
- } else
- prevo = 0;
- if (align_off != orig_off && align_off < prevo)
- align_off = prevo;
- /*
- * If the next block overlaps with this proposed allocation
- * then move the start back without adjusting the length,
- * but not before offset 0.
- * This may of course make the start overlap previous block,
- * and if we hit the offset 0 limit then the next block
- * can still overlap too.
- */
- if (!eof && gotp->br_startoff != NULLFILEOFF) {
- if ((delay && gotp->br_startblock == HOLESTARTBLOCK) ||
- (!delay && gotp->br_startblock == DELAYSTARTBLOCK))
- nexto = gotp->br_startoff + gotp->br_blockcount;
+ *first_unused = max;
+ return 0;
+}
+
+/*
+ * Returns the file-relative block number of the last block + 1 before
+ * last_block (input value) in the file.
+ * This is not based on i_size, it is based on the extent records.
+ * Returns 0 for local files, as they do not have extent records.
+ */
+int /* error */
+xfs_bmap_last_before(
+ xfs_trans_t *tp, /* transaction pointer */
+ xfs_inode_t *ip, /* incore inode */
+ xfs_fileoff_t *last_block, /* last block */
+ int whichfork) /* data or attr fork */
+{
+ xfs_fileoff_t bno; /* input file offset */
+ int eof; /* hit end of file */
+ xfs_bmbt_rec_host_t *ep; /* pointer to last extent */
+ int error; /* error return value */
+ xfs_bmbt_irec_t got; /* current extent value */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_extnum_t lastx; /* last extent used */
+ xfs_bmbt_irec_t prev; /* previous extent value */
+
+ if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL)
+ return XFS_ERROR(EIO);
+ if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
+ *last_block = 0;
+ return 0;
+ }
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ if (!(ifp->if_flags & XFS_IFEXTENTS) &&
+ (error = xfs_iread_extents(tp, ip, whichfork)))
+ return error;
+ bno = *last_block - 1;
+ ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
+ &prev);
+ if (eof || xfs_bmbt_get_startoff(ep) > bno) {
+ if (prev.br_startoff == NULLFILEOFF)
+ *last_block = 0;
else
- nexto = gotp->br_startoff;
- } else
- nexto = NULLFILEOFF;
- if (!eof &&
- align_off + align_alen != orig_end &&
- align_off + align_alen > nexto)
- align_off = nexto > align_alen ? nexto - align_alen : 0;
- /*
- * If we're now overlapping the next or previous extent that
- * means we can't fit an extsz piece in this hole. Just move
- * the start forward to the first valid spot and set
- * the length so we hit the end.
- */
- if (align_off != orig_off && align_off < prevo)
- align_off = prevo;
- if (align_off + align_alen != orig_end &&
- align_off + align_alen > nexto &&
- nexto != NULLFILEOFF) {
- ASSERT(nexto > prevo);
- align_alen = nexto - align_off;
+ *last_block = prev.br_startoff + prev.br_blockcount;
}
-
/*
- * If realtime, and the result isn't a multiple of the realtime
- * extent size we need to remove blocks until it is.
+ * Otherwise *last_block is already the right answer.
*/
- if (rt && (temp = (align_alen % mp->m_sb.sb_rextsize))) {
- /*
- * We're not covering the original request, or
- * we won't be able to once we fix the length.
- */
- if (orig_off < align_off ||
- orig_end > align_off + align_alen ||
- align_alen - temp < orig_alen)
- return XFS_ERROR(EINVAL);
- /*
- * Try to fix it by moving the start up.
- */
- if (align_off + temp <= orig_off) {
- align_alen -= temp;
- align_off += temp;
- }
- /*
- * Try to fix it by moving the end in.
- */
- else if (align_off + align_alen - temp >= orig_end)
- align_alen -= temp;
- /*
- * Set the start to the minimum then trim the length.
- */
- else {
- align_alen -= orig_off - align_off;
- align_off = orig_off;
- align_alen -= align_alen % mp->m_sb.sb_rextsize;
- }
- /*
- * Result doesn't cover the request, fail it.
- */
- if (orig_off < align_off || orig_end > align_off + align_alen)
- return XFS_ERROR(EINVAL);
- } else {
- ASSERT(orig_off >= align_off);
- ASSERT(orig_end <= align_off + align_alen);
+ return 0;
+}
+
+STATIC int
+xfs_bmap_last_extent(
+ struct xfs_trans *tp,
+ struct xfs_inode *ip,
+ int whichfork,
+ struct xfs_bmbt_irec *rec,
+ int *is_empty)
+{
+ struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, whichfork);
+ int error;
+ int nextents;
+
+ if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+ error = xfs_iread_extents(tp, ip, whichfork);
+ if (error)
+ return error;
}
-#ifdef DEBUG
- if (!eof && gotp->br_startoff != NULLFILEOFF)
- ASSERT(align_off + align_alen <= gotp->br_startoff);
- if (prevp->br_startoff != NULLFILEOFF)
- ASSERT(align_off >= prevp->br_startoff + prevp->br_blockcount);
-#endif
+ nextents = ifp->if_bytes / sizeof(xfs_bmbt_rec_t);
+ if (nextents == 0) {
+ *is_empty = 1;
+ return 0;
+ }
- *lenp = align_alen;
- *offp = align_off;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, nextents - 1), rec);
+ *is_empty = 0;
return 0;
}
-#define XFS_ALLOC_GAP_UNITS 4
-
-STATIC void
-xfs_bmap_adjacent(
- xfs_bmalloca_t *ap) /* bmap alloc argument struct */
+/*
+ * Check the last inode extent to determine whether this allocation will result
+ * in blocks being allocated at the end of the file. When we allocate new data
+ * blocks at the end of the file which do not start at the previous data block,
+ * we will try to align the new blocks at stripe unit boundaries.
+ *
+ * Returns 0 in bma->aeof if the file (fork) is empty as any new write will be
+ * at, or past the EOF.
+ */
+STATIC int
+xfs_bmap_isaeof(
+ struct xfs_bmalloca *bma,
+ int whichfork)
{
- xfs_fsblock_t adjust; /* adjustment to block numbers */
- xfs_agnumber_t fb_agno; /* ag number of ap->firstblock */
- xfs_mount_t *mp; /* mount point structure */
- int nullfb; /* true if ap->firstblock isn't set */
- int rt; /* true if inode is realtime */
+ struct xfs_bmbt_irec rec;
+ int is_empty;
+ int error;
-#define ISVALID(x,y) \
- (rt ? \
- (x) < mp->m_sb.sb_rblocks : \
- XFS_FSB_TO_AGNO(mp, x) == XFS_FSB_TO_AGNO(mp, y) && \
- XFS_FSB_TO_AGNO(mp, x) < mp->m_sb.sb_agcount && \
- XFS_FSB_TO_AGBNO(mp, x) < mp->m_sb.sb_agblocks)
+ bma->aeof = 0;
+ error = xfs_bmap_last_extent(NULL, bma->ip, whichfork, &rec,
+ &is_empty);
+ if (error || is_empty)
+ return error;
- mp = ap->ip->i_mount;
- nullfb = *ap->firstblock == NULLFSBLOCK;
- rt = XFS_IS_REALTIME_INODE(ap->ip) && ap->userdata;
- fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
- /*
- * If allocating at eof, and there's a previous real block,
- * try to use its last block as our starting point.
- */
- if (ap->eof && ap->prev.br_startoff != NULLFILEOFF &&
- !isnullstartblock(ap->prev.br_startblock) &&
- ISVALID(ap->prev.br_startblock + ap->prev.br_blockcount,
- ap->prev.br_startblock)) {
- ap->blkno = ap->prev.br_startblock + ap->prev.br_blockcount;
- /*
- * Adjust for the gap between prevp and us.
- */
- adjust = ap->offset -
- (ap->prev.br_startoff + ap->prev.br_blockcount);
- if (adjust &&
- ISVALID(ap->blkno + adjust, ap->prev.br_startblock))
- ap->blkno += adjust;
- }
/*
- * If not at eof, then compare the two neighbor blocks.
- * Figure out whether either one gives us a good starting point,
- * and pick the better one.
+ * Check if we are allocation or past the last extent, or at least into
+ * the last delayed allocated extent.
*/
- else if (!ap->eof) {
- xfs_fsblock_t gotbno; /* right side block number */
- xfs_fsblock_t gotdiff=0; /* right side difference */
- xfs_fsblock_t prevbno; /* left side block number */
- xfs_fsblock_t prevdiff=0; /* left side difference */
+ bma->aeof = bma->offset >= rec.br_startoff + rec.br_blockcount ||
+ (bma->offset >= rec.br_startoff &&
+ isnullstartblock(rec.br_startblock));
+ return 0;
+}
- /*
- * If there's a previous (left) block, select a requested
- * start block based on it.
- */
- if (ap->prev.br_startoff != NULLFILEOFF &&
- !isnullstartblock(ap->prev.br_startblock) &&
- (prevbno = ap->prev.br_startblock +
- ap->prev.br_blockcount) &&
- ISVALID(prevbno, ap->prev.br_startblock)) {
- /*
- * Calculate gap to end of previous block.
- */
- adjust = prevdiff = ap->offset -
- (ap->prev.br_startoff +
- ap->prev.br_blockcount);
- /*
- * Figure the startblock based on the previous block's
- * end and the gap size.
- * Heuristic!
- * If the gap is large relative to the piece we're
- * allocating, or using it gives us an invalid block
- * number, then just use the end of the previous block.
- */
- if (prevdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
- ISVALID(prevbno + prevdiff,
- ap->prev.br_startblock))
- prevbno += adjust;
- else
- prevdiff += adjust;
- /*
- * If the firstblock forbids it, can't use it,
- * must use default.
- */
- if (!rt && !nullfb &&
- XFS_FSB_TO_AGNO(mp, prevbno) != fb_agno)
- prevbno = NULLFSBLOCK;
- }
- /*
- * No previous block or can't follow it, just default.
- */
- else
- prevbno = NULLFSBLOCK;
- /*
- * If there's a following (right) block, select a requested
- * start block based on it.
- */
- if (!isnullstartblock(ap->got.br_startblock)) {
- /*
- * Calculate gap to start of next block.
- */
- adjust = gotdiff = ap->got.br_startoff - ap->offset;
- /*
- * Figure the startblock based on the next block's
- * start and the gap size.
- */
- gotbno = ap->got.br_startblock;
- /*
- * Heuristic!
- * If the gap is large relative to the piece we're
- * allocating, or using it gives us an invalid block
- * number, then just use the start of the next block
- * offset by our length.
- */
- if (gotdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
- ISVALID(gotbno - gotdiff, gotbno))
- gotbno -= adjust;
- else if (ISVALID(gotbno - ap->length, gotbno)) {
- gotbno -= ap->length;
- gotdiff += adjust - ap->length;
- } else
- gotdiff += adjust;
- /*
- * If the firstblock forbids it, can't use it,
- * must use default.
- */
- if (!rt && !nullfb &&
- XFS_FSB_TO_AGNO(mp, gotbno) != fb_agno)
- gotbno = NULLFSBLOCK;
- }
- /*
- * No next block, just default.
- */
- else
- gotbno = NULLFSBLOCK;
- /*
- * If both valid, pick the better one, else the only good
- * one, else ap->blkno is already set (to 0 or the inode block).
- */
- if (prevbno != NULLFSBLOCK && gotbno != NULLFSBLOCK)
- ap->blkno = prevdiff <= gotdiff ? prevbno : gotbno;
- else if (prevbno != NULLFSBLOCK)
- ap->blkno = prevbno;
- else if (gotbno != NULLFSBLOCK)
- ap->blkno = gotbno;
- }
-#undef ISVALID
-}
-
-STATIC int
-xfs_bmap_rtalloc(
- xfs_bmalloca_t *ap) /* bmap alloc argument struct */
+/*
+ * Check if the endoff is outside the last extent. If so the caller will grow
+ * the allocation to a stripe unit boundary. All offsets are considered outside
+ * the end of file for an empty fork, so 1 is returned in *eof in that case.
+ */
+int
+xfs_bmap_eof(
+ struct xfs_inode *ip,
+ xfs_fileoff_t endoff,
+ int whichfork,
+ int *eof)
{
- xfs_alloctype_t atype = 0; /* type for allocation routines */
- int error; /* error return value */
- xfs_mount_t *mp; /* mount point structure */
- xfs_extlen_t prod = 0; /* product factor for allocators */
- xfs_extlen_t ralen = 0; /* realtime allocation length */
- xfs_extlen_t align; /* minimum allocation alignment */
- xfs_rtblock_t rtb;
+ struct xfs_bmbt_irec rec;
+ int error;
- mp = ap->ip->i_mount;
- align = xfs_get_extsz_hint(ap->ip);
- prod = align / mp->m_sb.sb_rextsize;
- error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
- align, 1, ap->eof, 0,
- ap->conv, &ap->offset, &ap->length);
- if (error)
+ error = xfs_bmap_last_extent(NULL, ip, whichfork, &rec, eof);
+ if (error || *eof)
return error;
- ASSERT(ap->length);
- ASSERT(ap->length % mp->m_sb.sb_rextsize == 0);
- /*
- * If the offset & length are not perfectly aligned
- * then kill prod, it will just get us in trouble.
- */
- if (do_mod(ap->offset, align) || ap->length % align)
- prod = 1;
- /*
- * Set ralen to be the actual requested length in rtextents.
- */
- ralen = ap->length / mp->m_sb.sb_rextsize;
- /*
- * If the old value was close enough to MAXEXTLEN that
- * we rounded up to it, cut it back so it's valid again.
- * Note that if it's a really large request (bigger than
- * MAXEXTLEN), we don't hear about that number, and can't
- * adjust the starting point to match it.
- */
- if (ralen * mp->m_sb.sb_rextsize >= MAXEXTLEN)
- ralen = MAXEXTLEN / mp->m_sb.sb_rextsize;
+ *eof = endoff >= rec.br_startoff + rec.br_blockcount;
+ return 0;
+}
- /*
- * Lock out other modifications to the RT bitmap inode.
- */
- xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
- xfs_trans_ijoin(ap->tp, mp->m_rbmip, XFS_ILOCK_EXCL);
+/*
+ * Returns the file-relative block number of the first block past eof in
+ * the file. This is not based on i_size, it is based on the extent records.
+ * Returns 0 for local files, as they do not have extent records.
+ */
+int
+xfs_bmap_last_offset(
+ struct xfs_trans *tp,
+ struct xfs_inode *ip,
+ xfs_fileoff_t *last_block,
+ int whichfork)
+{
+ struct xfs_bmbt_irec rec;
+ int is_empty;
+ int error;
- /*
- * If it's an allocation to an empty file at offset 0,
- * pick an extent that will space things out in the rt area.
- */
- if (ap->eof && ap->offset == 0) {
- xfs_rtblock_t uninitialized_var(rtx); /* realtime extent no */
+ *last_block = 0;
- error = xfs_rtpick_extent(mp, ap->tp, ralen, &rtx);
- if (error)
- return error;
- ap->blkno = rtx * mp->m_sb.sb_rextsize;
- } else {
- ap->blkno = 0;
- }
+ if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL)
+ return 0;
- xfs_bmap_adjacent(ap);
+ if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
+ return XFS_ERROR(EIO);
- /*
- * Realtime allocation, done through xfs_rtallocate_extent.
- */
- atype = ap->blkno == 0 ? XFS_ALLOCTYPE_ANY_AG : XFS_ALLOCTYPE_NEAR_BNO;
- do_div(ap->blkno, mp->m_sb.sb_rextsize);
- rtb = ap->blkno;
- ap->length = ralen;
- if ((error = xfs_rtallocate_extent(ap->tp, ap->blkno, 1, ap->length,
- &ralen, atype, ap->wasdel, prod, &rtb)))
- return error;
- if (rtb == NULLFSBLOCK && prod > 1 &&
- (error = xfs_rtallocate_extent(ap->tp, ap->blkno, 1,
- ap->length, &ralen, atype,
- ap->wasdel, 1, &rtb)))
+ error = xfs_bmap_last_extent(NULL, ip, whichfork, &rec, &is_empty);
+ if (error || is_empty)
return error;
- ap->blkno = rtb;
- if (ap->blkno != NULLFSBLOCK) {
- ap->blkno *= mp->m_sb.sb_rextsize;
- ralen *= mp->m_sb.sb_rextsize;
- ap->length = ralen;
- ap->ip->i_d.di_nblocks += ralen;
- xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
- if (ap->wasdel)
- ap->ip->i_delayed_blks -= ralen;
- /*
- * Adjust the disk quota also. This was reserved
- * earlier.
- */
- xfs_trans_mod_dquot_byino(ap->tp, ap->ip,
- ap->wasdel ? XFS_TRANS_DQ_DELRTBCOUNT :
- XFS_TRANS_DQ_RTBCOUNT, (long) ralen);
- } else {
- ap->length = 0;
- }
+
+ *last_block = rec.br_startoff + rec.br_blockcount;
return 0;
}
-STATIC int
-xfs_bmap_btalloc_nullfb(
- struct xfs_bmalloca *ap,
- struct xfs_alloc_arg *args,
- xfs_extlen_t *blen)
+/*
+ * Returns whether the selected fork of the inode has exactly one
+ * block or not. For the data fork we check this matches di_size,
+ * implying the file's range is 0..bsize-1.
+ */
+int /* 1=>1 block, 0=>otherwise */
+xfs_bmap_one_block(
+ xfs_inode_t *ip, /* incore inode */
+ int whichfork) /* data or attr fork */
{
- struct xfs_mount *mp = ap->ip->i_mount;
- struct xfs_perag *pag;
- xfs_agnumber_t ag, startag;
- int notinit = 0;
- int error;
+ xfs_bmbt_rec_host_t *ep; /* ptr to fork's extent */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ int rval; /* return value */
+ xfs_bmbt_irec_t s; /* internal version of extent */
- if (ap->userdata && xfs_inode_is_filestream(ap->ip))
- args->type = XFS_ALLOCTYPE_NEAR_BNO;
- else
- args->type = XFS_ALLOCTYPE_START_BNO;
- args->total = ap->total;
+#ifndef DEBUG
+ if (whichfork == XFS_DATA_FORK)
+ return XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize;
+#endif /* !DEBUG */
+ if (XFS_IFORK_NEXTENTS(ip, whichfork) != 1)
+ return 0;
+ if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
+ return 0;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ ASSERT(ifp->if_flags & XFS_IFEXTENTS);
+ ep = xfs_iext_get_ext(ifp, 0);
+ xfs_bmbt_get_all(ep, &s);
+ rval = s.br_startoff == 0 && s.br_blockcount == 1;
+ if (rval && whichfork == XFS_DATA_FORK)
+ ASSERT(XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize);
+ return rval;
+}
- /*
- * Search for an allocation group with a single extent large enough
- * for the request. If one isn't found, then adjust the minimum
- * allocation size to the largest space found.
- */
- startag = ag = XFS_FSB_TO_AGNO(mp, args->fsbno);
- if (startag == NULLAGNUMBER)
- startag = ag = 0;
+/*
+ * Extent tree manipulation functions used during allocation.
+ */
- pag = xfs_perag_get(mp, ag);
- while (*blen < args->maxlen) {
- if (!pag->pagf_init) {
- error = xfs_alloc_pagf_init(mp, args->tp, ag,
- XFS_ALLOC_FLAG_TRYLOCK);
- if (error) {
- xfs_perag_put(pag);
- return error;
- }
- }
-
- /*
- * See xfs_alloc_fix_freelist...
- */
- if (pag->pagf_init) {
- xfs_extlen_t longest;
- longest = xfs_alloc_longest_free_extent(mp, pag);
- if (*blen < longest)
- *blen = longest;
- } else
- notinit = 1;
+/*
+ * Convert a delayed allocation to a real allocation.
+ */
+STATIC int /* error */
+xfs_bmap_add_extent_delay_real(
+ struct xfs_bmalloca *bma)
+{
+ struct xfs_bmbt_irec *new = &bma->got;
+ int diff; /* temp value */
+ xfs_bmbt_rec_host_t *ep; /* extent entry for idx */
+ int error; /* error return value */
+ int i; /* temp state */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_fileoff_t new_endoff; /* end offset of new entry */
+ xfs_bmbt_irec_t r[3]; /* neighbor extent entries */
+ /* left is 0, right is 1, prev is 2 */
+ int rval=0; /* return value (logging flags) */
+ int state = 0;/* state bits, accessed thru macros */
+ xfs_filblks_t da_new; /* new count del alloc blocks used */
+ xfs_filblks_t da_old; /* old count del alloc blocks used */
+ xfs_filblks_t temp=0; /* value for da_new calculations */
+ xfs_filblks_t temp2=0;/* value for da_new calculations */
+ int tmp_rval; /* partial logging flags */
- if (xfs_inode_is_filestream(ap->ip)) {
- if (*blen >= args->maxlen)
- break;
+ ifp = XFS_IFORK_PTR(bma->ip, XFS_DATA_FORK);
- if (ap->userdata) {
- /*
- * If startag is an invalid AG, we've
- * come here once before and
- * xfs_filestream_new_ag picked the
- * best currently available.
- *
- * Don't continue looping, since we
- * could loop forever.
- */
- if (startag == NULLAGNUMBER)
- break;
+ ASSERT(bma->idx >= 0);
+ ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
+ ASSERT(!isnullstartblock(new->br_startblock));
+ ASSERT(!bma->cur ||
+ (bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
- error = xfs_filestream_new_ag(ap, &ag);
- xfs_perag_put(pag);
- if (error)
- return error;
+ XFS_STATS_INC(xs_add_exlist);
- /* loop again to set 'blen'*/
- startag = NULLAGNUMBER;
- pag = xfs_perag_get(mp, ag);
- continue;
- }
- }
- if (++ag == mp->m_sb.sb_agcount)
- ag = 0;
- if (ag == startag)
- break;
- xfs_perag_put(pag);
- pag = xfs_perag_get(mp, ag);
- }
- xfs_perag_put(pag);
+#define LEFT r[0]
+#define RIGHT r[1]
+#define PREV r[2]
/*
- * Since the above loop did a BUF_TRYLOCK, it is
- * possible that there is space for this request.
- */
- if (notinit || *blen < ap->minlen)
- args->minlen = ap->minlen;
- /*
- * If the best seen length is less than the request
- * length, use the best as the minimum.
+ * Set up a bunch of variables to make the tests simpler.
*/
- else if (*blen < args->maxlen)
- args->minlen = *blen;
+ ep = xfs_iext_get_ext(ifp, bma->idx);
+ xfs_bmbt_get_all(ep, &PREV);
+ new_endoff = new->br_startoff + new->br_blockcount;
+ ASSERT(PREV.br_startoff <= new->br_startoff);
+ ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
+
+ da_old = startblockval(PREV.br_startblock);
+ da_new = 0;
+
/*
- * Otherwise we've seen an extent as big as maxlen,
- * use that as the minimum.
+ * Set flags determining what part of the previous delayed allocation
+ * extent is being replaced by a real allocation.
*/
- else
- args->minlen = args->maxlen;
+ if (PREV.br_startoff == new->br_startoff)
+ state |= BMAP_LEFT_FILLING;
+ if (PREV.br_startoff + PREV.br_blockcount == new_endoff)
+ state |= BMAP_RIGHT_FILLING;
/*
- * set the failure fallback case to look in the selected
- * AG as the stream may have moved.
+ * Check and set flags if this segment has a left neighbor.
+ * Don't set contiguous if the combined extent would be too large.
*/
- if (xfs_inode_is_filestream(ap->ip))
- ap->blkno = args->fsbno = XFS_AGB_TO_FSB(mp, ag, 0);
-
- return 0;
-}
-
-STATIC int
-xfs_bmap_btalloc(
- xfs_bmalloca_t *ap) /* bmap alloc argument struct */
-{
- xfs_mount_t *mp; /* mount point structure */
- xfs_alloctype_t atype = 0; /* type for allocation routines */
- xfs_extlen_t align; /* minimum allocation alignment */
- xfs_agnumber_t fb_agno; /* ag number of ap->firstblock */
- xfs_agnumber_t ag;
- xfs_alloc_arg_t args;
- xfs_extlen_t blen;
- xfs_extlen_t nextminlen = 0;
- int nullfb; /* true if ap->firstblock isn't set */
- int isaligned;
- int tryagain;
- int error;
-
- ASSERT(ap->length);
+ if (bma->idx > 0) {
+ state |= BMAP_LEFT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &LEFT);
- mp = ap->ip->i_mount;
- align = ap->userdata ? xfs_get_extsz_hint(ap->ip) : 0;
- if (unlikely(align)) {
- error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
- align, 0, ap->eof, 0, ap->conv,
- &ap->offset, &ap->length);
- ASSERT(!error);
- ASSERT(ap->length);
+ if (isnullstartblock(LEFT.br_startblock))
+ state |= BMAP_LEFT_DELAY;
}
- nullfb = *ap->firstblock == NULLFSBLOCK;
- fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
- if (nullfb) {
- if (ap->userdata && xfs_inode_is_filestream(ap->ip)) {
- ag = xfs_filestream_lookup_ag(ap->ip);
- ag = (ag != NULLAGNUMBER) ? ag : 0;
- ap->blkno = XFS_AGB_TO_FSB(mp, ag, 0);
- } else {
- ap->blkno = XFS_INO_TO_FSB(mp, ap->ip->i_ino);
- }
- } else
- ap->blkno = *ap->firstblock;
- xfs_bmap_adjacent(ap);
+ if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
+ LEFT.br_startoff + LEFT.br_blockcount == new->br_startoff &&
+ LEFT.br_startblock + LEFT.br_blockcount == new->br_startblock &&
+ LEFT.br_state == new->br_state &&
+ LEFT.br_blockcount + new->br_blockcount <= MAXEXTLEN)
+ state |= BMAP_LEFT_CONTIG;
/*
- * If allowed, use ap->blkno; otherwise must use firstblock since
- * it's in the right allocation group.
- */
- if (nullfb || XFS_FSB_TO_AGNO(mp, ap->blkno) == fb_agno)
- ;
- else
- ap->blkno = *ap->firstblock;
- /*
- * Normal allocation, done through xfs_alloc_vextent.
+ * Check and set flags if this segment has a right neighbor.
+ * Don't set contiguous if the combined extent would be too large.
+ * Also check for all-three-contiguous being too large.
*/
- tryagain = isaligned = 0;
- memset(&args, 0, sizeof(args));
- args.tp = ap->tp;
- args.mp = mp;
- args.fsbno = ap->blkno;
+ if (bma->idx < bma->ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
+ state |= BMAP_RIGHT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx + 1), &RIGHT);
- /* Trim the allocation back to the maximum an AG can fit. */
- args.maxlen = MIN(ap->length, XFS_ALLOC_AG_MAX_USABLE(mp));
- args.firstblock = *ap->firstblock;
- blen = 0;
- if (nullfb) {
- error = xfs_bmap_btalloc_nullfb(ap, &args, &blen);
- if (error)
- return error;
- } else if (ap->flist->xbf_low) {
- if (xfs_inode_is_filestream(ap->ip))
- args.type = XFS_ALLOCTYPE_FIRST_AG;
- else
- args.type = XFS_ALLOCTYPE_START_BNO;
- args.total = args.minlen = ap->minlen;
- } else {
- args.type = XFS_ALLOCTYPE_NEAR_BNO;
- args.total = ap->total;
- args.minlen = ap->minlen;
- }
- /* apply extent size hints if obtained earlier */
- if (unlikely(align)) {
- args.prod = align;
- if ((args.mod = (xfs_extlen_t)do_mod(ap->offset, args.prod)))
- args.mod = (xfs_extlen_t)(args.prod - args.mod);
- } else if (mp->m_sb.sb_blocksize >= PAGE_CACHE_SIZE) {
- args.prod = 1;
- args.mod = 0;
- } else {
- args.prod = PAGE_CACHE_SIZE >> mp->m_sb.sb_blocklog;
- if ((args.mod = (xfs_extlen_t)(do_mod(ap->offset, args.prod))))
- args.mod = (xfs_extlen_t)(args.prod - args.mod);
+ if (isnullstartblock(RIGHT.br_startblock))
+ state |= BMAP_RIGHT_DELAY;
}
+
+ if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
+ new_endoff == RIGHT.br_startoff &&
+ new->br_startblock + new->br_blockcount == RIGHT.br_startblock &&
+ new->br_state == RIGHT.br_state &&
+ new->br_blockcount + RIGHT.br_blockcount <= MAXEXTLEN &&
+ ((state & (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
+ BMAP_RIGHT_FILLING)) !=
+ (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
+ BMAP_RIGHT_FILLING) ||
+ LEFT.br_blockcount + new->br_blockcount + RIGHT.br_blockcount
+ <= MAXEXTLEN))
+ state |= BMAP_RIGHT_CONTIG;
+
+ error = 0;
/*
- * If we are not low on available data blocks, and the
- * underlying logical volume manager is a stripe, and
- * the file offset is zero then try to allocate data
- * blocks on stripe unit boundary.
- * NOTE: ap->aeof is only set if the allocation length
- * is >= the stripe unit and the allocation offset is
- * at the end of file.
+ * Switch out based on the FILLING and CONTIG state bits.
*/
- if (!ap->flist->xbf_low && ap->aeof) {
- if (!ap->offset) {
- args.alignment = mp->m_dalign;
- atype = args.type;
- isaligned = 1;
- /*
- * Adjust for alignment
- */
- if (blen > args.alignment && blen <= args.maxlen)
- args.minlen = blen - args.alignment;
- args.minalignslop = 0;
- } else {
- /*
- * First try an exact bno allocation.
- * If it fails then do a near or start bno
- * allocation with alignment turned on.
- */
- atype = args.type;
- tryagain = 1;
- args.type = XFS_ALLOCTYPE_THIS_BNO;
- args.alignment = 1;
- /*
- * Compute the minlen+alignment for the
- * next case. Set slop so that the value
- * of minlen+alignment+slop doesn't go up
- * between the calls.
- */
- if (blen > mp->m_dalign && blen <= args.maxlen)
- nextminlen = blen - mp->m_dalign;
- else
- nextminlen = args.minlen;
- if (nextminlen + mp->m_dalign > args.minlen + 1)
- args.minalignslop =
- nextminlen + mp->m_dalign -
- args.minlen - 1;
- else
- args.minalignslop = 0;
- }
- } else {
- args.alignment = 1;
- args.minalignslop = 0;
- }
- args.minleft = ap->minleft;
- args.wasdel = ap->wasdel;
- args.isfl = 0;
- args.userdata = ap->userdata;
- if ((error = xfs_alloc_vextent(&args)))
- return error;
- if (tryagain && args.fsbno == NULLFSBLOCK) {
+ switch (state & (BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
+ BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG)) {
+ case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
+ BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
/*
- * Exact allocation failed. Now try with alignment
- * turned on.
+ * Filling in all of a previously delayed allocation extent.
+ * The left and right neighbors are both contiguous with new.
*/
- args.type = atype;
- args.fsbno = ap->blkno;
- args.alignment = mp->m_dalign;
- args.minlen = nextminlen;
- args.minalignslop = 0;
- isaligned = 1;
- if ((error = xfs_alloc_vextent(&args)))
- return error;
- }
- if (isaligned && args.fsbno == NULLFSBLOCK) {
+ bma->idx--;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
+ LEFT.br_blockcount + PREV.br_blockcount +
+ RIGHT.br_blockcount);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+
+ xfs_iext_remove(bma->ip, bma->idx + 1, 2, state);
+ bma->ip->i_d.di_nextents--;
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
+ RIGHT.br_startblock,
+ RIGHT.br_blockcount, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_btree_delete(bma->cur, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_btree_decrement(bma->cur, 0, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
+ LEFT.br_startblock,
+ LEFT.br_blockcount +
+ PREV.br_blockcount +
+ RIGHT.br_blockcount, LEFT.br_state);
+ if (error)
+ goto done;
+ }
+ break;
+
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
/*
- * allocation failed, so turn off alignment and
- * try again.
+ * Filling in all of a previously delayed allocation extent.
+ * The left neighbor is contiguous, the right is not.
*/
- args.type = atype;
- args.fsbno = ap->blkno;
- args.alignment = 0;
- if ((error = xfs_alloc_vextent(&args)))
- return error;
- }
- if (args.fsbno == NULLFSBLOCK && nullfb &&
- args.minlen > ap->minlen) {
- args.minlen = ap->minlen;
- args.type = XFS_ALLOCTYPE_START_BNO;
- args.fsbno = ap->blkno;
- if ((error = xfs_alloc_vextent(&args)))
- return error;
- }
- if (args.fsbno == NULLFSBLOCK && nullfb) {
- args.fsbno = 0;
- args.type = XFS_ALLOCTYPE_FIRST_AG;
- args.total = ap->minlen;
- args.minleft = 0;
- if ((error = xfs_alloc_vextent(&args)))
- return error;
- ap->flist->xbf_low = 1;
- }
- if (args.fsbno != NULLFSBLOCK) {
+ bma->idx--;
+
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
+ LEFT.br_blockcount + PREV.br_blockcount);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+
+ xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_DEXT;
+ else {
+ rval = 0;
+ error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
+ LEFT.br_startblock, LEFT.br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
+ LEFT.br_startblock,
+ LEFT.br_blockcount +
+ PREV.br_blockcount, LEFT.br_state);
+ if (error)
+ goto done;
+ }
+ break;
+
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
/*
- * check the allocation happened at the same or higher AG than
- * the first block that was allocated.
+ * Filling in all of a previously delayed allocation extent.
+ * The right neighbor is contiguous, the left is not.
*/
- ASSERT(*ap->firstblock == NULLFSBLOCK ||
- XFS_FSB_TO_AGNO(mp, *ap->firstblock) ==
- XFS_FSB_TO_AGNO(mp, args.fsbno) ||
- (ap->flist->xbf_low &&
- XFS_FSB_TO_AGNO(mp, *ap->firstblock) <
- XFS_FSB_TO_AGNO(mp, args.fsbno)));
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_startblock(ep, new->br_startblock);
+ xfs_bmbt_set_blockcount(ep,
+ PREV.br_blockcount + RIGHT.br_blockcount);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
- ap->blkno = args.fsbno;
- if (*ap->firstblock == NULLFSBLOCK)
- *ap->firstblock = args.fsbno;
- ASSERT(nullfb || fb_agno == args.agno ||
- (ap->flist->xbf_low && fb_agno < args.agno));
- ap->length = args.len;
- ap->ip->i_d.di_nblocks += args.len;
- xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
- if (ap->wasdel)
- ap->ip->i_delayed_blks -= args.len;
+ xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_DEXT;
+ else {
+ rval = 0;
+ error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
+ RIGHT.br_startblock,
+ RIGHT.br_blockcount, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, PREV.br_startoff,
+ new->br_startblock,
+ PREV.br_blockcount +
+ RIGHT.br_blockcount, PREV.br_state);
+ if (error)
+ goto done;
+ }
+ break;
+
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
/*
- * Adjust the disk quota also. This was reserved
- * earlier.
+ * Filling in all of a previously delayed allocation extent.
+ * Neither the left nor right neighbors are contiguous with
+ * the new one.
*/
- xfs_trans_mod_dquot_byino(ap->tp, ap->ip,
- ap->wasdel ? XFS_TRANS_DQ_DELBCOUNT :
- XFS_TRANS_DQ_BCOUNT,
- (long) args.len);
- } else {
- ap->blkno = NULLFSBLOCK;
- ap->length = 0;
- }
- return 0;
-}
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_startblock(ep, new->br_startblock);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-/*
- * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file.
- * It figures out where to ask the underlying allocator to put the new extent.
- */
-STATIC int
-xfs_bmap_alloc(
- xfs_bmalloca_t *ap) /* bmap alloc argument struct */
-{
- if (XFS_IS_REALTIME_INODE(ap->ip) && ap->userdata)
- return xfs_bmap_rtalloc(ap);
- return xfs_bmap_btalloc(ap);
-}
+ bma->ip->i_d.di_nextents++;
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 0, done);
+ bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+ error = xfs_btree_insert(bma->cur, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ }
+ break;
-/*
- * Transform a btree format file with only one leaf node, where the
- * extents list will fit in the inode, into an extents format file.
- * Since the file extents are already in-core, all we have to do is
- * give up the space for the btree root and pitch the leaf block.
- */
-STATIC int /* error */
-xfs_bmap_btree_to_extents(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_btree_cur_t *cur, /* btree cursor */
- int *logflagsp, /* inode logging flags */
- int whichfork) /* data or attr fork */
-{
- /* REFERENCED */
- struct xfs_btree_block *cblock;/* child btree block */
- xfs_fsblock_t cbno; /* child block number */
- xfs_buf_t *cbp; /* child block's buffer */
- int error; /* error return value */
- xfs_ifork_t *ifp; /* inode fork data */
- xfs_mount_t *mp; /* mount point structure */
- __be64 *pp; /* ptr to block address */
- struct xfs_btree_block *rblock;/* root btree block */
-
- mp = ip->i_mount;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- ASSERT(ifp->if_flags & XFS_IFEXTENTS);
- ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE);
- rblock = ifp->if_broot;
- ASSERT(be16_to_cpu(rblock->bb_level) == 1);
- ASSERT(be16_to_cpu(rblock->bb_numrecs) == 1);
- ASSERT(xfs_bmbt_maxrecs(mp, ifp->if_broot_bytes, 0) == 1);
- pp = XFS_BMAP_BROOT_PTR_ADDR(mp, rblock, 1, ifp->if_broot_bytes);
- cbno = be64_to_cpu(*pp);
- *logflagsp = 0;
-#ifdef DEBUG
- if ((error = xfs_btree_check_lptr(cur, cbno, 1)))
- return error;
-#endif
- error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, XFS_BMAP_BTREE_REF,
- &xfs_bmbt_buf_ops);
- if (error)
- return error;
- cblock = XFS_BUF_TO_BLOCK(cbp);
- if ((error = xfs_btree_check_block(cur, cblock, 0, cbp)))
- return error;
- xfs_bmap_add_free(cbno, 1, cur->bc_private.b.flist, mp);
- ip->i_d.di_nblocks--;
- xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, -1L);
- xfs_trans_binval(tp, cbp);
- if (cur->bc_bufs[0] == cbp)
- cur->bc_bufs[0] = NULL;
- xfs_iroot_realloc(ip, -1, whichfork);
- ASSERT(ifp->if_broot == NULL);
- ASSERT((ifp->if_flags & XFS_IFBROOT) == 0);
- XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
- *logflagsp = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
- return 0;
-}
-
-/*
- * Called by xfs_bmapi to update file extent records and the btree
- * after removing space (or undoing a delayed allocation).
- */
-STATIC int /* error */
-xfs_bmap_del_extent(
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_trans_t *tp, /* current transaction pointer */
- xfs_extnum_t *idx, /* extent number to update/delete */
- xfs_bmap_free_t *flist, /* list of extents to be freed */
- xfs_btree_cur_t *cur, /* if null, not a btree */
- xfs_bmbt_irec_t *del, /* data to remove from extents */
- int *logflagsp, /* inode logging flags */
- int whichfork) /* data or attr fork */
-{
- xfs_filblks_t da_new; /* new delay-alloc indirect blocks */
- xfs_filblks_t da_old; /* old delay-alloc indirect blocks */
- xfs_fsblock_t del_endblock=0; /* first block past del */
- xfs_fileoff_t del_endoff; /* first offset past del */
- int delay; /* current block is delayed allocated */
- int do_fx; /* free extent at end of routine */
- xfs_bmbt_rec_host_t *ep; /* current extent entry pointer */
- int error; /* error return value */
- int flags; /* inode logging flags */
- xfs_bmbt_irec_t got; /* current extent entry */
- xfs_fileoff_t got_endoff; /* first offset past got */
- int i; /* temp state */
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_mount_t *mp; /* mount structure */
- xfs_filblks_t nblks; /* quota/sb block count */
- xfs_bmbt_irec_t new; /* new record to be inserted */
- /* REFERENCED */
- uint qfield; /* quota field to update */
- xfs_filblks_t temp; /* for indirect length calculations */
- xfs_filblks_t temp2; /* for indirect length calculations */
- int state = 0;
-
- XFS_STATS_INC(xs_del_exlist);
-
- if (whichfork == XFS_ATTR_FORK)
- state |= BMAP_ATTRFORK;
-
- mp = ip->i_mount;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- ASSERT((*idx >= 0) && (*idx < ifp->if_bytes /
- (uint)sizeof(xfs_bmbt_rec_t)));
- ASSERT(del->br_blockcount > 0);
- ep = xfs_iext_get_ext(ifp, *idx);
- xfs_bmbt_get_all(ep, &got);
- ASSERT(got.br_startoff <= del->br_startoff);
- del_endoff = del->br_startoff + del->br_blockcount;
- got_endoff = got.br_startoff + got.br_blockcount;
- ASSERT(got_endoff >= del_endoff);
- delay = isnullstartblock(got.br_startblock);
- ASSERT(isnullstartblock(del->br_startblock) == delay);
- flags = 0;
- qfield = 0;
- error = 0;
- /*
- * If deleting a real allocation, must free up the disk space.
- */
- if (!delay) {
- flags = XFS_ILOG_CORE;
+ case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG:
/*
- * Realtime allocation. Free it and record di_nblocks update.
+ * Filling in the first part of a previous delayed allocation.
+ * The left neighbor is contiguous.
*/
- if (whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip)) {
- xfs_fsblock_t bno;
- xfs_filblks_t len;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx - 1),
+ LEFT.br_blockcount + new->br_blockcount);
+ xfs_bmbt_set_startoff(ep,
+ PREV.br_startoff + new->br_blockcount);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx - 1, state, _THIS_IP_);
- ASSERT(do_mod(del->br_blockcount,
- mp->m_sb.sb_rextsize) == 0);
- ASSERT(do_mod(del->br_startblock,
- mp->m_sb.sb_rextsize) == 0);
- bno = del->br_startblock;
- len = del->br_blockcount;
- do_div(bno, mp->m_sb.sb_rextsize);
- do_div(len, mp->m_sb.sb_rextsize);
- error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
+ temp = PREV.br_blockcount - new->br_blockcount;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep, temp);
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_DEXT;
+ else {
+ rval = 0;
+ error = xfs_bmbt_lookup_eq(bma->cur, LEFT.br_startoff,
+ LEFT.br_startblock, LEFT.br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, LEFT.br_startoff,
+ LEFT.br_startblock,
+ LEFT.br_blockcount +
+ new->br_blockcount,
+ LEFT.br_state);
if (error)
goto done;
- do_fx = 0;
- nblks = len * mp->m_sb.sb_rextsize;
- qfield = XFS_TRANS_DQ_RTBCOUNT;
}
+ da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
+ startblockval(PREV.br_startblock));
+ xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+
+ bma->idx--;
+ break;
+
+ case BMAP_LEFT_FILLING:
/*
- * Ordinary allocation.
+ * Filling in the first part of a previous delayed allocation.
+ * The left neighbor is not contiguous.
*/
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_startoff(ep, new_endoff);
+ temp = PREV.br_blockcount - new->br_blockcount;
+ xfs_bmbt_set_blockcount(ep, temp);
+ xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
+ bma->ip->i_d.di_nextents++;
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
else {
- do_fx = 1;
- nblks = del->br_blockcount;
- qfield = XFS_TRANS_DQ_BCOUNT;
- }
- /*
- * Set up del_endblock and cur for later.
- */
- del_endblock = del->br_startblock + del->br_blockcount;
- if (cur) {
- if ((error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
- got.br_startblock, got.br_blockcount,
- &i)))
+ rval = XFS_ILOG_CORE;
+ error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 0, done);
+ bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+ error = xfs_btree_insert(bma->cur, &i);
+ if (error)
goto done;
XFS_WANT_CORRUPTED_GOTO(i == 1, done);
}
- da_old = da_new = 0;
- } else {
- da_old = startblockval(got.br_startblock);
- da_new = 0;
- nblks = 0;
- do_fx = 0;
- }
- /*
- * Set flag value to use in switch statement.
- * Left-contig is 2, right-contig is 1.
- */
- switch (((got.br_startoff == del->br_startoff) << 1) |
- (got_endoff == del_endoff)) {
- case 3:
- /*
- * Matches the whole extent. Delete the entry.
- */
- xfs_iext_remove(ip, *idx, 1,
- whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0);
- --*idx;
- if (delay)
- break;
- XFS_IFORK_NEXT_SET(ip, whichfork,
- XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
- flags |= XFS_ILOG_CORE;
- if (!cur) {
- flags |= xfs_ilog_fext(whichfork);
- break;
+ if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
+ error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+ bma->firstblock, bma->flist,
+ &bma->cur, 1, &tmp_rval, XFS_DATA_FORK);
+ rval |= tmp_rval;
+ if (error)
+ goto done;
}
- if ((error = xfs_btree_delete(cur, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
+ startblockval(PREV.br_startblock) -
+ (bma->cur ? bma->cur->bc_private.b.allocated : 0));
+ ep = xfs_iext_get_ext(ifp, bma->idx + 1);
+ xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
+ trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
break;
- case 2:
+ case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
/*
- * Deleting the first part of the extent.
+ * Filling in the last part of a previous delayed allocation.
+ * The right neighbor is contiguous with the new allocation.
*/
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_startoff(ep, del_endoff);
- temp = got.br_blockcount - del->br_blockcount;
+ temp = PREV.br_blockcount - new->br_blockcount;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
xfs_bmbt_set_blockcount(ep, temp);
- if (delay) {
- temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
- da_old);
- xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- da_new = temp;
- break;
- }
- xfs_bmbt_set_startblock(ep, del_endblock);
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- if (!cur) {
- flags |= xfs_ilog_fext(whichfork);
- break;
+ xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->idx + 1),
+ new->br_startoff, new->br_startblock,
+ new->br_blockcount + RIGHT.br_blockcount,
+ RIGHT.br_state);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx + 1, state, _THIS_IP_);
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_DEXT;
+ else {
+ rval = 0;
+ error = xfs_bmbt_lookup_eq(bma->cur, RIGHT.br_startoff,
+ RIGHT.br_startblock,
+ RIGHT.br_blockcount, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, new->br_startoff,
+ new->br_startblock,
+ new->br_blockcount +
+ RIGHT.br_blockcount,
+ RIGHT.br_state);
+ if (error)
+ goto done;
}
- if ((error = xfs_bmbt_update(cur, del_endoff, del_endblock,
- got.br_blockcount - del->br_blockcount,
- got.br_state)))
- goto done;
+
+ da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
+ startblockval(PREV.br_startblock));
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+
+ bma->idx++;
break;
- case 1:
+ case BMAP_RIGHT_FILLING:
/*
- * Deleting the last part of the extent.
+ * Filling in the last part of a previous delayed allocation.
+ * The right neighbor is not contiguous.
*/
- temp = got.br_blockcount - del->br_blockcount;
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ temp = PREV.br_blockcount - new->br_blockcount;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
xfs_bmbt_set_blockcount(ep, temp);
- if (delay) {
- temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
- da_old);
- xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- da_new = temp;
- break;
+ xfs_iext_insert(bma->ip, bma->idx + 1, 1, new, state);
+ bma->ip->i_d.di_nextents++;
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 0, done);
+ bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+ error = xfs_btree_insert(bma->cur, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
}
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- if (!cur) {
- flags |= xfs_ilog_fext(whichfork);
- break;
+
+ if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
+ error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+ bma->firstblock, bma->flist, &bma->cur, 1,
+ &tmp_rval, XFS_DATA_FORK);
+ rval |= tmp_rval;
+ if (error)
+ goto done;
}
- if ((error = xfs_bmbt_update(cur, got.br_startoff,
- got.br_startblock,
- got.br_blockcount - del->br_blockcount,
- got.br_state)))
- goto done;
+ da_new = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(bma->ip, temp),
+ startblockval(PREV.br_startblock) -
+ (bma->cur ? bma->cur->bc_private.b.allocated : 0));
+ ep = xfs_iext_get_ext(ifp, bma->idx);
+ xfs_bmbt_set_startblock(ep, nullstartblock(da_new));
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+
+ bma->idx++;
break;
case 0:
/*
- * Deleting the middle of the extent.
+ * Filling in the middle part of a previous delayed allocation.
+ * Contiguity is impossible here.
+ * This case is avoided almost all the time.
+ *
+ * We start with a delayed allocation:
+ *
+ * +ddddddddddddddddddddddddddddddddddddddddddddddddddddddd+
+ * PREV @ idx
+ *
+ * and we are allocating:
+ * +rrrrrrrrrrrrrrrrr+
+ * new
+ *
+ * and we set it up for insertion as:
+ * +ddddddddddddddddddd+rrrrrrrrrrrrrrrrr+ddddddddddddddddd+
+ * new
+ * PREV @ idx LEFT RIGHT
+ * inserted at idx + 1
*/
- temp = del->br_startoff - got.br_startoff;
- trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
- xfs_bmbt_set_blockcount(ep, temp);
- new.br_startoff = del_endoff;
- temp2 = got_endoff - del_endoff;
- new.br_blockcount = temp2;
- new.br_state = got.br_state;
- if (!delay) {
- new.br_startblock = del_endblock;
- flags |= XFS_ILOG_CORE;
- if (cur) {
- if ((error = xfs_bmbt_update(cur,
- got.br_startoff,
- got.br_startblock, temp,
- got.br_state)))
- goto done;
- if ((error = xfs_btree_increment(cur, 0, &i)))
- goto done;
- cur->bc_rec.b = new;
- error = xfs_btree_insert(cur, &i);
- if (error && error != ENOSPC)
- goto done;
- /*
- * If get no-space back from btree insert,
- * it tried a split, and we have a zero
- * block reservation.
- * Fix up our state and return the error.
- */
- if (error == ENOSPC) {
- /*
- * Reset the cursor, don't trust
- * it after any insert operation.
- */
- if ((error = xfs_bmbt_lookup_eq(cur,
- got.br_startoff,
- got.br_startblock,
- temp, &i)))
- goto done;
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- /*
- * Update the btree record back
- * to the original value.
- */
- if ((error = xfs_bmbt_update(cur,
- got.br_startoff,
- got.br_startblock,
- got.br_blockcount,
- got.br_state)))
- goto done;
- /*
- * Reset the extent record back
- * to the original value.
- */
- xfs_bmbt_set_blockcount(ep,
- got.br_blockcount);
- flags = 0;
- error = XFS_ERROR(ENOSPC);
- goto done;
- }
- XFS_WANT_CORRUPTED_GOTO(i == 1, done);
- } else
- flags |= xfs_ilog_fext(whichfork);
- XFS_IFORK_NEXT_SET(ip, whichfork,
- XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
- } else {
- ASSERT(whichfork == XFS_DATA_FORK);
- temp = xfs_bmap_worst_indlen(ip, temp);
- xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
- temp2 = xfs_bmap_worst_indlen(ip, temp2);
- new.br_startblock = nullstartblock((int)temp2);
- da_new = temp + temp2;
- while (da_new > da_old) {
- if (temp) {
- temp--;
- da_new--;
- xfs_bmbt_set_startblock(ep,
- nullstartblock((int)temp));
- }
- if (da_new == da_old)
- break;
- if (temp2) {
- temp2--;
- da_new--;
- new.br_startblock =
- nullstartblock((int)temp2);
- }
- }
+ temp = new->br_startoff - PREV.br_startoff;
+ temp2 = PREV.br_startoff + PREV.br_blockcount - new_endoff;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, 0, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep, temp); /* truncate PREV */
+ LEFT = *new;
+ RIGHT.br_state = PREV.br_state;
+ RIGHT.br_startblock = nullstartblock(
+ (int)xfs_bmap_worst_indlen(bma->ip, temp2));
+ RIGHT.br_startoff = new_endoff;
+ RIGHT.br_blockcount = temp2;
+ /* insert LEFT (r[0]) and RIGHT (r[1]) at the same time */
+ xfs_iext_insert(bma->ip, bma->idx + 1, 2, &LEFT, state);
+ bma->ip->i_d.di_nextents++;
+ if (bma->cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ error = xfs_bmbt_lookup_eq(bma->cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 0, done);
+ bma->cur->bc_rec.b.br_state = XFS_EXT_NORM;
+ error = xfs_btree_insert(bma->cur, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
}
- trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- xfs_iext_insert(ip, *idx + 1, 1, &new, state);
- ++*idx;
- break;
- }
- /*
- * If we need to, add to list of extents to delete.
- */
- if (do_fx)
- xfs_bmap_add_free(del->br_startblock, del->br_blockcount, flist,
- mp);
- /*
- * Adjust inode # blocks in the file.
- */
- if (nblks)
- ip->i_d.di_nblocks -= nblks;
- /*
- * Adjust quota data.
- */
- if (qfield)
- xfs_trans_mod_dquot_byino(tp, ip, qfield, (long)-nblks);
- /*
- * Account for change in delayed indirect blocks.
- * Nothing to do for disk quota accounting here.
- */
- ASSERT(da_old >= da_new);
- if (da_old > da_new) {
- xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
- (int64_t)(da_old - da_new), 0);
+ if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
+ error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+ bma->firstblock, bma->flist, &bma->cur,
+ 1, &tmp_rval, XFS_DATA_FORK);
+ rval |= tmp_rval;
+ if (error)
+ goto done;
+ }
+ temp = xfs_bmap_worst_indlen(bma->ip, temp);
+ temp2 = xfs_bmap_worst_indlen(bma->ip, temp2);
+ diff = (int)(temp + temp2 - startblockval(PREV.br_startblock) -
+ (bma->cur ? bma->cur->bc_private.b.allocated : 0));
+ if (diff > 0) {
+ error = xfs_icsb_modify_counters(bma->ip->i_mount,
+ XFS_SBS_FDBLOCKS,
+ -((int64_t)diff), 0);
+ ASSERT(!error);
+ if (error)
+ goto done;
+ }
+
+ ep = xfs_iext_get_ext(ifp, bma->idx);
+ xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
+ xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, bma->idx + 2),
+ nullstartblock((int)temp2));
+ trace_xfs_bmap_post_update(bma->ip, bma->idx + 2, state, _THIS_IP_);
+
+ bma->idx++;
+ da_new = temp + temp2;
+ break;
+
+ case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_CONTIG:
+ case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
+ case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ case BMAP_LEFT_CONTIG:
+ case BMAP_RIGHT_CONTIG:
+ /*
+ * These cases are all impossible.
+ */
+ ASSERT(0);
+ }
+
+ /* convert to a btree if necessary */
+ if (xfs_bmap_needs_btree(bma->ip, XFS_DATA_FORK)) {
+ int tmp_logflags; /* partial log flag return val */
+
+ ASSERT(bma->cur == NULL);
+ error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+ bma->firstblock, bma->flist, &bma->cur,
+ da_old > 0, &tmp_logflags, XFS_DATA_FORK);
+ bma->logflags |= tmp_logflags;
+ if (error)
+ goto done;
+ }
+
+ /* adjust for changes in reserved delayed indirect blocks */
+ if (da_old || da_new) {
+ temp = da_new;
+ if (bma->cur)
+ temp += bma->cur->bc_private.b.allocated;
+ ASSERT(temp <= da_old);
+ if (temp < da_old)
+ xfs_icsb_modify_counters(bma->ip->i_mount,
+ XFS_SBS_FDBLOCKS,
+ (int64_t)(da_old - temp), 0);
}
+
+ /* clear out the allocated field, done with it now in any case. */
+ if (bma->cur)
+ bma->cur->bc_private.b.allocated = 0;
+
+ xfs_bmap_check_leaf_extents(bma->cur, bma->ip, XFS_DATA_FORK);
done:
- *logflagsp = flags;
+ bma->logflags |= rval;
return error;
+#undef LEFT
+#undef RIGHT
+#undef PREV
}
/*
- * Remove the entry "free" from the free item list. Prev points to the
- * previous entry, unless "free" is the head of the list.
+ * Convert an unwritten allocation to a real allocation or vice versa.
*/
-STATIC void
-xfs_bmap_del_free(
- xfs_bmap_free_t *flist, /* free item list header */
- xfs_bmap_free_item_t *prev, /* previous item on list, if any */
- xfs_bmap_free_item_t *free) /* list item to be freed */
+STATIC int /* error */
+xfs_bmap_add_extent_unwritten_real(
+ struct xfs_trans *tp,
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_extnum_t *idx, /* extent number to update/insert */
+ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */
+ xfs_bmbt_irec_t *new, /* new data to add to file extents */
+ xfs_fsblock_t *first, /* pointer to firstblock variable */
+ xfs_bmap_free_t *flist, /* list of extents to be freed */
+ int *logflagsp) /* inode logging flags */
{
- if (prev)
- prev->xbfi_next = free->xbfi_next;
- else
- flist->xbf_first = free->xbfi_next;
- flist->xbf_count--;
- kmem_zone_free(xfs_bmap_free_item_zone, free);
-}
+ xfs_btree_cur_t *cur; /* btree cursor */
+ xfs_bmbt_rec_host_t *ep; /* extent entry for idx */
+ int error; /* error return value */
+ int i; /* temp state */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_fileoff_t new_endoff; /* end offset of new entry */
+ xfs_exntst_t newext; /* new extent state */
+ xfs_exntst_t oldext; /* old extent state */
+ xfs_bmbt_irec_t r[3]; /* neighbor extent entries */
+ /* left is 0, right is 1, prev is 2 */
+ int rval=0; /* return value (logging flags) */
+ int state = 0;/* state bits, accessed thru macros */
-/*
- * Convert an extents-format file into a btree-format file.
- * The new file will have a root block (in the inode) and a single child block.
- */
-STATIC int /* error */
-xfs_bmap_extents_to_btree(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first-block-allocated */
- xfs_bmap_free_t *flist, /* blocks freed in xaction */
- xfs_btree_cur_t **curp, /* cursor returned to caller */
- int wasdel, /* converting a delayed alloc */
- int *logflagsp, /* inode logging flags */
- int whichfork) /* data or attr fork */
-{
- struct xfs_btree_block *ablock; /* allocated (child) bt block */
- xfs_buf_t *abp; /* buffer for ablock */
- xfs_alloc_arg_t args; /* allocation arguments */
- xfs_bmbt_rec_t *arp; /* child record pointer */
- struct xfs_btree_block *block; /* btree root block */
- xfs_btree_cur_t *cur; /* bmap btree cursor */
- xfs_bmbt_rec_host_t *ep; /* extent record pointer */
- int error; /* error return value */
- xfs_extnum_t i, cnt; /* extent record index */
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_bmbt_key_t *kp; /* root block key pointer */
- xfs_mount_t *mp; /* mount structure */
- xfs_extnum_t nextents; /* number of file extents */
- xfs_bmbt_ptr_t *pp; /* root block address pointer */
+ *logflagsp = 0;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS);
+ cur = *curp;
+ ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
- /*
- * Make space in the inode incore.
- */
- xfs_iroot_realloc(ip, 1, whichfork);
- ifp->if_flags |= XFS_IFBROOT;
+ ASSERT(*idx >= 0);
+ ASSERT(*idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
+ ASSERT(!isnullstartblock(new->br_startblock));
+
+ XFS_STATS_INC(xs_add_exlist);
+
+#define LEFT r[0]
+#define RIGHT r[1]
+#define PREV r[2]
/*
- * Fill in the root.
+ * Set up a bunch of variables to make the tests simpler.
*/
- block = ifp->if_broot;
- block->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC);
- block->bb_level = cpu_to_be16(1);
- block->bb_numrecs = cpu_to_be16(1);
- block->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
- block->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
+ error = 0;
+ ep = xfs_iext_get_ext(ifp, *idx);
+ xfs_bmbt_get_all(ep, &PREV);
+ newext = new->br_state;
+ oldext = (newext == XFS_EXT_UNWRITTEN) ?
+ XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
+ ASSERT(PREV.br_state == oldext);
+ new_endoff = new->br_startoff + new->br_blockcount;
+ ASSERT(PREV.br_startoff <= new->br_startoff);
+ ASSERT(PREV.br_startoff + PREV.br_blockcount >= new_endoff);
/*
- * Need a cursor. Can't allocate until bb_level is filled in.
+ * Set flags determining what part of the previous oldext allocation
+ * extent is being replaced by a newext allocation.
*/
- mp = ip->i_mount;
- cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
- cur->bc_private.b.firstblock = *firstblock;
- cur->bc_private.b.flist = flist;
- cur->bc_private.b.flags = wasdel ? XFS_BTCUR_BPRV_WASDEL : 0;
+ if (PREV.br_startoff == new->br_startoff)
+ state |= BMAP_LEFT_FILLING;
+ if (PREV.br_startoff + PREV.br_blockcount == new_endoff)
+ state |= BMAP_RIGHT_FILLING;
+
/*
- * Convert to a btree with two levels, one record in root.
+ * Check and set flags if this segment has a left neighbor.
+ * Don't set contiguous if the combined extent would be too large.
*/
- XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_BTREE);
- memset(&args, 0, sizeof(args));
- args.tp = tp;
- args.mp = mp;
- args.firstblock = *firstblock;
- if (*firstblock == NULLFSBLOCK) {
- args.type = XFS_ALLOCTYPE_START_BNO;
- args.fsbno = XFS_INO_TO_FSB(mp, ip->i_ino);
- } else if (flist->xbf_low) {
- args.type = XFS_ALLOCTYPE_START_BNO;
- args.fsbno = *firstblock;
- } else {
- args.type = XFS_ALLOCTYPE_NEAR_BNO;
- args.fsbno = *firstblock;
- }
- args.minlen = args.maxlen = args.prod = 1;
- args.wasdel = wasdel;
- *logflagsp = 0;
- if ((error = xfs_alloc_vextent(&args))) {
- xfs_iroot_realloc(ip, -1, whichfork);
- xfs_btree_del_cursor(cur, XFS_BTREE_ERROR);
- return error;
+ if (*idx > 0) {
+ state |= BMAP_LEFT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &LEFT);
+
+ if (isnullstartblock(LEFT.br_startblock))
+ state |= BMAP_LEFT_DELAY;
}
+
+ if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
+ LEFT.br_startoff + LEFT.br_blockcount == new->br_startoff &&
+ LEFT.br_startblock + LEFT.br_blockcount == new->br_startblock &&
+ LEFT.br_state == newext &&
+ LEFT.br_blockcount + new->br_blockcount <= MAXEXTLEN)
+ state |= BMAP_LEFT_CONTIG;
+
/*
- * Allocation can't fail, the space was reserved.
+ * Check and set flags if this segment has a right neighbor.
+ * Don't set contiguous if the combined extent would be too large.
+ * Also check for all-three-contiguous being too large.
*/
- ASSERT(args.fsbno != NULLFSBLOCK);
- ASSERT(*firstblock == NULLFSBLOCK ||
- args.agno == XFS_FSB_TO_AGNO(mp, *firstblock) ||
- (flist->xbf_low &&
- args.agno > XFS_FSB_TO_AGNO(mp, *firstblock)));
- *firstblock = cur->bc_private.b.firstblock = args.fsbno;
- cur->bc_private.b.allocated++;
- ip->i_d.di_nblocks++;
- xfs_trans_mod_dquot_byino(tp, ip, XFS_TRANS_DQ_BCOUNT, 1L);
- abp = xfs_btree_get_bufl(mp, tp, args.fsbno, 0);
- /*
- * Fill in the child block.
- */
- abp->b_ops = &xfs_bmbt_buf_ops;
- ablock = XFS_BUF_TO_BLOCK(abp);
- ablock->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC);
- ablock->bb_level = 0;
- ablock->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
- ablock->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
- arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
- nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
- for (cnt = i = 0; i < nextents; i++) {
- ep = xfs_iext_get_ext(ifp, i);
- if (!isnullstartblock(xfs_bmbt_get_startblock(ep))) {
- arp->l0 = cpu_to_be64(ep->l0);
- arp->l1 = cpu_to_be64(ep->l1);
- arp++; cnt++;
- }
+ if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) {
+ state |= BMAP_RIGHT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT);
+ if (isnullstartblock(RIGHT.br_startblock))
+ state |= BMAP_RIGHT_DELAY;
}
- ASSERT(cnt == XFS_IFORK_NEXTENTS(ip, whichfork));
- xfs_btree_set_numrecs(ablock, cnt);
- /*
- * Fill in the root key and pointer.
- */
- kp = XFS_BMBT_KEY_ADDR(mp, block, 1);
- arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
- kp->br_startoff = cpu_to_be64(xfs_bmbt_disk_get_startoff(arp));
- pp = XFS_BMBT_PTR_ADDR(mp, block, 1, xfs_bmbt_get_maxrecs(cur,
- be16_to_cpu(block->bb_level)));
- *pp = cpu_to_be64(args.fsbno);
+ if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
+ new_endoff == RIGHT.br_startoff &&
+ new->br_startblock + new->br_blockcount == RIGHT.br_startblock &&
+ newext == RIGHT.br_state &&
+ new->br_blockcount + RIGHT.br_blockcount <= MAXEXTLEN &&
+ ((state & (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
+ BMAP_RIGHT_FILLING)) !=
+ (BMAP_LEFT_CONTIG | BMAP_LEFT_FILLING |
+ BMAP_RIGHT_FILLING) ||
+ LEFT.br_blockcount + new->br_blockcount + RIGHT.br_blockcount
+ <= MAXEXTLEN))
+ state |= BMAP_RIGHT_CONTIG;
/*
- * Do all this logging at the end so that
- * the root is at the right level.
+ * Switch out based on the FILLING and CONTIG state bits.
*/
- xfs_btree_log_block(cur, abp, XFS_BB_ALL_BITS);
- xfs_btree_log_recs(cur, abp, 1, be16_to_cpu(ablock->bb_numrecs));
- ASSERT(*curp == NULL);
- *curp = cur;
- *logflagsp = XFS_ILOG_CORE | xfs_ilog_fbroot(whichfork);
- return 0;
-}
-
-/*
- * Calculate the default attribute fork offset for newly created inodes.
- */
-uint
-xfs_default_attroffset(
- struct xfs_inode *ip)
-{
- struct xfs_mount *mp = ip->i_mount;
- uint offset;
-
- if (mp->m_sb.sb_inodesize == 256) {
- offset = XFS_LITINO(mp) -
- XFS_BMDR_SPACE_CALC(MINABTPTRS);
- } else {
- offset = XFS_BMDR_SPACE_CALC(6 * MINABTPTRS);
- }
+ switch (state & (BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
+ BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG)) {
+ case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG |
+ BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
+ /*
+ * Setting all of a previous oldext extent to newext.
+ * The left and right neighbors are both contiguous with new.
+ */
+ --*idx;
- ASSERT(offset < XFS_LITINO(mp));
- return offset;
-}
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
+ LEFT.br_blockcount + PREV.br_blockcount +
+ RIGHT.br_blockcount);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-/*
- * Helper routine to reset inode di_forkoff field when switching
- * attribute fork from local to extent format - we reset it where
- * possible to make space available for inline data fork extents.
- */
-STATIC void
-xfs_bmap_forkoff_reset(
- xfs_mount_t *mp,
- xfs_inode_t *ip,
- int whichfork)
-{
- if (whichfork == XFS_ATTR_FORK &&
- ip->i_d.di_format != XFS_DINODE_FMT_DEV &&
- ip->i_d.di_format != XFS_DINODE_FMT_UUID &&
- ip->i_d.di_format != XFS_DINODE_FMT_BTREE) {
- uint dfl_forkoff = xfs_default_attroffset(ip) >> 3;
+ xfs_iext_remove(ip, *idx + 1, 2, state);
+ ip->i_d.di_nextents -= 2;
+ if (cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
+ RIGHT.br_startblock,
+ RIGHT.br_blockcount, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_delete(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_decrement(cur, 0, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_delete(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_decrement(cur, 0, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
+ LEFT.br_startblock,
+ LEFT.br_blockcount + PREV.br_blockcount +
+ RIGHT.br_blockcount, LEFT.br_state)))
+ goto done;
+ }
+ break;
- if (dfl_forkoff > ip->i_d.di_forkoff)
- ip->i_d.di_forkoff = dfl_forkoff;
- }
-}
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
+ /*
+ * Setting all of a previous oldext extent to newext.
+ * The left neighbor is contiguous, the right is not.
+ */
+ --*idx;
-/*
- * Convert a local file to an extents file.
- * This code is out of bounds for data forks of regular files,
- * since the file data needs to get logged so things will stay consistent.
- * (The bmap-level manipulations are ok, though).
- */
-STATIC int /* error */
-xfs_bmap_local_to_extents(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fsblock_t *firstblock, /* first block allocated in xaction */
- xfs_extlen_t total, /* total blocks needed by transaction */
- int *logflagsp, /* inode logging flags */
- int whichfork,
- void (*init_fn)(struct xfs_buf *bp,
- struct xfs_inode *ip,
- struct xfs_ifork *ifp))
-{
- int error; /* error return value */
- int flags; /* logging flags returned */
- xfs_ifork_t *ifp; /* inode fork pointer */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx),
+ LEFT.br_blockcount + PREV.br_blockcount);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- /*
- * We don't want to deal with the case of keeping inode data inline yet.
- * So sending the data fork of a regular inode is invalid.
- */
- ASSERT(!(S_ISREG(ip->i_d.di_mode) && whichfork == XFS_DATA_FORK));
- ifp = XFS_IFORK_PTR(ip, whichfork);
- ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
- flags = 0;
- error = 0;
- if (ifp->if_bytes) {
- xfs_alloc_arg_t args; /* allocation arguments */
- xfs_buf_t *bp; /* buffer for extent block */
- xfs_bmbt_rec_host_t *ep;/* extent record pointer */
+ xfs_iext_remove(ip, *idx + 1, 1, state);
+ ip->i_d.di_nextents--;
+ if (cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
+ PREV.br_startblock, PREV.br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_delete(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_decrement(cur, 0, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur, LEFT.br_startoff,
+ LEFT.br_startblock,
+ LEFT.br_blockcount + PREV.br_blockcount,
+ LEFT.br_state)))
+ goto done;
+ }
+ break;
- ASSERT((ifp->if_flags &
- (XFS_IFINLINE|XFS_IFEXTENTS|XFS_IFEXTIREC)) == XFS_IFINLINE);
- memset(&args, 0, sizeof(args));
- args.tp = tp;
- args.mp = ip->i_mount;
- args.firstblock = *firstblock;
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
/*
- * Allocate a block. We know we need only one, since the
- * file currently fits in an inode.
+ * Setting all of a previous oldext extent to newext.
+ * The right neighbor is contiguous, the left is not.
*/
- if (*firstblock == NULLFSBLOCK) {
- args.fsbno = XFS_INO_TO_FSB(args.mp, ip->i_ino);
- args.type = XFS_ALLOCTYPE_START_BNO;
- } else {
- args.fsbno = *firstblock;
- args.type = XFS_ALLOCTYPE_NEAR_BNO;
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep,
+ PREV.br_blockcount + RIGHT.br_blockcount);
+ xfs_bmbt_set_state(ep, newext);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ xfs_iext_remove(ip, *idx + 1, 1, state);
+ ip->i_d.di_nextents--;
+ if (cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ if ((error = xfs_bmbt_lookup_eq(cur, RIGHT.br_startoff,
+ RIGHT.br_startblock,
+ RIGHT.br_blockcount, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_delete(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_btree_decrement(cur, 0, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur, new->br_startoff,
+ new->br_startblock,
+ new->br_blockcount + RIGHT.br_blockcount,
+ newext)))
+ goto done;
}
- args.total = total;
- args.minlen = args.maxlen = args.prod = 1;
- error = xfs_alloc_vextent(&args);
- if (error)
- goto done;
+ break;
- /* Can't fail, the space was reserved. */
- ASSERT(args.fsbno != NULLFSBLOCK);
- ASSERT(args.len == 1);
- *firstblock = args.fsbno;
- bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0);
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_FILLING:
+ /*
+ * Setting all of a previous oldext extent to newext.
+ * Neither the left nor right neighbors are contiguous with
+ * the new one.
+ */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_state(ep, newext);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- /* initialise the block and copy the data */
- init_fn(bp, ip, ifp);
+ if (cur == NULL)
+ rval = XFS_ILOG_DEXT;
+ else {
+ rval = 0;
+ if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ newext)))
+ goto done;
+ }
+ break;
- /* account for the change in fork size and log everything */
- xfs_trans_log_buf(tp, bp, 0, ifp->if_bytes - 1);
- xfs_bmap_forkoff_reset(args.mp, ip, whichfork);
- xfs_idata_realloc(ip, -ifp->if_bytes, whichfork);
- xfs_iext_add(ifp, 0, 1);
- ep = xfs_iext_get_ext(ifp, 0);
- xfs_bmbt_set_allf(ep, 0, args.fsbno, 1, XFS_EXT_NORM);
- trace_xfs_bmap_post_update(ip, 0,
- whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0,
- _THIS_IP_);
- XFS_IFORK_NEXT_SET(ip, whichfork, 1);
- ip->i_d.di_nblocks = 1;
- xfs_trans_mod_dquot_byino(tp, ip,
- XFS_TRANS_DQ_BCOUNT, 1L);
- flags |= xfs_ilog_fext(whichfork);
- } else {
- ASSERT(XFS_IFORK_NEXTENTS(ip, whichfork) == 0);
- xfs_bmap_forkoff_reset(ip->i_mount, ip, whichfork);
- }
- ifp->if_flags &= ~XFS_IFINLINE;
- ifp->if_flags |= XFS_IFEXTENTS;
- XFS_IFORK_FMT_SET(ip, whichfork, XFS_DINODE_FMT_EXTENTS);
- flags |= XFS_ILOG_CORE;
-done:
- *logflagsp = flags;
- return error;
-}
+ case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG:
+ /*
+ * Setting the first part of a previous oldext extent to newext.
+ * The left neighbor is contiguous.
+ */
+ trace_xfs_bmap_pre_update(ip, *idx - 1, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx - 1),
+ LEFT.br_blockcount + new->br_blockcount);
+ xfs_bmbt_set_startoff(ep,
+ PREV.br_startoff + new->br_blockcount);
+ trace_xfs_bmap_post_update(ip, *idx - 1, state, _THIS_IP_);
-/*
- * Search the extent records for the entry containing block bno.
- * If bno lies in a hole, point to the next entry. If bno lies
- * past eof, *eofp will be set, and *prevp will contain the last
- * entry (null if none). Else, *lastxp will be set to the index
- * of the found entry; *gotp will contain the entry.
- */
-STATIC xfs_bmbt_rec_host_t * /* pointer to found extent entry */
-xfs_bmap_search_multi_extents(
- xfs_ifork_t *ifp, /* inode fork pointer */
- xfs_fileoff_t bno, /* block number searched for */
- int *eofp, /* out: end of file found */
- xfs_extnum_t *lastxp, /* out: last extent index */
- xfs_bmbt_irec_t *gotp, /* out: extent entry found */
- xfs_bmbt_irec_t *prevp) /* out: previous extent entry found */
-{
- xfs_bmbt_rec_host_t *ep; /* extent record pointer */
- xfs_extnum_t lastx; /* last extent index */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_startblock(ep,
+ new->br_startblock + new->br_blockcount);
+ xfs_bmbt_set_blockcount(ep,
+ PREV.br_blockcount - new->br_blockcount);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- /*
- * Initialize the extent entry structure to catch access to
- * uninitialized br_startblock field.
- */
- gotp->br_startoff = 0xffa5a5a5a5a5a5a5LL;
- gotp->br_blockcount = 0xa55a5a5a5a5a5a5aLL;
- gotp->br_state = XFS_EXT_INVALID;
-#if XFS_BIG_BLKNOS
- gotp->br_startblock = 0xffffa5a5a5a5a5a5LL;
-#else
- gotp->br_startblock = 0xffffa5a5;
-#endif
- prevp->br_startoff = NULLFILEOFF;
+ --*idx;
- ep = xfs_iext_bno_to_ext(ifp, bno, &lastx);
- if (lastx > 0) {
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx - 1), prevp);
- }
- if (lastx < (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))) {
- xfs_bmbt_get_all(ep, gotp);
- *eofp = 0;
- } else {
- if (lastx > 0) {
- *gotp = *prevp;
+ if (cur == NULL)
+ rval = XFS_ILOG_DEXT;
+ else {
+ rval = 0;
+ if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
+ PREV.br_startblock, PREV.br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur,
+ PREV.br_startoff + new->br_blockcount,
+ PREV.br_startblock + new->br_blockcount,
+ PREV.br_blockcount - new->br_blockcount,
+ oldext)))
+ goto done;
+ if ((error = xfs_btree_decrement(cur, 0, &i)))
+ goto done;
+ error = xfs_bmbt_update(cur, LEFT.br_startoff,
+ LEFT.br_startblock,
+ LEFT.br_blockcount + new->br_blockcount,
+ LEFT.br_state);
+ if (error)
+ goto done;
}
- *eofp = 1;
- ep = NULL;
- }
- *lastxp = lastx;
- return ep;
-}
+ break;
-/*
- * Search the extents list for the inode, for the extent containing bno.
- * If bno lies in a hole, point to the next entry. If bno lies past eof,
- * *eofp will be set, and *prevp will contain the last entry (null if none).
- * Else, *lastxp will be set to the index of the found
- * entry; *gotp will contain the entry.
- */
-STATIC xfs_bmbt_rec_host_t * /* pointer to found extent entry */
-xfs_bmap_search_extents(
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_fileoff_t bno, /* block number searched for */
- int fork, /* data or attr fork */
- int *eofp, /* out: end of file found */
- xfs_extnum_t *lastxp, /* out: last extent index */
- xfs_bmbt_irec_t *gotp, /* out: extent entry found */
- xfs_bmbt_irec_t *prevp) /* out: previous extent entry found */
-{
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_bmbt_rec_host_t *ep; /* extent record pointer */
+ case BMAP_LEFT_FILLING:
+ /*
+ * Setting the first part of a previous oldext extent to newext.
+ * The left neighbor is not contiguous.
+ */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ ASSERT(ep && xfs_bmbt_get_state(ep) == oldext);
+ xfs_bmbt_set_startoff(ep, new_endoff);
+ xfs_bmbt_set_blockcount(ep,
+ PREV.br_blockcount - new->br_blockcount);
+ xfs_bmbt_set_startblock(ep,
+ new->br_startblock + new->br_blockcount);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- XFS_STATS_INC(xs_look_exlist);
- ifp = XFS_IFORK_PTR(ip, fork);
+ xfs_iext_insert(ip, *idx, 1, new, state);
+ ip->i_d.di_nextents++;
+ if (cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
+ PREV.br_startblock, PREV.br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur,
+ PREV.br_startoff + new->br_blockcount,
+ PREV.br_startblock + new->br_blockcount,
+ PREV.br_blockcount - new->br_blockcount,
+ oldext)))
+ goto done;
+ cur->bc_rec.b = *new;
+ if ((error = xfs_btree_insert(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ }
+ break;
- ep = xfs_bmap_search_multi_extents(ifp, bno, eofp, lastxp, gotp, prevp);
+ case BMAP_RIGHT_FILLING | BMAP_RIGHT_CONTIG:
+ /*
+ * Setting the last part of a previous oldext extent to newext.
+ * The right neighbor is contiguous with the new allocation.
+ */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep,
+ PREV.br_blockcount - new->br_blockcount);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- if (unlikely(!(gotp->br_startblock) && (*lastxp != NULLEXTNUM) &&
- !(XFS_IS_REALTIME_INODE(ip) && fork == XFS_DATA_FORK))) {
- xfs_alert_tag(ip->i_mount, XFS_PTAG_FSBLOCK_ZERO,
- "Access to block zero in inode %llu "
- "start_block: %llx start_off: %llx "
- "blkcnt: %llx extent-state: %x lastx: %x\n",
- (unsigned long long)ip->i_ino,
- (unsigned long long)gotp->br_startblock,
- (unsigned long long)gotp->br_startoff,
- (unsigned long long)gotp->br_blockcount,
- gotp->br_state, *lastxp);
- *lastxp = NULLEXTNUM;
- *eofp = 1;
- return NULL;
- }
- return ep;
-}
+ ++*idx;
-/*
- * Compute the worst-case number of indirect blocks that will be used
- * for ip's delayed extent of length "len".
- */
-STATIC xfs_filblks_t
-xfs_bmap_worst_indlen(
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_filblks_t len) /* delayed extent length */
-{
- int level; /* btree level number */
- int maxrecs; /* maximum record count at this level */
- xfs_mount_t *mp; /* mount structure */
- xfs_filblks_t rval; /* return value */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
+ new->br_startoff, new->br_startblock,
+ new->br_blockcount + RIGHT.br_blockcount, newext);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- mp = ip->i_mount;
- maxrecs = mp->m_bmap_dmxr[0];
- for (level = 0, rval = 0;
- level < XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK);
- level++) {
- len += maxrecs - 1;
- do_div(len, maxrecs);
- rval += len;
- if (len == 1)
- return rval + XFS_BM_MAXLEVELS(mp, XFS_DATA_FORK) -
- level - 1;
- if (level == 0)
- maxrecs = mp->m_bmap_dmxr[1];
- }
- return rval;
-}
+ if (cur == NULL)
+ rval = XFS_ILOG_DEXT;
+ else {
+ rval = 0;
+ if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
+ PREV.br_startblock,
+ PREV.br_blockcount, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
+ PREV.br_startblock,
+ PREV.br_blockcount - new->br_blockcount,
+ oldext)))
+ goto done;
+ if ((error = xfs_btree_increment(cur, 0, &i)))
+ goto done;
+ if ((error = xfs_bmbt_update(cur, new->br_startoff,
+ new->br_startblock,
+ new->br_blockcount + RIGHT.br_blockcount,
+ newext)))
+ goto done;
+ }
+ break;
-/*
- * Convert inode from non-attributed to attributed.
- * Must not be in a transaction, ip must not be locked.
- */
-int /* error code */
-xfs_bmap_add_attrfork(
- xfs_inode_t *ip, /* incore inode pointer */
- int size, /* space new attribute needs */
- int rsvd) /* xact may use reserved blks */
-{
- xfs_fsblock_t firstblock; /* 1st block/ag allocated */
- xfs_bmap_free_t flist; /* freed extent records */
- xfs_mount_t *mp; /* mount structure */
- xfs_trans_t *tp; /* transaction pointer */
- int blks; /* space reservation */
- int version = 1; /* superblock attr version */
- int committed; /* xaction was committed */
- int logflags; /* logging flags */
- int error; /* error return value */
+ case BMAP_RIGHT_FILLING:
+ /*
+ * Setting the last part of a previous oldext extent to newext.
+ * The right neighbor is not contiguous.
+ */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep,
+ PREV.br_blockcount - new->br_blockcount);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- ASSERT(XFS_IFORK_Q(ip) == 0);
+ ++*idx;
+ xfs_iext_insert(ip, *idx, 1, new, state);
- mp = ip->i_mount;
- ASSERT(!XFS_NOT_DQATTACHED(mp, ip));
- tp = xfs_trans_alloc(mp, XFS_TRANS_ADDAFORK);
- blks = XFS_ADDAFORK_SPACE_RES(mp);
- if (rsvd)
- tp->t_flags |= XFS_TRANS_RESERVE;
- if ((error = xfs_trans_reserve(tp, blks, XFS_ADDAFORK_LOG_RES(mp), 0,
- XFS_TRANS_PERM_LOG_RES, XFS_ADDAFORK_LOG_COUNT)))
- goto error0;
- xfs_ilock(ip, XFS_ILOCK_EXCL);
- error = xfs_trans_reserve_quota_nblks(tp, ip, blks, 0, rsvd ?
- XFS_QMOPT_RES_REGBLKS | XFS_QMOPT_FORCE_RES :
- XFS_QMOPT_RES_REGBLKS);
- if (error) {
- xfs_iunlock(ip, XFS_ILOCK_EXCL);
- xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES);
- return error;
- }
- if (XFS_IFORK_Q(ip))
- goto error1;
- if (ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS) {
+ ip->i_d.di_nextents++;
+ if (cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
+ PREV.br_startblock, PREV.br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ if ((error = xfs_bmbt_update(cur, PREV.br_startoff,
+ PREV.br_startblock,
+ PREV.br_blockcount - new->br_blockcount,
+ oldext)))
+ goto done;
+ if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 0, done);
+ cur->bc_rec.b.br_state = XFS_EXT_NORM;
+ if ((error = xfs_btree_insert(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ }
+ break;
+
+ case 0:
/*
- * For inodes coming from pre-6.2 filesystems.
+ * Setting the middle part of a previous oldext extent to
+ * newext. Contiguity is impossible here.
+ * One extent becomes three extents.
*/
- ASSERT(ip->i_d.di_aformat == 0);
- ip->i_d.di_aformat = XFS_DINODE_FMT_EXTENTS;
- }
- ASSERT(ip->i_d.di_anextents == 0);
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep,
+ new->br_startoff - PREV.br_startoff);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
- xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
- xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+ r[0] = *new;
+ r[1].br_startoff = new_endoff;
+ r[1].br_blockcount =
+ PREV.br_startoff + PREV.br_blockcount - new_endoff;
+ r[1].br_startblock = new->br_startblock + new->br_blockcount;
+ r[1].br_state = oldext;
- switch (ip->i_d.di_format) {
- case XFS_DINODE_FMT_DEV:
- ip->i_d.di_forkoff = roundup(sizeof(xfs_dev_t), 8) >> 3;
- break;
- case XFS_DINODE_FMT_UUID:
- ip->i_d.di_forkoff = roundup(sizeof(uuid_t), 8) >> 3;
- break;
- case XFS_DINODE_FMT_LOCAL:
- case XFS_DINODE_FMT_EXTENTS:
- case XFS_DINODE_FMT_BTREE:
- ip->i_d.di_forkoff = xfs_attr_shortform_bytesfit(ip, size);
- if (!ip->i_d.di_forkoff)
- ip->i_d.di_forkoff = xfs_default_attroffset(ip) >> 3;
- else if (mp->m_flags & XFS_MOUNT_ATTR2)
- version = 2;
+ ++*idx;
+ xfs_iext_insert(ip, *idx, 2, &r[0], state);
+
+ ip->i_d.di_nextents += 2;
+ if (cur == NULL)
+ rval = XFS_ILOG_CORE | XFS_ILOG_DEXT;
+ else {
+ rval = XFS_ILOG_CORE;
+ if ((error = xfs_bmbt_lookup_eq(cur, PREV.br_startoff,
+ PREV.br_startblock, PREV.br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ /* new right extent - oldext */
+ if ((error = xfs_bmbt_update(cur, r[1].br_startoff,
+ r[1].br_startblock, r[1].br_blockcount,
+ r[1].br_state)))
+ goto done;
+ /* new left extent - oldext */
+ cur->bc_rec.b = PREV;
+ cur->bc_rec.b.br_blockcount =
+ new->br_startoff - PREV.br_startoff;
+ if ((error = xfs_btree_insert(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ /*
+ * Reset the cursor to the position of the new extent
+ * we are about to insert as we can't trust it after
+ * the previous insert.
+ */
+ if ((error = xfs_bmbt_lookup_eq(cur, new->br_startoff,
+ new->br_startblock, new->br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 0, done);
+ /* new middle extent - newext */
+ cur->bc_rec.b.br_state = new->br_state;
+ if ((error = xfs_btree_insert(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ }
break;
- default:
+
+ case BMAP_LEFT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ case BMAP_LEFT_FILLING | BMAP_RIGHT_CONTIG:
+ case BMAP_RIGHT_FILLING | BMAP_LEFT_CONTIG:
+ case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ case BMAP_LEFT_CONTIG:
+ case BMAP_RIGHT_CONTIG:
+ /*
+ * These cases are all impossible.
+ */
ASSERT(0);
- error = XFS_ERROR(EINVAL);
- goto error1;
}
- ASSERT(ip->i_afp == NULL);
- ip->i_afp = kmem_zone_zalloc(xfs_ifork_zone, KM_SLEEP);
- ip->i_afp->if_flags = XFS_IFEXTENTS;
- logflags = 0;
- xfs_bmap_init(&flist, &firstblock);
- switch (ip->i_d.di_format) {
- case XFS_DINODE_FMT_LOCAL:
- error = xfs_bmap_add_attrfork_local(tp, ip, &firstblock, &flist,
- &logflags);
- break;
- case XFS_DINODE_FMT_EXTENTS:
- error = xfs_bmap_add_attrfork_extents(tp, ip, &firstblock,
- &flist, &logflags);
- break;
- case XFS_DINODE_FMT_BTREE:
- error = xfs_bmap_add_attrfork_btree(tp, ip, &firstblock, &flist,
- &logflags);
- break;
- default:
- error = 0;
- break;
+ /* convert to a btree if necessary */
+ if (xfs_bmap_needs_btree(ip, XFS_DATA_FORK)) {
+ int tmp_logflags; /* partial log flag return val */
+
+ ASSERT(cur == NULL);
+ error = xfs_bmap_extents_to_btree(tp, ip, first, flist, &cur,
+ 0, &tmp_logflags, XFS_DATA_FORK);
+ *logflagsp |= tmp_logflags;
+ if (error)
+ goto done;
}
- if (logflags)
- xfs_trans_log_inode(tp, ip, logflags);
- if (error)
- goto error2;
- if (!xfs_sb_version_hasattr(&mp->m_sb) ||
- (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2)) {
- __int64_t sbfields = 0;
- spin_lock(&mp->m_sb_lock);
- if (!xfs_sb_version_hasattr(&mp->m_sb)) {
- xfs_sb_version_addattr(&mp->m_sb);
- sbfields |= XFS_SB_VERSIONNUM;
- }
- if (!xfs_sb_version_hasattr2(&mp->m_sb) && version == 2) {
- xfs_sb_version_addattr2(&mp->m_sb);
- sbfields |= (XFS_SB_VERSIONNUM | XFS_SB_FEATURES2);
- }
- if (sbfields) {
- spin_unlock(&mp->m_sb_lock);
- xfs_mod_sb(tp, sbfields);
- } else
- spin_unlock(&mp->m_sb_lock);
+ /* clear out the allocated field, done with it now in any case. */
+ if (cur) {
+ cur->bc_private.b.allocated = 0;
+ *curp = cur;
}
- error = xfs_bmap_finish(&tp, &flist, &committed);
- if (error)
- goto error2;
- return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
-error2:
- xfs_bmap_cancel(&flist);
-error1:
- xfs_iunlock(ip, XFS_ILOCK_EXCL);
-error0:
- xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT);
+ xfs_bmap_check_leaf_extents(*curp, ip, XFS_DATA_FORK);
+done:
+ *logflagsp |= rval;
return error;
+#undef LEFT
+#undef RIGHT
+#undef PREV
}
/*
- * Add the extent to the list of extents to be free at transaction end.
- * The list is maintained sorted (by block number).
+ * Convert a hole to a delayed allocation.
*/
-/* ARGSUSED */
-void
-xfs_bmap_add_free(
- xfs_fsblock_t bno, /* fs block number of extent */
- xfs_filblks_t len, /* length of extent */
- xfs_bmap_free_t *flist, /* list of extents */
- xfs_mount_t *mp) /* mount point structure */
+STATIC void
+xfs_bmap_add_extent_hole_delay(
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_extnum_t *idx, /* extent number to update/insert */
+ xfs_bmbt_irec_t *new) /* new data to add to file extents */
{
- xfs_bmap_free_item_t *cur; /* current (next) element */
- xfs_bmap_free_item_t *new; /* new element */
- xfs_bmap_free_item_t *prev; /* previous element */
-#ifdef DEBUG
- xfs_agnumber_t agno;
- xfs_agblock_t agbno;
-
- ASSERT(bno != NULLFSBLOCK);
- ASSERT(len > 0);
- ASSERT(len <= MAXEXTLEN);
- ASSERT(!isnullstartblock(bno));
- agno = XFS_FSB_TO_AGNO(mp, bno);
- agbno = XFS_FSB_TO_AGBNO(mp, bno);
- ASSERT(agno < mp->m_sb.sb_agcount);
- ASSERT(agbno < mp->m_sb.sb_agblocks);
- ASSERT(len < mp->m_sb.sb_agblocks);
- ASSERT(agbno + len <= mp->m_sb.sb_agblocks);
-#endif
- ASSERT(xfs_bmap_free_item_zone != NULL);
- new = kmem_zone_alloc(xfs_bmap_free_item_zone, KM_SLEEP);
- new->xbfi_startblock = bno;
- new->xbfi_blockcount = (xfs_extlen_t)len;
- for (prev = NULL, cur = flist->xbf_first;
- cur != NULL;
- prev = cur, cur = cur->xbfi_next) {
- if (cur->xbfi_startblock >= bno)
- break;
- }
- if (prev)
- prev->xbfi_next = new;
- else
- flist->xbf_first = new;
- new->xbfi_next = cur;
- flist->xbf_count++;
-}
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_bmbt_irec_t left; /* left neighbor extent entry */
+ xfs_filblks_t newlen=0; /* new indirect size */
+ xfs_filblks_t oldlen=0; /* old indirect size */
+ xfs_bmbt_irec_t right; /* right neighbor extent entry */
+ int state; /* state bits, accessed thru macros */
+ xfs_filblks_t temp=0; /* temp for indirect calculations */
-/*
- * Compute and fill in the value of the maximum depth of a bmap btree
- * in this filesystem. Done once, during mount.
- */
-void
-xfs_bmap_compute_maxlevels(
- xfs_mount_t *mp, /* file system mount structure */
- int whichfork) /* data or attr fork */
-{
- int level; /* btree level */
- uint maxblocks; /* max blocks at this level */
- uint maxleafents; /* max leaf entries possible */
- int maxrootrecs; /* max records in root block */
- int minleafrecs; /* min records in leaf block */
- int minnoderecs; /* min records in node block */
- int sz; /* root block size */
+ ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+ state = 0;
+ ASSERT(isnullstartblock(new->br_startblock));
/*
- * The maximum number of extents in a file, hence the maximum
- * number of leaf entries, is controlled by the type of di_nextents
- * (a signed 32-bit number, xfs_extnum_t), or by di_anextents
- * (a signed 16-bit number, xfs_aextnum_t).
- *
- * Note that we can no longer assume that if we are in ATTR1 that
- * the fork offset of all the inodes will be
- * (xfs_default_attroffset(ip) >> 3) because we could have mounted
- * with ATTR2 and then mounted back with ATTR1, keeping the
- * di_forkoff's fixed but probably at various positions. Therefore,
- * for both ATTR1 and ATTR2 we have to assume the worst case scenario
- * of a minimum size available.
+ * Check and set flags if this segment has a left neighbor
*/
- if (whichfork == XFS_DATA_FORK) {
- maxleafents = MAXEXTNUM;
- sz = XFS_BMDR_SPACE_CALC(MINDBTPTRS);
- } else {
- maxleafents = MAXAEXTNUM;
- sz = XFS_BMDR_SPACE_CALC(MINABTPTRS);
- }
- maxrootrecs = xfs_bmdr_maxrecs(mp, sz, 0);
- minleafrecs = mp->m_bmap_dmnr[0];
- minnoderecs = mp->m_bmap_dmnr[1];
- maxblocks = (maxleafents + minleafrecs - 1) / minleafrecs;
- for (level = 1; maxblocks > 1; level++) {
- if (maxblocks <= maxrootrecs)
- maxblocks = 1;
- else
- maxblocks = (maxblocks + minnoderecs - 1) / minnoderecs;
+ if (*idx > 0) {
+ state |= BMAP_LEFT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left);
+
+ if (isnullstartblock(left.br_startblock))
+ state |= BMAP_LEFT_DELAY;
}
- mp->m_bm_maxlevels[whichfork] = level;
-}
-/*
- * Routine to be called at transaction's end by xfs_bmapi, xfs_bunmapi
- * caller. Frees all the extents that need freeing, which must be done
- * last due to locking considerations. We never free any extents in
- * the first transaction.
- *
- * Return 1 if the given transaction was committed and a new one
- * started, and 0 otherwise in the committed parameter.
- */
-int /* error */
-xfs_bmap_finish(
- xfs_trans_t **tp, /* transaction pointer addr */
- xfs_bmap_free_t *flist, /* i/o: list extents to free */
- int *committed) /* xact committed or not */
-{
- xfs_efd_log_item_t *efd; /* extent free data */
- xfs_efi_log_item_t *efi; /* extent free intention */
- int error; /* error return value */
- xfs_bmap_free_item_t *free; /* free extent item */
- unsigned int logres; /* new log reservation */
- unsigned int logcount; /* new log count */
- xfs_mount_t *mp; /* filesystem mount structure */
- xfs_bmap_free_item_t *next; /* next item on free list */
- xfs_trans_t *ntp; /* new transaction pointer */
+ /*
+ * Check and set flags if the current (right) segment exists.
+ * If it doesn't exist, we're converting the hole at end-of-file.
+ */
+ if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
+ state |= BMAP_RIGHT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right);
- ASSERT((*tp)->t_flags & XFS_TRANS_PERM_LOG_RES);
- if (flist->xbf_count == 0) {
- *committed = 0;
- return 0;
+ if (isnullstartblock(right.br_startblock))
+ state |= BMAP_RIGHT_DELAY;
}
- ntp = *tp;
- efi = xfs_trans_get_efi(ntp, flist->xbf_count);
- for (free = flist->xbf_first; free; free = free->xbfi_next)
- xfs_trans_log_efi_extent(ntp, efi, free->xbfi_startblock,
- free->xbfi_blockcount);
- logres = ntp->t_log_res;
- logcount = ntp->t_log_count;
- ntp = xfs_trans_dup(*tp);
- error = xfs_trans_commit(*tp, 0);
- *tp = ntp;
- *committed = 1;
+
/*
- * We have a new transaction, so we should return committed=1,
- * even though we're returning an error.
+ * Set contiguity flags on the left and right neighbors.
+ * Don't let extents get too large, even if the pieces are contiguous.
*/
- if (error)
- return error;
+ if ((state & BMAP_LEFT_VALID) && (state & BMAP_LEFT_DELAY) &&
+ left.br_startoff + left.br_blockcount == new->br_startoff &&
+ left.br_blockcount + new->br_blockcount <= MAXEXTLEN)
+ state |= BMAP_LEFT_CONTIG;
+
+ if ((state & BMAP_RIGHT_VALID) && (state & BMAP_RIGHT_DELAY) &&
+ new->br_startoff + new->br_blockcount == right.br_startoff &&
+ new->br_blockcount + right.br_blockcount <= MAXEXTLEN &&
+ (!(state & BMAP_LEFT_CONTIG) ||
+ (left.br_blockcount + new->br_blockcount +
+ right.br_blockcount <= MAXEXTLEN)))
+ state |= BMAP_RIGHT_CONTIG;
/*
- * transaction commit worked ok so we can drop the extra ticket
- * reference that we gained in xfs_trans_dup()
+ * Switch out based on the contiguity flags.
*/
- xfs_log_ticket_put(ntp->t_ticket);
+ switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
+ case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ /*
+ * New allocation is contiguous with delayed allocations
+ * on the left and on the right.
+ * Merge all three into a single extent record.
+ */
+ --*idx;
+ temp = left.br_blockcount + new->br_blockcount +
+ right.br_blockcount;
- if ((error = xfs_trans_reserve(ntp, 0, logres, 0, XFS_TRANS_PERM_LOG_RES,
- logcount)))
- return error;
- efd = xfs_trans_get_efd(ntp, efi, flist->xbf_count);
- for (free = flist->xbf_first; free != NULL; free = next) {
- next = free->xbfi_next;
- if ((error = xfs_free_extent(ntp, free->xbfi_startblock,
- free->xbfi_blockcount))) {
- /*
- * The bmap free list will be cleaned up at a
- * higher level. The EFI will be canceled when
- * this transaction is aborted.
- * Need to force shutdown here to make sure it
- * happens, since this transaction may not be
- * dirty yet.
- */
- mp = ntp->t_mountp;
- if (!XFS_FORCED_SHUTDOWN(mp))
- xfs_force_shutdown(mp,
- (error == EFSCORRUPTED) ?
- SHUTDOWN_CORRUPT_INCORE :
- SHUTDOWN_META_IO_ERROR);
- return error;
- }
- xfs_trans_log_efd_extent(ntp, efd, free->xbfi_startblock,
- free->xbfi_blockcount);
- xfs_bmap_del_free(flist, NULL, free);
- }
- return 0;
-}
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
+ oldlen = startblockval(left.br_startblock) +
+ startblockval(new->br_startblock) +
+ startblockval(right.br_startblock);
+ newlen = xfs_bmap_worst_indlen(ip, temp);
+ xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
+ nullstartblock((int)newlen));
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
-/*
- * Free up any items left in the list.
- */
-void
-xfs_bmap_cancel(
- xfs_bmap_free_t *flist) /* list of bmap_free_items */
-{
- xfs_bmap_free_item_t *free; /* free list item */
- xfs_bmap_free_item_t *next;
+ xfs_iext_remove(ip, *idx + 1, 1, state);
+ break;
- if (flist->xbf_count == 0)
- return;
- ASSERT(flist->xbf_first != NULL);
- for (free = flist->xbf_first; free; free = next) {
- next = free->xbfi_next;
- xfs_bmap_del_free(flist, NULL, free);
- }
- ASSERT(flist->xbf_count == 0);
-}
+ case BMAP_LEFT_CONTIG:
+ /*
+ * New allocation is contiguous with a delayed allocation
+ * on the left.
+ * Merge the new allocation with the left neighbor.
+ */
+ --*idx;
+ temp = left.br_blockcount + new->br_blockcount;
-/*
- * Returns the file-relative block number of the first unused block(s)
- * in the file with at least "len" logically contiguous blocks free.
- * This is the lowest-address hole if the file has holes, else the first block
- * past the end of file.
- * Return 0 if the file is currently local (in-inode).
- */
-int /* error */
-xfs_bmap_first_unused(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode */
- xfs_extlen_t len, /* size of hole to find */
- xfs_fileoff_t *first_unused, /* unused block */
- int whichfork) /* data or attr fork */
-{
- int error; /* error return value */
- int idx; /* extent record index */
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_fileoff_t lastaddr; /* last block number seen */
- xfs_fileoff_t lowest; /* lowest useful block */
- xfs_fileoff_t max; /* starting useful block */
- xfs_fileoff_t off; /* offset for this block */
- xfs_extnum_t nextents; /* number of extent entries */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp);
+ oldlen = startblockval(left.br_startblock) +
+ startblockval(new->br_startblock);
+ newlen = xfs_bmap_worst_indlen(ip, temp);
+ xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx),
+ nullstartblock((int)newlen));
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ break;
- ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE ||
- XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS ||
- XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL);
- if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
- *first_unused = 0;
- return 0;
+ case BMAP_RIGHT_CONTIG:
+ /*
+ * New allocation is contiguous with a delayed allocation
+ * on the right.
+ * Merge the new allocation with the right neighbor.
+ */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ temp = new->br_blockcount + right.br_blockcount;
+ oldlen = startblockval(new->br_startblock) +
+ startblockval(right.br_startblock);
+ newlen = xfs_bmap_worst_indlen(ip, temp);
+ xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx),
+ new->br_startoff,
+ nullstartblock((int)newlen), temp, right.br_state);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ break;
+
+ case 0:
+ /*
+ * New allocation is not contiguous with another
+ * delayed allocation.
+ * Insert a new entry.
+ */
+ oldlen = newlen = 0;
+ xfs_iext_insert(ip, *idx, 1, new, state);
+ break;
}
- ifp = XFS_IFORK_PTR(ip, whichfork);
- if (!(ifp->if_flags & XFS_IFEXTENTS) &&
- (error = xfs_iread_extents(tp, ip, whichfork)))
- return error;
- lowest = *first_unused;
- nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
- for (idx = 0, lastaddr = 0, max = lowest; idx < nextents; idx++) {
- xfs_bmbt_rec_host_t *ep = xfs_iext_get_ext(ifp, idx);
- off = xfs_bmbt_get_startoff(ep);
+ if (oldlen != newlen) {
+ ASSERT(oldlen > newlen);
+ xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS,
+ (int64_t)(oldlen - newlen), 0);
/*
- * See if the hole before this extent will work.
+ * Nothing to do for disk quota accounting here.
*/
- if (off >= lowest + len && off - max >= len) {
- *first_unused = max;
- return 0;
- }
- lastaddr = off + xfs_bmbt_get_blockcount(ep);
- max = XFS_FILEOFF_MAX(lastaddr, lowest);
}
- *first_unused = max;
- return 0;
}
/*
- * Returns the file-relative block number of the last block + 1 before
- * last_block (input value) in the file.
- * This is not based on i_size, it is based on the extent records.
- * Returns 0 for local files, as they do not have extent records.
+ * Convert a hole to a real allocation.
*/
-int /* error */
-xfs_bmap_last_before(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode */
- xfs_fileoff_t *last_block, /* last block */
- int whichfork) /* data or attr fork */
-{
- xfs_fileoff_t bno; /* input file offset */
- int eof; /* hit end of file */
- xfs_bmbt_rec_host_t *ep; /* pointer to last extent */
- int error; /* error return value */
- xfs_bmbt_irec_t got; /* current extent value */
- xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_extnum_t lastx; /* last extent used */
- xfs_bmbt_irec_t prev; /* previous extent value */
-
- if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL)
- return XFS_ERROR(EIO);
- if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
- *last_block = 0;
- return 0;
- }
- ifp = XFS_IFORK_PTR(ip, whichfork);
- if (!(ifp->if_flags & XFS_IFEXTENTS) &&
- (error = xfs_iread_extents(tp, ip, whichfork)))
- return error;
- bno = *last_block - 1;
- ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
- &prev);
- if (eof || xfs_bmbt_get_startoff(ep) > bno) {
- if (prev.br_startoff == NULLFILEOFF)
- *last_block = 0;
- else
- *last_block = prev.br_startoff + prev.br_blockcount;
- }
- /*
- * Otherwise *last_block is already the right answer.
- */
- return 0;
-}
-
-STATIC int
-xfs_bmap_last_extent(
- struct xfs_trans *tp,
- struct xfs_inode *ip,
- int whichfork,
- struct xfs_bmbt_irec *rec,
- int *is_empty)
+STATIC int /* error */
+xfs_bmap_add_extent_hole_real(
+ struct xfs_bmalloca *bma,
+ int whichfork)
{
- struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, whichfork);
- int error;
- int nextents;
+ struct xfs_bmbt_irec *new = &bma->got;
+ int error; /* error return value */
+ int i; /* temp state */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_bmbt_irec_t left; /* left neighbor extent entry */
+ xfs_bmbt_irec_t right; /* right neighbor extent entry */
+ int rval=0; /* return value (logging flags) */
+ int state; /* state bits, accessed thru macros */
- if (!(ifp->if_flags & XFS_IFEXTENTS)) {
- error = xfs_iread_extents(tp, ip, whichfork);
- if (error)
- return error;
- }
+ ifp = XFS_IFORK_PTR(bma->ip, whichfork);
- nextents = ifp->if_bytes / sizeof(xfs_bmbt_rec_t);
- if (nextents == 0) {
- *is_empty = 1;
- return 0;
- }
+ ASSERT(bma->idx >= 0);
+ ASSERT(bma->idx <= ifp->if_bytes / sizeof(struct xfs_bmbt_rec));
+ ASSERT(!isnullstartblock(new->br_startblock));
+ ASSERT(!bma->cur ||
+ !(bma->cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL));
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, nextents - 1), rec);
- *is_empty = 0;
- return 0;
-}
+ XFS_STATS_INC(xs_add_exlist);
-/*
- * Check the last inode extent to determine whether this allocation will result
- * in blocks being allocated at the end of the file. When we allocate new data
- * blocks at the end of the file which do not start at the previous data block,
- * we will try to align the new blocks at stripe unit boundaries.
- *
- * Returns 0 in bma->aeof if the file (fork) is empty as any new write will be
- * at, or past the EOF.
- */
-STATIC int
-xfs_bmap_isaeof(
- struct xfs_bmalloca *bma,
- int whichfork)
-{
- struct xfs_bmbt_irec rec;
- int is_empty;
- int error;
+ state = 0;
+ if (whichfork == XFS_ATTR_FORK)
+ state |= BMAP_ATTRFORK;
- bma->aeof = 0;
- error = xfs_bmap_last_extent(NULL, bma->ip, whichfork, &rec,
- &is_empty);
- if (error || is_empty)
- return error;
+ /*
+ * Check and set flags if this segment has a left neighbor.
+ */
+ if (bma->idx > 0) {
+ state |= BMAP_LEFT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1), &left);
+ if (isnullstartblock(left.br_startblock))
+ state |= BMAP_LEFT_DELAY;
+ }
/*
- * Check if we are allocation or past the last extent, or at least into
- * the last delayed allocated extent.
+ * Check and set flags if this segment has a current value.
+ * Not true if we're inserting into the "hole" at eof.
*/
- bma->aeof = bma->offset >= rec.br_startoff + rec.br_blockcount ||
- (bma->offset >= rec.br_startoff &&
- isnullstartblock(rec.br_startblock));
- return 0;
-}
+ if (bma->idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) {
+ state |= BMAP_RIGHT_VALID;
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &right);
+ if (isnullstartblock(right.br_startblock))
+ state |= BMAP_RIGHT_DELAY;
+ }
-/*
- * Check if the endoff is outside the last extent. If so the caller will grow
- * the allocation to a stripe unit boundary. All offsets are considered outside
- * the end of file for an empty fork, so 1 is returned in *eof in that case.
- */
-int
-xfs_bmap_eof(
- struct xfs_inode *ip,
- xfs_fileoff_t endoff,
- int whichfork,
- int *eof)
-{
- struct xfs_bmbt_irec rec;
- int error;
+ /*
+ * We're inserting a real allocation between "left" and "right".
+ * Set the contiguity flags. Don't let extents get too large.
+ */
+ if ((state & BMAP_LEFT_VALID) && !(state & BMAP_LEFT_DELAY) &&
+ left.br_startoff + left.br_blockcount == new->br_startoff &&
+ left.br_startblock + left.br_blockcount == new->br_startblock &&
+ left.br_state == new->br_state &&
+ left.br_blockcount + new->br_blockcount <= MAXEXTLEN)
+ state |= BMAP_LEFT_CONTIG;
- error = xfs_bmap_last_extent(NULL, ip, whichfork, &rec, eof);
- if (error || *eof)
- return error;
+ if ((state & BMAP_RIGHT_VALID) && !(state & BMAP_RIGHT_DELAY) &&
+ new->br_startoff + new->br_blockcount == right.br_startoff &&
+ new->br_startblock + new->br_blockcount == right.br_startblock &&
+ new->br_state == right.br_state &&
+ new->br_blockcount + right.br_blockcount <= MAXEXTLEN &&
+ (!(state & BMAP_LEFT_CONTIG) ||
+ left.br_blockcount + new->br_blockcount +
+ right.br_blockcount <= MAXEXTLEN))
+ state |= BMAP_RIGHT_CONTIG;
- *eof = endoff >= rec.br_startoff + rec.br_blockcount;
- return 0;
-}
+ error = 0;
+ /*
+ * Select which case we're in here, and implement it.
+ */
+ switch (state & (BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG)) {
+ case BMAP_LEFT_CONTIG | BMAP_RIGHT_CONTIG:
+ /*
+ * New allocation is contiguous with real allocations on the
+ * left and on the right.
+ * Merge all three into a single extent record.
+ */
+ --bma->idx;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
+ left.br_blockcount + new->br_blockcount +
+ right.br_blockcount);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
-/*
- * Returns the file-relative block number of the first block past eof in
- * the file. This is not based on i_size, it is based on the extent records.
- * Returns 0 for local files, as they do not have extent records.
- */
-int
-xfs_bmap_last_offset(
- struct xfs_trans *tp,
- struct xfs_inode *ip,
- xfs_fileoff_t *last_block,
- int whichfork)
-{
- struct xfs_bmbt_irec rec;
- int is_empty;
- int error;
+ xfs_iext_remove(bma->ip, bma->idx + 1, 1, state);
- *last_block = 0;
+ XFS_IFORK_NEXT_SET(bma->ip, whichfork,
+ XFS_IFORK_NEXTENTS(bma->ip, whichfork) - 1);
+ if (bma->cur == NULL) {
+ rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
+ } else {
+ rval = XFS_ILOG_CORE;
+ error = xfs_bmbt_lookup_eq(bma->cur, right.br_startoff,
+ right.br_startblock, right.br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_btree_delete(bma->cur, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_btree_decrement(bma->cur, 0, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, left.br_startoff,
+ left.br_startblock,
+ left.br_blockcount +
+ new->br_blockcount +
+ right.br_blockcount,
+ left.br_state);
+ if (error)
+ goto done;
+ }
+ break;
- if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL)
- return 0;
+ case BMAP_LEFT_CONTIG:
+ /*
+ * New allocation is contiguous with a real allocation
+ * on the left.
+ * Merge the new allocation with the left neighbor.
+ */
+ --bma->idx;
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, bma->idx),
+ left.br_blockcount + new->br_blockcount);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
- if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
- return XFS_ERROR(EIO);
+ if (bma->cur == NULL) {
+ rval = xfs_ilog_fext(whichfork);
+ } else {
+ rval = 0;
+ error = xfs_bmbt_lookup_eq(bma->cur, left.br_startoff,
+ left.br_startblock, left.br_blockcount,
+ &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, left.br_startoff,
+ left.br_startblock,
+ left.br_blockcount +
+ new->br_blockcount,
+ left.br_state);
+ if (error)
+ goto done;
+ }
+ break;
- error = xfs_bmap_last_extent(NULL, ip, whichfork, &rec, &is_empty);
- if (error || is_empty)
- return error;
+ case BMAP_RIGHT_CONTIG:
+ /*
+ * New allocation is contiguous with a real allocation
+ * on the right.
+ * Merge the new allocation with the right neighbor.
+ */
+ trace_xfs_bmap_pre_update(bma->ip, bma->idx, state, _THIS_IP_);
+ xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, bma->idx),
+ new->br_startoff, new->br_startblock,
+ new->br_blockcount + right.br_blockcount,
+ right.br_state);
+ trace_xfs_bmap_post_update(bma->ip, bma->idx, state, _THIS_IP_);
- *last_block = rec.br_startoff + rec.br_blockcount;
- return 0;
+ if (bma->cur == NULL) {
+ rval = xfs_ilog_fext(whichfork);
+ } else {
+ rval = 0;
+ error = xfs_bmbt_lookup_eq(bma->cur,
+ right.br_startoff,
+ right.br_startblock,
+ right.br_blockcount, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ error = xfs_bmbt_update(bma->cur, new->br_startoff,
+ new->br_startblock,
+ new->br_blockcount +
+ right.br_blockcount,
+ right.br_state);
+ if (error)
+ goto done;
+ }
+ break;
+
+ case 0:
+ /*
+ * New allocation is not contiguous with another
+ * real allocation.
+ * Insert a new entry.
+ */
+ xfs_iext_insert(bma->ip, bma->idx, 1, new, state);
+ XFS_IFORK_NEXT_SET(bma->ip, whichfork,
+ XFS_IFORK_NEXTENTS(bma->ip, whichfork) + 1);
+ if (bma->cur == NULL) {
+ rval = XFS_ILOG_CORE | xfs_ilog_fext(whichfork);
+ } else {
+ rval = XFS_ILOG_CORE;
+ error = xfs_bmbt_lookup_eq(bma->cur,
+ new->br_startoff,
+ new->br_startblock,
+ new->br_blockcount, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 0, done);
+ bma->cur->bc_rec.b.br_state = new->br_state;
+ error = xfs_btree_insert(bma->cur, &i);
+ if (error)
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ }
+ break;
+ }
+
+ /* convert to a btree if necessary */
+ if (xfs_bmap_needs_btree(bma->ip, whichfork)) {
+ int tmp_logflags; /* partial log flag return val */
+
+ ASSERT(bma->cur == NULL);
+ error = xfs_bmap_extents_to_btree(bma->tp, bma->ip,
+ bma->firstblock, bma->flist, &bma->cur,
+ 0, &tmp_logflags, whichfork);
+ bma->logflags |= tmp_logflags;
+ if (error)
+ goto done;
+ }
+
+ /* clear out the allocated field, done with it now in any case. */
+ if (bma->cur)
+ bma->cur->bc_private.b.allocated = 0;
+
+ xfs_bmap_check_leaf_extents(bma->cur, bma->ip, whichfork);
+done:
+ bma->logflags |= rval;
+ return error;
}
/*
- * Returns whether the selected fork of the inode has exactly one
- * block or not. For the data fork we check this matches di_size,
- * implying the file's range is 0..bsize-1.
+ * Functions used in the extent read, allocate and remove paths
*/
-int /* 1=>1 block, 0=>otherwise */
-xfs_bmap_one_block(
- xfs_inode_t *ip, /* incore inode */
- int whichfork) /* data or attr fork */
-{
- xfs_bmbt_rec_host_t *ep; /* ptr to fork's extent */
- xfs_ifork_t *ifp; /* inode fork pointer */
- int rval; /* return value */
- xfs_bmbt_irec_t s; /* internal version of extent */
-
-#ifndef DEBUG
- if (whichfork == XFS_DATA_FORK)
- return XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize;
-#endif /* !DEBUG */
- if (XFS_IFORK_NEXTENTS(ip, whichfork) != 1)
- return 0;
- if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
- return 0;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- ASSERT(ifp->if_flags & XFS_IFEXTENTS);
- ep = xfs_iext_get_ext(ifp, 0);
- xfs_bmbt_get_all(ep, &s);
- rval = s.br_startoff == 0 && s.br_blockcount == 1;
- if (rval && whichfork == XFS_DATA_FORK)
- ASSERT(XFS_ISIZE(ip) == ip->i_mount->m_sb.sb_blocksize);
- return rval;
-}
+/*
+ * Adjust the size of the new extent based on di_extsize and rt extsize.
+ */
STATIC int
-xfs_bmap_sanity_check(
- struct xfs_mount *mp,
- struct xfs_buf *bp,
- int level)
+xfs_bmap_extsize_align(
+ xfs_mount_t *mp,
+ xfs_bmbt_irec_t *gotp, /* next extent pointer */
+ xfs_bmbt_irec_t *prevp, /* previous extent pointer */
+ xfs_extlen_t extsz, /* align to this extent size */
+ int rt, /* is this a realtime inode? */
+ int eof, /* is extent at end-of-file? */
+ int delay, /* creating delalloc extent? */
+ int convert, /* overwriting unwritten extent? */
+ xfs_fileoff_t *offp, /* in/out: aligned offset */
+ xfs_extlen_t *lenp) /* in/out: aligned length */
{
- struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
+ xfs_fileoff_t orig_off; /* original offset */
+ xfs_extlen_t orig_alen; /* original length */
+ xfs_fileoff_t orig_end; /* original off+len */
+ xfs_fileoff_t nexto; /* next file offset */
+ xfs_fileoff_t prevo; /* previous file offset */
+ xfs_fileoff_t align_off; /* temp for offset */
+ xfs_extlen_t align_alen; /* temp for length */
+ xfs_extlen_t temp; /* temp for calculations */
- if (block->bb_magic != cpu_to_be32(XFS_BMAP_MAGIC) ||
- be16_to_cpu(block->bb_level) != level ||
- be16_to_cpu(block->bb_numrecs) == 0 ||
- be16_to_cpu(block->bb_numrecs) > mp->m_bmap_dmxr[level != 0])
+ if (convert)
return 0;
- return 1;
-}
-/*
- * Read in the extents to if_extents.
- * All inode fields are set up by caller, we just traverse the btree
- * and copy the records in. If the file system cannot contain unwritten
- * extents, the records are checked for no "state" flags.
- */
-int /* error */
-xfs_bmap_read_extents(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode */
- int whichfork) /* data or attr fork */
-{
- struct xfs_btree_block *block; /* current btree block */
- xfs_fsblock_t bno; /* block # of "block" */
- xfs_buf_t *bp; /* buffer for "block" */
- int error; /* error return value */
- xfs_exntfmt_t exntf; /* XFS_EXTFMT_NOSTATE, if checking */
- xfs_extnum_t i, j; /* index into the extents list */
- xfs_ifork_t *ifp; /* fork structure */
- int level; /* btree level, for checking */
- xfs_mount_t *mp; /* file system mount structure */
- __be64 *pp; /* pointer to block address */
- /* REFERENCED */
- xfs_extnum_t room; /* number of entries there's room for */
+ orig_off = align_off = *offp;
+ orig_alen = align_alen = *lenp;
+ orig_end = orig_off + orig_alen;
- bno = NULLFSBLOCK;
- mp = ip->i_mount;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- exntf = (whichfork != XFS_DATA_FORK) ? XFS_EXTFMT_NOSTATE :
- XFS_EXTFMT_INODE(ip);
- block = ifp->if_broot;
/*
- * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
+ * If this request overlaps an existing extent, then don't
+ * attempt to perform any additional alignment.
*/
- level = be16_to_cpu(block->bb_level);
- ASSERT(level > 0);
- pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
- bno = be64_to_cpu(*pp);
- ASSERT(bno != NULLDFSBNO);
- ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
- ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
+ if (!delay && !eof &&
+ (orig_off >= gotp->br_startoff) &&
+ (orig_end <= gotp->br_startoff + gotp->br_blockcount)) {
+ return 0;
+ }
+
/*
- * Go down the tree until leaf level is reached, following the first
- * pointer (leftmost) at each level.
+ * If the file offset is unaligned vs. the extent size
+ * we need to align it. This will be possible unless
+ * the file was previously written with a kernel that didn't
+ * perform this alignment, or if a truncate shot us in the
+ * foot.
*/
- while (level-- > 0) {
- error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
- XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
- if (error)
- return error;
- block = XFS_BUF_TO_BLOCK(bp);
- XFS_WANT_CORRUPTED_GOTO(
- xfs_bmap_sanity_check(mp, bp, level),
- error0);
- if (level == 0)
- break;
- pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
- bno = be64_to_cpu(*pp);
- XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0);
- xfs_trans_brelse(tp, bp);
+ temp = do_mod(orig_off, extsz);
+ if (temp) {
+ align_alen += temp;
+ align_off -= temp;
}
/*
- * Here with bp and block set to the leftmost leaf node in the tree.
+ * Same adjustment for the end of the requested area.
*/
- room = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
- i = 0;
+ if ((temp = (align_alen % extsz))) {
+ align_alen += extsz - temp;
+ }
/*
- * Loop over all leaf nodes. Copy information to the extent records.
+ * If the previous block overlaps with this proposed allocation
+ * then move the start forward without adjusting the length.
*/
- for (;;) {
- xfs_bmbt_rec_t *frp;
- xfs_fsblock_t nextbno;
- xfs_extnum_t num_recs;
- xfs_extnum_t start;
+ if (prevp->br_startoff != NULLFILEOFF) {
+ if (prevp->br_startblock == HOLESTARTBLOCK)
+ prevo = prevp->br_startoff;
+ else
+ prevo = prevp->br_startoff + prevp->br_blockcount;
+ } else
+ prevo = 0;
+ if (align_off != orig_off && align_off < prevo)
+ align_off = prevo;
+ /*
+ * If the next block overlaps with this proposed allocation
+ * then move the start back without adjusting the length,
+ * but not before offset 0.
+ * This may of course make the start overlap previous block,
+ * and if we hit the offset 0 limit then the next block
+ * can still overlap too.
+ */
+ if (!eof && gotp->br_startoff != NULLFILEOFF) {
+ if ((delay && gotp->br_startblock == HOLESTARTBLOCK) ||
+ (!delay && gotp->br_startblock == DELAYSTARTBLOCK))
+ nexto = gotp->br_startoff + gotp->br_blockcount;
+ else
+ nexto = gotp->br_startoff;
+ } else
+ nexto = NULLFILEOFF;
+ if (!eof &&
+ align_off + align_alen != orig_end &&
+ align_off + align_alen > nexto)
+ align_off = nexto > align_alen ? nexto - align_alen : 0;
+ /*
+ * If we're now overlapping the next or previous extent that
+ * means we can't fit an extsz piece in this hole. Just move
+ * the start forward to the first valid spot and set
+ * the length so we hit the end.
+ */
+ if (align_off != orig_off && align_off < prevo)
+ align_off = prevo;
+ if (align_off + align_alen != orig_end &&
+ align_off + align_alen > nexto &&
+ nexto != NULLFILEOFF) {
+ ASSERT(nexto > prevo);
+ align_alen = nexto - align_off;
+ }
- num_recs = xfs_btree_get_numrecs(block);
- if (unlikely(i + num_recs > room)) {
- ASSERT(i + num_recs <= room);
- xfs_warn(ip->i_mount,
- "corrupt dinode %Lu, (btree extents).",
- (unsigned long long) ip->i_ino);
- XFS_CORRUPTION_ERROR("xfs_bmap_read_extents(1)",
- XFS_ERRLEVEL_LOW, ip->i_mount, block);
- goto error0;
- }
- XFS_WANT_CORRUPTED_GOTO(
- xfs_bmap_sanity_check(mp, bp, 0),
- error0);
+ /*
+ * If realtime, and the result isn't a multiple of the realtime
+ * extent size we need to remove blocks until it is.
+ */
+ if (rt && (temp = (align_alen % mp->m_sb.sb_rextsize))) {
/*
- * Read-ahead the next leaf block, if any.
+ * We're not covering the original request, or
+ * we won't be able to once we fix the length.
*/
- nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
- if (nextbno != NULLFSBLOCK)
- xfs_btree_reada_bufl(mp, nextbno, 1,
- &xfs_bmbt_buf_ops);
+ if (orig_off < align_off ||
+ orig_end > align_off + align_alen ||
+ align_alen - temp < orig_alen)
+ return XFS_ERROR(EINVAL);
/*
- * Copy records into the extent records.
+ * Try to fix it by moving the start up.
*/
- frp = XFS_BMBT_REC_ADDR(mp, block, 1);
- start = i;
- for (j = 0; j < num_recs; j++, i++, frp++) {
- xfs_bmbt_rec_host_t *trp = xfs_iext_get_ext(ifp, i);
- trp->l0 = be64_to_cpu(frp->l0);
- trp->l1 = be64_to_cpu(frp->l1);
+ if (align_off + temp <= orig_off) {
+ align_alen -= temp;
+ align_off += temp;
}
- if (exntf == XFS_EXTFMT_NOSTATE) {
- /*
- * Check all attribute bmap btree records and
- * any "older" data bmap btree records for a
- * set bit in the "extent flag" position.
- */
- if (unlikely(xfs_check_nostate_extents(ifp,
- start, num_recs))) {
- XFS_ERROR_REPORT("xfs_bmap_read_extents(2)",
- XFS_ERRLEVEL_LOW,
- ip->i_mount);
- goto error0;
- }
+ /*
+ * Try to fix it by moving the end in.
+ */
+ else if (align_off + align_alen - temp >= orig_end)
+ align_alen -= temp;
+ /*
+ * Set the start to the minimum then trim the length.
+ */
+ else {
+ align_alen -= orig_off - align_off;
+ align_off = orig_off;
+ align_alen -= align_alen % mp->m_sb.sb_rextsize;
}
- xfs_trans_brelse(tp, bp);
- bno = nextbno;
/*
- * If we've reached the end, stop.
+ * Result doesn't cover the request, fail it.
*/
- if (bno == NULLFSBLOCK)
- break;
- error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
- XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops);
- if (error)
- return error;
- block = XFS_BUF_TO_BLOCK(bp);
+ if (orig_off < align_off || orig_end > align_off + align_alen)
+ return XFS_ERROR(EINVAL);
+ } else {
+ ASSERT(orig_off >= align_off);
+ ASSERT(orig_end <= align_off + align_alen);
}
- ASSERT(i == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)));
- ASSERT(i == XFS_IFORK_NEXTENTS(ip, whichfork));
- XFS_BMAP_TRACE_EXLIST(ip, i, whichfork);
- return 0;
-error0:
- xfs_trans_brelse(tp, bp);
- return XFS_ERROR(EFSCORRUPTED);
-}
#ifdef DEBUG
-/*
- * Add bmap trace insert entries for all the contents of the extent records.
- */
-void
-xfs_bmap_trace_exlist(
- xfs_inode_t *ip, /* incore inode pointer */
- xfs_extnum_t cnt, /* count of entries in the list */
- int whichfork, /* data or attr fork */
- unsigned long caller_ip)
-{
- xfs_extnum_t idx; /* extent record index */
- xfs_ifork_t *ifp; /* inode fork pointer */
- int state = 0;
-
- if (whichfork == XFS_ATTR_FORK)
- state |= BMAP_ATTRFORK;
+ if (!eof && gotp->br_startoff != NULLFILEOFF)
+ ASSERT(align_off + align_alen <= gotp->br_startoff);
+ if (prevp->br_startoff != NULLFILEOFF)
+ ASSERT(align_off >= prevp->br_startoff + prevp->br_blockcount);
+#endif
- ifp = XFS_IFORK_PTR(ip, whichfork);
- ASSERT(cnt == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)));
- for (idx = 0; idx < cnt; idx++)
- trace_xfs_extlist(ip, idx, whichfork, caller_ip);
+ *lenp = align_alen;
+ *offp = align_off;
+ return 0;
}
-/*
- * Validate that the bmbt_irecs being returned from bmapi are valid
- * given the callers original parameters. Specifically check the
- * ranges of the returned irecs to ensure that they only extent beyond
- * the given parameters if the XFS_BMAPI_ENTIRE flag was set.
- */
+#define XFS_ALLOC_GAP_UNITS 4
+
STATIC void
-xfs_bmap_validate_ret(
- xfs_fileoff_t bno,
- xfs_filblks_t len,
- int flags,
- xfs_bmbt_irec_t *mval,
- int nmap,
- int ret_nmap)
+xfs_bmap_adjacent(
+ xfs_bmalloca_t *ap) /* bmap alloc argument struct */
{
- int i; /* index to map values */
+ xfs_fsblock_t adjust; /* adjustment to block numbers */
+ xfs_agnumber_t fb_agno; /* ag number of ap->firstblock */
+ xfs_mount_t *mp; /* mount point structure */
+ int nullfb; /* true if ap->firstblock isn't set */
+ int rt; /* true if inode is realtime */
- ASSERT(ret_nmap <= nmap);
+#define ISVALID(x,y) \
+ (rt ? \
+ (x) < mp->m_sb.sb_rblocks : \
+ XFS_FSB_TO_AGNO(mp, x) == XFS_FSB_TO_AGNO(mp, y) && \
+ XFS_FSB_TO_AGNO(mp, x) < mp->m_sb.sb_agcount && \
+ XFS_FSB_TO_AGBNO(mp, x) < mp->m_sb.sb_agblocks)
- for (i = 0; i < ret_nmap; i++) {
- ASSERT(mval[i].br_blockcount > 0);
- if (!(flags & XFS_BMAPI_ENTIRE)) {
- ASSERT(mval[i].br_startoff >= bno);
- ASSERT(mval[i].br_blockcount <= len);
- ASSERT(mval[i].br_startoff + mval[i].br_blockcount <=
- bno + len);
- } else {
- ASSERT(mval[i].br_startoff < bno + len);
- ASSERT(mval[i].br_startoff + mval[i].br_blockcount >
- bno);
+ mp = ap->ip->i_mount;
+ nullfb = *ap->firstblock == NULLFSBLOCK;
+ rt = XFS_IS_REALTIME_INODE(ap->ip) && ap->userdata;
+ fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
+ /*
+ * If allocating at eof, and there's a previous real block,
+ * try to use its last block as our starting point.
+ */
+ if (ap->eof && ap->prev.br_startoff != NULLFILEOFF &&
+ !isnullstartblock(ap->prev.br_startblock) &&
+ ISVALID(ap->prev.br_startblock + ap->prev.br_blockcount,
+ ap->prev.br_startblock)) {
+ ap->blkno = ap->prev.br_startblock + ap->prev.br_blockcount;
+ /*
+ * Adjust for the gap between prevp and us.
+ */
+ adjust = ap->offset -
+ (ap->prev.br_startoff + ap->prev.br_blockcount);
+ if (adjust &&
+ ISVALID(ap->blkno + adjust, ap->prev.br_startblock))
+ ap->blkno += adjust;
+ }
+ /*
+ * If not at eof, then compare the two neighbor blocks.
+ * Figure out whether either one gives us a good starting point,
+ * and pick the better one.
+ */
+ else if (!ap->eof) {
+ xfs_fsblock_t gotbno; /* right side block number */
+ xfs_fsblock_t gotdiff=0; /* right side difference */
+ xfs_fsblock_t prevbno; /* left side block number */
+ xfs_fsblock_t prevdiff=0; /* left side difference */
+
+ /*
+ * If there's a previous (left) block, select a requested
+ * start block based on it.
+ */
+ if (ap->prev.br_startoff != NULLFILEOFF &&
+ !isnullstartblock(ap->prev.br_startblock) &&
+ (prevbno = ap->prev.br_startblock +
+ ap->prev.br_blockcount) &&
+ ISVALID(prevbno, ap->prev.br_startblock)) {
+ /*
+ * Calculate gap to end of previous block.
+ */
+ adjust = prevdiff = ap->offset -
+ (ap->prev.br_startoff +
+ ap->prev.br_blockcount);
+ /*
+ * Figure the startblock based on the previous block's
+ * end and the gap size.
+ * Heuristic!
+ * If the gap is large relative to the piece we're
+ * allocating, or using it gives us an invalid block
+ * number, then just use the end of the previous block.
+ */
+ if (prevdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
+ ISVALID(prevbno + prevdiff,
+ ap->prev.br_startblock))
+ prevbno += adjust;
+ else
+ prevdiff += adjust;
+ /*
+ * If the firstblock forbids it, can't use it,
+ * must use default.
+ */
+ if (!rt && !nullfb &&
+ XFS_FSB_TO_AGNO(mp, prevbno) != fb_agno)
+ prevbno = NULLFSBLOCK;
}
- ASSERT(i == 0 ||
- mval[i - 1].br_startoff + mval[i - 1].br_blockcount ==
- mval[i].br_startoff);
- ASSERT(mval[i].br_startblock != DELAYSTARTBLOCK &&
- mval[i].br_startblock != HOLESTARTBLOCK);
- ASSERT(mval[i].br_state == XFS_EXT_NORM ||
- mval[i].br_state == XFS_EXT_UNWRITTEN);
+ /*
+ * No previous block or can't follow it, just default.
+ */
+ else
+ prevbno = NULLFSBLOCK;
+ /*
+ * If there's a following (right) block, select a requested
+ * start block based on it.
+ */
+ if (!isnullstartblock(ap->got.br_startblock)) {
+ /*
+ * Calculate gap to start of next block.
+ */
+ adjust = gotdiff = ap->got.br_startoff - ap->offset;
+ /*
+ * Figure the startblock based on the next block's
+ * start and the gap size.
+ */
+ gotbno = ap->got.br_startblock;
+ /*
+ * Heuristic!
+ * If the gap is large relative to the piece we're
+ * allocating, or using it gives us an invalid block
+ * number, then just use the start of the next block
+ * offset by our length.
+ */
+ if (gotdiff <= XFS_ALLOC_GAP_UNITS * ap->length &&
+ ISVALID(gotbno - gotdiff, gotbno))
+ gotbno -= adjust;
+ else if (ISVALID(gotbno - ap->length, gotbno)) {
+ gotbno -= ap->length;
+ gotdiff += adjust - ap->length;
+ } else
+ gotdiff += adjust;
+ /*
+ * If the firstblock forbids it, can't use it,
+ * must use default.
+ */
+ if (!rt && !nullfb &&
+ XFS_FSB_TO_AGNO(mp, gotbno) != fb_agno)
+ gotbno = NULLFSBLOCK;
+ }
+ /*
+ * No next block, just default.
+ */
+ else
+ gotbno = NULLFSBLOCK;
+ /*
+ * If both valid, pick the better one, else the only good
+ * one, else ap->blkno is already set (to 0 or the inode block).
+ */
+ if (prevbno != NULLFSBLOCK && gotbno != NULLFSBLOCK)
+ ap->blkno = prevdiff <= gotdiff ? prevbno : gotbno;
+ else if (prevbno != NULLFSBLOCK)
+ ap->blkno = prevbno;
+ else if (gotbno != NULLFSBLOCK)
+ ap->blkno = gotbno;
}
+#undef ISVALID
}
-#endif /* DEBUG */
-
-/*
- * Trim the returned map to the required bounds
- */
-STATIC void
-xfs_bmapi_trim_map(
- struct xfs_bmbt_irec *mval,
- struct xfs_bmbt_irec *got,
- xfs_fileoff_t *bno,
- xfs_filblks_t len,
- xfs_fileoff_t obno,
- xfs_fileoff_t end,
- int n,
- int flags)
+STATIC int
+xfs_bmap_rtalloc(
+ xfs_bmalloca_t *ap) /* bmap alloc argument struct */
{
- if ((flags & XFS_BMAPI_ENTIRE) ||
- got->br_startoff + got->br_blockcount <= obno) {
- *mval = *got;
- if (isnullstartblock(got->br_startblock))
- mval->br_startblock = DELAYSTARTBLOCK;
- return;
- }
+ xfs_alloctype_t atype = 0; /* type for allocation routines */
+ int error; /* error return value */
+ xfs_mount_t *mp; /* mount point structure */
+ xfs_extlen_t prod = 0; /* product factor for allocators */
+ xfs_extlen_t ralen = 0; /* realtime allocation length */
+ xfs_extlen_t align; /* minimum allocation alignment */
+ xfs_rtblock_t rtb;
+
+ mp = ap->ip->i_mount;
+ align = xfs_get_extsz_hint(ap->ip);
+ prod = align / mp->m_sb.sb_rextsize;
+ error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
+ align, 1, ap->eof, 0,
+ ap->conv, &ap->offset, &ap->length);
+ if (error)
+ return error;
+ ASSERT(ap->length);
+ ASSERT(ap->length % mp->m_sb.sb_rextsize == 0);
- if (obno > *bno)
- *bno = obno;
- ASSERT((*bno >= obno) || (n == 0));
- ASSERT(*bno < end);
- mval->br_startoff = *bno;
- if (isnullstartblock(got->br_startblock))
- mval->br_startblock = DELAYSTARTBLOCK;
- else
- mval->br_startblock = got->br_startblock +
- (*bno - got->br_startoff);
/*
- * Return the minimum of what we got and what we asked for for
- * the length. We can use the len variable here because it is
- * modified below and we could have been there before coming
- * here if the first part of the allocation didn't overlap what
- * was asked for.
+ * If the offset & length are not perfectly aligned
+ * then kill prod, it will just get us in trouble.
*/
- mval->br_blockcount = XFS_FILBLKS_MIN(end - *bno,
- got->br_blockcount - (*bno - got->br_startoff));
- mval->br_state = got->br_state;
- ASSERT(mval->br_blockcount <= len);
- return;
-}
-
-/*
- * Update and validate the extent map to return
- */
-STATIC void
-xfs_bmapi_update_map(
- struct xfs_bmbt_irec **map,
- xfs_fileoff_t *bno,
- xfs_filblks_t *len,
- xfs_fileoff_t obno,
- xfs_fileoff_t end,
- int *n,
- int flags)
-{
- xfs_bmbt_irec_t *mval = *map;
-
- ASSERT((flags & XFS_BMAPI_ENTIRE) ||
- ((mval->br_startoff + mval->br_blockcount) <= end));
- ASSERT((flags & XFS_BMAPI_ENTIRE) || (mval->br_blockcount <= *len) ||
- (mval->br_startoff < obno));
-
- *bno = mval->br_startoff + mval->br_blockcount;
- *len = end - *bno;
- if (*n > 0 && mval->br_startoff == mval[-1].br_startoff) {
- /* update previous map with new information */
- ASSERT(mval->br_startblock == mval[-1].br_startblock);
- ASSERT(mval->br_blockcount > mval[-1].br_blockcount);
- ASSERT(mval->br_state == mval[-1].br_state);
- mval[-1].br_blockcount = mval->br_blockcount;
- mval[-1].br_state = mval->br_state;
- } else if (*n > 0 && mval->br_startblock != DELAYSTARTBLOCK &&
- mval[-1].br_startblock != DELAYSTARTBLOCK &&
- mval[-1].br_startblock != HOLESTARTBLOCK &&
- mval->br_startblock == mval[-1].br_startblock +
- mval[-1].br_blockcount &&
- ((flags & XFS_BMAPI_IGSTATE) ||
- mval[-1].br_state == mval->br_state)) {
- ASSERT(mval->br_startoff ==
- mval[-1].br_startoff + mval[-1].br_blockcount);
- mval[-1].br_blockcount += mval->br_blockcount;
- } else if (*n > 0 &&
- mval->br_startblock == DELAYSTARTBLOCK &&
- mval[-1].br_startblock == DELAYSTARTBLOCK &&
- mval->br_startoff ==
- mval[-1].br_startoff + mval[-1].br_blockcount) {
- mval[-1].br_blockcount += mval->br_blockcount;
- mval[-1].br_state = mval->br_state;
- } else if (!((*n == 0) &&
- ((mval->br_startoff + mval->br_blockcount) <=
- obno))) {
- mval++;
- (*n)++;
- }
- *map = mval;
-}
-
-/*
- * Map file blocks to filesystem blocks without allocation.
- */
-int
-xfs_bmapi_read(
- struct xfs_inode *ip,
- xfs_fileoff_t bno,
- xfs_filblks_t len,
- struct xfs_bmbt_irec *mval,
- int *nmap,
- int flags)
-{
- struct xfs_mount *mp = ip->i_mount;
- struct xfs_ifork *ifp;
- struct xfs_bmbt_irec got;
- struct xfs_bmbt_irec prev;
- xfs_fileoff_t obno;
- xfs_fileoff_t end;
- xfs_extnum_t lastx;
- int error;
- int eof;
- int n = 0;
- int whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
- XFS_ATTR_FORK : XFS_DATA_FORK;
-
- ASSERT(*nmap >= 1);
- ASSERT(!(flags & ~(XFS_BMAPI_ATTRFORK|XFS_BMAPI_ENTIRE|
- XFS_BMAPI_IGSTATE)));
-
- if (unlikely(XFS_TEST_ERROR(
- (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
- mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
- XFS_ERROR_REPORT("xfs_bmapi_read", XFS_ERRLEVEL_LOW, mp);
- return XFS_ERROR(EFSCORRUPTED);
- }
-
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(EIO);
+ if (do_mod(ap->offset, align) || ap->length % align)
+ prod = 1;
+ /*
+ * Set ralen to be the actual requested length in rtextents.
+ */
+ ralen = ap->length / mp->m_sb.sb_rextsize;
+ /*
+ * If the old value was close enough to MAXEXTLEN that
+ * we rounded up to it, cut it back so it's valid again.
+ * Note that if it's a really large request (bigger than
+ * MAXEXTLEN), we don't hear about that number, and can't
+ * adjust the starting point to match it.
+ */
+ if (ralen * mp->m_sb.sb_rextsize >= MAXEXTLEN)
+ ralen = MAXEXTLEN / mp->m_sb.sb_rextsize;
- XFS_STATS_INC(xs_blk_mapr);
+ /*
+ * Lock out other modifications to the RT bitmap inode.
+ */
+ xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(ap->tp, mp->m_rbmip, XFS_ILOCK_EXCL);
- ifp = XFS_IFORK_PTR(ip, whichfork);
+ /*
+ * If it's an allocation to an empty file at offset 0,
+ * pick an extent that will space things out in the rt area.
+ */
+ if (ap->eof && ap->offset == 0) {
+ xfs_rtblock_t uninitialized_var(rtx); /* realtime extent no */
- if (!(ifp->if_flags & XFS_IFEXTENTS)) {
- error = xfs_iread_extents(NULL, ip, whichfork);
+ error = xfs_rtpick_extent(mp, ap->tp, ralen, &rtx);
if (error)
return error;
+ ap->blkno = rtx * mp->m_sb.sb_rextsize;
+ } else {
+ ap->blkno = 0;
}
- xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got, &prev);
- end = bno + len;
- obno = bno;
-
- while (bno < end && n < *nmap) {
- /* Reading past eof, act as though there's a hole up to end. */
- if (eof)
- got.br_startoff = end;
- if (got.br_startoff > bno) {
- /* Reading in a hole. */
- mval->br_startoff = bno;
- mval->br_startblock = HOLESTARTBLOCK;
- mval->br_blockcount =
- XFS_FILBLKS_MIN(len, got.br_startoff - bno);
- mval->br_state = XFS_EXT_NORM;
- bno += mval->br_blockcount;
- len -= mval->br_blockcount;
- mval++;
- n++;
- continue;
- }
-
- /* set up the extent map to return. */
- xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
- xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
-
- /* If we're done, stop now. */
- if (bno >= end || n >= *nmap)
- break;
+ xfs_bmap_adjacent(ap);
- /* Else go on to the next record. */
- if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
- else
- eof = 1;
+ /*
+ * Realtime allocation, done through xfs_rtallocate_extent.
+ */
+ atype = ap->blkno == 0 ? XFS_ALLOCTYPE_ANY_AG : XFS_ALLOCTYPE_NEAR_BNO;
+ do_div(ap->blkno, mp->m_sb.sb_rextsize);
+ rtb = ap->blkno;
+ ap->length = ralen;
+ if ((error = xfs_rtallocate_extent(ap->tp, ap->blkno, 1, ap->length,
+ &ralen, atype, ap->wasdel, prod, &rtb)))
+ return error;
+ if (rtb == NULLFSBLOCK && prod > 1 &&
+ (error = xfs_rtallocate_extent(ap->tp, ap->blkno, 1,
+ ap->length, &ralen, atype,
+ ap->wasdel, 1, &rtb)))
+ return error;
+ ap->blkno = rtb;
+ if (ap->blkno != NULLFSBLOCK) {
+ ap->blkno *= mp->m_sb.sb_rextsize;
+ ralen *= mp->m_sb.sb_rextsize;
+ ap->length = ralen;
+ ap->ip->i_d.di_nblocks += ralen;
+ xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
+ if (ap->wasdel)
+ ap->ip->i_delayed_blks -= ralen;
+ /*
+ * Adjust the disk quota also. This was reserved
+ * earlier.
+ */
+ xfs_trans_mod_dquot_byino(ap->tp, ap->ip,
+ ap->wasdel ? XFS_TRANS_DQ_DELRTBCOUNT :
+ XFS_TRANS_DQ_RTBCOUNT, (long) ralen);
+ } else {
+ ap->length = 0;
}
- *nmap = n;
return 0;
}
STATIC int
-xfs_bmapi_reserve_delalloc(
- struct xfs_inode *ip,
- xfs_fileoff_t aoff,
- xfs_filblks_t len,
- struct xfs_bmbt_irec *got,
- struct xfs_bmbt_irec *prev,
- xfs_extnum_t *lastx,
- int eof)
+xfs_bmap_btalloc_nullfb(
+ struct xfs_bmalloca *ap,
+ struct xfs_alloc_arg *args,
+ xfs_extlen_t *blen)
{
- struct xfs_mount *mp = ip->i_mount;
- struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
- xfs_extlen_t alen;
- xfs_extlen_t indlen;
- char rt = XFS_IS_REALTIME_INODE(ip);
- xfs_extlen_t extsz;
+ struct xfs_mount *mp = ap->ip->i_mount;
+ struct xfs_perag *pag;
+ xfs_agnumber_t ag, startag;
+ int notinit = 0;
int error;
- alen = XFS_FILBLKS_MIN(len, MAXEXTLEN);
- if (!eof)
- alen = XFS_FILBLKS_MIN(alen, got->br_startoff - aoff);
-
- /* Figure out the extent size, adjust alen */
- extsz = xfs_get_extsz_hint(ip);
- if (extsz) {
- /*
- * Make sure we don't exceed a single extent length when we
- * align the extent by reducing length we are going to
- * allocate by the maximum amount extent size aligment may
- * require.
- */
- alen = XFS_FILBLKS_MIN(len, MAXEXTLEN - (2 * extsz - 1));
- error = xfs_bmap_extsize_align(mp, got, prev, extsz, rt, eof,
- 1, 0, &aoff, &alen);
- ASSERT(!error);
- }
-
- if (rt)
- extsz = alen / mp->m_sb.sb_rextsize;
+ if (ap->userdata && xfs_inode_is_filestream(ap->ip))
+ args->type = XFS_ALLOCTYPE_NEAR_BNO;
+ else
+ args->type = XFS_ALLOCTYPE_START_BNO;
+ args->total = ap->total;
/*
- * Make a transaction-less quota reservation for delayed allocation
- * blocks. This number gets adjusted later. We return if we haven't
- * allocated blocks already inside this loop.
+ * Search for an allocation group with a single extent large enough
+ * for the request. If one isn't found, then adjust the minimum
+ * allocation size to the largest space found.
*/
- error = xfs_trans_reserve_quota_nblks(NULL, ip, (long)alen, 0,
- rt ? XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
- if (error)
- return error;
+ startag = ag = XFS_FSB_TO_AGNO(mp, args->fsbno);
+ if (startag == NULLAGNUMBER)
+ startag = ag = 0;
- /*
- * Split changing sb for alen and indlen since they could be coming
- * from different places.
- */
- indlen = (xfs_extlen_t)xfs_bmap_worst_indlen(ip, alen);
- ASSERT(indlen > 0);
+ pag = xfs_perag_get(mp, ag);
+ while (*blen < args->maxlen) {
+ if (!pag->pagf_init) {
+ error = xfs_alloc_pagf_init(mp, args->tp, ag,
+ XFS_ALLOC_FLAG_TRYLOCK);
+ if (error) {
+ xfs_perag_put(pag);
+ return error;
+ }
+ }
- if (rt) {
- error = xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS,
- -((int64_t)extsz), 0);
- } else {
- error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
- -((int64_t)alen), 0);
- }
+ /*
+ * See xfs_alloc_fix_freelist...
+ */
+ if (pag->pagf_init) {
+ xfs_extlen_t longest;
+ longest = xfs_alloc_longest_free_extent(mp, pag);
+ if (*blen < longest)
+ *blen = longest;
+ } else
+ notinit = 1;
- if (error)
- goto out_unreserve_quota;
+ if (xfs_inode_is_filestream(ap->ip)) {
+ if (*blen >= args->maxlen)
+ break;
- error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
- -((int64_t)indlen), 0);
- if (error)
- goto out_unreserve_blocks;
+ if (ap->userdata) {
+ /*
+ * If startag is an invalid AG, we've
+ * come here once before and
+ * xfs_filestream_new_ag picked the
+ * best currently available.
+ *
+ * Don't continue looping, since we
+ * could loop forever.
+ */
+ if (startag == NULLAGNUMBER)
+ break;
+ error = xfs_filestream_new_ag(ap, &ag);
+ xfs_perag_put(pag);
+ if (error)
+ return error;
- ip->i_delayed_blks += alen;
+ /* loop again to set 'blen'*/
+ startag = NULLAGNUMBER;
+ pag = xfs_perag_get(mp, ag);
+ continue;
+ }
+ }
+ if (++ag == mp->m_sb.sb_agcount)
+ ag = 0;
+ if (ag == startag)
+ break;
+ xfs_perag_put(pag);
+ pag = xfs_perag_get(mp, ag);
+ }
+ xfs_perag_put(pag);
- got->br_startoff = aoff;
- got->br_startblock = nullstartblock(indlen);
- got->br_blockcount = alen;
- got->br_state = XFS_EXT_NORM;
- xfs_bmap_add_extent_hole_delay(ip, lastx, got);
+ /*
+ * Since the above loop did a BUF_TRYLOCK, it is
+ * possible that there is space for this request.
+ */
+ if (notinit || *blen < ap->minlen)
+ args->minlen = ap->minlen;
+ /*
+ * If the best seen length is less than the request
+ * length, use the best as the minimum.
+ */
+ else if (*blen < args->maxlen)
+ args->minlen = *blen;
+ /*
+ * Otherwise we've seen an extent as big as maxlen,
+ * use that as the minimum.
+ */
+ else
+ args->minlen = args->maxlen;
/*
- * Update our extent pointer, given that xfs_bmap_add_extent_hole_delay
- * might have merged it into one of the neighbouring ones.
+ * set the failure fallback case to look in the selected
+ * AG as the stream may have moved.
*/
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *lastx), got);
+ if (xfs_inode_is_filestream(ap->ip))
+ ap->blkno = args->fsbno = XFS_AGB_TO_FSB(mp, ag, 0);
- ASSERT(got->br_startoff <= aoff);
- ASSERT(got->br_startoff + got->br_blockcount >= aoff + alen);
- ASSERT(isnullstartblock(got->br_startblock));
- ASSERT(got->br_state == XFS_EXT_NORM);
return 0;
-
-out_unreserve_blocks:
- if (rt)
- xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS, extsz, 0);
- else
- xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, alen, 0);
-out_unreserve_quota:
- if (XFS_IS_QUOTA_ON(mp))
- xfs_trans_unreserve_quota_nblks(NULL, ip, (long)alen, 0, rt ?
- XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
- return error;
}
-/*
- * Map file blocks to filesystem blocks, adding delayed allocations as needed.
- */
-int
-xfs_bmapi_delay(
- struct xfs_inode *ip, /* incore inode */
- xfs_fileoff_t bno, /* starting file offs. mapped */
- xfs_filblks_t len, /* length to map in file */
- struct xfs_bmbt_irec *mval, /* output: map values */
- int *nmap, /* i/o: mval size/count */
- int flags) /* XFS_BMAPI_... */
+STATIC int
+xfs_bmap_btalloc(
+ xfs_bmalloca_t *ap) /* bmap alloc argument struct */
{
- struct xfs_mount *mp = ip->i_mount;
- struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
- struct xfs_bmbt_irec got; /* current file extent record */
- struct xfs_bmbt_irec prev; /* previous file extent record */
- xfs_fileoff_t obno; /* old block number (offset) */
- xfs_fileoff_t end; /* end of mapped file region */
- xfs_extnum_t lastx; /* last useful extent number */
- int eof; /* we've hit the end of extents */
- int n = 0; /* current extent index */
- int error = 0;
+ xfs_mount_t *mp; /* mount point structure */
+ xfs_alloctype_t atype = 0; /* type for allocation routines */
+ xfs_extlen_t align; /* minimum allocation alignment */
+ xfs_agnumber_t fb_agno; /* ag number of ap->firstblock */
+ xfs_agnumber_t ag;
+ xfs_alloc_arg_t args;
+ xfs_extlen_t blen;
+ xfs_extlen_t nextminlen = 0;
+ int nullfb; /* true if ap->firstblock isn't set */
+ int isaligned;
+ int tryagain;
+ int error;
- ASSERT(*nmap >= 1);
- ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
- ASSERT(!(flags & ~XFS_BMAPI_ENTIRE));
+ ASSERT(ap->length);
- if (unlikely(XFS_TEST_ERROR(
- (XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_EXTENTS &&
- XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_BTREE),
- mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
- XFS_ERROR_REPORT("xfs_bmapi_delay", XFS_ERRLEVEL_LOW, mp);
- return XFS_ERROR(EFSCORRUPTED);
+ mp = ap->ip->i_mount;
+ align = ap->userdata ? xfs_get_extsz_hint(ap->ip) : 0;
+ if (unlikely(align)) {
+ error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
+ align, 0, ap->eof, 0, ap->conv,
+ &ap->offset, &ap->length);
+ ASSERT(!error);
+ ASSERT(ap->length);
}
+ nullfb = *ap->firstblock == NULLFSBLOCK;
+ fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
+ if (nullfb) {
+ if (ap->userdata && xfs_inode_is_filestream(ap->ip)) {
+ ag = xfs_filestream_lookup_ag(ap->ip);
+ ag = (ag != NULLAGNUMBER) ? ag : 0;
+ ap->blkno = XFS_AGB_TO_FSB(mp, ag, 0);
+ } else {
+ ap->blkno = XFS_INO_TO_FSB(mp, ap->ip->i_ino);
+ }
+ } else
+ ap->blkno = *ap->firstblock;
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(EIO);
+ xfs_bmap_adjacent(ap);
- XFS_STATS_INC(xs_blk_mapw);
+ /*
+ * If allowed, use ap->blkno; otherwise must use firstblock since
+ * it's in the right allocation group.
+ */
+ if (nullfb || XFS_FSB_TO_AGNO(mp, ap->blkno) == fb_agno)
+ ;
+ else
+ ap->blkno = *ap->firstblock;
+ /*
+ * Normal allocation, done through xfs_alloc_vextent.
+ */
+ tryagain = isaligned = 0;
+ memset(&args, 0, sizeof(args));
+ args.tp = ap->tp;
+ args.mp = mp;
+ args.fsbno = ap->blkno;
- if (!(ifp->if_flags & XFS_IFEXTENTS)) {
- error = xfs_iread_extents(NULL, ip, XFS_DATA_FORK);
+ /* Trim the allocation back to the maximum an AG can fit. */
+ args.maxlen = MIN(ap->length, XFS_ALLOC_AG_MAX_USABLE(mp));
+ args.firstblock = *ap->firstblock;
+ blen = 0;
+ if (nullfb) {
+ error = xfs_bmap_btalloc_nullfb(ap, &args, &blen);
if (error)
return error;
+ } else if (ap->flist->xbf_low) {
+ if (xfs_inode_is_filestream(ap->ip))
+ args.type = XFS_ALLOCTYPE_FIRST_AG;
+ else
+ args.type = XFS_ALLOCTYPE_START_BNO;
+ args.total = args.minlen = ap->minlen;
+ } else {
+ args.type = XFS_ALLOCTYPE_NEAR_BNO;
+ args.total = ap->total;
+ args.minlen = ap->minlen;
}
-
- xfs_bmap_search_extents(ip, bno, XFS_DATA_FORK, &eof, &lastx, &got, &prev);
- end = bno + len;
- obno = bno;
-
- while (bno < end && n < *nmap) {
- if (eof || got.br_startoff > bno) {
- error = xfs_bmapi_reserve_delalloc(ip, bno, len, &got,
- &prev, &lastx, eof);
- if (error) {
- if (n == 0) {
- *nmap = 0;
- return error;
- }
- break;
- }
- }
-
- /* set up the extent map to return. */
- xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
- xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
-
- /* If we're done, stop now. */
- if (bno >= end || n >= *nmap)
- break;
-
- /* Else go on to the next record. */
- prev = got;
- if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
- else
- eof = 1;
+ /* apply extent size hints if obtained earlier */
+ if (unlikely(align)) {
+ args.prod = align;
+ if ((args.mod = (xfs_extlen_t)do_mod(ap->offset, args.prod)))
+ args.mod = (xfs_extlen_t)(args.prod - args.mod);
+ } else if (mp->m_sb.sb_blocksize >= PAGE_CACHE_SIZE) {
+ args.prod = 1;
+ args.mod = 0;
+ } else {
+ args.prod = PAGE_CACHE_SIZE >> mp->m_sb.sb_blocklog;
+ if ((args.mod = (xfs_extlen_t)(do_mod(ap->offset, args.prod))))
+ args.mod = (xfs_extlen_t)(args.prod - args.mod);
}
-
- *nmap = n;
- return 0;
-}
-
-
-STATIC int
-__xfs_bmapi_allocate(
- struct xfs_bmalloca *bma)
-{
- struct xfs_mount *mp = bma->ip->i_mount;
- int whichfork = (bma->flags & XFS_BMAPI_ATTRFORK) ?
- XFS_ATTR_FORK : XFS_DATA_FORK;
- struct xfs_ifork *ifp = XFS_IFORK_PTR(bma->ip, whichfork);
- int tmp_logflags = 0;
- int error;
- int rt;
-
- ASSERT(bma->length > 0);
-
- rt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(bma->ip);
-
/*
- * For the wasdelay case, we could also just allocate the stuff asked
- * for in this bmap call but that wouldn't be as good.
+ * If we are not low on available data blocks, and the
+ * underlying logical volume manager is a stripe, and
+ * the file offset is zero then try to allocate data
+ * blocks on stripe unit boundary.
+ * NOTE: ap->aeof is only set if the allocation length
+ * is >= the stripe unit and the allocation offset is
+ * at the end of file.
*/
- if (bma->wasdel) {
- bma->length = (xfs_extlen_t)bma->got.br_blockcount;
- bma->offset = bma->got.br_startoff;
- if (bma->idx != NULLEXTNUM && bma->idx) {
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1),
- &bma->prev);
+ if (!ap->flist->xbf_low && ap->aeof) {
+ if (!ap->offset) {
+ args.alignment = mp->m_dalign;
+ atype = args.type;
+ isaligned = 1;
+ /*
+ * Adjust for alignment
+ */
+ if (blen > args.alignment && blen <= args.maxlen)
+ args.minlen = blen - args.alignment;
+ args.minalignslop = 0;
+ } else {
+ /*
+ * First try an exact bno allocation.
+ * If it fails then do a near or start bno
+ * allocation with alignment turned on.
+ */
+ atype = args.type;
+ tryagain = 1;
+ args.type = XFS_ALLOCTYPE_THIS_BNO;
+ args.alignment = 1;
+ /*
+ * Compute the minlen+alignment for the
+ * next case. Set slop so that the value
+ * of minlen+alignment+slop doesn't go up
+ * between the calls.
+ */
+ if (blen > mp->m_dalign && blen <= args.maxlen)
+ nextminlen = blen - mp->m_dalign;
+ else
+ nextminlen = args.minlen;
+ if (nextminlen + mp->m_dalign > args.minlen + 1)
+ args.minalignslop =
+ nextminlen + mp->m_dalign -
+ args.minlen - 1;
+ else
+ args.minalignslop = 0;
}
} else {
- bma->length = XFS_FILBLKS_MIN(bma->length, MAXEXTLEN);
- if (!bma->eof)
- bma->length = XFS_FILBLKS_MIN(bma->length,
- bma->got.br_startoff - bma->offset);
+ args.alignment = 1;
+ args.minalignslop = 0;
}
-
- /*
- * Indicate if this is the first user data in the file, or just any
- * user data.
- */
- if (!(bma->flags & XFS_BMAPI_METADATA)) {
- bma->userdata = (bma->offset == 0) ?
- XFS_ALLOC_INITIAL_USER_DATA : XFS_ALLOC_USERDATA;
+ args.minleft = ap->minleft;
+ args.wasdel = ap->wasdel;
+ args.isfl = 0;
+ args.userdata = ap->userdata;
+ if ((error = xfs_alloc_vextent(&args)))
+ return error;
+ if (tryagain && args.fsbno == NULLFSBLOCK) {
+ /*
+ * Exact allocation failed. Now try with alignment
+ * turned on.
+ */
+ args.type = atype;
+ args.fsbno = ap->blkno;
+ args.alignment = mp->m_dalign;
+ args.minlen = nextminlen;
+ args.minalignslop = 0;
+ isaligned = 1;
+ if ((error = xfs_alloc_vextent(&args)))
+ return error;
}
-
- bma->minlen = (bma->flags & XFS_BMAPI_CONTIG) ? bma->length : 1;
-
- /*
- * Only want to do the alignment at the eof if it is userdata and
- * allocation length is larger than a stripe unit.
- */
- if (mp->m_dalign && bma->length >= mp->m_dalign &&
- !(bma->flags & XFS_BMAPI_METADATA) && whichfork == XFS_DATA_FORK) {
- error = xfs_bmap_isaeof(bma, whichfork);
- if (error)
+ if (isaligned && args.fsbno == NULLFSBLOCK) {
+ /*
+ * allocation failed, so turn off alignment and
+ * try again.
+ */
+ args.type = atype;
+ args.fsbno = ap->blkno;
+ args.alignment = 0;
+ if ((error = xfs_alloc_vextent(&args)))
return error;
}
-
- error = xfs_bmap_alloc(bma);
- if (error)
- return error;
-
- if (bma->flist->xbf_low)
- bma->minleft = 0;
- if (bma->cur)
- bma->cur->bc_private.b.firstblock = *bma->firstblock;
- if (bma->blkno == NULLFSBLOCK)
- return 0;
- if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
- bma->cur = xfs_bmbt_init_cursor(mp, bma->tp, bma->ip, whichfork);
- bma->cur->bc_private.b.firstblock = *bma->firstblock;
- bma->cur->bc_private.b.flist = bma->flist;
+ if (args.fsbno == NULLFSBLOCK && nullfb &&
+ args.minlen > ap->minlen) {
+ args.minlen = ap->minlen;
+ args.type = XFS_ALLOCTYPE_START_BNO;
+ args.fsbno = ap->blkno;
+ if ((error = xfs_alloc_vextent(&args)))
+ return error;
}
- /*
- * Bump the number of extents we've allocated
- * in this call.
- */
- bma->nallocs++;
-
- if (bma->cur)
- bma->cur->bc_private.b.flags =
- bma->wasdel ? XFS_BTCUR_BPRV_WASDEL : 0;
-
- bma->got.br_startoff = bma->offset;
- bma->got.br_startblock = bma->blkno;
- bma->got.br_blockcount = bma->length;
- bma->got.br_state = XFS_EXT_NORM;
+ if (args.fsbno == NULLFSBLOCK && nullfb) {
+ args.fsbno = 0;
+ args.type = XFS_ALLOCTYPE_FIRST_AG;
+ args.total = ap->minlen;
+ args.minleft = 0;
+ if ((error = xfs_alloc_vextent(&args)))
+ return error;
+ ap->flist->xbf_low = 1;
+ }
+ if (args.fsbno != NULLFSBLOCK) {
+ /*
+ * check the allocation happened at the same or higher AG than
+ * the first block that was allocated.
+ */
+ ASSERT(*ap->firstblock == NULLFSBLOCK ||
+ XFS_FSB_TO_AGNO(mp, *ap->firstblock) ==
+ XFS_FSB_TO_AGNO(mp, args.fsbno) ||
+ (ap->flist->xbf_low &&
+ XFS_FSB_TO_AGNO(mp, *ap->firstblock) <
+ XFS_FSB_TO_AGNO(mp, args.fsbno)));
- /*
- * A wasdelay extent has been initialized, so shouldn't be flagged
- * as unwritten.
- */
- if (!bma->wasdel && (bma->flags & XFS_BMAPI_PREALLOC) &&
- xfs_sb_version_hasextflgbit(&mp->m_sb))
- bma->got.br_state = XFS_EXT_UNWRITTEN;
+ ap->blkno = args.fsbno;
+ if (*ap->firstblock == NULLFSBLOCK)
+ *ap->firstblock = args.fsbno;
+ ASSERT(nullfb || fb_agno == args.agno ||
+ (ap->flist->xbf_low && fb_agno < args.agno));
+ ap->length = args.len;
+ ap->ip->i_d.di_nblocks += args.len;
+ xfs_trans_log_inode(ap->tp, ap->ip, XFS_ILOG_CORE);
+ if (ap->wasdel)
+ ap->ip->i_delayed_blks -= args.len;
+ /*
+ * Adjust the disk quota also. This was reserved
+ * earlier.
+ */
+ xfs_trans_mod_dquot_byino(ap->tp, ap->ip,
+ ap->wasdel ? XFS_TRANS_DQ_DELBCOUNT :
+ XFS_TRANS_DQ_BCOUNT,
+ (long) args.len);
+ } else {
+ ap->blkno = NULLFSBLOCK;
+ ap->length = 0;
+ }
+ return 0;
+}
- if (bma->wasdel)
- error = xfs_bmap_add_extent_delay_real(bma);
- else
- error = xfs_bmap_add_extent_hole_real(bma, whichfork);
+/*
+ * xfs_bmap_alloc is called by xfs_bmapi to allocate an extent for a file.
+ * It figures out where to ask the underlying allocator to put the new extent.
+ */
+STATIC int
+xfs_bmap_alloc(
+ xfs_bmalloca_t *ap) /* bmap alloc argument struct */
+{
+ if (XFS_IS_REALTIME_INODE(ap->ip) && ap->userdata)
+ return xfs_bmap_rtalloc(ap);
+ return xfs_bmap_btalloc(ap);
+}
- bma->logflags |= tmp_logflags;
- if (error)
- return error;
+/*
+ * Trim the returned map to the required bounds
+ */
+STATIC void
+xfs_bmapi_trim_map(
+ struct xfs_bmbt_irec *mval,
+ struct xfs_bmbt_irec *got,
+ xfs_fileoff_t *bno,
+ xfs_filblks_t len,
+ xfs_fileoff_t obno,
+ xfs_fileoff_t end,
+ int n,
+ int flags)
+{
+ if ((flags & XFS_BMAPI_ENTIRE) ||
+ got->br_startoff + got->br_blockcount <= obno) {
+ *mval = *got;
+ if (isnullstartblock(got->br_startblock))
+ mval->br_startblock = DELAYSTARTBLOCK;
+ return;
+ }
+ if (obno > *bno)
+ *bno = obno;
+ ASSERT((*bno >= obno) || (n == 0));
+ ASSERT(*bno < end);
+ mval->br_startoff = *bno;
+ if (isnullstartblock(got->br_startblock))
+ mval->br_startblock = DELAYSTARTBLOCK;
+ else
+ mval->br_startblock = got->br_startblock +
+ (*bno - got->br_startoff);
/*
- * Update our extent pointer, given that xfs_bmap_add_extent_delay_real
- * or xfs_bmap_add_extent_hole_real might have merged it into one of
- * the neighbouring ones.
+ * Return the minimum of what we got and what we asked for for
+ * the length. We can use the len variable here because it is
+ * modified below and we could have been there before coming
+ * here if the first part of the allocation didn't overlap what
+ * was asked for.
*/
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);
-
- ASSERT(bma->got.br_startoff <= bma->offset);
- ASSERT(bma->got.br_startoff + bma->got.br_blockcount >=
- bma->offset + bma->length);
- ASSERT(bma->got.br_state == XFS_EXT_NORM ||
- bma->got.br_state == XFS_EXT_UNWRITTEN);
- return 0;
+ mval->br_blockcount = XFS_FILBLKS_MIN(end - *bno,
+ got->br_blockcount - (*bno - got->br_startoff));
+ mval->br_state = got->br_state;
+ ASSERT(mval->br_blockcount <= len);
+ return;
}
-static void
-xfs_bmapi_allocate_worker(
- struct work_struct *work)
+/*
+ * Update and validate the extent map to return
+ */
+STATIC void
+xfs_bmapi_update_map(
+ struct xfs_bmbt_irec **map,
+ xfs_fileoff_t *bno,
+ xfs_filblks_t *len,
+ xfs_fileoff_t obno,
+ xfs_fileoff_t end,
+ int *n,
+ int flags)
{
- struct xfs_bmalloca *args = container_of(work,
- struct xfs_bmalloca, work);
- unsigned long pflags;
-
- /* we are in a transaction context here */
- current_set_flags_nested(&pflags, PF_FSTRANS);
+ xfs_bmbt_irec_t *mval = *map;
- args->result = __xfs_bmapi_allocate(args);
- complete(args->done);
+ ASSERT((flags & XFS_BMAPI_ENTIRE) ||
+ ((mval->br_startoff + mval->br_blockcount) <= end));
+ ASSERT((flags & XFS_BMAPI_ENTIRE) || (mval->br_blockcount <= *len) ||
+ (mval->br_startoff < obno));
- current_restore_flags_nested(&pflags, PF_FSTRANS);
+ *bno = mval->br_startoff + mval->br_blockcount;
+ *len = end - *bno;
+ if (*n > 0 && mval->br_startoff == mval[-1].br_startoff) {
+ /* update previous map with new information */
+ ASSERT(mval->br_startblock == mval[-1].br_startblock);
+ ASSERT(mval->br_blockcount > mval[-1].br_blockcount);
+ ASSERT(mval->br_state == mval[-1].br_state);
+ mval[-1].br_blockcount = mval->br_blockcount;
+ mval[-1].br_state = mval->br_state;
+ } else if (*n > 0 && mval->br_startblock != DELAYSTARTBLOCK &&
+ mval[-1].br_startblock != DELAYSTARTBLOCK &&
+ mval[-1].br_startblock != HOLESTARTBLOCK &&
+ mval->br_startblock == mval[-1].br_startblock +
+ mval[-1].br_blockcount &&
+ ((flags & XFS_BMAPI_IGSTATE) ||
+ mval[-1].br_state == mval->br_state)) {
+ ASSERT(mval->br_startoff ==
+ mval[-1].br_startoff + mval[-1].br_blockcount);
+ mval[-1].br_blockcount += mval->br_blockcount;
+ } else if (*n > 0 &&
+ mval->br_startblock == DELAYSTARTBLOCK &&
+ mval[-1].br_startblock == DELAYSTARTBLOCK &&
+ mval->br_startoff ==
+ mval[-1].br_startoff + mval[-1].br_blockcount) {
+ mval[-1].br_blockcount += mval->br_blockcount;
+ mval[-1].br_state = mval->br_state;
+ } else if (!((*n == 0) &&
+ ((mval->br_startoff + mval->br_blockcount) <=
+ obno))) {
+ mval++;
+ (*n)++;
+ }
+ *map = mval;
}
/*
- * Some allocation requests often come in with little stack to work on. Push
- * them off to a worker thread so there is lots of stack to use. Otherwise just
- * call directly to avoid the context switch overhead here.
+ * Map file blocks to filesystem blocks without allocation.
*/
int
-xfs_bmapi_allocate(
- struct xfs_bmalloca *args)
-{
- DECLARE_COMPLETION_ONSTACK(done);
-
- if (!args->stack_switch)
- return __xfs_bmapi_allocate(args);
-
-
- args->done = &done;
- INIT_WORK_ONSTACK(&args->work, xfs_bmapi_allocate_worker);
- queue_work(xfs_alloc_wq, &args->work);
- wait_for_completion(&done);
- return args->result;
-}
-
-STATIC int
-xfs_bmapi_convert_unwritten(
- struct xfs_bmalloca *bma,
- struct xfs_bmbt_irec *mval,
+xfs_bmapi_read(
+ struct xfs_inode *ip,
+ xfs_fileoff_t bno,
xfs_filblks_t len,
+ struct xfs_bmbt_irec *mval,
+ int *nmap,
int flags)
{
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_ifork *ifp;
+ struct xfs_bmbt_irec got;
+ struct xfs_bmbt_irec prev;
+ xfs_fileoff_t obno;
+ xfs_fileoff_t end;
+ xfs_extnum_t lastx;
+ int error;
+ int eof;
+ int n = 0;
int whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
XFS_ATTR_FORK : XFS_DATA_FORK;
- struct xfs_ifork *ifp = XFS_IFORK_PTR(bma->ip, whichfork);
- int tmp_logflags = 0;
- int error;
- /* check if we need to do unwritten->real conversion */
- if (mval->br_state == XFS_EXT_UNWRITTEN &&
- (flags & XFS_BMAPI_PREALLOC))
- return 0;
+ ASSERT(*nmap >= 1);
+ ASSERT(!(flags & ~(XFS_BMAPI_ATTRFORK|XFS_BMAPI_ENTIRE|
+ XFS_BMAPI_IGSTATE)));
- /* check if we need to do real->unwritten conversion */
- if (mval->br_state == XFS_EXT_NORM &&
- (flags & (XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT)) !=
- (XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT))
- return 0;
+ if (unlikely(XFS_TEST_ERROR(
+ (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE),
+ mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
+ XFS_ERROR_REPORT("xfs_bmapi_read", XFS_ERRLEVEL_LOW, mp);
+ return XFS_ERROR(EFSCORRUPTED);
+ }
- /*
- * Modify (by adding) the state flag, if writing.
- */
- ASSERT(mval->br_blockcount <= len);
- if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
- bma->cur = xfs_bmbt_init_cursor(bma->ip->i_mount, bma->tp,
- bma->ip, whichfork);
- bma->cur->bc_private.b.firstblock = *bma->firstblock;
- bma->cur->bc_private.b.flist = bma->flist;
+ if (XFS_FORCED_SHUTDOWN(mp))
+ return XFS_ERROR(EIO);
+
+ XFS_STATS_INC(xs_blk_mapr);
+
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+
+ if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+ error = xfs_iread_extents(NULL, ip, whichfork);
+ if (error)
+ return error;
}
- mval->br_state = (mval->br_state == XFS_EXT_UNWRITTEN)
- ? XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
- error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, &bma->idx,
- &bma->cur, mval, bma->firstblock, bma->flist,
- &tmp_logflags);
- bma->logflags |= tmp_logflags;
- if (error)
- return error;
+ xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got, &prev);
+ end = bno + len;
+ obno = bno;
- /*
- * Update our extent pointer, given that
- * xfs_bmap_add_extent_unwritten_real might have merged it into one
- * of the neighbouring ones.
- */
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);
+ while (bno < end && n < *nmap) {
+ /* Reading past eof, act as though there's a hole up to end. */
+ if (eof)
+ got.br_startoff = end;
+ if (got.br_startoff > bno) {
+ /* Reading in a hole. */
+ mval->br_startoff = bno;
+ mval->br_startblock = HOLESTARTBLOCK;
+ mval->br_blockcount =
+ XFS_FILBLKS_MIN(len, got.br_startoff - bno);
+ mval->br_state = XFS_EXT_NORM;
+ bno += mval->br_blockcount;
+ len -= mval->br_blockcount;
+ mval++;
+ n++;
+ continue;
+ }
- /*
- * We may have combined previously unwritten space with written space,
- * so generate another request.
- */
- if (mval->br_blockcount < len)
- return EAGAIN;
+ /* set up the extent map to return. */
+ xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
+ xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
+
+ /* If we're done, stop now. */
+ if (bno >= end || n >= *nmap)
+ break;
+
+ /* Else go on to the next record. */
+ if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
+ else
+ eof = 1;
+ }
+ *nmap = n;
return 0;
}
-/*
- * Map file blocks to filesystem blocks, and allocate blocks or convert the
- * extent state if necessary. Details behaviour is controlled by the flags
- * parameter. Only allocates blocks from a single allocation group, to avoid
- * locking problems.
- *
- * The returned value in "firstblock" from the first call in a transaction
- * must be remembered and presented to subsequent calls in "firstblock".
- * An upper bound for the number of blocks to be allocated is supplied to
- * the first call in "total"; if no allocation group has that many free
- * blocks then the call will fail (return NULLFSBLOCK in "firstblock").
- */
-int
-xfs_bmapi_write(
- struct xfs_trans *tp, /* transaction pointer */
- struct xfs_inode *ip, /* incore inode */
- xfs_fileoff_t bno, /* starting file offs. mapped */
- xfs_filblks_t len, /* length to map in file */
- int flags, /* XFS_BMAPI_... */
- xfs_fsblock_t *firstblock, /* first allocated block
- controls a.g. for allocs */
- xfs_extlen_t total, /* total blocks needed */
- struct xfs_bmbt_irec *mval, /* output: map values */
- int *nmap, /* i/o: mval size/count */
- struct xfs_bmap_free *flist) /* i/o: list extents to free */
+STATIC int
+xfs_bmapi_reserve_delalloc(
+ struct xfs_inode *ip,
+ xfs_fileoff_t aoff,
+ xfs_filblks_t len,
+ struct xfs_bmbt_irec *got,
+ struct xfs_bmbt_irec *prev,
+ xfs_extnum_t *lastx,
+ int eof)
{
struct xfs_mount *mp = ip->i_mount;
- struct xfs_ifork *ifp;
- struct xfs_bmalloca bma = { 0 }; /* args for xfs_bmap_alloc */
- xfs_fileoff_t end; /* end of mapped file region */
- int eof; /* after the end of extents */
- int error; /* error return */
- int n; /* current extent index */
- xfs_fileoff_t obno; /* old block number (offset) */
- int whichfork; /* data or attr fork */
- char inhole; /* current location is hole in file */
- char wasdelay; /* old extent was delayed */
+ struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+ xfs_extlen_t alen;
+ xfs_extlen_t indlen;
+ char rt = XFS_IS_REALTIME_INODE(ip);
+ xfs_extlen_t extsz;
+ int error;
-#ifdef DEBUG
- xfs_fileoff_t orig_bno; /* original block number value */
- int orig_flags; /* original flags arg value */
- xfs_filblks_t orig_len; /* original value of len arg */
- struct xfs_bmbt_irec *orig_mval; /* original value of mval */
- int orig_nmap; /* original value of *nmap */
+ alen = XFS_FILBLKS_MIN(len, MAXEXTLEN);
+ if (!eof)
+ alen = XFS_FILBLKS_MIN(alen, got->br_startoff - aoff);
- orig_bno = bno;
- orig_len = len;
- orig_flags = flags;
- orig_mval = mval;
- orig_nmap = *nmap;
-#endif
+ /* Figure out the extent size, adjust alen */
+ extsz = xfs_get_extsz_hint(ip);
+ if (extsz) {
+ /*
+ * Make sure we don't exceed a single extent length when we
+ * align the extent by reducing length we are going to
+ * allocate by the maximum amount extent size aligment may
+ * require.
+ */
+ alen = XFS_FILBLKS_MIN(len, MAXEXTLEN - (2 * extsz - 1));
+ error = xfs_bmap_extsize_align(mp, got, prev, extsz, rt, eof,
+ 1, 0, &aoff, &alen);
+ ASSERT(!error);
+ }
+
+ if (rt)
+ extsz = alen / mp->m_sb.sb_rextsize;
+
+ /*
+ * Make a transaction-less quota reservation for delayed allocation
+ * blocks. This number gets adjusted later. We return if we haven't
+ * allocated blocks already inside this loop.
+ */
+ error = xfs_trans_reserve_quota_nblks(NULL, ip, (long)alen, 0,
+ rt ? XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
+ if (error)
+ return error;
+
+ /*
+ * Split changing sb for alen and indlen since they could be coming
+ * from different places.
+ */
+ indlen = (xfs_extlen_t)xfs_bmap_worst_indlen(ip, alen);
+ ASSERT(indlen > 0);
+
+ if (rt) {
+ error = xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS,
+ -((int64_t)extsz), 0);
+ } else {
+ error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
+ -((int64_t)alen), 0);
+ }
+
+ if (error)
+ goto out_unreserve_quota;
+
+ error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
+ -((int64_t)indlen), 0);
+ if (error)
+ goto out_unreserve_blocks;
+
+
+ ip->i_delayed_blks += alen;
+
+ got->br_startoff = aoff;
+ got->br_startblock = nullstartblock(indlen);
+ got->br_blockcount = alen;
+ got->br_state = XFS_EXT_NORM;
+ xfs_bmap_add_extent_hole_delay(ip, lastx, got);
+
+ /*
+ * Update our extent pointer, given that xfs_bmap_add_extent_hole_delay
+ * might have merged it into one of the neighbouring ones.
+ */
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *lastx), got);
+
+ ASSERT(got->br_startoff <= aoff);
+ ASSERT(got->br_startoff + got->br_blockcount >= aoff + alen);
+ ASSERT(isnullstartblock(got->br_startblock));
+ ASSERT(got->br_state == XFS_EXT_NORM);
+ return 0;
+
+out_unreserve_blocks:
+ if (rt)
+ xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS, extsz, 0);
+ else
+ xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, alen, 0);
+out_unreserve_quota:
+ if (XFS_IS_QUOTA_ON(mp))
+ xfs_trans_unreserve_quota_nblks(NULL, ip, (long)alen, 0, rt ?
+ XFS_QMOPT_RES_RTBLKS : XFS_QMOPT_RES_REGBLKS);
+ return error;
+}
+
+/*
+ * Map file blocks to filesystem blocks, adding delayed allocations as needed.
+ */
+int
+xfs_bmapi_delay(
+ struct xfs_inode *ip, /* incore inode */
+ xfs_fileoff_t bno, /* starting file offs. mapped */
+ xfs_filblks_t len, /* length to map in file */
+ struct xfs_bmbt_irec *mval, /* output: map values */
+ int *nmap, /* i/o: mval size/count */
+ int flags) /* XFS_BMAPI_... */
+{
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_ifork *ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+ struct xfs_bmbt_irec got; /* current file extent record */
+ struct xfs_bmbt_irec prev; /* previous file extent record */
+ xfs_fileoff_t obno; /* old block number (offset) */
+ xfs_fileoff_t end; /* end of mapped file region */
+ xfs_extnum_t lastx; /* last useful extent number */
+ int eof; /* we've hit the end of extents */
+ int n = 0; /* current extent index */
+ int error = 0;
ASSERT(*nmap >= 1);
ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
- ASSERT(!(flags & XFS_BMAPI_IGSTATE));
- ASSERT(tp != NULL);
- ASSERT(len > 0);
-
- whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
- XFS_ATTR_FORK : XFS_DATA_FORK;
+ ASSERT(!(flags & ~XFS_BMAPI_ENTIRE));
if (unlikely(XFS_TEST_ERROR(
- (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL),
+ (XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_EXTENTS &&
+ XFS_IFORK_FORMAT(ip, XFS_DATA_FORK) != XFS_DINODE_FMT_BTREE),
mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
- XFS_ERROR_REPORT("xfs_bmapi_write", XFS_ERRLEVEL_LOW, mp);
+ XFS_ERROR_REPORT("xfs_bmapi_delay", XFS_ERRLEVEL_LOW, mp);
return XFS_ERROR(EFSCORRUPTED);
}
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
- ifp = XFS_IFORK_PTR(ip, whichfork);
-
XFS_STATS_INC(xs_blk_mapw);
- if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
- /*
- * XXX (dgc): This assumes we are only called for inodes that
- * contain content neutral data in local format. Anything that
- * contains caller-specific data in local format that needs
- * transformation to move to a block format needs to do the
- * conversion to extent format itself.
- *
- * Directory data forks and attribute forks handle this
- * themselves, but with the addition of metadata verifiers every
- * data fork in local format now contains caller specific data
- * and as such conversion through this function is likely to be
- * broken.
- *
- * The only likely user of this branch is for remote symlinks,
- * but we cannot overwrite the data fork contents of the symlink
- * (EEXIST occurs higher up the stack) and so it will never go
- * from local format to extent format here. Hence I don't think
- * this branch is ever executed intentionally and we should
- * consider removing it and asserting that xfs_bmapi_write()
- * cannot be called directly on local format forks. i.e. callers
- * are completely responsible for local to extent format
- * conversion, not xfs_bmapi_write().
- */
- error = xfs_bmap_local_to_extents(tp, ip, firstblock, total,
- &bma.logflags, whichfork,
- xfs_bmap_local_to_extents_init_fn);
- if (error)
- goto error0;
- }
-
- if (*firstblock == NULLFSBLOCK) {
- if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE)
- bma.minleft = be16_to_cpu(ifp->if_broot->bb_level) + 1;
- else
- bma.minleft = 1;
- } else {
- bma.minleft = 0;
- }
-
if (!(ifp->if_flags & XFS_IFEXTENTS)) {
- error = xfs_iread_extents(tp, ip, whichfork);
+ error = xfs_iread_extents(NULL, ip, XFS_DATA_FORK);
if (error)
- goto error0;
+ return error;
}
- xfs_bmap_search_extents(ip, bno, whichfork, &eof, &bma.idx, &bma.got,
- &bma.prev);
- n = 0;
+ xfs_bmap_search_extents(ip, bno, XFS_DATA_FORK, &eof, &lastx, &got, &prev);
end = bno + len;
obno = bno;
- bma.tp = tp;
- bma.ip = ip;
- bma.total = total;
- bma.userdata = 0;
- bma.flist = flist;
- bma.firstblock = firstblock;
-
- if (flags & XFS_BMAPI_STACK_SWITCH)
- bma.stack_switch = 1;
-
while (bno < end && n < *nmap) {
- inhole = eof || bma.got.br_startoff > bno;
- wasdelay = !inhole && isnullstartblock(bma.got.br_startblock);
+ if (eof || got.br_startoff > bno) {
+ error = xfs_bmapi_reserve_delalloc(ip, bno, len, &got,
+ &prev, &lastx, eof);
+ if (error) {
+ if (n == 0) {
+ *nmap = 0;
+ return error;
+ }
+ break;
+ }
+ }
- /*
- * First, deal with the hole before the allocated space
- * that we found, if any.
- */
- if (inhole || wasdelay) {
- bma.eof = eof;
- bma.conv = !!(flags & XFS_BMAPI_CONVERT);
- bma.wasdel = wasdelay;
- bma.offset = bno;
- bma.flags = flags;
-
- /*
- * There's a 32/64 bit type mismatch between the
- * allocation length request (which can be 64 bits in
- * length) and the bma length request, which is
- * xfs_extlen_t and therefore 32 bits. Hence we have to
- * check for 32-bit overflows and handle them here.
- */
- if (len > (xfs_filblks_t)MAXEXTLEN)
- bma.length = MAXEXTLEN;
- else
- bma.length = len;
-
- ASSERT(len > 0);
- ASSERT(bma.length > 0);
- error = xfs_bmapi_allocate(&bma);
- if (error)
- goto error0;
- if (bma.blkno == NULLFSBLOCK)
- break;
- }
-
- /* Deal with the allocated space we found. */
- xfs_bmapi_trim_map(mval, &bma.got, &bno, len, obno,
- end, n, flags);
-
- /* Execute unwritten extent conversion if necessary */
- error = xfs_bmapi_convert_unwritten(&bma, mval, len, flags);
- if (error == EAGAIN)
- continue;
- if (error)
- goto error0;
-
- /* update the extent map to return */
+ /* set up the extent map to return. */
+ xfs_bmapi_trim_map(mval, &got, &bno, len, obno, end, n, flags);
xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
- /*
- * If we're done, stop now. Stop when we've allocated
- * XFS_BMAP_MAX_NMAP extents no matter what. Otherwise
- * the transaction may get too big.
- */
- if (bno >= end || n >= *nmap || bma.nallocs >= *nmap)
+ /* If we're done, stop now. */
+ if (bno >= end || n >= *nmap)
break;
/* Else go on to the next record. */
- bma.prev = bma.got;
- if (++bma.idx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t)) {
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma.idx),
- &bma.got);
- } else
+ prev = got;
+ if (++lastx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t))
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, lastx), &got);
+ else
eof = 1;
}
+
*nmap = n;
+ return 0;
+}
- /*
- * Transform from btree to extents, give it cur.
- */
- if (xfs_bmap_wants_extents(ip, whichfork)) {
- int tmp_logflags = 0;
- ASSERT(bma.cur);
- error = xfs_bmap_btree_to_extents(tp, ip, bma.cur,
- &tmp_logflags, whichfork);
- bma.logflags |= tmp_logflags;
- if (error)
- goto error0;
- }
+STATIC int
+__xfs_bmapi_allocate(
+ struct xfs_bmalloca *bma)
+{
+ struct xfs_mount *mp = bma->ip->i_mount;
+ int whichfork = (bma->flags & XFS_BMAPI_ATTRFORK) ?
+ XFS_ATTR_FORK : XFS_DATA_FORK;
+ struct xfs_ifork *ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+ int tmp_logflags = 0;
+ int error;
+ int rt;
+
+ ASSERT(bma->length > 0);
+
+ rt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(bma->ip);
- ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE ||
- XFS_IFORK_NEXTENTS(ip, whichfork) >
- XFS_IFORK_MAXEXT(ip, whichfork));
- error = 0;
-error0:
- /*
- * Log everything. Do this after conversion, there's no point in
- * logging the extent records if we've converted to btree format.
- */
- if ((bma.logflags & xfs_ilog_fext(whichfork)) &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
- bma.logflags &= ~xfs_ilog_fext(whichfork);
- else if ((bma.logflags & xfs_ilog_fbroot(whichfork)) &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)
- bma.logflags &= ~xfs_ilog_fbroot(whichfork);
/*
- * Log whatever the flags say, even if error. Otherwise we might miss
- * detecting a case where the data is changed, there's an error,
- * and it's not logged so we don't shutdown when we should.
+ * For the wasdelay case, we could also just allocate the stuff asked
+ * for in this bmap call but that wouldn't be as good.
*/
- if (bma.logflags)
- xfs_trans_log_inode(tp, ip, bma.logflags);
-
- if (bma.cur) {
- if (!error) {
- ASSERT(*firstblock == NULLFSBLOCK ||
- XFS_FSB_TO_AGNO(mp, *firstblock) ==
- XFS_FSB_TO_AGNO(mp,
- bma.cur->bc_private.b.firstblock) ||
- (flist->xbf_low &&
- XFS_FSB_TO_AGNO(mp, *firstblock) <
- XFS_FSB_TO_AGNO(mp,
- bma.cur->bc_private.b.firstblock)));
- *firstblock = bma.cur->bc_private.b.firstblock;
+ if (bma->wasdel) {
+ bma->length = (xfs_extlen_t)bma->got.br_blockcount;
+ bma->offset = bma->got.br_startoff;
+ if (bma->idx != NULLEXTNUM && bma->idx) {
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx - 1),
+ &bma->prev);
}
- xfs_btree_del_cursor(bma.cur,
- error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+ } else {
+ bma->length = XFS_FILBLKS_MIN(bma->length, MAXEXTLEN);
+ if (!bma->eof)
+ bma->length = XFS_FILBLKS_MIN(bma->length,
+ bma->got.br_startoff - bma->offset);
}
- if (!error)
- xfs_bmap_validate_ret(orig_bno, orig_len, orig_flags, orig_mval,
- orig_nmap, *nmap);
- return error;
-}
-/*
- * Unmap (remove) blocks from a file.
- * If nexts is nonzero then the number of extents to remove is limited to
- * that value. If not all extents in the block range can be removed then
- * *done is set.
- */
-int /* error */
-xfs_bunmapi(
- xfs_trans_t *tp, /* transaction pointer */
- struct xfs_inode *ip, /* incore inode */
- xfs_fileoff_t bno, /* starting offset to unmap */
- xfs_filblks_t len, /* length to unmap in file */
- int flags, /* misc flags */
- xfs_extnum_t nexts, /* number of extents max */
- xfs_fsblock_t *firstblock, /* first allocated block
- controls a.g. for allocs */
- xfs_bmap_free_t *flist, /* i/o: list extents to free */
- int *done) /* set if not done yet */
-{
- xfs_btree_cur_t *cur; /* bmap btree cursor */
- xfs_bmbt_irec_t del; /* extent being deleted */
- int eof; /* is deleting at eof */
- xfs_bmbt_rec_host_t *ep; /* extent record pointer */
- int error; /* error return value */
- xfs_extnum_t extno; /* extent number in list */
- xfs_bmbt_irec_t got; /* current extent record */
- xfs_ifork_t *ifp; /* inode fork pointer */
- int isrt; /* freeing in rt area */
- xfs_extnum_t lastx; /* last extent index used */
- int logflags; /* transaction logging flags */
- xfs_extlen_t mod; /* rt extent offset */
- xfs_mount_t *mp; /* mount structure */
- xfs_extnum_t nextents; /* number of file extents */
- xfs_bmbt_irec_t prev; /* previous extent record */
- xfs_fileoff_t start; /* first file offset deleted */
- int tmp_logflags; /* partial logging flags */
- int wasdel; /* was a delayed alloc extent */
- int whichfork; /* data or attribute fork */
- xfs_fsblock_t sum;
+ /*
+ * Indicate if this is the first user data in the file, or just any
+ * user data.
+ */
+ if (!(bma->flags & XFS_BMAPI_METADATA)) {
+ bma->userdata = (bma->offset == 0) ?
+ XFS_ALLOC_INITIAL_USER_DATA : XFS_ALLOC_USERDATA;
+ }
- trace_xfs_bunmap(ip, bno, len, flags, _RET_IP_);
+ bma->minlen = (bma->flags & XFS_BMAPI_CONTIG) ? bma->length : 1;
- whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
- XFS_ATTR_FORK : XFS_DATA_FORK;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- if (unlikely(
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)) {
- XFS_ERROR_REPORT("xfs_bunmapi", XFS_ERRLEVEL_LOW,
- ip->i_mount);
- return XFS_ERROR(EFSCORRUPTED);
+ /*
+ * Only want to do the alignment at the eof if it is userdata and
+ * allocation length is larger than a stripe unit.
+ */
+ if (mp->m_dalign && bma->length >= mp->m_dalign &&
+ !(bma->flags & XFS_BMAPI_METADATA) && whichfork == XFS_DATA_FORK) {
+ error = xfs_bmap_isaeof(bma, whichfork);
+ if (error)
+ return error;
}
- mp = ip->i_mount;
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(EIO);
-
- ASSERT(len > 0);
- ASSERT(nexts >= 0);
- if (!(ifp->if_flags & XFS_IFEXTENTS) &&
- (error = xfs_iread_extents(tp, ip, whichfork)))
+ error = xfs_bmap_alloc(bma);
+ if (error)
return error;
- nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
- if (nextents == 0) {
- *done = 1;
+
+ if (bma->flist->xbf_low)
+ bma->minleft = 0;
+ if (bma->cur)
+ bma->cur->bc_private.b.firstblock = *bma->firstblock;
+ if (bma->blkno == NULLFSBLOCK)
return 0;
+ if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
+ bma->cur = xfs_bmbt_init_cursor(mp, bma->tp, bma->ip, whichfork);
+ bma->cur->bc_private.b.firstblock = *bma->firstblock;
+ bma->cur->bc_private.b.flist = bma->flist;
}
- XFS_STATS_INC(xs_blk_unmap);
- isrt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(ip);
- start = bno;
- bno = start + len - 1;
- ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
- &prev);
+ /*
+ * Bump the number of extents we've allocated
+ * in this call.
+ */
+ bma->nallocs++;
+
+ if (bma->cur)
+ bma->cur->bc_private.b.flags =
+ bma->wasdel ? XFS_BTCUR_BPRV_WASDEL : 0;
+
+ bma->got.br_startoff = bma->offset;
+ bma->got.br_startblock = bma->blkno;
+ bma->got.br_blockcount = bma->length;
+ bma->got.br_state = XFS_EXT_NORM;
/*
- * Check to see if the given block number is past the end of the
- * file, back up to the last block if so...
+ * A wasdelay extent has been initialized, so shouldn't be flagged
+ * as unwritten.
*/
- if (eof) {
- ep = xfs_iext_get_ext(ifp, --lastx);
- xfs_bmbt_get_all(ep, &got);
- bno = got.br_startoff + got.br_blockcount - 1;
+ if (!bma->wasdel && (bma->flags & XFS_BMAPI_PREALLOC) &&
+ xfs_sb_version_hasextflgbit(&mp->m_sb))
+ bma->got.br_state = XFS_EXT_UNWRITTEN;
+
+ if (bma->wasdel)
+ error = xfs_bmap_add_extent_delay_real(bma);
+ else
+ error = xfs_bmap_add_extent_hole_real(bma, whichfork);
+
+ bma->logflags |= tmp_logflags;
+ if (error)
+ return error;
+
+ /*
+ * Update our extent pointer, given that xfs_bmap_add_extent_delay_real
+ * or xfs_bmap_add_extent_hole_real might have merged it into one of
+ * the neighbouring ones.
+ */
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);
+
+ ASSERT(bma->got.br_startoff <= bma->offset);
+ ASSERT(bma->got.br_startoff + bma->got.br_blockcount >=
+ bma->offset + bma->length);
+ ASSERT(bma->got.br_state == XFS_EXT_NORM ||
+ bma->got.br_state == XFS_EXT_UNWRITTEN);
+ return 0;
+}
+
+static void
+xfs_bmapi_allocate_worker(
+ struct work_struct *work)
+{
+ struct xfs_bmalloca *args = container_of(work,
+ struct xfs_bmalloca, work);
+ unsigned long pflags;
+
+ /* we are in a transaction context here */
+ current_set_flags_nested(&pflags, PF_FSTRANS);
+
+ args->result = __xfs_bmapi_allocate(args);
+ complete(args->done);
+
+ current_restore_flags_nested(&pflags, PF_FSTRANS);
+}
+
+/*
+ * Some allocation requests often come in with little stack to work on. Push
+ * them off to a worker thread so there is lots of stack to use. Otherwise just
+ * call directly to avoid the context switch overhead here.
+ */
+int
+xfs_bmapi_allocate(
+ struct xfs_bmalloca *args)
+{
+ DECLARE_COMPLETION_ONSTACK(done);
+
+ if (!args->stack_switch)
+ return __xfs_bmapi_allocate(args);
+
+
+ args->done = &done;
+ INIT_WORK_ONSTACK(&args->work, xfs_bmapi_allocate_worker);
+ queue_work(xfs_alloc_wq, &args->work);
+ wait_for_completion(&done);
+ return args->result;
+}
+
+STATIC int
+xfs_bmapi_convert_unwritten(
+ struct xfs_bmalloca *bma,
+ struct xfs_bmbt_irec *mval,
+ xfs_filblks_t len,
+ int flags)
+{
+ int whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
+ XFS_ATTR_FORK : XFS_DATA_FORK;
+ struct xfs_ifork *ifp = XFS_IFORK_PTR(bma->ip, whichfork);
+ int tmp_logflags = 0;
+ int error;
+
+ /* check if we need to do unwritten->real conversion */
+ if (mval->br_state == XFS_EXT_UNWRITTEN &&
+ (flags & XFS_BMAPI_PREALLOC))
+ return 0;
+
+ /* check if we need to do real->unwritten conversion */
+ if (mval->br_state == XFS_EXT_NORM &&
+ (flags & (XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT)) !=
+ (XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT))
+ return 0;
+
+ /*
+ * Modify (by adding) the state flag, if writing.
+ */
+ ASSERT(mval->br_blockcount <= len);
+ if ((ifp->if_flags & XFS_IFBROOT) && !bma->cur) {
+ bma->cur = xfs_bmbt_init_cursor(bma->ip->i_mount, bma->tp,
+ bma->ip, whichfork);
+ bma->cur->bc_private.b.firstblock = *bma->firstblock;
+ bma->cur->bc_private.b.flist = bma->flist;
}
- logflags = 0;
- if (ifp->if_flags & XFS_IFBROOT) {
- ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE);
- cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
- cur->bc_private.b.firstblock = *firstblock;
- cur->bc_private.b.flist = flist;
- cur->bc_private.b.flags = 0;
- } else
- cur = NULL;
+ mval->br_state = (mval->br_state == XFS_EXT_UNWRITTEN)
+ ? XFS_EXT_NORM : XFS_EXT_UNWRITTEN;
- if (isrt) {
- /*
- * Synchronize by locking the bitmap inode.
- */
- xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
- xfs_trans_ijoin(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
+ error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, &bma->idx,
+ &bma->cur, mval, bma->firstblock, bma->flist,
+ &tmp_logflags);
+ bma->logflags |= tmp_logflags;
+ if (error)
+ return error;
+
+ /*
+ * Update our extent pointer, given that
+ * xfs_bmap_add_extent_unwritten_real might have merged it into one
+ * of the neighbouring ones.
+ */
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma->idx), &bma->got);
+
+ /*
+ * We may have combined previously unwritten space with written space,
+ * so generate another request.
+ */
+ if (mval->br_blockcount < len)
+ return EAGAIN;
+ return 0;
+}
+
+/*
+ * Map file blocks to filesystem blocks, and allocate blocks or convert the
+ * extent state if necessary. Details behaviour is controlled by the flags
+ * parameter. Only allocates blocks from a single allocation group, to avoid
+ * locking problems.
+ *
+ * The returned value in "firstblock" from the first call in a transaction
+ * must be remembered and presented to subsequent calls in "firstblock".
+ * An upper bound for the number of blocks to be allocated is supplied to
+ * the first call in "total"; if no allocation group has that many free
+ * blocks then the call will fail (return NULLFSBLOCK in "firstblock").
+ */
+int
+xfs_bmapi_write(
+ struct xfs_trans *tp, /* transaction pointer */
+ struct xfs_inode *ip, /* incore inode */
+ xfs_fileoff_t bno, /* starting file offs. mapped */
+ xfs_filblks_t len, /* length to map in file */
+ int flags, /* XFS_BMAPI_... */
+ xfs_fsblock_t *firstblock, /* first allocated block
+ controls a.g. for allocs */
+ xfs_extlen_t total, /* total blocks needed */
+ struct xfs_bmbt_irec *mval, /* output: map values */
+ int *nmap, /* i/o: mval size/count */
+ struct xfs_bmap_free *flist) /* i/o: list extents to free */
+{
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_ifork *ifp;
+ struct xfs_bmalloca bma = { 0 }; /* args for xfs_bmap_alloc */
+ xfs_fileoff_t end; /* end of mapped file region */
+ int eof; /* after the end of extents */
+ int error; /* error return */
+ int n; /* current extent index */
+ xfs_fileoff_t obno; /* old block number (offset) */
+ int whichfork; /* data or attr fork */
+ char inhole; /* current location is hole in file */
+ char wasdelay; /* old extent was delayed */
+
+#ifdef DEBUG
+ xfs_fileoff_t orig_bno; /* original block number value */
+ int orig_flags; /* original flags arg value */
+ xfs_filblks_t orig_len; /* original value of len arg */
+ struct xfs_bmbt_irec *orig_mval; /* original value of mval */
+ int orig_nmap; /* original value of *nmap */
+
+ orig_bno = bno;
+ orig_len = len;
+ orig_flags = flags;
+ orig_mval = mval;
+ orig_nmap = *nmap;
+#endif
+
+ ASSERT(*nmap >= 1);
+ ASSERT(*nmap <= XFS_BMAP_MAX_NMAP);
+ ASSERT(!(flags & XFS_BMAPI_IGSTATE));
+ ASSERT(tp != NULL);
+ ASSERT(len > 0);
+
+ whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
+ XFS_ATTR_FORK : XFS_DATA_FORK;
+
+ if (unlikely(XFS_TEST_ERROR(
+ (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_LOCAL),
+ mp, XFS_ERRTAG_BMAPIFORMAT, XFS_RANDOM_BMAPIFORMAT))) {
+ XFS_ERROR_REPORT("xfs_bmapi_write", XFS_ERRLEVEL_LOW, mp);
+ return XFS_ERROR(EFSCORRUPTED);
}
- extno = 0;
- while (bno != (xfs_fileoff_t)-1 && bno >= start && lastx >= 0 &&
- (nexts == 0 || extno < nexts)) {
- /*
- * Is the found extent after a hole in which bno lives?
- * Just back up to the previous extent, if so.
- */
- if (got.br_startoff > bno) {
- if (--lastx < 0)
- break;
- ep = xfs_iext_get_ext(ifp, lastx);
- xfs_bmbt_get_all(ep, &got);
- }
+ if (XFS_FORCED_SHUTDOWN(mp))
+ return XFS_ERROR(EIO);
+
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+
+ XFS_STATS_INC(xs_blk_mapw);
+
+ if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_LOCAL) {
/*
- * Is the last block of this extent before the range
- * we're supposed to delete? If so, we're done.
+ * XXX (dgc): This assumes we are only called for inodes that
+ * contain content neutral data in local format. Anything that
+ * contains caller-specific data in local format that needs
+ * transformation to move to a block format needs to do the
+ * conversion to extent format itself.
+ *
+ * Directory data forks and attribute forks handle this
+ * themselves, but with the addition of metadata verifiers every
+ * data fork in local format now contains caller specific data
+ * and as such conversion through this function is likely to be
+ * broken.
+ *
+ * The only likely user of this branch is for remote symlinks,
+ * but we cannot overwrite the data fork contents of the symlink
+ * (EEXIST occurs higher up the stack) and so it will never go
+ * from local format to extent format here. Hence I don't think
+ * this branch is ever executed intentionally and we should
+ * consider removing it and asserting that xfs_bmapi_write()
+ * cannot be called directly on local format forks. i.e. callers
+ * are completely responsible for local to extent format
+ * conversion, not xfs_bmapi_write().
*/
- bno = XFS_FILEOFF_MIN(bno,
- got.br_startoff + got.br_blockcount - 1);
- if (bno < start)
- break;
+ error = xfs_bmap_local_to_extents(tp, ip, firstblock, total,
+ &bma.logflags, whichfork,
+ xfs_bmap_local_to_extents_init_fn);
+ if (error)
+ goto error0;
+ }
+
+ if (*firstblock == NULLFSBLOCK) {
+ if (XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE)
+ bma.minleft = be16_to_cpu(ifp->if_broot->bb_level) + 1;
+ else
+ bma.minleft = 1;
+ } else {
+ bma.minleft = 0;
+ }
+
+ if (!(ifp->if_flags & XFS_IFEXTENTS)) {
+ error = xfs_iread_extents(tp, ip, whichfork);
+ if (error)
+ goto error0;
+ }
+
+ xfs_bmap_search_extents(ip, bno, whichfork, &eof, &bma.idx, &bma.got,
+ &bma.prev);
+ n = 0;
+ end = bno + len;
+ obno = bno;
+
+ bma.tp = tp;
+ bma.ip = ip;
+ bma.total = total;
+ bma.userdata = 0;
+ bma.flist = flist;
+ bma.firstblock = firstblock;
+
+ if (flags & XFS_BMAPI_STACK_SWITCH)
+ bma.stack_switch = 1;
+
+ while (bno < end && n < *nmap) {
+ inhole = eof || bma.got.br_startoff > bno;
+ wasdelay = !inhole && isnullstartblock(bma.got.br_startblock);
+
/*
- * Then deal with the (possibly delayed) allocated space
- * we found.
+ * First, deal with the hole before the allocated space
+ * that we found, if any.
*/
- ASSERT(ep != NULL);
- del = got;
- wasdel = isnullstartblock(del.br_startblock);
- if (got.br_startoff < start) {
- del.br_startoff = start;
- del.br_blockcount -= start - got.br_startoff;
- if (!wasdel)
- del.br_startblock += start - got.br_startoff;
- }
- if (del.br_startoff + del.br_blockcount > bno + 1)
- del.br_blockcount = bno + 1 - del.br_startoff;
- sum = del.br_startblock + del.br_blockcount;
- if (isrt &&
- (mod = do_mod(sum, mp->m_sb.sb_rextsize))) {
- /*
- * Realtime extent not lined up at the end.
- * The extent could have been split into written
- * and unwritten pieces, or we could just be
- * unmapping part of it. But we can't really
- * get rid of part of a realtime extent.
- */
- if (del.br_state == XFS_EXT_UNWRITTEN ||
- !xfs_sb_version_hasextflgbit(&mp->m_sb)) {
- /*
- * This piece is unwritten, or we're not
- * using unwritten extents. Skip over it.
- */
- ASSERT(bno >= mod);
- bno -= mod > del.br_blockcount ?
- del.br_blockcount : mod;
- if (bno < got.br_startoff) {
- if (--lastx >= 0)
- xfs_bmbt_get_all(xfs_iext_get_ext(
- ifp, lastx), &got);
- }
- continue;
- }
- /*
- * It's written, turn it unwritten.
- * This is better than zeroing it.
- */
- ASSERT(del.br_state == XFS_EXT_NORM);
- ASSERT(xfs_trans_get_block_res(tp) > 0);
+ if (inhole || wasdelay) {
+ bma.eof = eof;
+ bma.conv = !!(flags & XFS_BMAPI_CONVERT);
+ bma.wasdel = wasdelay;
+ bma.offset = bno;
+ bma.flags = flags;
+
/*
- * If this spans a realtime extent boundary,
- * chop it back to the start of the one we end at.
+ * There's a 32/64 bit type mismatch between the
+ * allocation length request (which can be 64 bits in
+ * length) and the bma length request, which is
+ * xfs_extlen_t and therefore 32 bits. Hence we have to
+ * check for 32-bit overflows and handle them here.
*/
- if (del.br_blockcount > mod) {
- del.br_startoff += del.br_blockcount - mod;
- del.br_startblock += del.br_blockcount - mod;
- del.br_blockcount = mod;
- }
- del.br_state = XFS_EXT_UNWRITTEN;
- error = xfs_bmap_add_extent_unwritten_real(tp, ip,
- &lastx, &cur, &del, firstblock, flist,
- &logflags);
+ if (len > (xfs_filblks_t)MAXEXTLEN)
+ bma.length = MAXEXTLEN;
+ else
+ bma.length = len;
+
+ ASSERT(len > 0);
+ ASSERT(bma.length > 0);
+ error = xfs_bmapi_allocate(&bma);
if (error)
goto error0;
- goto nodelete;
+ if (bma.blkno == NULLFSBLOCK)
+ break;
}
- if (isrt && (mod = do_mod(del.br_startblock, mp->m_sb.sb_rextsize))) {
- /*
- * Realtime extent is lined up at the end but not
- * at the front. We'll get rid of full extents if
- * we can.
- */
- mod = mp->m_sb.sb_rextsize - mod;
- if (del.br_blockcount > mod) {
- del.br_blockcount -= mod;
- del.br_startoff += mod;
- del.br_startblock += mod;
- } else if ((del.br_startoff == start &&
- (del.br_state == XFS_EXT_UNWRITTEN ||
- xfs_trans_get_block_res(tp) == 0)) ||
- !xfs_sb_version_hasextflgbit(&mp->m_sb)) {
- /*
- * Can't make it unwritten. There isn't
- * a full extent here so just skip it.
- */
- ASSERT(bno >= del.br_blockcount);
- bno -= del.br_blockcount;
- if (got.br_startoff > bno) {
- if (--lastx >= 0) {
- ep = xfs_iext_get_ext(ifp,
- lastx);
- xfs_bmbt_get_all(ep, &got);
- }
- }
- continue;
- } else if (del.br_state == XFS_EXT_UNWRITTEN) {
- /*
- * This one is already unwritten.
- * It must have a written left neighbor.
- * Unwrite the killed part of that one and
- * try again.
- */
- ASSERT(lastx > 0);
- xfs_bmbt_get_all(xfs_iext_get_ext(ifp,
- lastx - 1), &prev);
- ASSERT(prev.br_state == XFS_EXT_NORM);
- ASSERT(!isnullstartblock(prev.br_startblock));
- ASSERT(del.br_startblock ==
- prev.br_startblock + prev.br_blockcount);
- if (prev.br_startoff < start) {
- mod = start - prev.br_startoff;
- prev.br_blockcount -= mod;
- prev.br_startblock += mod;
- prev.br_startoff = start;
- }
- prev.br_state = XFS_EXT_UNWRITTEN;
- lastx--;
- error = xfs_bmap_add_extent_unwritten_real(tp,
- ip, &lastx, &cur, &prev,
- firstblock, flist, &logflags);
- if (error)
- goto error0;
- goto nodelete;
- } else {
- ASSERT(del.br_state == XFS_EXT_NORM);
- del.br_state = XFS_EXT_UNWRITTEN;
- error = xfs_bmap_add_extent_unwritten_real(tp,
- ip, &lastx, &cur, &del,
- firstblock, flist, &logflags);
- if (error)
- goto error0;
- goto nodelete;
- }
+
+ /* Deal with the allocated space we found. */
+ xfs_bmapi_trim_map(mval, &bma.got, &bno, len, obno,
+ end, n, flags);
+
+ /* Execute unwritten extent conversion if necessary */
+ error = xfs_bmapi_convert_unwritten(&bma, mval, len, flags);
+ if (error == EAGAIN)
+ continue;
+ if (error)
+ goto error0;
+
+ /* update the extent map to return */
+ xfs_bmapi_update_map(&mval, &bno, &len, obno, end, &n, flags);
+
+ /*
+ * If we're done, stop now. Stop when we've allocated
+ * XFS_BMAP_MAX_NMAP extents no matter what. Otherwise
+ * the transaction may get too big.
+ */
+ if (bno >= end || n >= *nmap || bma.nallocs >= *nmap)
+ break;
+
+ /* Else go on to the next record. */
+ bma.prev = bma.got;
+ if (++bma.idx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t)) {
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp, bma.idx),
+ &bma.got);
+ } else
+ eof = 1;
+ }
+ *nmap = n;
+
+ /*
+ * Transform from btree to extents, give it cur.
+ */
+ if (xfs_bmap_wants_extents(ip, whichfork)) {
+ int tmp_logflags = 0;
+
+ ASSERT(bma.cur);
+ error = xfs_bmap_btree_to_extents(tp, ip, bma.cur,
+ &tmp_logflags, whichfork);
+ bma.logflags |= tmp_logflags;
+ if (error)
+ goto error0;
+ }
+
+ ASSERT(XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE ||
+ XFS_IFORK_NEXTENTS(ip, whichfork) >
+ XFS_IFORK_MAXEXT(ip, whichfork));
+ error = 0;
+error0:
+ /*
+ * Log everything. Do this after conversion, there's no point in
+ * logging the extent records if we've converted to btree format.
+ */
+ if ((bma.logflags & xfs_ilog_fext(whichfork)) &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
+ bma.logflags &= ~xfs_ilog_fext(whichfork);
+ else if ((bma.logflags & xfs_ilog_fbroot(whichfork)) &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)
+ bma.logflags &= ~xfs_ilog_fbroot(whichfork);
+ /*
+ * Log whatever the flags say, even if error. Otherwise we might miss
+ * detecting a case where the data is changed, there's an error,
+ * and it's not logged so we don't shutdown when we should.
+ */
+ if (bma.logflags)
+ xfs_trans_log_inode(tp, ip, bma.logflags);
+
+ if (bma.cur) {
+ if (!error) {
+ ASSERT(*firstblock == NULLFSBLOCK ||
+ XFS_FSB_TO_AGNO(mp, *firstblock) ==
+ XFS_FSB_TO_AGNO(mp,
+ bma.cur->bc_private.b.firstblock) ||
+ (flist->xbf_low &&
+ XFS_FSB_TO_AGNO(mp, *firstblock) <
+ XFS_FSB_TO_AGNO(mp,
+ bma.cur->bc_private.b.firstblock)));
+ *firstblock = bma.cur->bc_private.b.firstblock;
}
- if (wasdel) {
- ASSERT(startblockval(del.br_startblock) > 0);
- /* Update realtime/data freespace, unreserve quota */
- if (isrt) {
- xfs_filblks_t rtexts;
+ xfs_btree_del_cursor(bma.cur,
+ error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+ }
+ if (!error)
+ xfs_bmap_validate_ret(orig_bno, orig_len, orig_flags, orig_mval,
+ orig_nmap, *nmap);
+ return error;
+}
- rtexts = XFS_FSB_TO_B(mp, del.br_blockcount);
- do_div(rtexts, mp->m_sb.sb_rextsize);
- xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS,
- (int64_t)rtexts, 0);
- (void)xfs_trans_reserve_quota_nblks(NULL,
- ip, -((long)del.br_blockcount), 0,
- XFS_QMOPT_RES_RTBLKS);
- } else {
- xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
- (int64_t)del.br_blockcount, 0);
- (void)xfs_trans_reserve_quota_nblks(NULL,
- ip, -((long)del.br_blockcount), 0,
- XFS_QMOPT_RES_REGBLKS);
- }
- ip->i_delayed_blks -= del.br_blockcount;
- if (cur)
- cur->bc_private.b.flags |=
- XFS_BTCUR_BPRV_WASDEL;
- } else if (cur)
- cur->bc_private.b.flags &= ~XFS_BTCUR_BPRV_WASDEL;
+/*
+ * Called by xfs_bmapi to update file extent records and the btree
+ * after removing space (or undoing a delayed allocation).
+ */
+STATIC int /* error */
+xfs_bmap_del_extent(
+ xfs_inode_t *ip, /* incore inode pointer */
+ xfs_trans_t *tp, /* current transaction pointer */
+ xfs_extnum_t *idx, /* extent number to update/delete */
+ xfs_bmap_free_t *flist, /* list of extents to be freed */
+ xfs_btree_cur_t *cur, /* if null, not a btree */
+ xfs_bmbt_irec_t *del, /* data to remove from extents */
+ int *logflagsp, /* inode logging flags */
+ int whichfork) /* data or attr fork */
+{
+ xfs_filblks_t da_new; /* new delay-alloc indirect blocks */
+ xfs_filblks_t da_old; /* old delay-alloc indirect blocks */
+ xfs_fsblock_t del_endblock=0; /* first block past del */
+ xfs_fileoff_t del_endoff; /* first offset past del */
+ int delay; /* current block is delayed allocated */
+ int do_fx; /* free extent at end of routine */
+ xfs_bmbt_rec_host_t *ep; /* current extent entry pointer */
+ int error; /* error return value */
+ int flags; /* inode logging flags */
+ xfs_bmbt_irec_t got; /* current extent entry */
+ xfs_fileoff_t got_endoff; /* first offset past got */
+ int i; /* temp state */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_mount_t *mp; /* mount structure */
+ xfs_filblks_t nblks; /* quota/sb block count */
+ xfs_bmbt_irec_t new; /* new record to be inserted */
+ /* REFERENCED */
+ uint qfield; /* quota field to update */
+ xfs_filblks_t temp; /* for indirect length calculations */
+ xfs_filblks_t temp2; /* for indirect length calculations */
+ int state = 0;
+
+ XFS_STATS_INC(xs_del_exlist);
+
+ if (whichfork == XFS_ATTR_FORK)
+ state |= BMAP_ATTRFORK;
+
+ mp = ip->i_mount;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ ASSERT((*idx >= 0) && (*idx < ifp->if_bytes /
+ (uint)sizeof(xfs_bmbt_rec_t)));
+ ASSERT(del->br_blockcount > 0);
+ ep = xfs_iext_get_ext(ifp, *idx);
+ xfs_bmbt_get_all(ep, &got);
+ ASSERT(got.br_startoff <= del->br_startoff);
+ del_endoff = del->br_startoff + del->br_blockcount;
+ got_endoff = got.br_startoff + got.br_blockcount;
+ ASSERT(got_endoff >= del_endoff);
+ delay = isnullstartblock(got.br_startblock);
+ ASSERT(isnullstartblock(del->br_startblock) == delay);
+ flags = 0;
+ qfield = 0;
+ error = 0;
+ /*
+ * If deleting a real allocation, must free up the disk space.
+ */
+ if (!delay) {
+ flags = XFS_ILOG_CORE;
+ /*
+ * Realtime allocation. Free it and record di_nblocks update.
+ */
+ if (whichfork == XFS_DATA_FORK && XFS_IS_REALTIME_INODE(ip)) {
+ xfs_fsblock_t bno;
+ xfs_filblks_t len;
+
+ ASSERT(do_mod(del->br_blockcount,
+ mp->m_sb.sb_rextsize) == 0);
+ ASSERT(do_mod(del->br_startblock,
+ mp->m_sb.sb_rextsize) == 0);
+ bno = del->br_startblock;
+ len = del->br_blockcount;
+ do_div(bno, mp->m_sb.sb_rextsize);
+ do_div(len, mp->m_sb.sb_rextsize);
+ error = xfs_rtfree_extent(tp, bno, (xfs_extlen_t)len);
+ if (error)
+ goto done;
+ do_fx = 0;
+ nblks = len * mp->m_sb.sb_rextsize;
+ qfield = XFS_TRANS_DQ_RTBCOUNT;
+ }
/*
- * If it's the case where the directory code is running
- * with no block reservation, and the deleted block is in
- * the middle of its extent, and the resulting insert
- * of an extent would cause transformation to btree format,
- * then reject it. The calling code will then swap
- * blocks around instead.
- * We have to do this now, rather than waiting for the
- * conversion to btree format, since the transaction
- * will be dirty.
+ * Ordinary allocation.
*/
- if (!wasdel && xfs_trans_get_block_res(tp) == 0 &&
- XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
- XFS_IFORK_NEXTENTS(ip, whichfork) >= /* Note the >= */
- XFS_IFORK_MAXEXT(ip, whichfork) &&
- del.br_startoff > got.br_startoff &&
- del.br_startoff + del.br_blockcount <
- got.br_startoff + got.br_blockcount) {
- error = XFS_ERROR(ENOSPC);
- goto error0;
+ else {
+ do_fx = 1;
+ nblks = del->br_blockcount;
+ qfield = XFS_TRANS_DQ_BCOUNT;
}
- error = xfs_bmap_del_extent(ip, tp, &lastx, flist, cur, &del,
- &tmp_logflags, whichfork);
- logflags |= tmp_logflags;
- if (error)
- goto error0;
- bno = del.br_startoff - 1;
-nodelete:
/*
- * If not done go on to the next (previous) record.
+ * Set up del_endblock and cur for later.
*/
- if (bno != (xfs_fileoff_t)-1 && bno >= start) {
- if (lastx >= 0) {
- ep = xfs_iext_get_ext(ifp, lastx);
- if (xfs_bmbt_get_startoff(ep) > bno) {
- if (--lastx >= 0)
- ep = xfs_iext_get_ext(ifp,
- lastx);
- }
- xfs_bmbt_get_all(ep, &got);
- }
- extno++;
+ del_endblock = del->br_startblock + del->br_blockcount;
+ if (cur) {
+ if ((error = xfs_bmbt_lookup_eq(cur, got.br_startoff,
+ got.br_startblock, got.br_blockcount,
+ &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
}
- }
- *done = bno == (xfs_fileoff_t)-1 || bno < start || lastx < 0;
-
- /*
- * Convert to a btree if necessary.
- */
- if (xfs_bmap_needs_btree(ip, whichfork)) {
- ASSERT(cur == NULL);
- error = xfs_bmap_extents_to_btree(tp, ip, firstblock, flist,
- &cur, 0, &tmp_logflags, whichfork);
- logflags |= tmp_logflags;
- if (error)
- goto error0;
+ da_old = da_new = 0;
+ } else {
+ da_old = startblockval(got.br_startblock);
+ da_new = 0;
+ nblks = 0;
+ do_fx = 0;
}
/*
- * transform from btree to extents, give it cur
+ * Set flag value to use in switch statement.
+ * Left-contig is 2, right-contig is 1.
*/
- else if (xfs_bmap_wants_extents(ip, whichfork)) {
- ASSERT(cur != NULL);
- error = xfs_bmap_btree_to_extents(tp, ip, cur, &tmp_logflags,
- whichfork);
- logflags |= tmp_logflags;
- if (error)
- goto error0;
+ switch (((got.br_startoff == del->br_startoff) << 1) |
+ (got_endoff == del_endoff)) {
+ case 3:
+ /*
+ * Matches the whole extent. Delete the entry.
+ */
+ xfs_iext_remove(ip, *idx, 1,
+ whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0);
+ --*idx;
+ if (delay)
+ break;
+
+ XFS_IFORK_NEXT_SET(ip, whichfork,
+ XFS_IFORK_NEXTENTS(ip, whichfork) - 1);
+ flags |= XFS_ILOG_CORE;
+ if (!cur) {
+ flags |= xfs_ilog_fext(whichfork);
+ break;
+ }
+ if ((error = xfs_btree_delete(cur, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ break;
+
+ case 2:
+ /*
+ * Deleting the first part of the extent.
+ */
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_startoff(ep, del_endoff);
+ temp = got.br_blockcount - del->br_blockcount;
+ xfs_bmbt_set_blockcount(ep, temp);
+ if (delay) {
+ temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+ da_old);
+ xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ da_new = temp;
+ break;
+ }
+ xfs_bmbt_set_startblock(ep, del_endblock);
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ if (!cur) {
+ flags |= xfs_ilog_fext(whichfork);
+ break;
+ }
+ if ((error = xfs_bmbt_update(cur, del_endoff, del_endblock,
+ got.br_blockcount - del->br_blockcount,
+ got.br_state)))
+ goto done;
+ break;
+
+ case 1:
+ /*
+ * Deleting the last part of the extent.
+ */
+ temp = got.br_blockcount - del->br_blockcount;
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep, temp);
+ if (delay) {
+ temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp),
+ da_old);
+ xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ da_new = temp;
+ break;
+ }
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ if (!cur) {
+ flags |= xfs_ilog_fext(whichfork);
+ break;
+ }
+ if ((error = xfs_bmbt_update(cur, got.br_startoff,
+ got.br_startblock,
+ got.br_blockcount - del->br_blockcount,
+ got.br_state)))
+ goto done;
+ break;
+
+ case 0:
+ /*
+ * Deleting the middle of the extent.
+ */
+ temp = del->br_startoff - got.br_startoff;
+ trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_);
+ xfs_bmbt_set_blockcount(ep, temp);
+ new.br_startoff = del_endoff;
+ temp2 = got_endoff - del_endoff;
+ new.br_blockcount = temp2;
+ new.br_state = got.br_state;
+ if (!delay) {
+ new.br_startblock = del_endblock;
+ flags |= XFS_ILOG_CORE;
+ if (cur) {
+ if ((error = xfs_bmbt_update(cur,
+ got.br_startoff,
+ got.br_startblock, temp,
+ got.br_state)))
+ goto done;
+ if ((error = xfs_btree_increment(cur, 0, &i)))
+ goto done;
+ cur->bc_rec.b = new;
+ error = xfs_btree_insert(cur, &i);
+ if (error && error != ENOSPC)
+ goto done;
+ /*
+ * If get no-space back from btree insert,
+ * it tried a split, and we have a zero
+ * block reservation.
+ * Fix up our state and return the error.
+ */
+ if (error == ENOSPC) {
+ /*
+ * Reset the cursor, don't trust
+ * it after any insert operation.
+ */
+ if ((error = xfs_bmbt_lookup_eq(cur,
+ got.br_startoff,
+ got.br_startblock,
+ temp, &i)))
+ goto done;
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ /*
+ * Update the btree record back
+ * to the original value.
+ */
+ if ((error = xfs_bmbt_update(cur,
+ got.br_startoff,
+ got.br_startblock,
+ got.br_blockcount,
+ got.br_state)))
+ goto done;
+ /*
+ * Reset the extent record back
+ * to the original value.
+ */
+ xfs_bmbt_set_blockcount(ep,
+ got.br_blockcount);
+ flags = 0;
+ error = XFS_ERROR(ENOSPC);
+ goto done;
+ }
+ XFS_WANT_CORRUPTED_GOTO(i == 1, done);
+ } else
+ flags |= xfs_ilog_fext(whichfork);
+ XFS_IFORK_NEXT_SET(ip, whichfork,
+ XFS_IFORK_NEXTENTS(ip, whichfork) + 1);
+ } else {
+ ASSERT(whichfork == XFS_DATA_FORK);
+ temp = xfs_bmap_worst_indlen(ip, temp);
+ xfs_bmbt_set_startblock(ep, nullstartblock((int)temp));
+ temp2 = xfs_bmap_worst_indlen(ip, temp2);
+ new.br_startblock = nullstartblock((int)temp2);
+ da_new = temp + temp2;
+ while (da_new > da_old) {
+ if (temp) {
+ temp--;
+ da_new--;
+ xfs_bmbt_set_startblock(ep,
+ nullstartblock((int)temp));
+ }
+ if (da_new == da_old)
+ break;
+ if (temp2) {
+ temp2--;
+ da_new--;
+ new.br_startblock =
+ nullstartblock((int)temp2);
+ }
+ }
+ }
+ trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_);
+ xfs_iext_insert(ip, *idx + 1, 1, &new, state);
+ ++*idx;
+ break;
}
/*
- * transform from extents to local?
+ * If we need to, add to list of extents to delete.
*/
- error = 0;
-error0:
+ if (do_fx)
+ xfs_bmap_add_free(del->br_startblock, del->br_blockcount, flist,
+ mp);
/*
- * Log everything. Do this after conversion, there's no point in
- * logging the extent records if we've converted to btree format.
+ * Adjust inode # blocks in the file.
*/
- if ((logflags & xfs_ilog_fext(whichfork)) &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
- logflags &= ~xfs_ilog_fext(whichfork);
- else if ((logflags & xfs_ilog_fbroot(whichfork)) &&
- XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)
- logflags &= ~xfs_ilog_fbroot(whichfork);
+ if (nblks)
+ ip->i_d.di_nblocks -= nblks;
/*
- * Log inode even in the error case, if the transaction
- * is dirty we'll need to shut down the filesystem.
+ * Adjust quota data.
*/
- if (logflags)
- xfs_trans_log_inode(tp, ip, logflags);
- if (cur) {
- if (!error) {
- *firstblock = cur->bc_private.b.firstblock;
- cur->bc_private.b.allocated = 0;
- }
- xfs_btree_del_cursor(cur,
- error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+ if (qfield)
+ xfs_trans_mod_dquot_byino(tp, ip, qfield, (long)-nblks);
+
+ /*
+ * Account for change in delayed indirect blocks.
+ * Nothing to do for disk quota accounting here.
+ */
+ ASSERT(da_old >= da_new);
+ if (da_old > da_new) {
+ xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
+ (int64_t)(da_old - da_new), 0);
}
+done:
+ *logflagsp = flags;
return error;
}
/*
- * returns 1 for success, 0 if we failed to map the extent.
+ * Unmap (remove) blocks from a file.
+ * If nexts is nonzero then the number of extents to remove is limited to
+ * that value. If not all extents in the block range can be removed then
+ * *done is set.
*/
-STATIC int
-xfs_getbmapx_fix_eof_hole(
- xfs_inode_t *ip, /* xfs incore inode pointer */
- struct getbmapx *out, /* output structure */
- int prealloced, /* this is a file with
- * preallocated data space */
- __int64_t end, /* last block requested */
- xfs_fsblock_t startblock)
+int /* error */
+xfs_bunmapi(
+ xfs_trans_t *tp, /* transaction pointer */
+ struct xfs_inode *ip, /* incore inode */
+ xfs_fileoff_t bno, /* starting offset to unmap */
+ xfs_filblks_t len, /* length to unmap in file */
+ int flags, /* misc flags */
+ xfs_extnum_t nexts, /* number of extents max */
+ xfs_fsblock_t *firstblock, /* first allocated block
+ controls a.g. for allocs */
+ xfs_bmap_free_t *flist, /* i/o: list extents to free */
+ int *done) /* set if not done yet */
{
- __int64_t fixlen;
- xfs_mount_t *mp; /* file system mount point */
+ xfs_btree_cur_t *cur; /* bmap btree cursor */
+ xfs_bmbt_irec_t del; /* extent being deleted */
+ int eof; /* is deleting at eof */
+ xfs_bmbt_rec_host_t *ep; /* extent record pointer */
+ int error; /* error return value */
+ xfs_extnum_t extno; /* extent number in list */
+ xfs_bmbt_irec_t got; /* current extent record */
xfs_ifork_t *ifp; /* inode fork pointer */
- xfs_extnum_t lastx; /* last extent pointer */
- xfs_fileoff_t fileblock;
-
- if (startblock == HOLESTARTBLOCK) {
- mp = ip->i_mount;
- out->bmv_block = -1;
- fixlen = XFS_FSB_TO_BB(mp, XFS_B_TO_FSB(mp, XFS_ISIZE(ip)));
- fixlen -= out->bmv_offset;
- if (prealloced && out->bmv_offset + out->bmv_length == end) {
- /* Came to hole at EOF. Trim it. */
- if (fixlen <= 0)
- return 0;
- out->bmv_length = fixlen;
- }
- } else {
- if (startblock == DELAYSTARTBLOCK)
- out->bmv_block = -2;
- else
- out->bmv_block = xfs_fsb_to_db(ip, startblock);
- fileblock = XFS_BB_TO_FSB(ip->i_mount, out->bmv_offset);
- ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
- if (xfs_iext_bno_to_ext(ifp, fileblock, &lastx) &&
- (lastx == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))-1))
- out->bmv_oflags |= BMV_OF_LAST;
- }
-
- return 1;
-}
-
-/*
- * Get inode's extents as described in bmv, and format for output.
- * Calls formatter to fill the user's buffer until all extents
- * are mapped, until the passed-in bmv->bmv_count slots have
- * been filled, or until the formatter short-circuits the loop,
- * if it is tracking filled-in extents on its own.
- */
-int /* error code */
-xfs_getbmap(
- xfs_inode_t *ip,
- struct getbmapx *bmv, /* user bmap structure */
- xfs_bmap_format_t formatter, /* format to user */
- void *arg) /* formatter arg */
-{
- __int64_t bmvend; /* last block requested */
- int error = 0; /* return value */
- __int64_t fixlen; /* length for -1 case */
- int i; /* extent number */
- int lock; /* lock state */
- xfs_bmbt_irec_t *map; /* buffer for user's data */
- xfs_mount_t *mp; /* file system mount point */
- int nex; /* # of user extents can do */
- int nexleft; /* # of user extents left */
- int subnex; /* # of bmapi's can do */
- int nmap; /* number of map entries */
- struct getbmapx *out; /* output structure */
- int whichfork; /* data or attr fork */
- int prealloced; /* this is a file with
- * preallocated data space */
- int iflags; /* interface flags */
- int bmapi_flags; /* flags for xfs_bmapi */
- int cur_ext = 0;
-
- mp = ip->i_mount;
- iflags = bmv->bmv_iflags;
- whichfork = iflags & BMV_IF_ATTRFORK ? XFS_ATTR_FORK : XFS_DATA_FORK;
-
- if (whichfork == XFS_ATTR_FORK) {
- if (XFS_IFORK_Q(ip)) {
- if (ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS &&
- ip->i_d.di_aformat != XFS_DINODE_FMT_BTREE &&
- ip->i_d.di_aformat != XFS_DINODE_FMT_LOCAL)
- return XFS_ERROR(EINVAL);
- } else if (unlikely(
- ip->i_d.di_aformat != 0 &&
- ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS)) {
- XFS_ERROR_REPORT("xfs_getbmap", XFS_ERRLEVEL_LOW,
- ip->i_mount);
- return XFS_ERROR(EFSCORRUPTED);
- }
-
- prealloced = 0;
- fixlen = 1LL << 32;
- } else {
- if (ip->i_d.di_format != XFS_DINODE_FMT_EXTENTS &&
- ip->i_d.di_format != XFS_DINODE_FMT_BTREE &&
- ip->i_d.di_format != XFS_DINODE_FMT_LOCAL)
- return XFS_ERROR(EINVAL);
-
- if (xfs_get_extsz_hint(ip) ||
- ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC|XFS_DIFLAG_APPEND)){
- prealloced = 1;
- fixlen = mp->m_super->s_maxbytes;
- } else {
- prealloced = 0;
- fixlen = XFS_ISIZE(ip);
- }
- }
-
- if (bmv->bmv_length == -1) {
- fixlen = XFS_FSB_TO_BB(mp, XFS_B_TO_FSB(mp, fixlen));
- bmv->bmv_length =
- max_t(__int64_t, fixlen - bmv->bmv_offset, 0);
- } else if (bmv->bmv_length == 0) {
- bmv->bmv_entries = 0;
- return 0;
- } else if (bmv->bmv_length < 0) {
- return XFS_ERROR(EINVAL);
- }
-
- nex = bmv->bmv_count - 1;
- if (nex <= 0)
- return XFS_ERROR(EINVAL);
- bmvend = bmv->bmv_offset + bmv->bmv_length;
-
-
- if (bmv->bmv_count > ULONG_MAX / sizeof(struct getbmapx))
- return XFS_ERROR(ENOMEM);
- out = kmem_zalloc(bmv->bmv_count * sizeof(struct getbmapx), KM_MAYFAIL);
- if (!out) {
- out = kmem_zalloc_large(bmv->bmv_count *
- sizeof(struct getbmapx));
- if (!out)
- return XFS_ERROR(ENOMEM);
- }
-
- xfs_ilock(ip, XFS_IOLOCK_SHARED);
- if (whichfork == XFS_DATA_FORK && !(iflags & BMV_IF_DELALLOC)) {
- if (ip->i_delayed_blks || XFS_ISIZE(ip) > ip->i_d.di_size) {
- error = -filemap_write_and_wait(VFS_I(ip)->i_mapping);
- if (error)
- goto out_unlock_iolock;
- }
- /*
- * even after flushing the inode, there can still be delalloc
- * blocks on the inode beyond EOF due to speculative
- * preallocation. These are not removed until the release
- * function is called or the inode is inactivated. Hence we
- * cannot assert here that ip->i_delayed_blks == 0.
- */
- }
-
- lock = xfs_ilock_map_shared(ip);
-
- /*
- * Don't let nex be bigger than the number of extents
- * we can have assuming alternating holes and real extents.
- */
- if (nex > XFS_IFORK_NEXTENTS(ip, whichfork) * 2 + 1)
- nex = XFS_IFORK_NEXTENTS(ip, whichfork) * 2 + 1;
-
- bmapi_flags = xfs_bmapi_aflag(whichfork);
- if (!(iflags & BMV_IF_PREALLOC))
- bmapi_flags |= XFS_BMAPI_IGSTATE;
-
- /*
- * Allocate enough space to handle "subnex" maps at a time.
- */
- error = ENOMEM;
- subnex = 16;
- map = kmem_alloc(subnex * sizeof(*map), KM_MAYFAIL | KM_NOFS);
- if (!map)
- goto out_unlock_ilock;
+ int isrt; /* freeing in rt area */
+ xfs_extnum_t lastx; /* last extent index used */
+ int logflags; /* transaction logging flags */
+ xfs_extlen_t mod; /* rt extent offset */
+ xfs_mount_t *mp; /* mount structure */
+ xfs_extnum_t nextents; /* number of file extents */
+ xfs_bmbt_irec_t prev; /* previous extent record */
+ xfs_fileoff_t start; /* first file offset deleted */
+ int tmp_logflags; /* partial logging flags */
+ int wasdel; /* was a delayed alloc extent */
+ int whichfork; /* data or attribute fork */
+ xfs_fsblock_t sum;
- bmv->bmv_entries = 0;
+ trace_xfs_bunmap(ip, bno, len, flags, _RET_IP_);
- if (XFS_IFORK_NEXTENTS(ip, whichfork) == 0 &&
- (whichfork == XFS_ATTR_FORK || !(iflags & BMV_IF_DELALLOC))) {
- error = 0;
- goto out_free_map;
+ whichfork = (flags & XFS_BMAPI_ATTRFORK) ?
+ XFS_ATTR_FORK : XFS_DATA_FORK;
+ ifp = XFS_IFORK_PTR(ip, whichfork);
+ if (unlikely(
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)) {
+ XFS_ERROR_REPORT("xfs_bunmapi", XFS_ERRLEVEL_LOW,
+ ip->i_mount);
+ return XFS_ERROR(EFSCORRUPTED);
}
+ mp = ip->i_mount;
+ if (XFS_FORCED_SHUTDOWN(mp))
+ return XFS_ERROR(EIO);
- nexleft = nex;
+ ASSERT(len > 0);
+ ASSERT(nexts >= 0);
- do {
- nmap = (nexleft > subnex) ? subnex : nexleft;
- error = xfs_bmapi_read(ip, XFS_BB_TO_FSBT(mp, bmv->bmv_offset),
- XFS_BB_TO_FSB(mp, bmv->bmv_length),
- map, &nmap, bmapi_flags);
- if (error)
- goto out_free_map;
- ASSERT(nmap <= subnex);
+ if (!(ifp->if_flags & XFS_IFEXTENTS) &&
+ (error = xfs_iread_extents(tp, ip, whichfork)))
+ return error;
+ nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
+ if (nextents == 0) {
+ *done = 1;
+ return 0;
+ }
+ XFS_STATS_INC(xs_blk_unmap);
+ isrt = (whichfork == XFS_DATA_FORK) && XFS_IS_REALTIME_INODE(ip);
+ start = bno;
+ bno = start + len - 1;
+ ep = xfs_bmap_search_extents(ip, bno, whichfork, &eof, &lastx, &got,
+ &prev);
- for (i = 0; i < nmap && nexleft && bmv->bmv_length; i++) {
- out[cur_ext].bmv_oflags = 0;
- if (map[i].br_state == XFS_EXT_UNWRITTEN)
- out[cur_ext].bmv_oflags |= BMV_OF_PREALLOC;
- else if (map[i].br_startblock == DELAYSTARTBLOCK)
- out[cur_ext].bmv_oflags |= BMV_OF_DELALLOC;
- out[cur_ext].bmv_offset =
- XFS_FSB_TO_BB(mp, map[i].br_startoff);
- out[cur_ext].bmv_length =
- XFS_FSB_TO_BB(mp, map[i].br_blockcount);
- out[cur_ext].bmv_unused1 = 0;
- out[cur_ext].bmv_unused2 = 0;
+ /*
+ * Check to see if the given block number is past the end of the
+ * file, back up to the last block if so...
+ */
+ if (eof) {
+ ep = xfs_iext_get_ext(ifp, --lastx);
+ xfs_bmbt_get_all(ep, &got);
+ bno = got.br_startoff + got.br_blockcount - 1;
+ }
+ logflags = 0;
+ if (ifp->if_flags & XFS_IFBROOT) {
+ ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_BTREE);
+ cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
+ cur->bc_private.b.firstblock = *firstblock;
+ cur->bc_private.b.flist = flist;
+ cur->bc_private.b.flags = 0;
+ } else
+ cur = NULL;
+
+ if (isrt) {
+ /*
+ * Synchronize by locking the bitmap inode.
+ */
+ xfs_ilock(mp->m_rbmip, XFS_ILOCK_EXCL);
+ xfs_trans_ijoin(tp, mp->m_rbmip, XFS_ILOCK_EXCL);
+ }
+ extno = 0;
+ while (bno != (xfs_fileoff_t)-1 && bno >= start && lastx >= 0 &&
+ (nexts == 0 || extno < nexts)) {
+ /*
+ * Is the found extent after a hole in which bno lives?
+ * Just back up to the previous extent, if so.
+ */
+ if (got.br_startoff > bno) {
+ if (--lastx < 0)
+ break;
+ ep = xfs_iext_get_ext(ifp, lastx);
+ xfs_bmbt_get_all(ep, &got);
+ }
+ /*
+ * Is the last block of this extent before the range
+ * we're supposed to delete? If so, we're done.
+ */
+ bno = XFS_FILEOFF_MIN(bno,
+ got.br_startoff + got.br_blockcount - 1);
+ if (bno < start)
+ break;
+ /*
+ * Then deal with the (possibly delayed) allocated space
+ * we found.
+ */
+ ASSERT(ep != NULL);
+ del = got;
+ wasdel = isnullstartblock(del.br_startblock);
+ if (got.br_startoff < start) {
+ del.br_startoff = start;
+ del.br_blockcount -= start - got.br_startoff;
+ if (!wasdel)
+ del.br_startblock += start - got.br_startoff;
+ }
+ if (del.br_startoff + del.br_blockcount > bno + 1)
+ del.br_blockcount = bno + 1 - del.br_startoff;
+ sum = del.br_startblock + del.br_blockcount;
+ if (isrt &&
+ (mod = do_mod(sum, mp->m_sb.sb_rextsize))) {
/*
- * delayed allocation extents that start beyond EOF can
- * occur due to speculative EOF allocation when the
- * delalloc extent is larger than the largest freespace
- * extent at conversion time. These extents cannot be
- * converted by data writeback, so can exist here even
- * if we are not supposed to be finding delalloc
- * extents.
+ * Realtime extent not lined up at the end.
+ * The extent could have been split into written
+ * and unwritten pieces, or we could just be
+ * unmapping part of it. But we can't really
+ * get rid of part of a realtime extent.
*/
- if (map[i].br_startblock == DELAYSTARTBLOCK &&
- map[i].br_startoff <= XFS_B_TO_FSB(mp, XFS_ISIZE(ip)))
- ASSERT((iflags & BMV_IF_DELALLOC) != 0);
-
- if (map[i].br_startblock == HOLESTARTBLOCK &&
- whichfork == XFS_ATTR_FORK) {
- /* came to the end of attribute fork */
- out[cur_ext].bmv_oflags |= BMV_OF_LAST;
- goto out_free_map;
+ if (del.br_state == XFS_EXT_UNWRITTEN ||
+ !xfs_sb_version_hasextflgbit(&mp->m_sb)) {
+ /*
+ * This piece is unwritten, or we're not
+ * using unwritten extents. Skip over it.
+ */
+ ASSERT(bno >= mod);
+ bno -= mod > del.br_blockcount ?
+ del.br_blockcount : mod;
+ if (bno < got.br_startoff) {
+ if (--lastx >= 0)
+ xfs_bmbt_get_all(xfs_iext_get_ext(
+ ifp, lastx), &got);
+ }
+ continue;
}
-
- if (!xfs_getbmapx_fix_eof_hole(ip, &out[cur_ext],
- prealloced, bmvend,
- map[i].br_startblock))
- goto out_free_map;
-
- bmv->bmv_offset =
- out[cur_ext].bmv_offset +
- out[cur_ext].bmv_length;
- bmv->bmv_length =
- max_t(__int64_t, 0, bmvend - bmv->bmv_offset);
-
/*
- * In case we don't want to return the hole,
- * don't increase cur_ext so that we can reuse
- * it in the next loop.
+ * It's written, turn it unwritten.
+ * This is better than zeroing it.
*/
- if ((iflags & BMV_IF_NO_HOLES) &&
- map[i].br_startblock == HOLESTARTBLOCK) {
- memset(&out[cur_ext], 0, sizeof(out[cur_ext]));
+ ASSERT(del.br_state == XFS_EXT_NORM);
+ ASSERT(xfs_trans_get_block_res(tp) > 0);
+ /*
+ * If this spans a realtime extent boundary,
+ * chop it back to the start of the one we end at.
+ */
+ if (del.br_blockcount > mod) {
+ del.br_startoff += del.br_blockcount - mod;
+ del.br_startblock += del.br_blockcount - mod;
+ del.br_blockcount = mod;
+ }
+ del.br_state = XFS_EXT_UNWRITTEN;
+ error = xfs_bmap_add_extent_unwritten_real(tp, ip,
+ &lastx, &cur, &del, firstblock, flist,
+ &logflags);
+ if (error)
+ goto error0;
+ goto nodelete;
+ }
+ if (isrt && (mod = do_mod(del.br_startblock, mp->m_sb.sb_rextsize))) {
+ /*
+ * Realtime extent is lined up at the end but not
+ * at the front. We'll get rid of full extents if
+ * we can.
+ */
+ mod = mp->m_sb.sb_rextsize - mod;
+ if (del.br_blockcount > mod) {
+ del.br_blockcount -= mod;
+ del.br_startoff += mod;
+ del.br_startblock += mod;
+ } else if ((del.br_startoff == start &&
+ (del.br_state == XFS_EXT_UNWRITTEN ||
+ xfs_trans_get_block_res(tp) == 0)) ||
+ !xfs_sb_version_hasextflgbit(&mp->m_sb)) {
+ /*
+ * Can't make it unwritten. There isn't
+ * a full extent here so just skip it.
+ */
+ ASSERT(bno >= del.br_blockcount);
+ bno -= del.br_blockcount;
+ if (got.br_startoff > bno) {
+ if (--lastx >= 0) {
+ ep = xfs_iext_get_ext(ifp,
+ lastx);
+ xfs_bmbt_get_all(ep, &got);
+ }
+ }
continue;
+ } else if (del.br_state == XFS_EXT_UNWRITTEN) {
+ /*
+ * This one is already unwritten.
+ * It must have a written left neighbor.
+ * Unwrite the killed part of that one and
+ * try again.
+ */
+ ASSERT(lastx > 0);
+ xfs_bmbt_get_all(xfs_iext_get_ext(ifp,
+ lastx - 1), &prev);
+ ASSERT(prev.br_state == XFS_EXT_NORM);
+ ASSERT(!isnullstartblock(prev.br_startblock));
+ ASSERT(del.br_startblock ==
+ prev.br_startblock + prev.br_blockcount);
+ if (prev.br_startoff < start) {
+ mod = start - prev.br_startoff;
+ prev.br_blockcount -= mod;
+ prev.br_startblock += mod;
+ prev.br_startoff = start;
+ }
+ prev.br_state = XFS_EXT_UNWRITTEN;
+ lastx--;
+ error = xfs_bmap_add_extent_unwritten_real(tp,
+ ip, &lastx, &cur, &prev,
+ firstblock, flist, &logflags);
+ if (error)
+ goto error0;
+ goto nodelete;
+ } else {
+ ASSERT(del.br_state == XFS_EXT_NORM);
+ del.br_state = XFS_EXT_UNWRITTEN;
+ error = xfs_bmap_add_extent_unwritten_real(tp,
+ ip, &lastx, &cur, &del,
+ firstblock, flist, &logflags);
+ if (error)
+ goto error0;
+ goto nodelete;
}
-
- nexleft--;
- bmv->bmv_entries++;
- cur_ext++;
}
- } while (nmap && nexleft && bmv->bmv_length);
-
- out_free_map:
- kmem_free(map);
- out_unlock_ilock:
- xfs_iunlock_map_shared(ip, lock);
- out_unlock_iolock:
- xfs_iunlock(ip, XFS_IOLOCK_SHARED);
-
- for (i = 0; i < cur_ext; i++) {
- int full = 0; /* user array is full */
-
- /* format results & advance arg */
- error = formatter(&arg, &out[i], &full);
- if (error || full)
- break;
- }
-
- if (is_vmalloc_addr(out))
- kmem_free_large(out);
- else
- kmem_free(out);
- return error;
-}
-
-#ifdef DEBUG
-STATIC struct xfs_buf *
-xfs_bmap_get_bp(
- struct xfs_btree_cur *cur,
- xfs_fsblock_t bno)
-{
- struct xfs_log_item_desc *lidp;
- int i;
-
- if (!cur)
- return NULL;
-
- for (i = 0; i < XFS_BTREE_MAXLEVELS; i++) {
- if (!cur->bc_bufs[i])
- break;
- if (XFS_BUF_ADDR(cur->bc_bufs[i]) == bno)
- return cur->bc_bufs[i];
- }
-
- /* Chase down all the log items to see if the bp is there */
- list_for_each_entry(lidp, &cur->bc_tp->t_items, lid_trans) {
- struct xfs_buf_log_item *bip;
- bip = (struct xfs_buf_log_item *)lidp->lid_item;
- if (bip->bli_item.li_type == XFS_LI_BUF &&
- XFS_BUF_ADDR(bip->bli_buf) == bno)
- return bip->bli_buf;
- }
-
- return NULL;
-}
-
-STATIC void
-xfs_check_block(
- struct xfs_btree_block *block,
- xfs_mount_t *mp,
- int root,
- short sz)
-{
- int i, j, dmxr;
- __be64 *pp, *thispa; /* pointer to block address */
- xfs_bmbt_key_t *prevp, *keyp;
-
- ASSERT(be16_to_cpu(block->bb_level) > 0);
-
- prevp = NULL;
- for( i = 1; i <= xfs_btree_get_numrecs(block); i++) {
- dmxr = mp->m_bmap_dmxr[0];
- keyp = XFS_BMBT_KEY_ADDR(mp, block, i);
+ if (wasdel) {
+ ASSERT(startblockval(del.br_startblock) > 0);
+ /* Update realtime/data freespace, unreserve quota */
+ if (isrt) {
+ xfs_filblks_t rtexts;
- if (prevp) {
- ASSERT(be64_to_cpu(prevp->br_startoff) <
- be64_to_cpu(keyp->br_startoff));
+ rtexts = XFS_FSB_TO_B(mp, del.br_blockcount);
+ do_div(rtexts, mp->m_sb.sb_rextsize);
+ xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS,
+ (int64_t)rtexts, 0);
+ (void)xfs_trans_reserve_quota_nblks(NULL,
+ ip, -((long)del.br_blockcount), 0,
+ XFS_QMOPT_RES_RTBLKS);
+ } else {
+ xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS,
+ (int64_t)del.br_blockcount, 0);
+ (void)xfs_trans_reserve_quota_nblks(NULL,
+ ip, -((long)del.br_blockcount), 0,
+ XFS_QMOPT_RES_REGBLKS);
+ }
+ ip->i_delayed_blks -= del.br_blockcount;
+ if (cur)
+ cur->bc_private.b.flags |=
+ XFS_BTCUR_BPRV_WASDEL;
+ } else if (cur)
+ cur->bc_private.b.flags &= ~XFS_BTCUR_BPRV_WASDEL;
+ /*
+ * If it's the case where the directory code is running
+ * with no block reservation, and the deleted block is in
+ * the middle of its extent, and the resulting insert
+ * of an extent would cause transformation to btree format,
+ * then reject it. The calling code will then swap
+ * blocks around instead.
+ * We have to do this now, rather than waiting for the
+ * conversion to btree format, since the transaction
+ * will be dirty.
+ */
+ if (!wasdel && xfs_trans_get_block_res(tp) == 0 &&
+ XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS &&
+ XFS_IFORK_NEXTENTS(ip, whichfork) >= /* Note the >= */
+ XFS_IFORK_MAXEXT(ip, whichfork) &&
+ del.br_startoff > got.br_startoff &&
+ del.br_startoff + del.br_blockcount <
+ got.br_startoff + got.br_blockcount) {
+ error = XFS_ERROR(ENOSPC);
+ goto error0;
}
- prevp = keyp;
-
+ error = xfs_bmap_del_extent(ip, tp, &lastx, flist, cur, &del,
+ &tmp_logflags, whichfork);
+ logflags |= tmp_logflags;
+ if (error)
+ goto error0;
+ bno = del.br_startoff - 1;
+nodelete:
/*
- * Compare the block numbers to see if there are dups.
+ * If not done go on to the next (previous) record.
*/
- if (root)
- pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, i, sz);
- else
- pp = XFS_BMBT_PTR_ADDR(mp, block, i, dmxr);
-
- for (j = i+1; j <= be16_to_cpu(block->bb_numrecs); j++) {
- if (root)
- thispa = XFS_BMAP_BROOT_PTR_ADDR(mp, block, j, sz);
- else
- thispa = XFS_BMBT_PTR_ADDR(mp, block, j, dmxr);
- if (*thispa == *pp) {
- xfs_warn(mp, "%s: thispa(%d) == pp(%d) %Ld",
- __func__, j, i,
- (unsigned long long)be64_to_cpu(*thispa));
- panic("%s: ptrs are equal in node\n",
- __func__);
+ if (bno != (xfs_fileoff_t)-1 && bno >= start) {
+ if (lastx >= 0) {
+ ep = xfs_iext_get_ext(ifp, lastx);
+ if (xfs_bmbt_get_startoff(ep) > bno) {
+ if (--lastx >= 0)
+ ep = xfs_iext_get_ext(ifp,
+ lastx);
+ }
+ xfs_bmbt_get_all(ep, &got);
}
+ extno++;
}
}
-}
-
-/*
- * Check that the extents for the inode ip are in the right order in all
- * btree leaves.
- */
-
-STATIC void
-xfs_bmap_check_leaf_extents(
- xfs_btree_cur_t *cur, /* btree cursor or null */
- xfs_inode_t *ip, /* incore inode pointer */
- int whichfork) /* data or attr fork */
-{
- struct xfs_btree_block *block; /* current btree block */
- xfs_fsblock_t bno; /* block # of "block" */
- xfs_buf_t *bp; /* buffer for "block" */
- int error; /* error return value */
- xfs_extnum_t i=0, j; /* index into the extents list */
- xfs_ifork_t *ifp; /* fork structure */
- int level; /* btree level, for checking */
- xfs_mount_t *mp; /* file system mount structure */
- __be64 *pp; /* pointer to block address */
- xfs_bmbt_rec_t *ep; /* pointer to current extent */
- xfs_bmbt_rec_t last = {0, 0}; /* last extent in prev block */
- xfs_bmbt_rec_t *nextp; /* pointer to next extent */
- int bp_release = 0;
-
- if (XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE) {
- return;
- }
-
- bno = NULLFSBLOCK;
- mp = ip->i_mount;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- block = ifp->if_broot;
- /*
- * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
- */
- level = be16_to_cpu(block->bb_level);
- ASSERT(level > 0);
- xfs_check_block(block, mp, 1, ifp->if_broot_bytes);
- pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
- bno = be64_to_cpu(*pp);
-
- ASSERT(bno != NULLDFSBNO);
- ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
- ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
+ *done = bno == (xfs_fileoff_t)-1 || bno < start || lastx < 0;
/*
- * Go down the tree until leaf level is reached, following the first
- * pointer (leftmost) at each level.
+ * Convert to a btree if necessary.
*/
- while (level-- > 0) {
- /* See if buf is in cur first */
- bp_release = 0;
- bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno));
- if (!bp) {
- bp_release = 1;
- error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp,
- XFS_BMAP_BTREE_REF,
- &xfs_bmbt_buf_ops);
- if (error)
- goto error_norelse;
- }
- block = XFS_BUF_TO_BLOCK(bp);
- XFS_WANT_CORRUPTED_GOTO(
- xfs_bmap_sanity_check(mp, bp, level),
- error0);
- if (level == 0)
- break;
-
- /*
- * Check this block for basic sanity (increasing keys and
- * no duplicate blocks).
- */
-
- xfs_check_block(block, mp, 0, 0);
- pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
- bno = be64_to_cpu(*pp);
- XFS_WANT_CORRUPTED_GOTO(XFS_FSB_SANITY_CHECK(mp, bno), error0);
- if (bp_release) {
- bp_release = 0;
- xfs_trans_brelse(NULL, bp);
- }
+ if (xfs_bmap_needs_btree(ip, whichfork)) {
+ ASSERT(cur == NULL);
+ error = xfs_bmap_extents_to_btree(tp, ip, firstblock, flist,
+ &cur, 0, &tmp_logflags, whichfork);
+ logflags |= tmp_logflags;
+ if (error)
+ goto error0;
}
-
/*
- * Here with bp and block set to the leftmost leaf node in the tree.
+ * transform from btree to extents, give it cur
*/
- i = 0;
-
+ else if (xfs_bmap_wants_extents(ip, whichfork)) {
+ ASSERT(cur != NULL);
+ error = xfs_bmap_btree_to_extents(tp, ip, cur, &tmp_logflags,
+ whichfork);
+ logflags |= tmp_logflags;
+ if (error)
+ goto error0;
+ }
/*
- * Loop over all leaf nodes checking that all extents are in the right order.
+ * transform from extents to local?
*/
- for (;;) {
- xfs_fsblock_t nextbno;
- xfs_extnum_t num_recs;
+ error = 0;
+error0:
+ /*
+ * Log everything. Do this after conversion, there's no point in
+ * logging the extent records if we've converted to btree format.
+ */
+ if ((logflags & xfs_ilog_fext(whichfork)) &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_EXTENTS)
+ logflags &= ~xfs_ilog_fext(whichfork);
+ else if ((logflags & xfs_ilog_fbroot(whichfork)) &&
+ XFS_IFORK_FORMAT(ip, whichfork) != XFS_DINODE_FMT_BTREE)
+ logflags &= ~xfs_ilog_fbroot(whichfork);
+ /*
+ * Log inode even in the error case, if the transaction
+ * is dirty we'll need to shut down the filesystem.
+ */
+ if (logflags)
+ xfs_trans_log_inode(tp, ip, logflags);
+ if (cur) {
+ if (!error) {
+ *firstblock = cur->bc_private.b.firstblock;
+ cur->bc_private.b.allocated = 0;
+ }
+ xfs_btree_del_cursor(cur,
+ error ? XFS_BTREE_ERROR : XFS_BTREE_NOERROR);
+ }
+ return error;
+}
+/*
+ * returns 1 for success, 0 if we failed to map the extent.
+ */
+STATIC int
+xfs_getbmapx_fix_eof_hole(
+ xfs_inode_t *ip, /* xfs incore inode pointer */
+ struct getbmapx *out, /* output structure */
+ int prealloced, /* this is a file with
+ * preallocated data space */
+ __int64_t end, /* last block requested */
+ xfs_fsblock_t startblock)
+{
+ __int64_t fixlen;
+ xfs_mount_t *mp; /* file system mount point */
+ xfs_ifork_t *ifp; /* inode fork pointer */
+ xfs_extnum_t lastx; /* last extent pointer */
+ xfs_fileoff_t fileblock;
- num_recs = xfs_btree_get_numrecs(block);
+ if (startblock == HOLESTARTBLOCK) {
+ mp = ip->i_mount;
+ out->bmv_block = -1;
+ fixlen = XFS_FSB_TO_BB(mp, XFS_B_TO_FSB(mp, XFS_ISIZE(ip)));
+ fixlen -= out->bmv_offset;
+ if (prealloced && out->bmv_offset + out->bmv_length == end) {
+ /* Came to hole at EOF. Trim it. */
+ if (fixlen <= 0)
+ return 0;
+ out->bmv_length = fixlen;
+ }
+ } else {
+ if (startblock == DELAYSTARTBLOCK)
+ out->bmv_block = -2;
+ else
+ out->bmv_block = xfs_fsb_to_db(ip, startblock);
+ fileblock = XFS_BB_TO_FSB(ip->i_mount, out->bmv_offset);
+ ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK);
+ if (xfs_iext_bno_to_ext(ifp, fileblock, &lastx) &&
+ (lastx == (ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))-1))
+ out->bmv_oflags |= BMV_OF_LAST;
+ }
- /*
- * Read-ahead the next leaf block, if any.
- */
+ return 1;
+}
- nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
+/*
+ * Get inode's extents as described in bmv, and format for output.
+ * Calls formatter to fill the user's buffer until all extents
+ * are mapped, until the passed-in bmv->bmv_count slots have
+ * been filled, or until the formatter short-circuits the loop,
+ * if it is tracking filled-in extents on its own.
+ */
+int /* error code */
+xfs_getbmap(
+ xfs_inode_t *ip,
+ struct getbmapx *bmv, /* user bmap structure */
+ xfs_bmap_format_t formatter, /* format to user */
+ void *arg) /* formatter arg */
+{
+ __int64_t bmvend; /* last block requested */
+ int error = 0; /* return value */
+ __int64_t fixlen; /* length for -1 case */
+ int i; /* extent number */
+ int lock; /* lock state */
+ xfs_bmbt_irec_t *map; /* buffer for user's data */
+ xfs_mount_t *mp; /* file system mount point */
+ int nex; /* # of user extents can do */
+ int nexleft; /* # of user extents left */
+ int subnex; /* # of bmapi's can do */
+ int nmap; /* number of map entries */
+ struct getbmapx *out; /* output structure */
+ int whichfork; /* data or attr fork */
+ int prealloced; /* this is a file with
+ * preallocated data space */
+ int iflags; /* interface flags */
+ int bmapi_flags; /* flags for xfs_bmapi */
+ int cur_ext = 0;
- /*
- * Check all the extents to make sure they are OK.
- * If we had a previous block, the last entry should
- * conform with the first entry in this one.
- */
+ mp = ip->i_mount;
+ iflags = bmv->bmv_iflags;
+ whichfork = iflags & BMV_IF_ATTRFORK ? XFS_ATTR_FORK : XFS_DATA_FORK;
- ep = XFS_BMBT_REC_ADDR(mp, block, 1);
- if (i) {
- ASSERT(xfs_bmbt_disk_get_startoff(&last) +
- xfs_bmbt_disk_get_blockcount(&last) <=
- xfs_bmbt_disk_get_startoff(ep));
- }
- for (j = 1; j < num_recs; j++) {
- nextp = XFS_BMBT_REC_ADDR(mp, block, j + 1);
- ASSERT(xfs_bmbt_disk_get_startoff(ep) +
- xfs_bmbt_disk_get_blockcount(ep) <=
- xfs_bmbt_disk_get_startoff(nextp));
- ep = nextp;
+ if (whichfork == XFS_ATTR_FORK) {
+ if (XFS_IFORK_Q(ip)) {
+ if (ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS &&
+ ip->i_d.di_aformat != XFS_DINODE_FMT_BTREE &&
+ ip->i_d.di_aformat != XFS_DINODE_FMT_LOCAL)
+ return XFS_ERROR(EINVAL);
+ } else if (unlikely(
+ ip->i_d.di_aformat != 0 &&
+ ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS)) {
+ XFS_ERROR_REPORT("xfs_getbmap", XFS_ERRLEVEL_LOW,
+ ip->i_mount);
+ return XFS_ERROR(EFSCORRUPTED);
}
- last = *ep;
- i += num_recs;
- if (bp_release) {
- bp_release = 0;
- xfs_trans_brelse(NULL, bp);
- }
- bno = nextbno;
- /*
- * If we've reached the end, stop.
- */
- if (bno == NULLFSBLOCK)
- break;
+ prealloced = 0;
+ fixlen = 1LL << 32;
+ } else {
+ if (ip->i_d.di_format != XFS_DINODE_FMT_EXTENTS &&
+ ip->i_d.di_format != XFS_DINODE_FMT_BTREE &&
+ ip->i_d.di_format != XFS_DINODE_FMT_LOCAL)
+ return XFS_ERROR(EINVAL);
- bp_release = 0;
- bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno));
- if (!bp) {
- bp_release = 1;
- error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp,
- XFS_BMAP_BTREE_REF,
- &xfs_bmbt_buf_ops);
- if (error)
- goto error_norelse;
+ if (xfs_get_extsz_hint(ip) ||
+ ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC|XFS_DIFLAG_APPEND)){
+ prealloced = 1;
+ fixlen = mp->m_super->s_maxbytes;
+ } else {
+ prealloced = 0;
+ fixlen = XFS_ISIZE(ip);
}
- block = XFS_BUF_TO_BLOCK(bp);
}
- if (bp_release) {
- bp_release = 0;
- xfs_trans_brelse(NULL, bp);
+
+ if (bmv->bmv_length == -1) {
+ fixlen = XFS_FSB_TO_BB(mp, XFS_B_TO_FSB(mp, fixlen));
+ bmv->bmv_length =
+ max_t(__int64_t, fixlen - bmv->bmv_offset, 0);
+ } else if (bmv->bmv_length == 0) {
+ bmv->bmv_entries = 0;
+ return 0;
+ } else if (bmv->bmv_length < 0) {
+ return XFS_ERROR(EINVAL);
}
- return;
-error0:
- xfs_warn(mp, "%s: at error0", __func__);
- if (bp_release)
- xfs_trans_brelse(NULL, bp);
-error_norelse:
- xfs_warn(mp, "%s: BAD after btree leaves for %d extents",
- __func__, i);
- panic("%s: CORRUPTED BTREE OR SOMETHING", __func__);
- return;
-}
-#endif
+ nex = bmv->bmv_count - 1;
+ if (nex <= 0)
+ return XFS_ERROR(EINVAL);
+ bmvend = bmv->bmv_offset + bmv->bmv_length;
-/*
- * Count fsblocks of the given fork.
- */
-int /* error */
-xfs_bmap_count_blocks(
- xfs_trans_t *tp, /* transaction pointer */
- xfs_inode_t *ip, /* incore inode */
- int whichfork, /* data or attr fork */
- int *count) /* out: count of blocks */
-{
- struct xfs_btree_block *block; /* current btree block */
- xfs_fsblock_t bno; /* block # of "block" */
- xfs_ifork_t *ifp; /* fork structure */
- int level; /* btree level, for checking */
- xfs_mount_t *mp; /* file system mount structure */
- __be64 *pp; /* pointer to block address */
- bno = NULLFSBLOCK;
- mp = ip->i_mount;
- ifp = XFS_IFORK_PTR(ip, whichfork);
- if ( XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS ) {
- xfs_bmap_count_leaves(ifp, 0,
- ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t),
- count);
- return 0;
+ if (bmv->bmv_count > ULONG_MAX / sizeof(struct getbmapx))
+ return XFS_ERROR(ENOMEM);
+ out = kmem_zalloc(bmv->bmv_count * sizeof(struct getbmapx), KM_MAYFAIL);
+ if (!out) {
+ out = kmem_zalloc_large(bmv->bmv_count *
+ sizeof(struct getbmapx));
+ if (!out)
+ return XFS_ERROR(ENOMEM);
+ }
+
+ xfs_ilock(ip, XFS_IOLOCK_SHARED);
+ if (whichfork == XFS_DATA_FORK && !(iflags & BMV_IF_DELALLOC)) {
+ if (ip->i_delayed_blks || XFS_ISIZE(ip) > ip->i_d.di_size) {
+ error = -filemap_write_and_wait(VFS_I(ip)->i_mapping);
+ if (error)
+ goto out_unlock_iolock;
+ }
+ /*
+ * even after flushing the inode, there can still be delalloc
+ * blocks on the inode beyond EOF due to speculative
+ * preallocation. These are not removed until the release
+ * function is called or the inode is inactivated. Hence we
+ * cannot assert here that ip->i_delayed_blks == 0.
+ */
}
+ lock = xfs_ilock_map_shared(ip);
+
/*
- * Root level must use BMAP_BROOT_PTR_ADDR macro to get ptr out.
+ * Don't let nex be bigger than the number of extents
+ * we can have assuming alternating holes and real extents.
*/
- block = ifp->if_broot;
- level = be16_to_cpu(block->bb_level);
- ASSERT(level > 0);
- pp = XFS_BMAP_BROOT_PTR_ADDR(mp, block, 1, ifp->if_broot_bytes);
- bno = be64_to_cpu(*pp);
- ASSERT(bno != NULLDFSBNO);
- ASSERT(XFS_FSB_TO_AGNO(mp, bno) < mp->m_sb.sb_agcount);
- ASSERT(XFS_FSB_TO_AGBNO(mp, bno) < mp->m_sb.sb_agblocks);
+ if (nex > XFS_IFORK_NEXTENTS(ip, whichfork) * 2 + 1)
+ nex = XFS_IFORK_NEXTENTS(ip, whichfork) * 2 + 1;
- if (unlikely(xfs_bmap_count_tree(mp, tp, ifp, bno, level, count) < 0)) {
- XFS_ERROR_REPORT("xfs_bmap_count_blocks(2)", XFS_ERRLEVEL_LOW,
- mp);
- return XFS_ERROR(EFSCORRUPTED);
+ bmapi_flags = xfs_bmapi_aflag(whichfork);
+ if (!(iflags & BMV_IF_PREALLOC))
+ bmapi_flags |= XFS_BMAPI_IGSTATE;
+
+ /*
+ * Allocate enough space to handle "subnex" maps at a time.
+ */
+ error = ENOMEM;
+ subnex = 16;
+ map = kmem_alloc(subnex * sizeof(*map), KM_MAYFAIL | KM_NOFS);
+ if (!map)
+ goto out_unlock_ilock;
+
+ bmv->bmv_entries = 0;
+
+ if (XFS_IFORK_NEXTENTS(ip, whichfork) == 0 &&
+ (whichfork == XFS_ATTR_FORK || !(iflags & BMV_IF_DELALLOC))) {
+ error = 0;
+ goto out_free_map;
}
- return 0;
-}
+ nexleft = nex;
-/*
- * Recursively walks each level of a btree
- * to count total fsblocks is use.
- */
-STATIC int /* error */
-xfs_bmap_count_tree(
- xfs_mount_t *mp, /* file system mount point */
- xfs_trans_t *tp, /* transaction pointer */
- xfs_ifork_t *ifp, /* inode fork pointer */
- xfs_fsblock_t blockno, /* file system block number */
- int levelin, /* level in btree */
- int *count) /* Count of blocks */
-{
- int error;
- xfs_buf_t *bp, *nbp;
- int level = levelin;
- __be64 *pp;
- xfs_fsblock_t bno = blockno;
- xfs_fsblock_t nextbno;
- struct xfs_btree_block *block, *nextblock;
- int numrecs;
+ do {
+ nmap = (nexleft > subnex) ? subnex : nexleft;
+ error = xfs_bmapi_read(ip, XFS_BB_TO_FSBT(mp, bmv->bmv_offset),
+ XFS_BB_TO_FSB(mp, bmv->bmv_length),
+ map, &nmap, bmapi_flags);
+ if (error)
+ goto out_free_map;
+ ASSERT(nmap <= subnex);
- error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, XFS_BMAP_BTREE_REF,
- &xfs_bmbt_buf_ops);
- if (error)
- return error;
- *count += 1;
- block = XFS_BUF_TO_BLOCK(bp);
+ for (i = 0; i < nmap && nexleft && bmv->bmv_length; i++) {
+ out[cur_ext].bmv_oflags = 0;
+ if (map[i].br_state == XFS_EXT_UNWRITTEN)
+ out[cur_ext].bmv_oflags |= BMV_OF_PREALLOC;
+ else if (map[i].br_startblock == DELAYSTARTBLOCK)
+ out[cur_ext].bmv_oflags |= BMV_OF_DELALLOC;
+ out[cur_ext].bmv_offset =
+ XFS_FSB_TO_BB(mp, map[i].br_startoff);
+ out[cur_ext].bmv_length =
+ XFS_FSB_TO_BB(mp, map[i].br_blockcount);
+ out[cur_ext].bmv_unused1 = 0;
+ out[cur_ext].bmv_unused2 = 0;
- if (--level) {
- /* Not at node above leaves, count this level of nodes */
- nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
- while (nextbno != NULLFSBLOCK) {
- error = xfs_btree_read_bufl(mp, tp, nextbno, 0, &nbp,
- XFS_BMAP_BTREE_REF,
- &xfs_bmbt_buf_ops);
- if (error)
- return error;
- *count += 1;
- nextblock = XFS_BUF_TO_BLOCK(nbp);
- nextbno = be64_to_cpu(nextblock->bb_u.l.bb_rightsib);
- xfs_trans_brelse(tp, nbp);
- }
+ /*
+ * delayed allocation extents that start beyond EOF can
+ * occur due to speculative EOF allocation when the
+ * delalloc extent is larger than the largest freespace
+ * extent at conversion time. These extents cannot be
+ * converted by data writeback, so can exist here even
+ * if we are not supposed to be finding delalloc
+ * extents.
+ */
+ if (map[i].br_startblock == DELAYSTARTBLOCK &&
+ map[i].br_startoff <= XFS_B_TO_FSB(mp, XFS_ISIZE(ip)))
+ ASSERT((iflags & BMV_IF_DELALLOC) != 0);
- /* Dive to the next level */
- pp = XFS_BMBT_PTR_ADDR(mp, block, 1, mp->m_bmap_dmxr[1]);
- bno = be64_to_cpu(*pp);
- if (unlikely((error =
- xfs_bmap_count_tree(mp, tp, ifp, bno, level, count)) < 0)) {
- xfs_trans_brelse(tp, bp);
- XFS_ERROR_REPORT("xfs_bmap_count_tree(1)",
- XFS_ERRLEVEL_LOW, mp);
- return XFS_ERROR(EFSCORRUPTED);
- }
- xfs_trans_brelse(tp, bp);
- } else {
- /* count all level 1 nodes and their leaves */
- for (;;) {
- nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib);
- numrecs = be16_to_cpu(block->bb_numrecs);
- xfs_bmap_disk_count_leaves(mp, block, numrecs, count);
- xfs_trans_brelse(tp, bp);
- if (nextbno == NULLFSBLOCK)
- break;
- bno = nextbno;
- error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp,
- XFS_BMAP_BTREE_REF,
- &xfs_bmbt_buf_ops);
- if (error)
- return error;
- *count += 1;
- block = XFS_BUF_TO_BLOCK(bp);
- }
- }
- return 0;
-}
+ if (map[i].br_startblock == HOLESTARTBLOCK &&
+ whichfork == XFS_ATTR_FORK) {
+ /* came to the end of attribute fork */
+ out[cur_ext].bmv_oflags |= BMV_OF_LAST;
+ goto out_free_map;
+ }
-/*
- * Count leaf blocks given a range of extent records.
- */
-STATIC void
-xfs_bmap_count_leaves(
- xfs_ifork_t *ifp,
- xfs_extnum_t idx,
- int numrecs,
- int *count)
-{
- int b;
+ if (!xfs_getbmapx_fix_eof_hole(ip, &out[cur_ext],
+ prealloced, bmvend,
+ map[i].br_startblock))
+ goto out_free_map;
- for (b = 0; b < numrecs; b++) {
- xfs_bmbt_rec_host_t *frp = xfs_iext_get_ext(ifp, idx + b);
- *count += xfs_bmbt_get_blockcount(frp);
- }
-}
+ bmv->bmv_offset =
+ out[cur_ext].bmv_offset +
+ out[cur_ext].bmv_length;
+ bmv->bmv_length =
+ max_t(__int64_t, 0, bmvend - bmv->bmv_offset);
-/*
- * Count leaf blocks given a range of extent records originally
- * in btree format.
- */
-STATIC void
-xfs_bmap_disk_count_leaves(
- struct xfs_mount *mp,
- struct xfs_btree_block *block,
- int numrecs,
- int *count)
-{
- int b;
- xfs_bmbt_rec_t *frp;
+ /*
+ * In case we don't want to return the hole,
+ * don't increase cur_ext so that we can reuse
+ * it in the next loop.
+ */
+ if ((iflags & BMV_IF_NO_HOLES) &&
+ map[i].br_startblock == HOLESTARTBLOCK) {
+ memset(&out[cur_ext], 0, sizeof(out[cur_ext]));
+ continue;
+ }
- for (b = 1; b <= numrecs; b++) {
- frp = XFS_BMBT_REC_ADDR(mp, block, b);
- *count += xfs_bmbt_disk_get_blockcount(frp);
+ nexleft--;
+ bmv->bmv_entries++;
+ cur_ext++;
+ }
+ } while (nmap && nexleft && bmv->bmv_length);
+
+ out_free_map:
+ kmem_free(map);
+ out_unlock_ilock:
+ xfs_iunlock_map_shared(ip, lock);
+ out_unlock_iolock:
+ xfs_iunlock(ip, XFS_IOLOCK_SHARED);
+
+ for (i = 0; i < cur_ext; i++) {
+ int full = 0; /* user array is full */
+
+ /* format results & advance arg */
+ error = formatter(&arg, &out[i], &full);
+ if (error || full)
+ break;
}
+
+ if (is_vmalloc_addr(out))
+ kmem_free_large(out);
+ else
+ kmem_free(out);
+ return error;
}
/*
@@ -6295,16 +6171,3 @@ next_block:
return error;
}
-
-/*
- * Convert the given file system block to a disk block. We have to treat it
- * differently based on whether the file is a real time file or not, because the
- * bmap code does.
- */
-xfs_daddr_t
-xfs_fsb_to_db(struct xfs_inode *ip, xfs_fsblock_t fsb)
-{
- return (XFS_IS_REALTIME_INODE(ip) ? \
- (xfs_daddr_t)XFS_FSB_TO_BB((ip)->i_mount, (fsb)) : \
- XFS_FSB_TO_DADDR((ip)->i_mount, (fsb)));
-}
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH 01/16] xfs: rearrange some code in xfs_bmap for better locality
2013-02-25 1:31 ` [PATCH 01/16] xfs: rearrange some code in xfs_bmap for better locality Dave Chinner
@ 2013-02-25 14:44 ` Mark Tinguely
2013-03-07 18:48 ` Ben Myers
1 sibling, 0 replies; 23+ messages in thread
From: Mark Tinguely @ 2013-02-25 14:44 UTC (permalink / raw)
To: Dave Chinner; +Cc: xfs
On 02/24/13 19:31, Dave Chinner wrote:
> From: Dave Chinner<dchinner@redhat.com>
>
> xfs_bmap.c is a big file, and some of the related code is spread all
> throughout the file requiring function prototypes for static
> function and jumping all through the file to follow a single call
> path. Rearrange the code so that:
>
> a) related functionality is grouped together; and
> b) functions are grouped in call dependency order
>
> While the diffstat is large, there are no code changes in the patch;
> it is just moving the functionality around and removing the function
> prototypes at the top of the file. The resulting layout of the code
> is as follows (top of file to bottom):
>
> - miscellaneous helper functions
> - extent tree block counting routines
> - debug/sanity checking code
> - bmap free list manipulation functions
> - inode fork format manipulation functions
> - internal/external extent tree seach functions
> - extent tree manipulation functions used during allocation
> - functions used during extent read/allocate/removal
> operations (i.e. xfs_bmapi_write, xfs_bmapi_read,
> xfs_bunmapi and xfs_getbmap)
>
> This means that following logic paths through the bmapi code is much
> simpler - most of the code relevant to a specific operation is now
> clustered together rather than spread all over the file....
>
> Signed-off-by: Dave Chinner<dchinner@redhat.com>
> ---
Same as the previous post:
http://oss.sgi.com/archives/xfs/2013-02/msg00132.html
Reviewed-by: Mark Tinguely
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH 01/16] xfs: rearrange some code in xfs_bmap for better locality
2013-02-25 1:31 ` [PATCH 01/16] xfs: rearrange some code in xfs_bmap for better locality Dave Chinner
2013-02-25 14:44 ` Mark Tinguely
@ 2013-03-07 18:48 ` Ben Myers
1 sibling, 0 replies; 23+ messages in thread
From: Ben Myers @ 2013-03-07 18:48 UTC (permalink / raw)
To: Dave Chinner; +Cc: xfs
On Mon, Feb 25, 2013 at 12:31:26PM +1100, Dave Chinner wrote:
> From: Dave Chinner <dchinner@redhat.com>
>
> xfs_bmap.c is a big file, and some of the related code is spread all
> throughout the file requiring function prototypes for static
> function and jumping all through the file to follow a single call
> path. Rearrange the code so that:
>
> a) related functionality is grouped together; and
> b) functions are grouped in call dependency order
>
> While the diffstat is large, there are no code changes in the patch;
> it is just moving the functionality around and removing the function
> prototypes at the top of the file. The resulting layout of the code
> is as follows (top of file to bottom):
>
> - miscellaneous helper functions
> - extent tree block counting routines
> - debug/sanity checking code
> - bmap free list manipulation functions
> - inode fork format manipulation functions
> - internal/external extent tree seach functions
> - extent tree manipulation functions used during allocation
> - functions used during extent read/allocate/removal
> operations (i.e. xfs_bmapi_write, xfs_bmapi_read,
> xfs_bunmapi and xfs_getbmap)
>
> This means that following logic paths through the bmapi code is much
> simpler - most of the code relevant to a specific operation is now
> clustered together rather than spread all over the file....
>
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
Applied.
-Ben
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 23+ messages in thread
* [PATCH 02/16] xfs: take inode version into account in XFS_LITINO
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
2013-02-25 1:31 ` [PATCH 01/16] xfs: rearrange some code in xfs_bmap for better locality Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 03/16] xfs: add support for large btree blocks Dave Chinner
` (13 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Christoph Hellwig <hch@lst.de>
Add a version argument to XFS_LITINO so that it can return different values
depending on the inode version. This is required for the upcoming v3 inodes
with a larger fixed layout dinode.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Ben Myers <bpm@sgi.com>
---
fs/xfs/xfs_attr_leaf.c | 6 ++++--
fs/xfs/xfs_bmap.c | 4 ++--
fs/xfs/xfs_dinode.h | 6 +++---
fs/xfs/xfs_inode.h | 5 +++--
fs/xfs/xfs_vnodeops.c | 2 +-
5 files changed, 13 insertions(+), 10 deletions(-)
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index ee24993..f96a734 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -172,7 +172,8 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
int dsize;
xfs_mount_t *mp = dp->i_mount;
- offset = (XFS_LITINO(mp) - bytes) >> 3; /* rounded down */
+ /* rounded down */
+ offset = (XFS_LITINO(mp, dp->i_d.di_version) - bytes) >> 3;
switch (dp->i_d.di_format) {
case XFS_DINODE_FMT_DEV:
@@ -243,7 +244,8 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
minforkoff = roundup(minforkoff, 8) >> 3;
/* attr fork btree root can have at least this many key/ptr pairs */
- maxforkoff = XFS_LITINO(mp) - XFS_BMDR_SPACE_CALC(MINABTPTRS);
+ maxforkoff = XFS_LITINO(mp, dp->i_d.di_version) -
+ XFS_BMDR_SPACE_CALC(MINABTPTRS);
maxforkoff = maxforkoff >> 3; /* rounded down */
if (offset >= maxforkoff)
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index d490fe8..20efb39 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -228,13 +228,13 @@ xfs_default_attroffset(
uint offset;
if (mp->m_sb.sb_inodesize == 256) {
- offset = XFS_LITINO(mp) -
+ offset = XFS_LITINO(mp, ip->i_d.di_version) -
XFS_BMDR_SPACE_CALC(MINABTPTRS);
} else {
offset = XFS_BMDR_SPACE_CALC(6 * MINABTPTRS);
}
- ASSERT(offset < XFS_LITINO(mp));
+ ASSERT(offset < XFS_LITINO(mp, ip->i_d.di_version));
return offset;
}
diff --git a/fs/xfs/xfs_dinode.h b/fs/xfs/xfs_dinode.h
index 1d9643b..88a3368 100644
--- a/fs/xfs/xfs_dinode.h
+++ b/fs/xfs/xfs_dinode.h
@@ -104,7 +104,7 @@ typedef enum xfs_dinode_fmt {
/*
* Inode size for given fs.
*/
-#define XFS_LITINO(mp) \
+#define XFS_LITINO(mp, version) \
((int)(((mp)->m_sb.sb_inodesize) - sizeof(struct xfs_dinode)))
#define XFS_BROOT_SIZE_ADJ \
@@ -119,10 +119,10 @@ typedef enum xfs_dinode_fmt {
#define XFS_DFORK_DSIZE(dip,mp) \
(XFS_DFORK_Q(dip) ? \
XFS_DFORK_BOFF(dip) : \
- XFS_LITINO(mp))
+ XFS_LITINO(mp, (dip)->di_version))
#define XFS_DFORK_ASIZE(dip,mp) \
(XFS_DFORK_Q(dip) ? \
- XFS_LITINO(mp) - XFS_DFORK_BOFF(dip) : \
+ XFS_LITINO(mp, (dip)->di_version) - XFS_DFORK_BOFF(dip) : \
0)
#define XFS_DFORK_SIZE(dip,mp,w) \
((w) == XFS_DATA_FORK ? \
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index 237e7f6..b8520b5 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -180,10 +180,11 @@ typedef struct xfs_icdinode {
#define XFS_IFORK_DSIZE(ip) \
(XFS_IFORK_Q(ip) ? \
XFS_IFORK_BOFF(ip) : \
- XFS_LITINO((ip)->i_mount))
+ XFS_LITINO((ip)->i_mount, (ip)->i_d.di_version))
#define XFS_IFORK_ASIZE(ip) \
(XFS_IFORK_Q(ip) ? \
- XFS_LITINO((ip)->i_mount) - XFS_IFORK_BOFF(ip) : \
+ XFS_LITINO((ip)->i_mount, (ip)->i_d.di_version) - \
+ XFS_IFORK_BOFF(ip) : \
0)
#define XFS_IFORK_SIZE(ip,w) \
((w) == XFS_DATA_FORK ? \
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index 77ad748..aa0c066 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -1420,7 +1420,7 @@ xfs_symlink(
* The symlink will fit into the inode data fork?
* There can't be any attributes so we get the whole variable part.
*/
- if (pathlen <= XFS_LITINO(mp))
+ if (pathlen <= XFS_LITINO(mp, dp->i_d.di_version))
fs_blocks = 0;
else
fs_blocks = XFS_B_TO_FSB(mp, pathlen);
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 03/16] xfs: add support for large btree blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
2013-02-25 1:31 ` [PATCH 01/16] xfs: rearrange some code in xfs_bmap for better locality Dave Chinner
2013-02-25 1:31 ` [PATCH 02/16] xfs: take inode version into account in XFS_LITINO Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 04/16] xfs: add CRC checks to the AGF Dave Chinner
` (12 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Christoph Hellwig <hch@lst.de>
Add support for larger btree blocks that contains a CRC32C checksum,
a filesystem uuid and block number for detecting filesystem
consistency and out of place writes.
[dchinner@redhat.com] Also include an owner field to allow reverse
mappings to be implemented for improved repairability and a LSN
field to so that log recovery can easily determine the last
modification that made it to disk for each buffer.
[dchinner@redhat.com] Add buffer log format flags to indicate the
type of buffer to recovery so that we don't have to do blind magic
number tests to determine what the buffer is.
[dchinner@redhat.com] Modified to fit into the verifier structure.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_alloc_btree.c | 105 +++++++++++++------
fs/xfs/xfs_alloc_btree.h | 12 ++-
fs/xfs/xfs_attr_leaf.c | 2 +-
fs/xfs/xfs_bmap.c | 47 ++++++---
fs/xfs/xfs_bmap_btree.c | 110 +++++++++++++------
fs/xfs/xfs_bmap_btree.h | 19 ++--
fs/xfs/xfs_btree.c | 256 +++++++++++++++++++++++++++++++++++++--------
fs/xfs/xfs_btree.h | 64 ++++++++++--
fs/xfs/xfs_buf_item.h | 24 ++++-
fs/xfs/xfs_dinode.h | 4 +-
fs/xfs/xfs_fsops.c | 23 +++-
fs/xfs/xfs_ialloc_btree.c | 87 ++++++++++-----
fs/xfs/xfs_ialloc_btree.h | 9 +-
fs/xfs/xfs_inode.c | 33 +++---
fs/xfs/xfs_log_recover.c | 28 +++++
fs/xfs/xfs_trans.h | 2 +
fs/xfs/xfs_trans_buf.c | 29 +++--
17 files changed, 645 insertions(+), 209 deletions(-)
diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c
index b1ddef6..30c4c14 100644
--- a/fs/xfs/xfs_alloc_btree.c
+++ b/fs/xfs/xfs_alloc_btree.c
@@ -33,6 +33,7 @@
#include "xfs_extent_busy.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_cksum.h"
STATIC struct xfs_btree_cur *
@@ -272,7 +273,7 @@ xfs_allocbt_key_diff(
return (__int64_t)be32_to_cpu(kp->ar_startblock) - rec->ar_startblock;
}
-static void
+static bool
xfs_allocbt_verify(
struct xfs_buf *bp)
{
@@ -280,66 +281,103 @@ xfs_allocbt_verify(
struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
struct xfs_perag *pag = bp->b_pag;
unsigned int level;
- int sblock_ok; /* block passes checks */
/*
* magic number and level verification
*
- * During growfs operations, we can't verify the exact level as the
- * perag is not fully initialised and hence not attached to the buffer.
- * In this case, check against the maximum tree depth.
+ * 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.
*/
level = be16_to_cpu(block->bb_level);
switch (block->bb_magic) {
+ case cpu_to_be32(XFS_ABTB_CRC_MAGIC):
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return false;
+ if (!uuid_equal(&block->bb_u.s.bb_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (block->bb_u.s.bb_blkno != cpu_to_be64(bp->b_bn))
+ return false;
+ if (pag &&
+ be32_to_cpu(block->bb_u.s.bb_owner) != pag->pag_agno)
+ return false;
+ /* fall through */
case cpu_to_be32(XFS_ABTB_MAGIC):
- if (pag)
- sblock_ok = level < pag->pagf_levels[XFS_BTNUM_BNOi];
- else
- sblock_ok = level < mp->m_ag_maxlevels;
+ if (pag && pag->pagf_init) {
+ if (level >= pag->pagf_levels[XFS_BTNUM_BNOi])
+ return false;
+ } else if (level >= mp->m_ag_maxlevels)
+ return false;
break;
+ case cpu_to_be32(XFS_ABTC_CRC_MAGIC):
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return false;
+ if (!uuid_equal(&block->bb_u.s.bb_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (block->bb_u.s.bb_blkno != cpu_to_be64(bp->b_bn))
+ return false;
+ if (pag &&
+ be32_to_cpu(block->bb_u.s.bb_owner) != pag->pag_agno)
+ return false;
+ /* fall through */
case cpu_to_be32(XFS_ABTC_MAGIC):
- if (pag)
- sblock_ok = level < pag->pagf_levels[XFS_BTNUM_CNTi];
- else
- sblock_ok = level < mp->m_ag_maxlevels;
+ if (pag && pag->pagf_init) {
+ if (level >= pag->pagf_levels[XFS_BTNUM_CNTi])
+ return false;
+ } else if (level >= mp->m_ag_maxlevels)
+ return false;
break;
default:
- sblock_ok = 0;
- break;
+ return false;
}
/* numrecs verification */
- sblock_ok = sblock_ok &&
- be16_to_cpu(block->bb_numrecs) <= mp->m_alloc_mxr[level != 0];
+ if (be16_to_cpu(block->bb_numrecs) > mp->m_alloc_mxr[level != 0])
+ return false;
/* sibling pointer verification */
- sblock_ok = sblock_ok &&
- (block->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK) ||
- be32_to_cpu(block->bb_u.s.bb_leftsib) < mp->m_sb.sb_agblocks) &&
- block->bb_u.s.bb_leftsib &&
- (block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) ||
- be32_to_cpu(block->bb_u.s.bb_rightsib) < mp->m_sb.sb_agblocks) &&
- block->bb_u.s.bb_rightsib;
-
- if (!sblock_ok) {
- trace_xfs_btree_corrupt(bp, _RET_IP_);
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
- }
+ if (!block->bb_u.s.bb_leftsib ||
+ (be32_to_cpu(block->bb_u.s.bb_leftsib) >= mp->m_sb.sb_agblocks &&
+ block->bb_u.s.bb_leftsib != cpu_to_be32(NULLAGBLOCK)))
+ return false;
+ if (!block->bb_u.s.bb_rightsib ||
+ (be32_to_cpu(block->bb_u.s.bb_rightsib) >= mp->m_sb.sb_agblocks &&
+ block->bb_u.s.bb_rightsib != cpu_to_be32(NULLAGBLOCK)))
+ return false;
+
+ return true;
}
static void
xfs_allocbt_read_verify(
struct xfs_buf *bp)
{
- xfs_allocbt_verify(bp);
+ if (!(xfs_btree_sblock_verify_crc(bp) &&
+ xfs_allocbt_verify(bp))) {
+ trace_xfs_btree_corrupt(bp, _RET_IP_);
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+ bp->b_target->bt_mount, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
static void
xfs_allocbt_write_verify(
struct xfs_buf *bp)
{
- xfs_allocbt_verify(bp);
+ if (!xfs_allocbt_verify(bp)) {
+ trace_xfs_btree_corrupt(bp, _RET_IP_);
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+ bp->b_target->bt_mount, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
+ xfs_btree_sblock_calc_crc(bp);
+
}
const struct xfs_buf_ops xfs_allocbt_buf_ops = {
@@ -444,6 +482,9 @@ xfs_allocbt_init_cursor(
cur->bc_private.a.agbp = agbp;
cur->bc_private.a.agno = agno;
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
+
return cur;
}
diff --git a/fs/xfs/xfs_alloc_btree.h b/fs/xfs/xfs_alloc_btree.h
index 7e89a2b..e3a3f74 100644
--- a/fs/xfs/xfs_alloc_btree.h
+++ b/fs/xfs/xfs_alloc_btree.h
@@ -31,8 +31,10 @@ struct xfs_mount;
* by blockcount and blockno. All blocks look the same to make the code
* simpler; if we have time later, we'll make the optimizations.
*/
-#define XFS_ABTB_MAGIC 0x41425442 /* 'ABTB' for bno tree */
-#define XFS_ABTC_MAGIC 0x41425443 /* 'ABTC' for cnt tree */
+#define XFS_ABTB_MAGIC 0x41425442 /* 'ABTB' for bno tree */
+#define XFS_ABTB_CRC_MAGIC 0x41423342 /* 'AB3B' */
+#define XFS_ABTC_MAGIC 0x41425443 /* 'ABTC' for cnt tree */
+#define XFS_ABTC_CRC_MAGIC 0x41423343 /* 'AB3C' */
/*
* Data record/key structure
@@ -59,10 +61,10 @@ typedef __be32 xfs_alloc_ptr_t;
/*
* Btree block header size depends on a superblock flag.
- *
- * (not quite yet, but soon)
*/
-#define XFS_ALLOC_BLOCK_LEN(mp) XFS_BTREE_SBLOCK_LEN
+#define XFS_ALLOC_BLOCK_LEN(mp) \
+ (xfs_sb_version_hascrc(&((mp)->m_sb)) ? \
+ XFS_BTREE_SBLOCK_CRC_LEN : XFS_BTREE_SBLOCK_LEN)
/*
* Record, key, and pointer address macros for btree blocks.
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index f96a734..aa4765f 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -232,7 +232,7 @@ xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes)
return 0;
return dp->i_d.di_forkoff;
}
- dsize = XFS_BMAP_BROOT_SPACE(dp->i_df.if_broot);
+ dsize = XFS_BMAP_BROOT_SPACE(mp, dp->i_df.if_broot);
break;
}
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index 20efb39..0531cd3 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -439,11 +439,15 @@ xfs_bmap_sanity_check(
{
struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
- if (block->bb_magic != cpu_to_be32(XFS_BMAP_MAGIC) ||
- be16_to_cpu(block->bb_level) != level ||
+ if (block->bb_magic != cpu_to_be32(XFS_BMAP_CRC_MAGIC) &&
+ block->bb_magic != cpu_to_be32(XFS_BMAP_MAGIC))
+ return 0;
+
+ if (be16_to_cpu(block->bb_level) != level ||
be16_to_cpu(block->bb_numrecs) == 0 ||
be16_to_cpu(block->bb_numrecs) > mp->m_bmap_dmxr[level != 0])
return 0;
+
return 1;
}
@@ -1031,6 +1035,7 @@ xfs_bmap_extents_to_btree(
xfs_extnum_t nextents; /* number of file extents */
xfs_bmbt_ptr_t *pp; /* root block address pointer */
+ mp = ip->i_mount;
ifp = XFS_IFORK_PTR(ip, whichfork);
ASSERT(XFS_IFORK_FORMAT(ip, whichfork) == XFS_DINODE_FMT_EXTENTS);
@@ -1044,16 +1049,18 @@ xfs_bmap_extents_to_btree(
* Fill in the root.
*/
block = ifp->if_broot;
- block->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC);
- block->bb_level = cpu_to_be16(1);
- block->bb_numrecs = cpu_to_be16(1);
- block->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
- block->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ xfs_btree_init_block_int(mp, block, XFS_BUF_DADDR_NULL,
+ XFS_BMAP_CRC_MAGIC, 1, 1, ip->i_ino,
+ XFS_BTREE_LONG_PTRS | XFS_BTREE_CRC_BLOCKS);
+ else
+ xfs_btree_init_block_int(mp, block, XFS_BUF_DADDR_NULL,
+ XFS_BMAP_MAGIC, 1, 1, ip->i_ino,
+ XFS_BTREE_LONG_PTRS);
/*
* Need a cursor. Can't allocate until bb_level is filled in.
*/
- mp = ip->i_mount;
cur = xfs_bmbt_init_cursor(mp, tp, ip, whichfork);
cur->bc_private.b.firstblock = *firstblock;
cur->bc_private.b.flist = flist;
@@ -1102,10 +1109,15 @@ xfs_bmap_extents_to_btree(
*/
abp->b_ops = &xfs_bmbt_buf_ops;
ablock = XFS_BUF_TO_BLOCK(abp);
- ablock->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC);
- ablock->bb_level = 0;
- ablock->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
- ablock->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ xfs_btree_init_block_int(mp, ablock, abp->b_bn,
+ XFS_BMAP_CRC_MAGIC, 0, 0, ip->i_ino,
+ XFS_BTREE_LONG_PTRS | XFS_BTREE_CRC_BLOCKS);
+ else
+ xfs_btree_init_block_int(mp, ablock, abp->b_bn,
+ XFS_BMAP_MAGIC, 0, 0, ip->i_ino,
+ XFS_BTREE_LONG_PTRS);
+
arp = XFS_BMBT_REC_ADDR(mp, ablock, 1);
nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t);
for (cnt = i = 0; i < nextents; i++) {
@@ -1155,7 +1167,8 @@ xfs_bmap_local_to_extents(
xfs_extlen_t total, /* total blocks needed by transaction */
int *logflagsp, /* inode logging flags */
int whichfork,
- void (*init_fn)(struct xfs_buf *bp,
+ void (*init_fn)(struct xfs_trans *tp,
+ struct xfs_buf *bp,
struct xfs_inode *ip,
struct xfs_ifork *ifp))
{
@@ -1207,7 +1220,7 @@ xfs_bmap_local_to_extents(
bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0);
/* initialise the block and copy the data */
- init_fn(bp, ip, ifp);
+ init_fn(tp, bp, ip, ifp);
/* account for the change in fork size and log everything */
xfs_trans_log_buf(tp, bp, 0, ifp->if_bytes - 1);
@@ -1314,16 +1327,19 @@ xfs_bmap_add_attrfork_extents(
*/
STATIC void
xfs_bmap_local_to_extents_init_fn(
+ struct xfs_trans *tp,
struct xfs_buf *bp,
struct xfs_inode *ip,
struct xfs_ifork *ifp)
{
bp->b_ops = &xfs_bmbt_buf_ops;
memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+ xfs_trans_buf_set_type(tp, bp, XFS_BLF_BTREE_BUF);
}
STATIC void
xfs_symlink_local_to_remote(
+ struct xfs_trans *tp,
struct xfs_buf *bp,
struct xfs_inode *ip,
struct xfs_ifork *ifp)
@@ -1342,8 +1358,7 @@ xfs_symlink_local_to_remote(
*
* XXX (dgc): investigate whether directory conversion can use the generic
* formatting callout. It should be possible - it's just a very complex
- * formatter. it would also require passing the transaction through to the init
- * function.
+ * formatter.
*/
STATIC int /* error */
xfs_bmap_add_attrfork_local(
diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c
index 061b45c..3a86c3f 100644
--- a/fs/xfs/xfs_bmap_btree.c
+++ b/fs/xfs/xfs_bmap_btree.c
@@ -37,6 +37,7 @@
#include "xfs_error.h"
#include "xfs_quota.h"
#include "xfs_trace.h"
+#include "xfs_cksum.h"
/*
* Determine the extent state.
@@ -59,24 +60,31 @@ xfs_extent_state(
*/
void
xfs_bmdr_to_bmbt(
- struct xfs_mount *mp,
+ struct xfs_inode *ip,
xfs_bmdr_block_t *dblock,
int dblocklen,
struct xfs_btree_block *rblock,
int rblocklen)
{
+ struct xfs_mount *mp = ip->i_mount;
int dmxr;
xfs_bmbt_key_t *fkp;
__be64 *fpp;
xfs_bmbt_key_t *tkp;
__be64 *tpp;
- rblock->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ xfs_btree_init_block_int(mp, rblock, XFS_BUF_DADDR_NULL,
+ XFS_BMAP_CRC_MAGIC, 0, 0, ip->i_ino,
+ XFS_BTREE_LONG_PTRS | XFS_BTREE_CRC_BLOCKS);
+ else
+ xfs_btree_init_block_int(mp, rblock, XFS_BUF_DADDR_NULL,
+ XFS_BMAP_MAGIC, 0, 0, ip->i_ino,
+ XFS_BTREE_LONG_PTRS);
+
rblock->bb_level = dblock->bb_level;
ASSERT(be16_to_cpu(rblock->bb_level) > 0);
rblock->bb_numrecs = dblock->bb_numrecs;
- rblock->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
- rblock->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
dmxr = xfs_bmdr_maxrecs(mp, dblocklen, 0);
fkp = XFS_BMDR_KEY_ADDR(dblock, 1);
tkp = XFS_BMBT_KEY_ADDR(mp, rblock, 1);
@@ -424,7 +432,13 @@ xfs_bmbt_to_bmdr(
xfs_bmbt_key_t *tkp;
__be64 *tpp;
- ASSERT(rblock->bb_magic == cpu_to_be32(XFS_BMAP_MAGIC));
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ ASSERT(rblock->bb_magic == cpu_to_be32(XFS_BMAP_CRC_MAGIC));
+ ASSERT(uuid_equal(&rblock->bb_u.l.bb_uuid, &mp->m_sb.sb_uuid));
+ ASSERT(rblock->bb_u.l.bb_blkno ==
+ cpu_to_be64(XFS_BUF_DADDR_NULL));
+ } else
+ ASSERT(rblock->bb_magic == cpu_to_be32(XFS_BMAP_MAGIC));
ASSERT(rblock->bb_u.l.bb_leftsib == cpu_to_be64(NULLDFSBNO));
ASSERT(rblock->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO));
ASSERT(rblock->bb_level != 0);
@@ -708,59 +722,89 @@ xfs_bmbt_key_diff(
cur->bc_rec.b.br_startoff;
}
-static void
+static int
xfs_bmbt_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
unsigned int level;
- int lblock_ok; /* block passes checks */
- /* magic number and level verification.
+ switch (block->bb_magic) {
+ case cpu_to_be32(XFS_BMAP_CRC_MAGIC):
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return false;
+ if (!uuid_equal(&block->bb_u.l.bb_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be64_to_cpu(block->bb_u.l.bb_blkno) != bp->b_bn)
+ return false;
+ /*
+ * XXX: need a better way of verifying the owner here. Right now
+ * just make sure there has been one set.
+ */
+ if (be64_to_cpu(block->bb_u.l.bb_owner) == 0)
+ return false;
+ /* fall through */
+ case cpu_to_be32(XFS_BMAP_MAGIC):
+ break;
+ default:
+ return false;
+ }
+
+ /*
+ * numrecs and level verification.
*
- * We don't know waht fork we belong to, so just verify that the level
+ * We don't know what fork we belong to, so just verify that the level
* is less than the maximum of the two. Later checks will be more
* precise.
*/
level = be16_to_cpu(block->bb_level);
- lblock_ok = block->bb_magic == cpu_to_be32(XFS_BMAP_MAGIC) &&
- level < max(mp->m_bm_maxlevels[0], mp->m_bm_maxlevels[1]);
-
- /* numrecs verification */
- lblock_ok = lblock_ok &&
- be16_to_cpu(block->bb_numrecs) <= mp->m_bmap_dmxr[level != 0];
+ if (level > max(mp->m_bm_maxlevels[0], mp->m_bm_maxlevels[1]))
+ return false;
+ if (be16_to_cpu(block->bb_numrecs) > mp->m_bmap_dmxr[level != 0])
+ return false;
/* sibling pointer verification */
- lblock_ok = lblock_ok &&
- block->bb_u.l.bb_leftsib &&
- (block->bb_u.l.bb_leftsib == cpu_to_be64(NULLDFSBNO) ||
- XFS_FSB_SANITY_CHECK(mp,
- be64_to_cpu(block->bb_u.l.bb_leftsib))) &&
- block->bb_u.l.bb_rightsib &&
- (block->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO) ||
- XFS_FSB_SANITY_CHECK(mp,
- be64_to_cpu(block->bb_u.l.bb_rightsib)));
-
- if (!lblock_ok) {
- trace_xfs_btree_corrupt(bp, _RET_IP_);
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
- }
+ if (!block->bb_u.l.bb_leftsib ||
+ (block->bb_u.l.bb_leftsib != cpu_to_be64(NULLDFSBNO) &&
+ !XFS_FSB_SANITY_CHECK(mp, be64_to_cpu(block->bb_u.l.bb_leftsib))))
+ return false;
+ if (!block->bb_u.l.bb_rightsib ||
+ (block->bb_u.l.bb_rightsib != cpu_to_be64(NULLDFSBNO) &&
+ !XFS_FSB_SANITY_CHECK(mp, be64_to_cpu(block->bb_u.l.bb_rightsib))))
+ return false;
+
+ return true;
+
}
static void
xfs_bmbt_read_verify(
struct xfs_buf *bp)
{
- xfs_bmbt_verify(bp);
+ if (!(xfs_btree_lblock_verify_crc(bp) &&
+ xfs_bmbt_verify(bp))) {
+ trace_xfs_btree_corrupt(bp, _RET_IP_);
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+ bp->b_target->bt_mount, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
+
}
static void
xfs_bmbt_write_verify(
struct xfs_buf *bp)
{
- xfs_bmbt_verify(bp);
+ if (!xfs_bmbt_verify(bp)) {
+ xfs_warn(bp->b_target->bt_mount, "bmbt daddr 0x%llx failed", bp->b_bn);
+ trace_xfs_btree_corrupt(bp, _RET_IP_);
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+ bp->b_target->bt_mount, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+ xfs_btree_lblock_calc_crc(bp);
}
const struct xfs_buf_ops xfs_bmbt_buf_ops = {
@@ -838,6 +882,8 @@ xfs_bmbt_init_cursor(
cur->bc_ops = &xfs_bmbt_ops;
cur->bc_flags = XFS_BTREE_LONG_PTRS | XFS_BTREE_ROOT_IN_INODE;
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
cur->bc_private.b.forksize = XFS_IFORK_SIZE(ip, whichfork);
cur->bc_private.b.ip = ip;
diff --git a/fs/xfs/xfs_bmap_btree.h b/fs/xfs/xfs_bmap_btree.h
index 88469ca..70c43d9 100644
--- a/fs/xfs/xfs_bmap_btree.h
+++ b/fs/xfs/xfs_bmap_btree.h
@@ -18,7 +18,8 @@
#ifndef __XFS_BMAP_BTREE_H__
#define __XFS_BMAP_BTREE_H__
-#define XFS_BMAP_MAGIC 0x424d4150 /* 'BMAP' */
+#define XFS_BMAP_MAGIC 0x424d4150 /* 'BMAP' */
+#define XFS_BMAP_CRC_MAGIC 0x424d4133 /* 'BMA3' */
struct xfs_btree_cur;
struct xfs_btree_block;
@@ -136,10 +137,10 @@ typedef __be64 xfs_bmbt_ptr_t, xfs_bmdr_ptr_t;
/*
* Btree block header size depends on a superblock flag.
- *
- * (not quite yet, but soon)
*/
-#define XFS_BMBT_BLOCK_LEN(mp) XFS_BTREE_LBLOCK_LEN
+#define XFS_BMBT_BLOCK_LEN(mp) \
+ (xfs_sb_version_hascrc(&((mp)->m_sb)) ? \
+ XFS_BTREE_LBLOCK_CRC_LEN : XFS_BTREE_LBLOCK_LEN)
#define XFS_BMBT_REC_ADDR(mp, block, index) \
((xfs_bmbt_rec_t *) \
@@ -186,12 +187,12 @@ typedef __be64 xfs_bmbt_ptr_t, xfs_bmdr_ptr_t;
#define XFS_BMAP_BROOT_PTR_ADDR(mp, bb, i, sz) \
XFS_BMBT_PTR_ADDR(mp, bb, i, xfs_bmbt_maxrecs(mp, sz, 0))
-#define XFS_BMAP_BROOT_SPACE_CALC(nrecs) \
- (int)(XFS_BTREE_LBLOCK_LEN + \
+#define XFS_BMAP_BROOT_SPACE_CALC(mp, nrecs) \
+ (int)(XFS_BMBT_BLOCK_LEN(mp) + \
((nrecs) * (sizeof(xfs_bmbt_key_t) + sizeof(xfs_bmbt_ptr_t))))
-#define XFS_BMAP_BROOT_SPACE(bb) \
- (XFS_BMAP_BROOT_SPACE_CALC(be16_to_cpu((bb)->bb_numrecs)))
+#define XFS_BMAP_BROOT_SPACE(mp, bb) \
+ (XFS_BMAP_BROOT_SPACE_CALC(mp, be16_to_cpu((bb)->bb_numrecs)))
#define XFS_BMDR_SPACE_CALC(nrecs) \
(int)(sizeof(xfs_bmdr_block_t) + \
((nrecs) * (sizeof(xfs_bmbt_key_t) + sizeof(xfs_bmbt_ptr_t))))
@@ -204,7 +205,7 @@ typedef __be64 xfs_bmbt_ptr_t, xfs_bmdr_ptr_t;
/*
* Prototypes for xfs_bmap.c to call.
*/
-extern void xfs_bmdr_to_bmbt(struct xfs_mount *, xfs_bmdr_block_t *, int,
+extern void xfs_bmdr_to_bmbt(struct xfs_inode *, xfs_bmdr_block_t *, int,
struct xfs_btree_block *, int);
extern void xfs_bmbt_get_all(xfs_bmbt_rec_host_t *r, xfs_bmbt_irec_t *s);
extern xfs_filblks_t xfs_bmbt_get_blockcount(xfs_bmbt_rec_host_t *r);
diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c
index db01040..ec77036 100644
--- a/fs/xfs/xfs_btree.c
+++ b/fs/xfs/xfs_btree.c
@@ -30,9 +30,11 @@
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_inode_item.h"
+#include "xfs_buf_item.h"
#include "xfs_btree.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_cksum.h"
/*
* Cursor allocation zone.
@@ -42,9 +44,13 @@ kmem_zone_t *xfs_btree_cur_zone;
/*
* Btree magic numbers.
*/
-const __uint32_t xfs_magics[XFS_BTNUM_MAX] = {
- XFS_ABTB_MAGIC, XFS_ABTC_MAGIC, XFS_BMAP_MAGIC, XFS_IBT_MAGIC
+static const __uint32_t xfs_magics[2][XFS_BTNUM_MAX] = {
+ { XFS_ABTB_MAGIC, XFS_ABTC_MAGIC, XFS_BMAP_MAGIC, XFS_IBT_MAGIC },
+ { XFS_ABTB_CRC_MAGIC, XFS_ABTC_CRC_MAGIC,
+ XFS_BMAP_CRC_MAGIC, XFS_IBT_CRC_MAGIC }
};
+#define xfs_btree_magic(cur) \
+ xfs_magics[!!((cur)->bc_flags & XFS_BTREE_CRC_BLOCKS)][cur->bc_btnum]
STATIC int /* error (0 or EFSCORRUPTED) */
@@ -54,30 +60,38 @@ xfs_btree_check_lblock(
int level, /* level of the btree block */
struct xfs_buf *bp) /* buffer for block, if any */
{
- int lblock_ok; /* block passes checks */
+ int lblock_ok = 1; /* block passes checks */
struct xfs_mount *mp; /* file system mount point */
mp = cur->bc_mp;
- lblock_ok =
- be32_to_cpu(block->bb_magic) == xfs_magics[cur->bc_btnum] &&
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ lblock_ok = lblock_ok &&
+ uuid_equal(&block->bb_u.l.bb_uuid, &mp->m_sb.sb_uuid) &&
+ block->bb_u.l.bb_blkno == cpu_to_be64(
+ bp ? bp->b_bn : XFS_BUF_DADDR_NULL);
+ }
+
+ lblock_ok = lblock_ok &&
+ be32_to_cpu(block->bb_magic) == xfs_btree_magic(cur) &&
be16_to_cpu(block->bb_level) == level &&
be16_to_cpu(block->bb_numrecs) <=
cur->bc_ops->get_maxrecs(cur, level) &&
block->bb_u.l.bb_leftsib &&
(block->bb_u.l.bb_leftsib == cpu_to_be64(NULLDFSBNO) ||
XFS_FSB_SANITY_CHECK(mp,
- be64_to_cpu(block->bb_u.l.bb_leftsib))) &&
+ be64_to_cpu(block->bb_u.l.bb_leftsib))) &&
block->bb_u.l.bb_rightsib &&
(block->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO) ||
XFS_FSB_SANITY_CHECK(mp,
- be64_to_cpu(block->bb_u.l.bb_rightsib)));
+ be64_to_cpu(block->bb_u.l.bb_rightsib)));
+
if (unlikely(XFS_TEST_ERROR(!lblock_ok, mp,
XFS_ERRTAG_BTREE_CHECK_LBLOCK,
XFS_RANDOM_BTREE_CHECK_LBLOCK))) {
if (bp)
trace_xfs_btree_corrupt(bp, _RET_IP_);
- XFS_ERROR_REPORT("xfs_btree_check_lblock", XFS_ERRLEVEL_LOW,
- mp);
+ XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_LOW, mp);
return XFS_ERROR(EFSCORRUPTED);
}
return 0;
@@ -90,16 +104,26 @@ xfs_btree_check_sblock(
int level, /* level of the btree block */
struct xfs_buf *bp) /* buffer containing block */
{
+ struct xfs_mount *mp; /* file system mount point */
struct xfs_buf *agbp; /* buffer for ag. freespace struct */
struct xfs_agf *agf; /* ag. freespace structure */
xfs_agblock_t agflen; /* native ag. freespace length */
- int sblock_ok; /* block passes checks */
+ int sblock_ok = 1; /* block passes checks */
+ mp = cur->bc_mp;
agbp = cur->bc_private.a.agbp;
agf = XFS_BUF_TO_AGF(agbp);
agflen = be32_to_cpu(agf->agf_length);
- sblock_ok =
- be32_to_cpu(block->bb_magic) == xfs_magics[cur->bc_btnum] &&
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ sblock_ok = sblock_ok &&
+ uuid_equal(&block->bb_u.s.bb_uuid, &mp->m_sb.sb_uuid) &&
+ block->bb_u.s.bb_blkno == cpu_to_be64(
+ bp ? bp->b_bn : XFS_BUF_DADDR_NULL);
+ }
+
+ sblock_ok = sblock_ok &&
+ be32_to_cpu(block->bb_magic) == xfs_btree_magic(cur) &&
be16_to_cpu(block->bb_level) == level &&
be16_to_cpu(block->bb_numrecs) <=
cur->bc_ops->get_maxrecs(cur, level) &&
@@ -109,13 +133,13 @@ xfs_btree_check_sblock(
(block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) ||
be32_to_cpu(block->bb_u.s.bb_rightsib) < agflen) &&
block->bb_u.s.bb_rightsib;
- if (unlikely(XFS_TEST_ERROR(!sblock_ok, cur->bc_mp,
+
+ if (unlikely(XFS_TEST_ERROR(!sblock_ok, mp,
XFS_ERRTAG_BTREE_CHECK_SBLOCK,
XFS_RANDOM_BTREE_CHECK_SBLOCK))) {
if (bp)
trace_xfs_btree_corrupt(bp, _RET_IP_);
- XFS_CORRUPTION_ERROR("xfs_btree_check_sblock",
- XFS_ERRLEVEL_LOW, cur->bc_mp, block);
+ XFS_ERROR_REPORT(__func__, XFS_ERRLEVEL_LOW, mp);
return XFS_ERROR(EFSCORRUPTED);
}
return 0;
@@ -194,6 +218,72 @@ xfs_btree_check_ptr(
#endif
/*
+ * Calculate CRC on the whole btree block and stuff it into the
+ * long-form btree header.
+ *
+ * Prior to calculting the CRC, pull the LSN out of the buffer log item and put
+ * it into the buffer so recovery knows what the last modifcation was that made
+ * it to disk.
+ */
+void
+xfs_btree_lblock_calc_crc(
+ struct xfs_buf *bp)
+{
+ struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ if (!xfs_sb_version_hascrc(&bp->b_target->bt_mount->m_sb))
+ return;
+ if (bip)
+ block->bb_u.l.bb_lsn = cpu_to_be64(bip->bli_item.li_lsn);
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_BTREE_LBLOCK_CRC_OFF);
+}
+
+bool
+xfs_btree_lblock_verify_crc(
+ struct xfs_buf *bp)
+{
+ if (xfs_sb_version_hascrc(&bp->b_target->bt_mount->m_sb))
+ return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_BTREE_LBLOCK_CRC_OFF);
+ return true;
+}
+
+/*
+ * Calculate CRC on the whole btree block and stuff it into the
+ * short-form btree header.
+ *
+ * Prior to calculting the CRC, pull the LSN out of the buffer log item and put
+ * it into the buffer so recovery knows what the last modifcation was that made
+ * it to disk.
+ */
+void
+xfs_btree_sblock_calc_crc(
+ struct xfs_buf *bp)
+{
+ struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ if (!xfs_sb_version_hascrc(&bp->b_target->bt_mount->m_sb))
+ return;
+ if (bip)
+ block->bb_u.s.bb_lsn = cpu_to_be64(bip->bli_item.li_lsn);
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_BTREE_SBLOCK_CRC_OFF);
+}
+
+bool
+xfs_btree_sblock_verify_crc(
+ struct xfs_buf *bp)
+{
+ if (xfs_sb_version_hascrc(&bp->b_target->bt_mount->m_sb))
+ return xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_BTREE_SBLOCK_CRC_OFF);
+ return true;
+}
+
+/*
* Delete the btree cursor.
*/
void
@@ -277,10 +367,8 @@ xfs_btree_dup_cursor(
*ncur = NULL;
return error;
}
- new->bc_bufs[i] = bp;
- ASSERT(!xfs_buf_geterror(bp));
- } else
- new->bc_bufs[i] = NULL;
+ }
+ new->bc_bufs[i] = bp;
}
*ncur = new;
return 0;
@@ -321,9 +409,14 @@ xfs_btree_dup_cursor(
*/
static inline size_t xfs_btree_block_len(struct xfs_btree_cur *cur)
{
- return (cur->bc_flags & XFS_BTREE_LONG_PTRS) ?
- XFS_BTREE_LBLOCK_LEN :
- XFS_BTREE_SBLOCK_LEN;
+ if (cur->bc_flags & XFS_BTREE_LONG_PTRS) {
+ if (cur->bc_flags & XFS_BTREE_CRC_BLOCKS)
+ return XFS_BTREE_LBLOCK_CRC_LEN;
+ return XFS_BTREE_LBLOCK_LEN;
+ }
+ if (cur->bc_flags & XFS_BTREE_CRC_BLOCKS)
+ return XFS_BTREE_SBLOCK_CRC_LEN;
+ return XFS_BTREE_SBLOCK_LEN;
}
/*
@@ -863,43 +956,85 @@ xfs_btree_set_sibling(
}
void
+xfs_btree_init_block_int(
+ struct xfs_mount *mp,
+ struct xfs_btree_block *buf,
+ xfs_daddr_t blkno,
+ __u32 magic,
+ __u16 level,
+ __u16 numrecs,
+ __u64 owner,
+ unsigned int flags)
+{
+ buf->bb_magic = cpu_to_be32(magic);
+ buf->bb_level = cpu_to_be16(level);
+ buf->bb_numrecs = cpu_to_be16(numrecs);
+
+ if (flags & XFS_BTREE_LONG_PTRS) {
+ buf->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
+ buf->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
+ if (flags & XFS_BTREE_CRC_BLOCKS) {
+ buf->bb_u.l.bb_blkno = cpu_to_be64(blkno);
+ buf->bb_u.l.bb_owner = cpu_to_be64(owner);
+ uuid_copy(&buf->bb_u.l.bb_uuid, &mp->m_sb.sb_uuid);
+ buf->bb_u.l.bb_pad = 0;
+ }
+ } else {
+ /* owner is a 32 bit value on short blocks */
+ __u32 __owner = (__u32)owner;
+
+ buf->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK);
+ buf->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK);
+ if (flags & XFS_BTREE_CRC_BLOCKS) {
+ buf->bb_u.s.bb_blkno = cpu_to_be64(blkno);
+ buf->bb_u.s.bb_owner = cpu_to_be32(__owner);
+ uuid_copy(&buf->bb_u.s.bb_uuid, &mp->m_sb.sb_uuid);
+ }
+ }
+}
+
+void
xfs_btree_init_block(
struct xfs_mount *mp,
struct xfs_buf *bp,
__u32 magic,
__u16 level,
__u16 numrecs,
+ __u64 owner,
unsigned int flags)
{
- struct xfs_btree_block *new = XFS_BUF_TO_BLOCK(bp);
-
- new->bb_magic = cpu_to_be32(magic);
- new->bb_level = cpu_to_be16(level);
- new->bb_numrecs = cpu_to_be16(numrecs);
-
- if (flags & XFS_BTREE_LONG_PTRS) {
- new->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO);
- new->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO);
- } else {
- new->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK);
- new->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK);
- }
+ xfs_btree_init_block_int(mp, XFS_BUF_TO_BLOCK(bp), bp->b_bn,
+ magic, level, numrecs, owner, flags);
}
STATIC void
xfs_btree_init_block_cur(
struct xfs_btree_cur *cur,
+ struct xfs_buf *bp,
int level,
- int numrecs,
- struct xfs_buf *bp)
+ int numrecs)
{
- xfs_btree_init_block(cur->bc_mp, bp, xfs_magics[cur->bc_btnum],
- level, numrecs, cur->bc_flags);
+ __u64 owner;
+
+ /*
+ * we can pull the owner from the cursor right now as the different
+ * owners align directly with the pointer size of the btree. This may
+ * change in future, but is safe for current users of the generic btree
+ * code.
+ */
+ if (cur->bc_flags & XFS_BTREE_LONG_PTRS)
+ owner = cur->bc_private.b.ip->i_ino;
+ else
+ owner = cur->bc_private.a.agno;
+
+ xfs_btree_init_block_int(cur->bc_mp, XFS_BUF_TO_BLOCK(bp), bp->b_bn,
+ xfs_btree_magic(cur), level, numrecs,
+ owner, cur->bc_flags);
}
/*
* Return true if ptr is the last record in the btree and
- * we need to track updateѕ to this record. The decision
+ * we need to track updates to this record. The decision
* will be further refined in the update_lastrec method.
*/
STATIC int
@@ -1147,6 +1282,7 @@ xfs_btree_log_keys(
XFS_BTREE_TRACE_ARGBII(cur, bp, first, last);
if (bp) {
+ xfs_trans_buf_set_type(cur->bc_tp, bp, XFS_BLF_BTREE_BUF);
xfs_trans_log_buf(cur->bc_tp, bp,
xfs_btree_key_offset(cur, first),
xfs_btree_key_offset(cur, last + 1) - 1);
@@ -1171,6 +1307,7 @@ xfs_btree_log_recs(
XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY);
XFS_BTREE_TRACE_ARGBII(cur, bp, first, last);
+ xfs_trans_buf_set_type(cur->bc_tp, bp, XFS_BLF_BTREE_BUF);
xfs_trans_log_buf(cur->bc_tp, bp,
xfs_btree_rec_offset(cur, first),
xfs_btree_rec_offset(cur, last + 1) - 1);
@@ -1195,6 +1332,7 @@ xfs_btree_log_ptrs(
struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp);
int level = xfs_btree_get_level(block);
+ xfs_trans_buf_set_type(cur->bc_tp, bp, XFS_BLF_BTREE_BUF);
xfs_trans_log_buf(cur->bc_tp, bp,
xfs_btree_ptr_offset(cur, first, level),
xfs_btree_ptr_offset(cur, last + 1, level) - 1);
@@ -1223,7 +1361,12 @@ xfs_btree_log_block(
offsetof(struct xfs_btree_block, bb_numrecs),
offsetof(struct xfs_btree_block, bb_u.s.bb_leftsib),
offsetof(struct xfs_btree_block, bb_u.s.bb_rightsib),
- XFS_BTREE_SBLOCK_LEN
+ offsetof(struct xfs_btree_block, bb_u.s.bb_blkno),
+ offsetof(struct xfs_btree_block, bb_u.s.bb_lsn),
+ offsetof(struct xfs_btree_block, bb_u.s.bb_uuid),
+ offsetof(struct xfs_btree_block, bb_u.s.bb_owner),
+ offsetof(struct xfs_btree_block, bb_u.s.bb_crc),
+ XFS_BTREE_SBLOCK_CRC_LEN
};
static const short loffsets[] = { /* table of offsets (long) */
offsetof(struct xfs_btree_block, bb_magic),
@@ -1231,17 +1374,40 @@ xfs_btree_log_block(
offsetof(struct xfs_btree_block, bb_numrecs),
offsetof(struct xfs_btree_block, bb_u.l.bb_leftsib),
offsetof(struct xfs_btree_block, bb_u.l.bb_rightsib),
- XFS_BTREE_LBLOCK_LEN
+ offsetof(struct xfs_btree_block, bb_u.l.bb_blkno),
+ offsetof(struct xfs_btree_block, bb_u.l.bb_lsn),
+ offsetof(struct xfs_btree_block, bb_u.l.bb_uuid),
+ offsetof(struct xfs_btree_block, bb_u.l.bb_owner),
+ offsetof(struct xfs_btree_block, bb_u.l.bb_crc),
+ offsetof(struct xfs_btree_block, bb_u.l.bb_pad),
+ XFS_BTREE_LBLOCK_CRC_LEN
};
XFS_BTREE_TRACE_CURSOR(cur, XBT_ENTRY);
XFS_BTREE_TRACE_ARGBI(cur, bp, fields);
if (bp) {
+ int nbits;
+
+ if (cur->bc_flags & XFS_BTREE_CRC_BLOCKS) {
+ /*
+ * We don't log the CRC when updating a btree
+ * block but instead recreate it during log
+ * recovery. As the log buffers have checksums
+ * of their own this is safe and avoids logging a crc
+ * update in a lot of places.
+ */
+ if (fields == XFS_BB_ALL_BITS)
+ fields = XFS_BB_ALL_BITS_CRC;
+ nbits = XFS_BB_NUM_BITS_CRC;
+ } else {
+ nbits = XFS_BB_NUM_BITS;
+ }
xfs_btree_offsets(fields,
(cur->bc_flags & XFS_BTREE_LONG_PTRS) ?
loffsets : soffsets,
- XFS_BB_NUM_BITS, &first, &last);
+ nbits, &first, &last);
+ xfs_trans_buf_set_type(cur->bc_tp, bp, XFS_BLF_BTREE_BUF);
xfs_trans_log_buf(cur->bc_tp, bp, first, last);
} else {
xfs_trans_log_inode(cur->bc_tp, cur->bc_private.b.ip,
@@ -2204,7 +2370,7 @@ xfs_btree_split(
goto error0;
/* Fill in the btree header for the new right block. */
- xfs_btree_init_block_cur(cur, xfs_btree_get_level(left), 0, rbp);
+ xfs_btree_init_block_cur(cur, rbp, xfs_btree_get_level(left), 0);
/*
* Split the entries between the old and the new block evenly.
@@ -2513,7 +2679,7 @@ xfs_btree_new_root(
nptr = 2;
}
/* Fill in the new block's btree header and log it. */
- xfs_btree_init_block_cur(cur, cur->bc_nlevels, 2, nbp);
+ xfs_btree_init_block_cur(cur, nbp, cur->bc_nlevels, 2);
xfs_btree_log_block(cur, nbp, XFS_BB_ALL_BITS);
ASSERT(!xfs_btree_ptr_is_null(cur, &lptr) &&
!xfs_btree_ptr_is_null(cur, &rptr));
diff --git a/fs/xfs/xfs_btree.h b/fs/xfs/xfs_btree.h
index f932897..6e6c915 100644
--- a/fs/xfs/xfs_btree.h
+++ b/fs/xfs/xfs_btree.h
@@ -42,11 +42,15 @@ extern kmem_zone_t *xfs_btree_cur_zone;
* Generic btree header.
*
* This is a combination of the actual format used on disk for short and long
- * format btrees. The first three fields are shared by both format, but
- * the pointers are different and should be used with care.
+ * format btrees. The first three fields are shared by both format, but the
+ * pointers are different and should be used with care.
*
- * To get the size of the actual short or long form headers please use
- * the size macros below. Never use sizeof(xfs_btree_block).
+ * To get the size of the actual short or long form headers please use the size
+ * macros below. Never use sizeof(xfs_btree_block).
+ *
+ * The blkno, crc, lsn, owner and uuid fields are only available in filesystems
+ * with the crc feature bit, and all accesses to them must be conditional on
+ * that flag.
*/
struct xfs_btree_block {
__be32 bb_magic; /* magic number for block type */
@@ -56,10 +60,23 @@ struct xfs_btree_block {
struct {
__be32 bb_leftsib;
__be32 bb_rightsib;
+
+ __be64 bb_blkno;
+ __be64 bb_lsn;
+ uuid_t bb_uuid;
+ __be32 bb_owner;
+ __le32 bb_crc;
} s; /* short form pointers */
struct {
__be64 bb_leftsib;
__be64 bb_rightsib;
+
+ __be64 bb_blkno;
+ __be64 bb_lsn;
+ uuid_t bb_uuid;
+ __be64 bb_owner;
+ __le32 bb_crc;
+ __be32 bb_pad; /* padding for alignment */
} l; /* long form pointers */
} bb_u; /* rest */
};
@@ -67,6 +84,16 @@ struct xfs_btree_block {
#define XFS_BTREE_SBLOCK_LEN 16 /* size of a short form block */
#define XFS_BTREE_LBLOCK_LEN 24 /* size of a long form block */
+/* sizes of CRC enabled btree blocks */
+#define XFS_BTREE_SBLOCK_CRC_LEN (XFS_BTREE_SBLOCK_LEN + 40)
+#define XFS_BTREE_LBLOCK_CRC_LEN (XFS_BTREE_LBLOCK_LEN + 48)
+
+
+#define XFS_BTREE_SBLOCK_CRC_OFF \
+ offsetof(struct xfs_btree_block, bb_u.s.bb_crc)
+#define XFS_BTREE_LBLOCK_CRC_OFF \
+ offsetof(struct xfs_btree_block, bb_u.l.bb_crc)
+
/*
* Generic key, ptr and record wrapper structures.
@@ -101,13 +128,11 @@ union xfs_btree_rec {
#define XFS_BB_NUMRECS 0x04
#define XFS_BB_LEFTSIB 0x08
#define XFS_BB_RIGHTSIB 0x10
+#define XFS_BB_BLKNO 0x20
#define XFS_BB_NUM_BITS 5
#define XFS_BB_ALL_BITS ((1 << XFS_BB_NUM_BITS) - 1)
-
-/*
- * Magic numbers for btree blocks.
- */
-extern const __uint32_t xfs_magics[];
+#define XFS_BB_NUM_BITS_CRC 8
+#define XFS_BB_ALL_BITS_CRC ((1 << XFS_BB_NUM_BITS_CRC) - 1)
/*
* Generic stats interface
@@ -256,6 +281,7 @@ typedef struct xfs_btree_cur
#define XFS_BTREE_LONG_PTRS (1<<0) /* pointers are 64bits long */
#define XFS_BTREE_ROOT_IN_INODE (1<<1) /* root may be variable size */
#define XFS_BTREE_LASTREC_UPDATE (1<<2) /* track last rec externally */
+#define XFS_BTREE_CRC_BLOCKS (1<<3) /* uses extended btree blocks */
#define XFS_BTREE_NOERROR 0
@@ -393,8 +419,20 @@ xfs_btree_init_block(
__u32 magic,
__u16 level,
__u16 numrecs,
+ __u64 owner,
unsigned int flags);
+void
+xfs_btree_init_block_int(
+ struct xfs_mount *mp,
+ struct xfs_btree_block *buf,
+ xfs_daddr_t blkno,
+ __u32 magic,
+ __u16 level,
+ __u16 numrecs,
+ __u64 owner,
+ unsigned int flags);
+
/*
* Common btree core entry points.
*/
@@ -408,6 +446,14 @@ int xfs_btree_delete(struct xfs_btree_cur *, int *);
int xfs_btree_get_rec(struct xfs_btree_cur *, union xfs_btree_rec **, int *);
/*
+ * btree block CRC helpers
+ */
+void xfs_btree_lblock_calc_crc(struct xfs_buf *);
+bool xfs_btree_lblock_verify_crc(struct xfs_buf *);
+void xfs_btree_sblock_calc_crc(struct xfs_buf *);
+bool xfs_btree_sblock_verify_crc(struct xfs_buf *);
+
+/*
* Internal btree helpers also used by xfs_bmap.c.
*/
void xfs_btree_log_block(struct xfs_btree_cur *, struct xfs_buf *, int);
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index ee36c88..101ef83 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -24,19 +24,33 @@ extern kmem_zone_t *xfs_buf_item_zone;
* This flag indicates that the buffer contains on disk inodes
* and requires special recovery handling.
*/
-#define XFS_BLF_INODE_BUF 0x1
+#define XFS_BLF_INODE_BUF (1<<0)
/*
* This flag indicates that the buffer should not be replayed
* during recovery because its blocks are being freed.
*/
-#define XFS_BLF_CANCEL 0x2
+#define XFS_BLF_CANCEL (1<<1)
+
/*
* This flag indicates that the buffer contains on disk
* user or group dquots and may require special recovery handling.
*/
-#define XFS_BLF_UDQUOT_BUF 0x4
-#define XFS_BLF_PDQUOT_BUF 0x8
-#define XFS_BLF_GDQUOT_BUF 0x10
+#define XFS_BLF_UDQUOT_BUF (1<<2)
+#define XFS_BLF_PDQUOT_BUF (1<<3)
+#define XFS_BLF_GDQUOT_BUF (1<<4)
+
+/*
+ * all buffers now need flags to tell recovery where the magic number
+ * is so that it can verify and calculate the CRCs on the buffer correctly
+ * once the changes have been replayed into the buffer.
+ */
+#define XFS_BLF_BTREE_BUF (1<<5)
+
+#define XFS_BLF_TYPE_MASK \
+ (XFS_BLF_UDQUOT_BUF | \
+ XFS_BLF_PDQUOT_BUF | \
+ XFS_BLF_GDQUOT_BUF | \
+ XFS_BLF_BTREE_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_dinode.h b/fs/xfs/xfs_dinode.h
index 88a3368..6b5bd17 100644
--- a/fs/xfs/xfs_dinode.h
+++ b/fs/xfs/xfs_dinode.h
@@ -107,8 +107,8 @@ typedef enum xfs_dinode_fmt {
#define XFS_LITINO(mp, version) \
((int)(((mp)->m_sb.sb_inodesize) - sizeof(struct xfs_dinode)))
-#define XFS_BROOT_SIZE_ADJ \
- (XFS_BTREE_LBLOCK_LEN - sizeof(xfs_bmdr_block_t))
+#define XFS_BROOT_SIZE_ADJ(ip) \
+ (XFS_BMBT_BLOCK_LEN((ip)->i_mount) - sizeof(xfs_bmdr_block_t))
/*
* Inode data & attribute fork sizes, per inode.
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index 2866b8c..6fe286a 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -316,7 +316,13 @@ xfs_growfs_data_private(
goto error0;
}
- xfs_btree_init_block(mp, bp, XFS_ABTB_MAGIC, 0, 1, 0);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ xfs_btree_init_block(mp, bp, XFS_ABTB_CRC_MAGIC, 0, 1,
+ agno, XFS_BTREE_CRC_BLOCKS);
+ else
+ xfs_btree_init_block(mp, bp, XFS_ABTB_MAGIC, 0, 1,
+ agno, 0);
+
arec = XFS_ALLOC_REC_ADDR(mp, XFS_BUF_TO_BLOCK(bp), 1);
arec->ar_startblock = cpu_to_be32(XFS_PREALLOC_BLOCKS(mp));
arec->ar_blockcount = cpu_to_be32(
@@ -339,7 +345,13 @@ xfs_growfs_data_private(
goto error0;
}
- xfs_btree_init_block(mp, bp, XFS_ABTC_MAGIC, 0, 1, 0);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ xfs_btree_init_block(mp, bp, XFS_ABTC_CRC_MAGIC, 0, 1,
+ agno, XFS_BTREE_CRC_BLOCKS);
+ else
+ xfs_btree_init_block(mp, bp, XFS_ABTC_MAGIC, 0, 1,
+ agno, 0);
+
arec = XFS_ALLOC_REC_ADDR(mp, XFS_BUF_TO_BLOCK(bp), 1);
arec->ar_startblock = cpu_to_be32(XFS_PREALLOC_BLOCKS(mp));
arec->ar_blockcount = cpu_to_be32(
@@ -363,7 +375,12 @@ xfs_growfs_data_private(
goto error0;
}
- xfs_btree_init_block(mp, bp, XFS_IBT_MAGIC, 0, 0, 0);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ xfs_btree_init_block(mp, bp, XFS_IBT_CRC_MAGIC, 0, 0,
+ agno, XFS_BTREE_CRC_BLOCKS);
+ else
+ xfs_btree_init_block(mp, bp, XFS_IBT_MAGIC, 0, 0,
+ agno, 0);
error = xfs_bwrite(bp);
xfs_buf_relse(bp);
diff --git a/fs/xfs/xfs_ialloc_btree.c b/fs/xfs/xfs_ialloc_btree.c
index bec344b..c82ac88 100644
--- a/fs/xfs/xfs_ialloc_btree.c
+++ b/fs/xfs/xfs_ialloc_btree.c
@@ -34,6 +34,7 @@
#include "xfs_alloc.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_cksum.h"
STATIC int
@@ -182,52 +183,88 @@ xfs_inobt_key_diff(
cur->bc_rec.i.ir_startino;
}
-void
+static int
xfs_inobt_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;
- int sblock_ok; /* block passes checks */
- /* magic number and level verification */
- level = be16_to_cpu(block->bb_level);
- sblock_ok = block->bb_magic == cpu_to_be32(XFS_IBT_MAGIC) &&
- level < mp->m_in_maxlevels;
+ /*
+ * During growfs operations, we can't verify the exact owner as the
+ * perag is not fully initialised and hence not attached to the buffer.
+ *
+ * Similarly, during log recovery we will have a perag structure
+ * attached, but the agi information will not yet have been initialised
+ * from the on disk AGI. We don't currently use any of this information,
+ * but beware of the landmine (i.e. need to check pag->pagi_init) if we
+ * ever do.
+ */
+ switch (block->bb_magic) {
+ case cpu_to_be32(XFS_IBT_CRC_MAGIC):
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return false;
+ if (!uuid_equal(&block->bb_u.s.bb_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (block->bb_u.s.bb_blkno != cpu_to_be64(bp->b_bn))
+ return false;
+ if (pag &&
+ be32_to_cpu(block->bb_u.s.bb_owner) != pag->pag_agno)
+ return false;
+ /* fall through */
+ case cpu_to_be32(XFS_IBT_MAGIC):
+ break;
+ default:
+ return 0;
+ }
- /* numrecs verification */
- sblock_ok = sblock_ok &&
- be16_to_cpu(block->bb_numrecs) <= mp->m_inobt_mxr[level != 0];
+ /* numrecs and level verification */
+ level = be16_to_cpu(block->bb_level);
+ if (level >= mp->m_in_maxlevels)
+ return false;
+ if (be16_to_cpu(block->bb_numrecs) > mp->m_inobt_mxr[level != 0])
+ return false;
/* sibling pointer verification */
- sblock_ok = sblock_ok &&
- (block->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK) ||
- be32_to_cpu(block->bb_u.s.bb_leftsib) < mp->m_sb.sb_agblocks) &&
- block->bb_u.s.bb_leftsib &&
- (block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) ||
- be32_to_cpu(block->bb_u.s.bb_rightsib) < mp->m_sb.sb_agblocks) &&
- block->bb_u.s.bb_rightsib;
-
- if (!sblock_ok) {
- trace_xfs_btree_corrupt(bp, _RET_IP_);
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
- }
+ if (!block->bb_u.s.bb_leftsib ||
+ (be32_to_cpu(block->bb_u.s.bb_leftsib) >= mp->m_sb.sb_agblocks &&
+ block->bb_u.s.bb_leftsib != cpu_to_be32(NULLAGBLOCK)))
+ return false;
+ if (!block->bb_u.s.bb_rightsib ||
+ (be32_to_cpu(block->bb_u.s.bb_rightsib) >= mp->m_sb.sb_agblocks &&
+ block->bb_u.s.bb_rightsib != cpu_to_be32(NULLAGBLOCK)))
+ return false;
+
+ return true;
}
static void
xfs_inobt_read_verify(
struct xfs_buf *bp)
{
- xfs_inobt_verify(bp);
+ if (!(xfs_btree_sblock_verify_crc(bp) &&
+ xfs_inobt_verify(bp))) {
+ trace_xfs_btree_corrupt(bp, _RET_IP_);
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+ bp->b_target->bt_mount, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
static void
xfs_inobt_write_verify(
struct xfs_buf *bp)
{
- xfs_inobt_verify(bp);
+ if (!xfs_inobt_verify(bp)) {
+ trace_xfs_btree_corrupt(bp, _RET_IP_);
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
+ bp->b_target->bt_mount, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
+ xfs_btree_sblock_calc_crc(bp);
+
}
const struct xfs_buf_ops xfs_inobt_buf_ops = {
@@ -301,6 +338,8 @@ xfs_inobt_init_cursor(
cur->bc_blocklog = mp->m_sb.sb_blocklog;
cur->bc_ops = &xfs_inobt_ops;
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ cur->bc_flags |= XFS_BTREE_CRC_BLOCKS;
cur->bc_private.a.agbp = agbp;
cur->bc_private.a.agno = agno;
diff --git a/fs/xfs/xfs_ialloc_btree.h b/fs/xfs/xfs_ialloc_btree.h
index 25c0239..3ac36b76 100644
--- a/fs/xfs/xfs_ialloc_btree.h
+++ b/fs/xfs/xfs_ialloc_btree.h
@@ -29,7 +29,8 @@ struct xfs_mount;
/*
* There is a btree for the inode map per allocation group.
*/
-#define XFS_IBT_MAGIC 0x49414254 /* 'IABT' */
+#define XFS_IBT_MAGIC 0x49414254 /* 'IABT' */
+#define XFS_IBT_CRC_MAGIC 0x49414233 /* 'IAB3' */
typedef __uint64_t xfs_inofree_t;
#define XFS_INODES_PER_CHUNK (NBBY * sizeof(xfs_inofree_t))
@@ -76,10 +77,10 @@ typedef __be32 xfs_inobt_ptr_t;
/*
* Btree block header size depends on a superblock flag.
- *
- * (not quite yet, but soon)
*/
-#define XFS_INOBT_BLOCK_LEN(mp) XFS_BTREE_SBLOCK_LEN
+#define XFS_INOBT_BLOCK_LEN(mp) \
+ (xfs_sb_version_hascrc(&((mp)->m_sb)) ? \
+ XFS_BTREE_SBLOCK_CRC_LEN : XFS_BTREE_SBLOCK_LEN)
/*
* Record, key, and pointer address macros for btree blocks.
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 4f20165..202ce37 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -786,6 +786,7 @@ xfs_iformat_btree(
xfs_dinode_t *dip,
int whichfork)
{
+ struct xfs_mount *mp = ip->i_mount;
xfs_bmdr_block_t *dfp;
xfs_ifork_t *ifp;
/* REFERENCED */
@@ -794,7 +795,7 @@ xfs_iformat_btree(
ifp = XFS_IFORK_PTR(ip, whichfork);
dfp = (xfs_bmdr_block_t *)XFS_DFORK_PTR(dip, whichfork);
- size = XFS_BMAP_BROOT_SPACE(dfp);
+ size = XFS_BMAP_BROOT_SPACE(mp, dfp);
nrecs = be16_to_cpu(dfp->bb_numrecs);
/*
@@ -805,14 +806,14 @@ xfs_iformat_btree(
* blocks.
*/
if (unlikely(XFS_IFORK_NEXTENTS(ip, whichfork) <=
- XFS_IFORK_MAXEXT(ip, whichfork) ||
+ XFS_IFORK_MAXEXT(ip, whichfork) ||
XFS_BMDR_SPACE_CALC(nrecs) >
- XFS_DFORK_SIZE(dip, ip->i_mount, whichfork) ||
+ XFS_DFORK_SIZE(dip, mp, whichfork) ||
XFS_IFORK_NEXTENTS(ip, whichfork) > ip->i_d.di_nblocks)) {
- xfs_warn(ip->i_mount, "corrupt inode %Lu (btree).",
- (unsigned long long) ip->i_ino);
+ xfs_warn(mp, "corrupt inode %Lu (btree).",
+ (unsigned long long) ip->i_ino);
XFS_CORRUPTION_ERROR("xfs_iformat_btree", XFS_ERRLEVEL_LOW,
- ip->i_mount, dip);
+ mp, dip);
return XFS_ERROR(EFSCORRUPTED);
}
@@ -823,8 +824,7 @@ xfs_iformat_btree(
* Copy and convert from the on-disk structure
* to the in-memory structure.
*/
- xfs_bmdr_to_bmbt(ip->i_mount, dfp,
- XFS_DFORK_SIZE(dip, ip->i_mount, whichfork),
+ xfs_bmdr_to_bmbt(ip, dfp, XFS_DFORK_SIZE(dip, ip->i_mount, whichfork),
ifp->if_broot, size);
ifp->if_flags &= ~XFS_IFEXTENTS;
ifp->if_flags |= XFS_IFBROOT;
@@ -2037,7 +2037,7 @@ xfs_iroot_realloc(
* allocate it now and get out.
*/
if (ifp->if_broot_bytes == 0) {
- new_size = (size_t)XFS_BMAP_BROOT_SPACE_CALC(rec_diff);
+ new_size = XFS_BMAP_BROOT_SPACE_CALC(mp, rec_diff);
ifp->if_broot = kmem_alloc(new_size, KM_SLEEP | KM_NOFS);
ifp->if_broot_bytes = (int)new_size;
return;
@@ -2051,9 +2051,9 @@ xfs_iroot_realloc(
*/
cur_max = xfs_bmbt_maxrecs(mp, ifp->if_broot_bytes, 0);
new_max = cur_max + rec_diff;
- new_size = (size_t)XFS_BMAP_BROOT_SPACE_CALC(new_max);
+ new_size = XFS_BMAP_BROOT_SPACE_CALC(mp, new_max);
ifp->if_broot = kmem_realloc(ifp->if_broot, new_size,
- (size_t)XFS_BMAP_BROOT_SPACE_CALC(cur_max), /* old size */
+ XFS_BMAP_BROOT_SPACE_CALC(mp, cur_max),
KM_SLEEP | KM_NOFS);
op = (char *)XFS_BMAP_BROOT_PTR_ADDR(mp, ifp->if_broot, 1,
ifp->if_broot_bytes);
@@ -2061,7 +2061,7 @@ xfs_iroot_realloc(
(int)new_size);
ifp->if_broot_bytes = (int)new_size;
ASSERT(ifp->if_broot_bytes <=
- XFS_IFORK_SIZE(ip, whichfork) + XFS_BROOT_SIZE_ADJ);
+ XFS_IFORK_SIZE(ip, whichfork) + XFS_BROOT_SIZE_ADJ(ip));
memmove(np, op, cur_max * (uint)sizeof(xfs_dfsbno_t));
return;
}
@@ -2076,7 +2076,7 @@ xfs_iroot_realloc(
new_max = cur_max + rec_diff;
ASSERT(new_max >= 0);
if (new_max > 0)
- new_size = (size_t)XFS_BMAP_BROOT_SPACE_CALC(new_max);
+ new_size = XFS_BMAP_BROOT_SPACE_CALC(mp, new_max);
else
new_size = 0;
if (new_size > 0) {
@@ -2084,7 +2084,8 @@ xfs_iroot_realloc(
/*
* First copy over the btree block header.
*/
- memcpy(new_broot, ifp->if_broot, XFS_BTREE_LBLOCK_LEN);
+ memcpy(new_broot, ifp->if_broot,
+ XFS_BMBT_BLOCK_LEN(ip->i_mount));
} else {
new_broot = NULL;
ifp->if_flags &= ~XFS_IFBROOT;
@@ -2114,7 +2115,7 @@ xfs_iroot_realloc(
ifp->if_broot = new_broot;
ifp->if_broot_bytes = (int)new_size;
ASSERT(ifp->if_broot_bytes <=
- XFS_IFORK_SIZE(ip, whichfork) + XFS_BROOT_SIZE_ADJ);
+ XFS_IFORK_SIZE(ip, whichfork) + XFS_BROOT_SIZE_ADJ(ip));
return;
}
@@ -2427,7 +2428,7 @@ xfs_iflush_fork(
ASSERT(ifp->if_broot != NULL);
ASSERT(ifp->if_broot_bytes <=
(XFS_IFORK_SIZE(ip, whichfork) +
- XFS_BROOT_SIZE_ADJ));
+ XFS_BROOT_SIZE_ADJ(ip)));
xfs_bmbt_to_bmdr(mp, ifp->if_broot, ifp->if_broot_bytes,
(xfs_bmdr_block_t *)cp,
XFS_DFORK_SIZE(dip, mp, whichfork));
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 96fcbb8..c57a987 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -29,6 +29,7 @@
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
+#include "xfs_btree.h"
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_inode_item.h"
@@ -1929,6 +1930,33 @@ xlog_recover_do_reg_buffer(
/* Shouldn't be any more regions */
ASSERT(i == item->ri_total);
+
+ switch (buf_f->blf_flags & XFS_BLF_TYPE_MASK) {
+ case XFS_BLF_BTREE_BUF:
+ switch (be32_to_cpu(*(__be32 *)bp->b_addr)) {
+ case XFS_ABTB_CRC_MAGIC:
+ case XFS_ABTC_CRC_MAGIC:
+ case XFS_ABTB_MAGIC:
+ case XFS_ABTC_MAGIC:
+ bp->b_ops = &xfs_allocbt_buf_ops;
+ break;
+ case XFS_IBT_CRC_MAGIC:
+ case XFS_IBT_MAGIC:
+ bp->b_ops = &xfs_inobt_buf_ops;
+ break;
+ case XFS_BMAP_CRC_MAGIC:
+ case XFS_BMAP_MAGIC:
+ bp->b_ops = &xfs_bmbt_buf_ops;
+ break;
+ default:
+ xfs_warn(mp, "Bad btree block magic!");
+ ASSERT(0);
+ break;
+ }
+ break;
+ default:
+ break;
+ }
}
/*
diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h
index cd29f61..1b04fe5 100644
--- a/fs/xfs/xfs_trans.h
+++ b/fs/xfs/xfs_trans.h
@@ -505,6 +505,8 @@ void xfs_trans_inode_buf(xfs_trans_t *, struct xfs_buf *);
void xfs_trans_stale_inode_buf(xfs_trans_t *, struct xfs_buf *);
void xfs_trans_dquot_buf(xfs_trans_t *, struct xfs_buf *, uint);
void xfs_trans_inode_alloc_buf(xfs_trans_t *, struct xfs_buf *);
+void xfs_trans_buf_set_type(struct xfs_trans *, struct xfs_buf *,
+ uint);
void xfs_trans_ichgtime(struct xfs_trans *, struct xfs_inode *, int);
void xfs_trans_ijoin(struct xfs_trans *, struct xfs_inode *, uint);
void xfs_trans_log_buf(xfs_trans_t *, struct xfs_buf *, uint, uint);
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index 3edf5db..f950edd 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -659,6 +659,7 @@ xfs_trans_binval(
ASSERT(XFS_BUF_ISSTALE(bp));
ASSERT(!(bip->bli_flags & (XFS_BLI_LOGGED | XFS_BLI_DIRTY)));
ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_INODE_BUF));
+ ASSERT(!(bip->__bli_format.blf_flags & XFS_BLF_TYPE_MASK));
ASSERT(bip->__bli_format.blf_flags & XFS_BLF_CANCEL);
ASSERT(bip->bli_item.li_desc->lid_flags & XFS_LID_DIRTY);
ASSERT(tp->t_flags & XFS_TRANS_DIRTY);
@@ -671,6 +672,7 @@ xfs_trans_binval(
bip->bli_flags &= ~(XFS_BLI_INODE_BUF | XFS_BLI_LOGGED | XFS_BLI_DIRTY);
bip->__bli_format.blf_flags &= ~XFS_BLF_INODE_BUF;
bip->__bli_format.blf_flags |= XFS_BLF_CANCEL;
+ bip->__bli_format.blf_flags &= ~XFS_BLF_TYPE_MASK;
for (i = 0; i < bip->bli_format_count; i++) {
memset(bip->bli_formats[i].blf_data_map, 0,
(bip->bli_formats[i].blf_map_size * sizeof(uint)));
@@ -751,6 +753,26 @@ xfs_trans_inode_alloc_buf(
bip->bli_flags |= XFS_BLI_INODE_ALLOC_BUF;
}
+/*
+ * Set the type of the buffer for log recovery so that it can correctly identify
+ * and hence attach the correct buffer ops to the buffer after replay.
+ */
+void
+xfs_trans_buf_set_type(
+ struct xfs_trans *tp,
+ struct xfs_buf *bp,
+ uint type)
+{
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ ASSERT(bp->b_transp == tp);
+ ASSERT(bip != NULL);
+ ASSERT(atomic_read(&bip->bli_refcount) > 0);
+ ASSERT((type & XFS_BLF_TYPE_MASK) != 0);
+
+ bip->__bli_format.blf_flags &= ~XFS_BLF_TYPE_MASK;
+ bip->__bli_format.blf_flags |= type;
+}
/*
* Similar to xfs_trans_inode_buf(), this marks the buffer as a cluster of
@@ -769,14 +791,9 @@ xfs_trans_dquot_buf(
xfs_buf_t *bp,
uint type)
{
- xfs_buf_log_item_t *bip = bp->b_fspriv;
-
- ASSERT(bp->b_transp == tp);
- ASSERT(bip != NULL);
ASSERT(type == XFS_BLF_UDQUOT_BUF ||
type == XFS_BLF_PDQUOT_BUF ||
type == XFS_BLF_GDQUOT_BUF);
- ASSERT(atomic_read(&bip->bli_refcount) > 0);
- bip->__bli_format.blf_flags |= type;
+ xfs_trans_buf_set_type(tp, bp, type);
}
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 04/16] xfs: add CRC checks to the AGF
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (2 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 03/16] xfs: add support for large btree blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 05/16] xfs: add CRC checks to the AGFL Dave Chinner
` (11 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
The AGF already has some self identifying fields (e.g. the sequence
number) so we only need to add the uuid to it to identify the
filesystem it belongs to. The location is fixed based on the
sequence number, so there's no need to add a block number, either.
Hence the only additional fields are the CRC and LSN fields. These
are unlogged, so place some space between the end of the logged
fields and them so that future expansion of the AGF for logged
fields can be placed adjacent to the existing logged fields and
hence not complicate the field-derived range based logging we
currently have.
Based originally on a patch from myself, modified further by
Christoph Hellwig and then modified again to fit into the
verifier structure with additional fields by myself. The multiple
signed-off-by tags indicate the age and history of this patch.
Signed-off-by: Dave Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_ag.h | 23 +++++++++++--
fs/xfs/xfs_alloc.c | 80 ++++++++++++++++++++++++++++++++--------------
fs/xfs/xfs_buf_item.h | 4 ++-
fs/xfs/xfs_fsops.c | 3 ++
fs/xfs/xfs_log_recover.c | 8 +++++
5 files changed, 91 insertions(+), 27 deletions(-)
diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h
index f2aeedb..9f4815e 100644
--- a/fs/xfs/xfs_ag.h
+++ b/fs/xfs/xfs_ag.h
@@ -63,12 +63,29 @@ typedef struct xfs_agf {
__be32 agf_spare0; /* spare field */
__be32 agf_levels[XFS_BTNUM_AGF]; /* btree levels */
__be32 agf_spare1; /* spare field */
+
__be32 agf_flfirst; /* first freelist block's index */
__be32 agf_fllast; /* last freelist block's index */
__be32 agf_flcount; /* count of blocks in freelist */
__be32 agf_freeblks; /* total free blocks */
+
__be32 agf_longest; /* longest free space */
__be32 agf_btreeblks; /* # of blocks held in AGF btrees */
+ uuid_t agf_uuid; /* uuid of filesystem */
+
+ /*
+ * reserve some contiguous space for future logged fields before we add
+ * the unlogged fields. This makes the range logging via flags and
+ * structure offsets much simpler.
+ */
+ __be64 agf_spare64[16];
+
+ /* unlogged fields, written during buffer writeback. */
+ __be64 agf_lsn; /* last write sequence */
+ __be32 agf_crc; /* crc of agf sector */
+ __be32 agf_spare2;
+
+ /* structure must be padded to 64 bit alignment */
} xfs_agf_t;
#define XFS_AGF_MAGICNUM 0x00000001
@@ -83,7 +100,8 @@ typedef struct xfs_agf {
#define XFS_AGF_FREEBLKS 0x00000200
#define XFS_AGF_LONGEST 0x00000400
#define XFS_AGF_BTREEBLKS 0x00000800
-#define XFS_AGF_NUM_BITS 12
+#define XFS_AGF_UUID 0x00001000
+#define XFS_AGF_NUM_BITS 13
#define XFS_AGF_ALL_BITS ((1 << XFS_AGF_NUM_BITS) - 1)
#define XFS_AGF_FLAGS \
@@ -98,7 +116,8 @@ typedef struct xfs_agf {
{ XFS_AGF_FLCOUNT, "FLCOUNT" }, \
{ XFS_AGF_FREEBLKS, "FREEBLKS" }, \
{ XFS_AGF_LONGEST, "LONGEST" }, \
- { XFS_AGF_BTREEBLKS, "BTREEBLKS" }
+ { XFS_AGF_BTREEBLKS, "BTREEBLKS" }, \
+ { XFS_AGF_UUID, "UUID" }
/* disk block (xfs_daddr_t) in the AG */
#define XFS_AGF_DADDR(mp) ((xfs_daddr_t)(1 << (mp)->m_sectbb_log))
diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 0ad2325..1d2e9c3 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -33,7 +33,9 @@
#include "xfs_alloc.h"
#include "xfs_extent_busy.h"
#include "xfs_error.h"
+#include "xfs_cksum.h"
#include "xfs_trace.h"
+#include "xfs_buf_item.h"
struct workqueue_struct *xfs_alloc_wq;
@@ -2058,11 +2060,14 @@ xfs_alloc_log_agf(
offsetof(xfs_agf_t, agf_freeblks),
offsetof(xfs_agf_t, agf_longest),
offsetof(xfs_agf_t, agf_btreeblks),
+ offsetof(xfs_agf_t, agf_uuid),
sizeof(xfs_agf_t)
};
trace_xfs_agf(tp->t_mountp, XFS_BUF_TO_AGF(bp), fields, _RET_IP_);
+ xfs_trans_buf_set_type(tp, bp, XFS_BLF_AGF_BUF);
+
xfs_btree_offsets(fields, offsets, XFS_AGF_NUM_BITS, &first, &last);
xfs_trans_log_buf(tp, bp, (uint)first, (uint)last);
}
@@ -2143,22 +2148,24 @@ xfs_alloc_put_freelist(
return 0;
}
-static void
+static bool
xfs_agf_verify(
+ struct xfs_mount *mp,
struct xfs_buf *bp)
{
- struct xfs_mount *mp = bp->b_target->bt_mount;
- struct xfs_agf *agf;
- int agf_ok;
+ struct xfs_agf *agf = XFS_BUF_TO_AGF(bp);
- agf = XFS_BUF_TO_AGF(bp);
+ if (xfs_sb_version_hascrc(&mp->m_sb) &&
+ !uuid_equal(&agf->agf_uuid, &mp->m_sb.sb_uuid))
+ return false;
- agf_ok = agf->agf_magicnum == cpu_to_be32(XFS_AGF_MAGIC) &&
- XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) &&
- be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) &&
- be32_to_cpu(agf->agf_flfirst) < XFS_AGFL_SIZE(mp) &&
- be32_to_cpu(agf->agf_fllast) < XFS_AGFL_SIZE(mp) &&
- be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp);
+ if (!(agf->agf_magicnum == cpu_to_be32(XFS_AGF_MAGIC) &&
+ XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) &&
+ be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) &&
+ be32_to_cpu(agf->agf_flfirst) < XFS_AGFL_SIZE(mp) &&
+ be32_to_cpu(agf->agf_fllast) < XFS_AGFL_SIZE(mp) &&
+ be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp)))
+ return false;
/*
* during growfs operations, the perag is not fully initialised,
@@ -2166,33 +2173,58 @@ xfs_agf_verify(
* use it by using uncached buffers that don't have the perag attached
* so we can detect and avoid this problem.
*/
- if (bp->b_pag)
- agf_ok = agf_ok && be32_to_cpu(agf->agf_seqno) ==
- bp->b_pag->pag_agno;
+ if (bp->b_pag && be32_to_cpu(agf->agf_seqno) != bp->b_pag->pag_agno)
+ return false;
- if (xfs_sb_version_haslazysbcount(&mp->m_sb))
- agf_ok = agf_ok && be32_to_cpu(agf->agf_btreeblks) <=
- be32_to_cpu(agf->agf_length);
+ if (xfs_sb_version_haslazysbcount(&mp->m_sb) &&
+ be32_to_cpu(agf->agf_btreeblks) > be32_to_cpu(agf->agf_length))
+ return false;
+
+ return true;;
- if (unlikely(XFS_TEST_ERROR(!agf_ok, mp, XFS_ERRTAG_ALLOC_READ_AGF,
- XFS_RANDOM_ALLOC_READ_AGF))) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, agf);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
- }
}
static void
xfs_agf_read_verify(
struct xfs_buf *bp)
{
- xfs_agf_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ int agf_ok = 1;
+
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ agf_ok = xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_agf, agf_crc));
+
+ agf_ok = agf_ok && xfs_agf_verify(mp, bp);
+
+ if (unlikely(XFS_TEST_ERROR(!agf_ok, mp, XFS_ERRTAG_ALLOC_READ_AGF,
+ XFS_RANDOM_ALLOC_READ_AGF))) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
static void
xfs_agf_write_verify(
struct xfs_buf *bp)
{
- xfs_agf_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ if (!xfs_agf_verify(mp, bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ XFS_BUF_TO_AGF(bp)->agf_lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_agf, agf_crc));
}
const struct xfs_buf_ops xfs_agf_buf_ops = {
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index 101ef83..76bd7a1 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -45,12 +45,14 @@ extern kmem_zone_t *xfs_buf_item_zone;
* once the changes have been replayed into the buffer.
*/
#define XFS_BLF_BTREE_BUF (1<<5)
+#define XFS_BLF_AGF_BUF (1<<6)
#define XFS_BLF_TYPE_MASK \
(XFS_BLF_UDQUOT_BUF | \
XFS_BLF_PDQUOT_BUF | \
XFS_BLF_GDQUOT_BUF | \
- XFS_BLF_BTREE_BUF)
+ XFS_BLF_BTREE_BUF | \
+ XFS_BLF_AGF_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index 6fe286a..b6cca27 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -247,6 +247,9 @@ xfs_growfs_data_private(
tmpsize = agsize - XFS_PREALLOC_BLOCKS(mp);
agf->agf_freeblks = cpu_to_be32(tmpsize);
agf->agf_longest = cpu_to_be32(tmpsize);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ uuid_copy(&agf->agf_uuid, &mp->m_sb.sb_uuid);
+
error = xfs_bwrite(bp);
xfs_buf_relse(bp);
if (error)
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index c57a987..c9f1c6d 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1954,6 +1954,14 @@ xlog_recover_do_reg_buffer(
break;
}
break;
+ case XFS_BLF_AGF_BUF:
+ if (*(__be32 *)bp->b_addr != cpu_to_be32(XFS_AGF_MAGIC)) {
+ xfs_warn(mp, "Bad AGF block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_agf_buf_ops;
+ break;
default:
break;
}
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 05/16] xfs: add CRC checks to the AGFL
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (3 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 04/16] xfs: add CRC checks to the AGF Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-03-01 23:56 ` Ben Myers
2013-02-25 1:31 ` [PATCH 06/16] xfs: add CRC checks to the AGI Dave Chinner
` (10 subsequent siblings)
15 siblings, 1 reply; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Christoph Hellwig <hch@lst.de>
Add CRC checks, location information and a magic number to the AGFL.
Previously the AGFL was just a block containing nothing but the
free block pointers. The new AGFL has a real header with the usual
boilerplate instead, so that we can verify it's not corrupted and
written into the right place.
[dchinner@redhat.com] Added LSN field, reworked significantly to fit
into new verifier structure and growfs structure, enabled full
verifier functionality now there is a header to verify and we can
guarantee an initialised AGFL.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_ag.h | 25 +++++++++-
fs/xfs/xfs_alloc.c | 119 ++++++++++++++++++++++++++++++----------------
fs/xfs/xfs_buf_item.h | 4 +-
fs/xfs/xfs_fsops.c | 5 ++
fs/xfs/xfs_log_recover.c | 10 ++++
5 files changed, 119 insertions(+), 44 deletions(-)
diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h
index 9f4815e..272ef09 100644
--- a/fs/xfs/xfs_ag.h
+++ b/fs/xfs/xfs_ag.h
@@ -30,6 +30,7 @@ struct xfs_trans;
#define XFS_AGF_MAGIC 0x58414746 /* 'XAGF' */
#define XFS_AGI_MAGIC 0x58414749 /* 'XAGI' */
+#define XFS_AGFL_MAGIC 0x5841464c /* 'XAFL' */
#define XFS_AGF_VERSION 1
#define XFS_AGI_VERSION 1
@@ -190,11 +191,31 @@ extern const struct xfs_buf_ops xfs_agi_buf_ops;
*/
#define XFS_AGFL_DADDR(mp) ((xfs_daddr_t)(3 << (mp)->m_sectbb_log))
#define XFS_AGFL_BLOCK(mp) XFS_HDR_BLOCK(mp, XFS_AGFL_DADDR(mp))
-#define XFS_AGFL_SIZE(mp) ((mp)->m_sb.sb_sectsize / sizeof(xfs_agblock_t))
#define XFS_BUF_TO_AGFL(bp) ((xfs_agfl_t *)((bp)->b_addr))
+#define XFS_BUF_TO_AGFL_BNO(mp, bp) \
+ (xfs_sb_version_hascrc(&((mp)->m_sb)) ? \
+ &(XFS_BUF_TO_AGFL(bp)->agfl_bno[0]) : \
+ (__be32 *)(bp)->b_addr)
+
+/*
+ * Size of the AGFL. For CRC-enabled filesystes we steal a couple of
+ * slots in the beginning of the block for a proper header with the
+ * location information and CRC.
+ */
+#define XFS_AGFL_SIZE(mp) \
+ (((mp)->m_sb.sb_sectsize - \
+ (xfs_sb_version_hascrc(&((mp)->m_sb)) ? \
+ sizeof(struct xfs_agfl) : 0)) / \
+ sizeof(xfs_agblock_t))
+
typedef struct xfs_agfl {
- __be32 agfl_bno[1]; /* actually XFS_AGFL_SIZE(mp) */
+ __be32 agfl_magicnum;
+ __be32 agfl_seqno;
+ uuid_t agfl_uuid;
+ __be64 agfl_lsn;
+ __be32 agfl_crc;
+ __be32 agfl_bno[]; /* actually XFS_AGFL_SIZE(mp) */
} xfs_agfl_t;
/*
diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 1d2e9c3..53e699e 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -432,53 +432,84 @@ xfs_alloc_fixup_trees(
return 0;
}
-static void
+static bool
xfs_agfl_verify(
struct xfs_buf *bp)
{
-#ifdef WHEN_CRCS_COME_ALONG
- /*
- * we cannot actually do any verification of the AGFL because mkfs does
- * not initialise the AGFL to zero or NULL. Hence the only valid part of
- * the AGFL is what the AGF says is active. We can't get to the AGF, so
- * we can't verify just those entries are valid.
- *
- * This problem goes away when the CRC format change comes along as that
- * requires the AGFL to be initialised by mkfs. At that point, we can
- * verify the blocks in the agfl -active or not- lie within the bounds
- * of the AG. Until then, just leave this check ifdef'd out.
- */
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_agfl *agfl = XFS_BUF_TO_AGFL(bp);
- int agfl_ok = 1;
-
int i;
+ if (!uuid_equal(&agfl->agfl_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be32_to_cpu(agfl->agfl_magicnum) != XFS_AGFL_MAGIC)
+ return false;
+ /*
+ * during growfs operations, the perag is not fully initialised,
+ * so we can't use it for any useful checking. growfs ensures we can't
+ * use it by using uncached buffers that don't have the perag attached
+ * so we can detect and avoid this problem.
+ */
+ if (bp->b_pag && be32_to_cpu(agfl->agfl_seqno) != bp->b_pag->pag_agno)
+ return false;
+
for (i = 0; i < XFS_AGFL_SIZE(mp); i++) {
- if (be32_to_cpu(agfl->agfl_bno[i]) == NULLAGBLOCK ||
+ if (be32_to_cpu(agfl->agfl_bno[i]) != NULLAGBLOCK &&
be32_to_cpu(agfl->agfl_bno[i]) >= mp->m_sb.sb_agblocks)
- agfl_ok = 0;
+ return false;
}
+ return true;
+}
+
+static void
+xfs_agfl_read_verify(
+ struct xfs_buf *bp)
+{
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ int agfl_ok = 1;
+
+ /*
+ * There is no verification of non-crc AGFLs because mkfs does not
+ * initialise the AGFL to zero or NULL. Hence the only valid part of the
+ * AGFL is what the AGF says is active. We can't get to the AGF, so we
+ * can't verify just those entries are valid.
+ */
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ agfl_ok = xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_agfl, agfl_crc));
+
+ agfl_ok = agfl_ok && xfs_agfl_verify(bp);
if (!agfl_ok) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, agfl);
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
}
-#endif
}
static void
xfs_agfl_write_verify(
struct xfs_buf *bp)
{
- xfs_agfl_verify(bp);
-}
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
-static void
-xfs_agfl_read_verify(
- struct xfs_buf *bp)
-{
- xfs_agfl_verify(bp);
+ /* no verification of non-crc AGFLs */
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (!xfs_agfl_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (bip)
+ XFS_BUF_TO_AGFL(bp)->agfl_lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_agfl, agfl_crc));
}
const struct xfs_buf_ops xfs_agfl_buf_ops = {
@@ -1984,18 +2015,18 @@ xfs_alloc_get_freelist(
int btreeblk) /* destination is a AGF btree */
{
xfs_agf_t *agf; /* a.g. freespace structure */
- xfs_agfl_t *agfl; /* a.g. freelist structure */
xfs_buf_t *agflbp;/* buffer for a.g. freelist structure */
xfs_agblock_t bno; /* block number returned */
+ __be32 *agfl_bno;
int error;
int logflags;
- xfs_mount_t *mp; /* mount structure */
+ xfs_mount_t *mp = tp->t_mountp;
xfs_perag_t *pag; /* per allocation group data */
- agf = XFS_BUF_TO_AGF(agbp);
/*
* Freelist is empty, give up.
*/
+ agf = XFS_BUF_TO_AGF(agbp);
if (!agf->agf_flcount) {
*bnop = NULLAGBLOCK;
return 0;
@@ -2003,15 +2034,17 @@ xfs_alloc_get_freelist(
/*
* Read the array of free blocks.
*/
- mp = tp->t_mountp;
- if ((error = xfs_alloc_read_agfl(mp, tp,
- be32_to_cpu(agf->agf_seqno), &agflbp)))
+ error = xfs_alloc_read_agfl(mp, tp, be32_to_cpu(agf->agf_seqno),
+ &agflbp);
+ if (error)
return error;
- agfl = XFS_BUF_TO_AGFL(agflbp);
+
+
/*
* Get the block number and update the data structures.
*/
- bno = be32_to_cpu(agfl->agfl_bno[be32_to_cpu(agf->agf_flfirst)]);
+ agfl_bno = XFS_BUF_TO_AGFL_BNO(mp, agflbp);
+ bno = be32_to_cpu(agfl_bno[be32_to_cpu(agf->agf_flfirst)]);
be32_add_cpu(&agf->agf_flfirst, 1);
xfs_trans_brelse(tp, agflbp);
if (be32_to_cpu(agf->agf_flfirst) == XFS_AGFL_SIZE(mp))
@@ -2104,12 +2137,13 @@ xfs_alloc_put_freelist(
int btreeblk) /* block came from a AGF btree */
{
xfs_agf_t *agf; /* a.g. freespace structure */
- xfs_agfl_t *agfl; /* a.g. free block array */
__be32 *blockp;/* pointer to array entry */
int error;
int logflags;
xfs_mount_t *mp; /* mount structure */
xfs_perag_t *pag; /* per allocation group data */
+ __be32 *agfl_bno;
+ int startoff;
agf = XFS_BUF_TO_AGF(agbp);
mp = tp->t_mountp;
@@ -2117,7 +2151,6 @@ xfs_alloc_put_freelist(
if (!agflbp && (error = xfs_alloc_read_agfl(mp, tp,
be32_to_cpu(agf->agf_seqno), &agflbp)))
return error;
- agfl = XFS_BUF_TO_AGFL(agflbp);
be32_add_cpu(&agf->agf_fllast, 1);
if (be32_to_cpu(agf->agf_fllast) == XFS_AGFL_SIZE(mp))
agf->agf_fllast = 0;
@@ -2138,13 +2171,17 @@ xfs_alloc_put_freelist(
xfs_alloc_log_agf(tp, agbp, logflags);
ASSERT(be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp));
- blockp = &agfl->agfl_bno[be32_to_cpu(agf->agf_fllast)];
+
+ agfl_bno = XFS_BUF_TO_AGFL_BNO(mp, agflbp);
+ blockp = &agfl_bno[be32_to_cpu(agf->agf_fllast)];
*blockp = cpu_to_be32(bno);
+ startoff = (char *)blockp - (char *)agflbp->b_addr;
+
xfs_alloc_log_agf(tp, agbp, logflags);
- xfs_trans_log_buf(tp, agflbp,
- (int)((xfs_caddr_t)blockp - (xfs_caddr_t)agfl),
- (int)((xfs_caddr_t)blockp - (xfs_caddr_t)agfl +
- sizeof(xfs_agblock_t) - 1));
+
+ xfs_trans_buf_set_type(tp, agflbp, XFS_BLF_AGFL_BUF);
+ xfs_trans_log_buf(tp, agflbp, startoff,
+ startoff + sizeof(xfs_agblock_t) - 1);
return 0;
}
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index 76bd7a1..067d5f0 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -46,13 +46,15 @@ extern kmem_zone_t *xfs_buf_item_zone;
*/
#define XFS_BLF_BTREE_BUF (1<<5)
#define XFS_BLF_AGF_BUF (1<<6)
+#define XFS_BLF_AGFL_BUF (1<<7)
#define XFS_BLF_TYPE_MASK \
(XFS_BLF_UDQUOT_BUF | \
XFS_BLF_PDQUOT_BUF | \
XFS_BLF_GDQUOT_BUF | \
XFS_BLF_BTREE_BUF | \
- XFS_BLF_AGF_BUF)
+ XFS_BLF_AGF_BUF | \
+ XFS_BLF_AGFL_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index b6cca27..a693a54 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -268,6 +268,11 @@ xfs_growfs_data_private(
}
agfl = XFS_BUF_TO_AGFL(bp);
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ agfl->agfl_magicnum = cpu_to_be32(XFS_AGFL_MAGIC);
+ agfl->agfl_seqno = cpu_to_be32(agno);
+ uuid_copy(&agfl->agfl_uuid, &mp->m_sb.sb_uuid);
+ }
for (bucket = 0; bucket < XFS_AGFL_SIZE(mp); bucket++)
agfl->agfl_bno[bucket] = cpu_to_be32(NULLAGBLOCK);
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index c9f1c6d..b256cc7 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1962,6 +1962,16 @@ xlog_recover_do_reg_buffer(
}
bp->b_ops = &xfs_agf_buf_ops;
break;
+ case XFS_BLF_AGFL_BUF:
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ break;
+ if (*(__be32 *)bp->b_addr != cpu_to_be32(XFS_AGFL_MAGIC)) {
+ xfs_warn(mp, "Bad AGFL block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_agfl_buf_ops;
+ break;
default:
break;
}
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH 05/16] xfs: add CRC checks to the AGFL
2013-02-25 1:31 ` [PATCH 05/16] xfs: add CRC checks to the AGFL Dave Chinner
@ 2013-03-01 23:56 ` Ben Myers
0 siblings, 0 replies; 23+ messages in thread
From: Ben Myers @ 2013-03-01 23:56 UTC (permalink / raw)
To: Dave Chinner; +Cc: xfs
Hi Dave,
On Mon, Feb 25, 2013 at 12:31:30PM +1100, Dave Chinner wrote:
> From: Christoph Hellwig <hch@lst.de>
>
> Add CRC checks, location information and a magic number to the AGFL.
> Previously the AGFL was just a block containing nothing but the
> free block pointers. The new AGFL has a real header with the usual
> boilerplate instead, so that we can verify it's not corrupted and
> written into the right place.
>
> [dchinner@redhat.com] Added LSN field, reworked significantly to fit
> into new verifier structure and growfs structure, enabled full
> verifier functionality now there is a header to verify and we can
> guarantee an initialised AGFL.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
> ---
> fs/xfs/xfs_ag.h | 25 +++++++++-
> fs/xfs/xfs_alloc.c | 119 ++++++++++++++++++++++++++++++----------------
> fs/xfs/xfs_buf_item.h | 4 +-
> fs/xfs/xfs_fsops.c | 5 ++
> fs/xfs/xfs_log_recover.c | 10 ++++
> 5 files changed, 119 insertions(+), 44 deletions(-)
>
> diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h
> index 9f4815e..272ef09 100644
> --- a/fs/xfs/xfs_ag.h
> +++ b/fs/xfs/xfs_ag.h
> @@ -30,6 +30,7 @@ struct xfs_trans;
>
> #define XFS_AGF_MAGIC 0x58414746 /* 'XAGF' */
> #define XFS_AGI_MAGIC 0x58414749 /* 'XAGI' */
> +#define XFS_AGFL_MAGIC 0x5841464c /* 'XAFL' */
> #define XFS_AGF_VERSION 1
> #define XFS_AGI_VERSION 1
>
> @@ -190,11 +191,31 @@ extern const struct xfs_buf_ops xfs_agi_buf_ops;
> */
> #define XFS_AGFL_DADDR(mp) ((xfs_daddr_t)(3 << (mp)->m_sectbb_log))
> #define XFS_AGFL_BLOCK(mp) XFS_HDR_BLOCK(mp, XFS_AGFL_DADDR(mp))
> -#define XFS_AGFL_SIZE(mp) ((mp)->m_sb.sb_sectsize / sizeof(xfs_agblock_t))
> #define XFS_BUF_TO_AGFL(bp) ((xfs_agfl_t *)((bp)->b_addr))
>
> +#define XFS_BUF_TO_AGFL_BNO(mp, bp) \
> + (xfs_sb_version_hascrc(&((mp)->m_sb)) ? \
> + &(XFS_BUF_TO_AGFL(bp)->agfl_bno[0]) : \
> + (__be32 *)(bp)->b_addr)
> +
> +/*
> + * Size of the AGFL. For CRC-enabled filesystes we steal a couple of
> + * slots in the beginning of the block for a proper header with the
> + * location information and CRC.
> + */
> +#define XFS_AGFL_SIZE(mp) \
> + (((mp)->m_sb.sb_sectsize - \
> + (xfs_sb_version_hascrc(&((mp)->m_sb)) ? \
> + sizeof(struct xfs_agfl) : 0)) / \
> + sizeof(xfs_agblock_t))
> +
> typedef struct xfs_agfl {
> - __be32 agfl_bno[1]; /* actually XFS_AGFL_SIZE(mp) */
> + __be32 agfl_magicnum;
+ __be32 agfl_versionnum;
Now would be a good time to add this, and it would be consistent with the AGF
and AGI structures. There is no shortage of agfl slots according to my
calculation, and this might be helpful for xfsrepair later.
-Ben
> + __be32 agfl_seqno;
> + uuid_t agfl_uuid;
> + __be64 agfl_lsn;
> + __be32 agfl_crc;
> + __be32 agfl_bno[]; /* actually XFS_AGFL_SIZE(mp) */
> } xfs_agfl_t;
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 23+ messages in thread
* [PATCH 06/16] xfs: add CRC checks to the AGI
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (4 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 05/16] xfs: add CRC checks to the AGFL Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 07/16] xfs: add CRC checks for quota blocks Dave Chinner
` (9 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dgc@sgi.com>
Same set of changes made to the AGF need to be made to the AGI.
This patch has a similar history to the AGF, hence a similar
sign-off chain.
Signed-off-by: Dave Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dgc@redhat.com>
---
fs/xfs/xfs_ag.h | 8 +++++++
fs/xfs/xfs_buf_item.h | 4 +++-
fs/xfs/xfs_fsops.c | 3 +++
fs/xfs/xfs_ialloc.c | 57 ++++++++++++++++++++++++++++++++++------------
fs/xfs/xfs_log_recover.c | 8 +++++++
5 files changed, 65 insertions(+), 15 deletions(-)
diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h
index 272ef09..317aa86 100644
--- a/fs/xfs/xfs_ag.h
+++ b/fs/xfs/xfs_ag.h
@@ -152,6 +152,7 @@ typedef struct xfs_agi {
__be32 agi_root; /* root of inode btree */
__be32 agi_level; /* levels in inode btree */
__be32 agi_freecount; /* number of free inodes */
+
__be32 agi_newino; /* new inode just allocated */
__be32 agi_dirino; /* last directory inode chunk */
/*
@@ -159,6 +160,13 @@ typedef struct xfs_agi {
* still being referenced.
*/
__be32 agi_unlinked[XFS_AGI_UNLINKED_BUCKETS];
+
+ uuid_t agi_uuid; /* uuid of filesystem */
+ __be32 agi_crc; /* crc of agi sector */
+ __be32 agi_pad32;
+ __be64 agi_lsn; /* last write sequence */
+
+ /* structure must be padded to 64 bit alignment */
} xfs_agi_t;
#define XFS_AGI_MAGICNUM 0x00000001
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index 067d5f0..c256606 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -47,6 +47,7 @@ extern kmem_zone_t *xfs_buf_item_zone;
#define XFS_BLF_BTREE_BUF (1<<5)
#define XFS_BLF_AGF_BUF (1<<6)
#define XFS_BLF_AGFL_BUF (1<<7)
+#define XFS_BLF_AGI_BUF (1<<8)
#define XFS_BLF_TYPE_MASK \
(XFS_BLF_UDQUOT_BUF | \
@@ -54,7 +55,8 @@ extern kmem_zone_t *xfs_buf_item_zone;
XFS_BLF_GDQUOT_BUF | \
XFS_BLF_BTREE_BUF | \
XFS_BLF_AGF_BUF | \
- XFS_BLF_AGFL_BUF)
+ XFS_BLF_AGFL_BUF | \
+ XFS_BLF_AGI_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c
index a693a54..87595b2 100644
--- a/fs/xfs/xfs_fsops.c
+++ b/fs/xfs/xfs_fsops.c
@@ -304,8 +304,11 @@ xfs_growfs_data_private(
agi->agi_freecount = 0;
agi->agi_newino = cpu_to_be32(NULLAGINO);
agi->agi_dirino = cpu_to_be32(NULLAGINO);
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ uuid_copy(&agi->agi_uuid, &mp->m_sb.sb_uuid);
for (bucket = 0; bucket < XFS_AGI_UNLINKED_BUCKETS; bucket++)
agi->agi_unlinked[bucket] = cpu_to_be32(NULLAGINO);
+
error = xfs_bwrite(bp);
xfs_buf_relse(bp);
if (error)
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c
index 515bf71..41c76d6 100644
--- a/fs/xfs/xfs_ialloc.c
+++ b/fs/xfs/xfs_ialloc.c
@@ -36,6 +36,8 @@
#include "xfs_rtalloc.h"
#include "xfs_error.h"
#include "xfs_bmap.h"
+#include "xfs_cksum.h"
+#include "xfs_buf_item.h"
/*
@@ -1453,6 +1455,7 @@ xfs_ialloc_log_agi(
/*
* Log the allocation group inode header buffer.
*/
+ xfs_trans_buf_set_type(tp, bp, XFS_BLF_AGI_BUF);
xfs_trans_log_buf(tp, bp, first, last);
}
@@ -1470,19 +1473,23 @@ xfs_check_agi_unlinked(
#define xfs_check_agi_unlinked(agi)
#endif
-static void
+static bool
xfs_agi_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_agi *agi = XFS_BUF_TO_AGI(bp);
- int agi_ok;
+ if (xfs_sb_version_hascrc(&mp->m_sb) &&
+ !uuid_equal(&agi->agi_uuid, &mp->m_sb.sb_uuid))
+ return false;
/*
* Validate the magic number of the agi block.
*/
- agi_ok = agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC) &&
- XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum));
+ if (agi->agi_magicnum != cpu_to_be32(XFS_AGI_MAGIC))
+ return false;
+ if (!XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)))
+ return false;
/*
* during growfs operations, the perag is not fully initialised,
@@ -1490,30 +1497,52 @@ xfs_agi_verify(
* use it by using uncached buffers that don't have the perag attached
* so we can detect and avoid this problem.
*/
- if (bp->b_pag)
- agi_ok = agi_ok && be32_to_cpu(agi->agi_seqno) ==
- bp->b_pag->pag_agno;
+ if (bp->b_pag && be32_to_cpu(agi->agi_seqno) != bp->b_pag->pag_agno)
+ return false;
- if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI,
- XFS_RANDOM_IALLOC_READ_AGI))) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, agi);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
- }
xfs_check_agi_unlinked(agi);
+ return true;
}
static void
xfs_agi_read_verify(
struct xfs_buf *bp)
{
- xfs_agi_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ int agi_ok = 1;
+
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ agi_ok = xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_agi, agi_crc));
+ agi_ok = agi_ok && xfs_agi_verify(bp);
+
+ if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI,
+ XFS_RANDOM_IALLOC_READ_AGI))) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
static void
xfs_agi_write_verify(
struct xfs_buf *bp)
{
- xfs_agi_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ if (!xfs_agi_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ XFS_BUF_TO_AGI(bp)->agi_lsn = cpu_to_be64(bip->bli_item.li_lsn);
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_agi, agi_crc));
}
const struct xfs_buf_ops xfs_agi_buf_ops = {
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index b256cc7..8b6864a 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1972,6 +1972,14 @@ xlog_recover_do_reg_buffer(
}
bp->b_ops = &xfs_agfl_buf_ops;
break;
+ case XFS_BLF_AGI_BUF:
+ if (*(__be32 *)bp->b_addr != cpu_to_be32(XFS_AGI_MAGIC)) {
+ xfs_warn(mp, "Bad AGI block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_agi_buf_ops;
+ break;
default:
break;
}
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 07/16] xfs: add CRC checks for quota blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (5 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 06/16] xfs: add CRC checks to the AGI Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-03-05 20:36 ` Ben Myers
2013-02-25 1:31 ` [PATCH 08/16] xfs: add version 3 inode format with CRCs Dave Chinner
` (8 subsequent siblings)
15 siblings, 1 reply; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Christoph Hellwig <hch@lst.de>
Use the reserved space in struct xfs_dqblk to store a UUID and a crc
for the quota blocks.
[dchinner@redhat.com] Add a LSN field and update for current verifier
infrastructure.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_dquot.c | 112 ++++++++++++++++++++++++++++++++++++++++------
fs/xfs/xfs_log_recover.c | 10 +++++
fs/xfs/xfs_qm.c | 23 ++++++++--
fs/xfs/xfs_qm.h | 2 +
fs/xfs/xfs_quota.h | 11 ++++-
5 files changed, 141 insertions(+), 17 deletions(-)
diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c
index 8025eb2..aedf29c 100644
--- a/fs/xfs/xfs_dquot.c
+++ b/fs/xfs/xfs_dquot.c
@@ -36,6 +36,7 @@
#include "xfs_trans_space.h"
#include "xfs_trans_priv.h"
#include "xfs_qm.h"
+#include "xfs_cksum.h"
#include "xfs_trace.h"
/*
@@ -239,6 +240,8 @@ xfs_qm_init_dquot_blk(
d->dd_diskdq.d_version = XFS_DQUOT_VERSION;
d->dd_diskdq.d_id = cpu_to_be32(curid);
d->dd_diskdq.d_flags = type;
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ uuid_copy(&d->dd_uuid, &mp->m_sb.sb_uuid);
}
xfs_trans_dquot_buf(tp, bp,
@@ -248,25 +251,87 @@ xfs_qm_init_dquot_blk(
xfs_trans_log_buf(tp, bp, 0, BBTOB(q->qi_dqchunklen) - 1);
}
-static void
+STATIC void
+xfs_dquot_buf_calc_crc(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp)
+{
+ struct xfs_dqblk *d = (struct xfs_dqblk *)bp->b_addr;
+ int i;
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ for (i = 0; i < mp->m_quotainfo->qi_dqperchunk; i++, d++) {
+ xfs_update_cksum((char *)d, sizeof(struct xfs_dqblk),
+ offsetof(struct xfs_dqblk, dd_crc));
+ }
+}
+
+STATIC bool
+xfs_dquot_buf_verify_crc(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp)
+{
+ struct xfs_dqblk *d = (struct xfs_dqblk *)bp->b_addr;
+ int ndquots;
+ int i;
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return true;
+
+ /*
+ * if we are in log recovery, the quota subsystem has not been
+ * initialised so we have no quotainfo structure. In that case, we need
+ * to manually calculate the number of dquots in the buffer.
+ */
+ if (mp->m_quotainfo)
+ ndquots = mp->m_quotainfo->qi_dqperchunk;
+ else
+ ndquots = xfs_qm_calc_dquots_per_chunk(mp,
+ XFS_BB_TO_FSB(mp, bp->b_length));
+
+ for (i = 0; i < ndquots; i++, d++) {
+ if (!xfs_verify_cksum((char *)d, sizeof(struct xfs_dqblk),
+ offsetof(struct xfs_dqblk, dd_crc)))
+ return false;
+ if (!uuid_equal(&d->dd_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ }
+
+ return true;
+}
+
+STATIC bool
xfs_dquot_buf_verify(
+ struct xfs_mount *mp,
struct xfs_buf *bp)
{
- struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_dqblk *d = (struct xfs_dqblk *)bp->b_addr;
- struct xfs_disk_dquot *ddq;
xfs_dqid_t id = 0;
+ int ndquots;
int i;
/*
+ * if we are in log recovery, the quota subsystem has not been
+ * initialised so we have no quotainfo structure. In that case, we need
+ * to manually calculate the number of dquots in the buffer.
+ */
+ if (mp->m_quotainfo)
+ ndquots = mp->m_quotainfo->qi_dqperchunk;
+ else
+ ndquots = xfs_qm_calc_dquots_per_chunk(mp, bp->b_length);
+
+ /*
* On the first read of the buffer, verify that each dquot is valid.
* We don't know what the id of the dquot is supposed to be, just that
* they should be increasing monotonically within the buffer. If the
* first id is corrupt, then it will fail on the second dquot in the
* buffer so corruptions could point to the wrong dquot in this case.
*/
- for (i = 0; i < mp->m_quotainfo->qi_dqperchunk; i++) {
- int error;
+ for (i = 0; i < ndquots; i++) {
+ struct xfs_disk_dquot *ddq;
+ int error;
ddq = &d[i].dd_diskdq;
@@ -274,27 +339,37 @@ xfs_dquot_buf_verify(
id = be32_to_cpu(ddq->d_id);
error = xfs_qm_dqcheck(mp, ddq, id + i, 0, XFS_QMOPT_DOWARN,
- "xfs_dquot_read_verify");
- if (error) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, d);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
- break;
- }
+ "xfs_dquot_buf_verify");
+ if (error)
+ return false;
}
+ return true;
}
static void
xfs_dquot_buf_read_verify(
struct xfs_buf *bp)
{
- xfs_dquot_buf_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ if (!xfs_dquot_buf_verify_crc(mp, bp) || !xfs_dquot_buf_verify(mp, bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
void
xfs_dquot_buf_write_verify(
struct xfs_buf *bp)
{
- xfs_dquot_buf_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ if (!xfs_dquot_buf_verify(mp, bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+ xfs_dquot_buf_calc_crc(mp, bp);
}
const struct xfs_buf_ops xfs_dquot_buf_ops = {
@@ -1035,6 +1110,17 @@ xfs_qm_dqflush(
&dqp->q_logitem.qli_item.li_lsn);
/*
+ * copy the lsn into the on-disk dquot now while we have the in memory
+ * dquot here. This can't be done later in the write verifier as we
+ * can't get access to the log item at that point in time.
+ */
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_dqblk *dqb = (struct xfs_dqblk *)ddqp;
+
+ dqb->dd_lsn = cpu_to_be64(dqp->q_logitem.qli_item.li_lsn);
+ }
+
+ /*
* Attach an iodone routine so that we can remove this dquot from the
* AIL and release the flush lock once the dquot is synced to disk.
*/
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 8b6864a..e77c770 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1980,6 +1980,16 @@ xlog_recover_do_reg_buffer(
}
bp->b_ops = &xfs_agi_buf_ops;
break;
+ case XFS_BLF_UDQUOT_BUF:
+ case XFS_BLF_PDQUOT_BUF:
+ case XFS_BLF_GDQUOT_BUF:
+ if (*(__be16 *)bp->b_addr != cpu_to_be16(XFS_DQUOT_MAGIC)) {
+ xfs_warn(mp, "Bad DQUOT block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_dquot_buf_ops;
+ break;
default:
break;
}
diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c
index e5b5cf9..9d00500 100644
--- a/fs/xfs/xfs_qm.c
+++ b/fs/xfs/xfs_qm.c
@@ -617,6 +617,20 @@ xfs_qm_dqdetach(
}
}
+int
+xfs_qm_calc_dquots_per_chunk(
+ struct xfs_mount *mp,
+ unsigned int nbblks) /* basic block units */
+{
+ unsigned int ndquots;
+
+ ASSERT(nbblks > 0);
+ ndquots = BBTOB(nbblks);
+ do_div(ndquots, sizeof(xfs_dqblk_t));
+
+ return ndquots;
+}
+
/*
* This initializes all the quota information that's kept in the
* mount structure
@@ -656,9 +670,8 @@ xfs_qm_init_quotainfo(
/* Precalc some constants */
qinf->qi_dqchunklen = XFS_FSB_TO_BB(mp, XFS_DQUOT_CLUSTER_SIZE_FSB);
- ASSERT(qinf->qi_dqchunklen);
- qinf->qi_dqperchunk = BBTOB(qinf->qi_dqchunklen);
- do_div(qinf->qi_dqperchunk, sizeof(xfs_dqblk_t));
+ qinf->qi_dqperchunk = xfs_qm_calc_dquots_per_chunk(mp,
+ qinf->qi_dqchunklen);
mp->m_qflags |= (mp->m_sb.sb_qflags & XFS_ALL_QUOTA_CHKD);
@@ -897,6 +910,10 @@ xfs_qm_dqiter_bufs(
if (error)
break;
+ /*
+ * XXX(hch): need to figure out if it makes sense to validate
+ * the CRC here.
+ */
xfs_qm_reset_dqcounts(mp, bp, firstid, type);
xfs_buf_delwri_queue(bp, buffer_list);
xfs_buf_relse(bp);
diff --git a/fs/xfs/xfs_qm.h b/fs/xfs/xfs_qm.h
index 44b858b..45d1656 100644
--- a/fs/xfs/xfs_qm.h
+++ b/fs/xfs/xfs_qm.h
@@ -75,6 +75,8 @@ typedef struct xfs_quotainfo {
&((qi)->qi_gquota_tree))
+extern int xfs_qm_calc_dquots_per_chunk(struct xfs_mount *mp,
+ unsigned int nbblks);
extern void xfs_trans_mod_dquot(xfs_trans_t *, xfs_dquot_t *, uint, long);
extern int xfs_trans_reserve_quota_bydquots(xfs_trans_t *, xfs_mount_t *,
xfs_dquot_t *, xfs_dquot_t *, long, long, uint);
diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h
index b50ec5b..c61e31c 100644
--- a/fs/xfs/xfs_quota.h
+++ b/fs/xfs/xfs_quota.h
@@ -77,7 +77,14 @@ typedef struct xfs_disk_dquot {
*/
typedef struct xfs_dqblk {
xfs_disk_dquot_t dd_diskdq; /* portion that lives incore as well */
- char dd_fill[32]; /* filling for posterity */
+ char dd_fill[4]; /* filling for posterity */
+
+ /*
+ * These two are only present on filesystems with the CRC bits set.
+ */
+ __be32 dd_crc; /* checksum */
+ __be64 dd_lsn; /* last modification in log */
+ uuid_t dd_uuid; /* location information */
} xfs_dqblk_t;
/*
@@ -380,5 +387,7 @@ extern int xfs_qm_dqcheck(struct xfs_mount *, xfs_disk_dquot_t *,
xfs_dqid_t, uint, uint, char *);
extern int xfs_mount_reset_sbqflags(struct xfs_mount *);
+extern const struct xfs_buf_ops xfs_dquot_buf_ops;
+
#endif /* __KERNEL__ */
#endif /* __XFS_QUOTA_H__ */
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH 07/16] xfs: add CRC checks for quota blocks
2013-02-25 1:31 ` [PATCH 07/16] xfs: add CRC checks for quota blocks Dave Chinner
@ 2013-03-05 20:36 ` Ben Myers
2013-03-05 22:49 ` Dave Chinner
0 siblings, 1 reply; 23+ messages in thread
From: Ben Myers @ 2013-03-05 20:36 UTC (permalink / raw)
To: Dave Chinner; +Cc: xfs
Hi Dave,
On Mon, Feb 25, 2013 at 12:31:32PM +1100, Dave Chinner wrote:
> From: Christoph Hellwig <hch@lst.de>
>
> Use the reserved space in struct xfs_dqblk to store a UUID and a crc
> for the quota blocks.
>
> [dchinner@redhat.com] Add a LSN field and update for current verifier
> infrastructure.
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
Been over this and it looks fine.
> @@ -897,6 +910,10 @@ xfs_qm_dqiter_bufs(
> if (error)
> break;
>
> + /*
> + * XXX(hch): need to figure out if it makes sense to validate
> + * the CRC here.
> + */
What's you're opinion on this?
-Ben
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH 07/16] xfs: add CRC checks for quota blocks
2013-03-05 20:36 ` Ben Myers
@ 2013-03-05 22:49 ` Dave Chinner
0 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-03-05 22:49 UTC (permalink / raw)
To: Ben Myers; +Cc: xfs
On Tue, Mar 05, 2013 at 02:36:15PM -0600, Ben Myers wrote:
> Hi Dave,
>
> On Mon, Feb 25, 2013 at 12:31:32PM +1100, Dave Chinner wrote:
> > From: Christoph Hellwig <hch@lst.de>
> >
> > Use the reserved space in struct xfs_dqblk to store a UUID and a crc
> > for the quota blocks.
> >
> > [dchinner@redhat.com] Add a LSN field and update for current verifier
> > infrastructure.
> >
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> > Signed-off-by: Dave Chinner <dchinner@redhat.com>
>
> Been over this and it looks fine.
>
> > @@ -897,6 +910,10 @@ xfs_qm_dqiter_bufs(
> > if (error)
> > break;
> >
> > + /*
> > + * XXX(hch): need to figure out if it makes sense to validate
> > + * the CRC here.
> > + */
>
> What's you're opinion on this?
I did actually change the code to validate the CRC here via
the call to xfs_trans_read_buf(.... &xfs_dquot_buf_ops) above. i.e.
the read verifier will do CRC validation of the buffer.
The comment is still relevant, however, because it's not exactly
clear what the best approach to do here if we get a CRC failure.
That would indicate some kind of corruption that we weren't
expecting, and quite possibly a corruption that rewriting the dquot
can't fix. So without knowing what kind of corruption is typical
here, I don't know what the best approach to take here is.
So effectively what I've done is add CRC validation so that we'll
get people telling us about problems (because the quotacheck will
abort and there will be a stack trace in the log). If it turns out
that corrupted quota files is a common problem that quotacheck
encounters we can gather the corpses, do post-mortem analysis of the
failures and then revisit the code appropriately with that knowledge
in hand.
Cheers,
Dave.
--
Dave Chinner
david@fromorbit.com
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply [flat|nested] 23+ messages in thread
* [PATCH 08/16] xfs: add version 3 inode format with CRCs
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (6 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 07/16] xfs: add CRC checks for quota blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 09/16] xfs: add CRC checks to remote symlinks Dave Chinner
` (7 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Christoph Hellwig <hch@lst.de>
Add a new inode version with a larger core. The primary objective is
to allow for a crc of the inode, and location information (uuid and ino)
to verify it was written in the right place. We also extend it by:
a creation time (for Samba);
a changecount (for NFSv4);
a flush sequence (in LSN format for recovery);
an additional inode flags field; and
some additional padding.
These additional fields are not implemented yet, but already laid
out in the structure.
[dchinner@redhat.com] Added LSN and flags field, some factoring and rework to
capture all the necessary information in the crc calculation.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_buf_item.h | 4 +-
fs/xfs/xfs_dinode.h | 33 +++++++--
fs/xfs/xfs_ialloc.c | 23 +++++-
fs/xfs/xfs_inode.c | 179 ++++++++++++++++++++++++++++++++--------------
fs/xfs/xfs_inode.h | 26 +++++++
fs/xfs/xfs_inode_item.c | 2 +-
fs/xfs/xfs_log_recover.c | 35 +++++++--
fs/xfs/xfs_trans_buf.c | 5 +-
8 files changed, 237 insertions(+), 70 deletions(-)
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index c256606..abae8c8 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -48,6 +48,7 @@ extern kmem_zone_t *xfs_buf_item_zone;
#define XFS_BLF_AGF_BUF (1<<6)
#define XFS_BLF_AGFL_BUF (1<<7)
#define XFS_BLF_AGI_BUF (1<<8)
+#define XFS_BLF_DINO_BUF (1<<9)
#define XFS_BLF_TYPE_MASK \
(XFS_BLF_UDQUOT_BUF | \
@@ -56,7 +57,8 @@ extern kmem_zone_t *xfs_buf_item_zone;
XFS_BLF_BTREE_BUF | \
XFS_BLF_AGF_BUF | \
XFS_BLF_AGFL_BUF | \
- XFS_BLF_AGI_BUF)
+ XFS_BLF_AGI_BUF | \
+ XFS_BLF_DINO_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_dinode.h b/fs/xfs/xfs_dinode.h
index 6b5bd17..bdc946a 100644
--- a/fs/xfs/xfs_dinode.h
+++ b/fs/xfs/xfs_dinode.h
@@ -19,7 +19,7 @@
#define __XFS_DINODE_H__
#define XFS_DINODE_MAGIC 0x494e /* 'IN' */
-#define XFS_DINODE_GOOD_VERSION(v) (((v) == 1 || (v) == 2))
+#define XFS_DINODE_GOOD_VERSION(v) ((v) >= 1 && (v) <= 3)
typedef struct xfs_timestamp {
__be32 t_sec; /* timestamp seconds */
@@ -70,11 +70,36 @@ typedef struct xfs_dinode {
/* di_next_unlinked is the only non-core field in the old dinode */
__be32 di_next_unlinked;/* agi unlinked list ptr */
-} __attribute__((packed)) xfs_dinode_t;
+
+ /* start of the extended dinode, writable fields */
+ __be32 di_crc; /* CRC of the inode */
+ __be64 di_changecount; /* number of attribute changes */
+ __be64 di_lsn; /* flush sequence */
+ __be64 di_flags2; /* more random flags */
+ __u8 di_pad2[16]; /* more padding for future expansion */
+
+ /* fields only written to during inode creation */
+ xfs_timestamp_t di_crtime; /* time created */
+ __be64 di_ino; /* inode number */
+ uuid_t di_uuid; /* UUID of the filesystem */
+
+ /* structure must be padded to 64 bit alignment */
+} xfs_dinode_t;
#define DI_MAX_FLUSH 0xffff
/*
+ * Size of the core inode on disk. Version 1 and 2 inodes have
+ * the same size, but version 3 has grown a few additional fields.
+ */
+static inline uint xfs_dinode_size(int version)
+{
+ if (version == 3)
+ return sizeof(struct xfs_dinode);
+ return offsetof(struct xfs_dinode, di_crc);
+}
+
+/*
* The 32 bit link count in the inode theoretically maxes out at UINT_MAX.
* Since the pathconf interface is signed, we use 2^31 - 1 instead.
* The old inode format had a 16 bit link count, so its maximum is USHRT_MAX.
@@ -105,7 +130,7 @@ typedef enum xfs_dinode_fmt {
* Inode size for given fs.
*/
#define XFS_LITINO(mp, version) \
- ((int)(((mp)->m_sb.sb_inodesize) - sizeof(struct xfs_dinode)))
+ ((int)(((mp)->m_sb.sb_inodesize) - xfs_dinode_size(version)))
#define XFS_BROOT_SIZE_ADJ(ip) \
(XFS_BMBT_BLOCK_LEN((ip)->i_mount) - sizeof(xfs_bmdr_block_t))
@@ -133,7 +158,7 @@ typedef enum xfs_dinode_fmt {
* Return pointers to the data or attribute forks.
*/
#define XFS_DFORK_DPTR(dip) \
- ((char *)(dip) + sizeof(struct xfs_dinode))
+ ((char *)dip + xfs_dinode_size(dip->di_version))
#define XFS_DFORK_APTR(dip) \
(XFS_DFORK_DPTR(dip) + XFS_DFORK_BOFF(dip))
#define XFS_DFORK_PTR(dip,w) \
diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c
index 41c76d6..25e4c17 100644
--- a/fs/xfs/xfs_ialloc.c
+++ b/fs/xfs/xfs_ialloc.c
@@ -167,6 +167,7 @@ xfs_ialloc_inode_init(
int version;
int i, j;
xfs_daddr_t d;
+ xfs_ino_t ino = 0;
/*
* Loop over the new block(s), filling in the inodes.
@@ -190,8 +191,18 @@ xfs_ialloc_inode_init(
* the new inode format, then use the new inode version. Otherwise
* use the old version so that old kernels will continue to be
* able to use the file system.
+ *
+ * For v3 inodes, we also need to write the inode number into the inode,
+ * so calculate the first inode number of the chunk here as
+ * XFS_OFFBNO_TO_AGINO() only works on filesystem block boundaries, not
+ * cluster boundaries and so cannot be used in the cluster buffer loop
+ * below.
*/
- if (xfs_sb_version_hasnlink(&mp->m_sb))
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ version = 3;
+ ino = XFS_AGINO_TO_INO(mp, agno,
+ XFS_OFFBNO_TO_AGINO(mp, agbno, 0));
+ } else if (xfs_sb_version_hasnlink(&mp->m_sb))
version = 2;
else
version = 1;
@@ -217,13 +228,21 @@ xfs_ialloc_inode_init(
xfs_buf_zero(fbuf, 0, ninodes << mp->m_sb.sb_inodelog);
for (i = 0; i < ninodes; i++) {
int ioffset = i << mp->m_sb.sb_inodelog;
- uint isize = sizeof(struct xfs_dinode);
+ uint isize = xfs_dinode_size(version);
free = xfs_make_iptr(mp, fbuf, i);
free->di_magic = cpu_to_be16(XFS_DINODE_MAGIC);
free->di_version = version;
free->di_gen = cpu_to_be32(gen);
free->di_next_unlinked = cpu_to_be32(NULLAGINO);
+
+ if (version == 3) {
+ free->di_ino = cpu_to_be64(ino);
+ ino++;
+ uuid_copy(&free->di_uuid, &mp->m_sb.sb_uuid);
+ xfs_dinode_calc_crc(mp, free);
+ }
+
xfs_trans_log_buf(tp, fbuf, ioffset, ioffset + isize - 1);
}
xfs_trans_inode_alloc_buf(tp, fbuf);
diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
index 202ce37..558ef49 100644
--- a/fs/xfs/xfs_inode.c
+++ b/fs/xfs/xfs_inode.c
@@ -44,6 +44,7 @@
#include "xfs_quota.h"
#include "xfs_filestream.h"
#include "xfs_vnodeops.h"
+#include "xfs_cksum.h"
#include "xfs_trace.h"
#include "xfs_icache.h"
@@ -866,6 +867,17 @@ xfs_dinode_from_disk(
to->di_dmstate = be16_to_cpu(from->di_dmstate);
to->di_flags = be16_to_cpu(from->di_flags);
to->di_gen = be32_to_cpu(from->di_gen);
+
+ if (to->di_version == 3) {
+ to->di_changecount = be64_to_cpu(from->di_changecount);
+ to->di_crtime.t_sec = be32_to_cpu(from->di_crtime.t_sec);
+ to->di_crtime.t_nsec = be32_to_cpu(from->di_crtime.t_nsec);
+ to->di_flags2 = be64_to_cpu(from->di_flags2);
+ to->di_ino = be64_to_cpu(from->di_ino);
+ to->di_lsn = be64_to_cpu(from->di_lsn);
+ memcpy(to->di_pad2, from->di_pad2, sizeof(to->di_pad2));
+ uuid_copy(&to->di_uuid, &from->di_uuid);
+ }
}
void
@@ -902,6 +914,17 @@ xfs_dinode_to_disk(
to->di_dmstate = cpu_to_be16(from->di_dmstate);
to->di_flags = cpu_to_be16(from->di_flags);
to->di_gen = cpu_to_be32(from->di_gen);
+
+ if (from->di_version == 3) {
+ to->di_changecount = cpu_to_be64(from->di_changecount);
+ to->di_crtime.t_sec = cpu_to_be32(from->di_crtime.t_sec);
+ to->di_crtime.t_nsec = cpu_to_be32(from->di_crtime.t_nsec);
+ to->di_flags2 = cpu_to_be64(from->di_flags2);
+ to->di_ino = cpu_to_be64(from->di_ino);
+ to->di_lsn = cpu_to_be64(from->di_lsn);
+ memcpy(to->di_pad2, from->di_pad2, sizeof(to->di_pad2));
+ uuid_copy(&to->di_uuid, &from->di_uuid);
+ }
}
STATIC uint
@@ -962,6 +985,47 @@ xfs_dic2xflags(
(XFS_DFORK_Q(dip) ? XFS_XFLAG_HASATTR : 0);
}
+static bool
+xfs_dinode_verify(
+ struct xfs_mount *mp,
+ struct xfs_inode *ip,
+ struct xfs_dinode *dip)
+{
+ if (dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC))
+ return false;
+
+ /* only version 3 or greater inodes are extensively verified here */
+ if (dip->di_version < 3)
+ return true;
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return false;
+ if (!xfs_verify_cksum((char *)dip, mp->m_sb.sb_inodesize,
+ offsetof(struct xfs_dinode, di_crc)))
+ return false;
+ if (be64_to_cpu(dip->di_ino) != ip->i_ino)
+ return false;
+ if (!uuid_equal(&dip->di_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ return true;
+}
+
+void
+xfs_dinode_calc_crc(
+ struct xfs_mount *mp,
+ struct xfs_dinode *dip)
+{
+ __uint32_t crc;
+
+ if (dip->di_version < 3)
+ return;
+
+ ASSERT(xfs_sb_version_hascrc(&mp->m_sb));
+ crc = xfs_start_cksum((char *)dip, mp->m_sb.sb_inodesize,
+ offsetof(struct xfs_dinode, di_crc));
+ dip->di_crc = xfs_end_cksum(crc);
+}
+
/*
* Read the disk inode attributes into the in-core inode structure.
*/
@@ -990,17 +1054,13 @@ xfs_iread(
if (error)
return error;
- /*
- * If we got something that isn't an inode it means someone
- * (nfs or dmi) has a stale handle.
- */
- if (dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC)) {
-#ifdef DEBUG
- xfs_alert(mp,
- "%s: dip->di_magic (0x%x) != XFS_DINODE_MAGIC (0x%x)",
- __func__, be16_to_cpu(dip->di_magic), XFS_DINODE_MAGIC);
-#endif /* DEBUG */
- error = XFS_ERROR(EINVAL);
+ /* even unallocated inodes are verified */
+ if (!xfs_dinode_verify(mp, ip, dip)) {
+ xfs_alert(mp, "%s: validation failed for inode %lld failed",
+ __func__, ip->i_ino);
+
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, dip);
+ error = XFS_ERROR(EFSCORRUPTED);
goto out_brelse;
}
@@ -1022,10 +1082,20 @@ xfs_iread(
goto out_brelse;
}
} else {
+ /*
+ * Partial initialisation of the in-core inode. Just the bits
+ * that xfs_ialloc won't overwrite or relies on being correct.
+ */
ip->i_d.di_magic = be16_to_cpu(dip->di_magic);
ip->i_d.di_version = dip->di_version;
ip->i_d.di_gen = be32_to_cpu(dip->di_gen);
ip->i_d.di_flushiter = be16_to_cpu(dip->di_flushiter);
+
+ if (dip->di_version == 3) {
+ ip->i_d.di_ino = be64_to_cpu(dip->di_ino);
+ uuid_copy(&ip->i_d.di_uuid, &dip->di_uuid);
+ }
+
/*
* Make sure to pull in the mode here as well in
* case the inode is released without being used.
@@ -1161,6 +1231,7 @@ xfs_ialloc(
xfs_buf_t **ialloc_context,
xfs_inode_t **ipp)
{
+ struct xfs_mount *mp = tp->t_mountp;
xfs_ino_t ino;
xfs_inode_t *ip;
uint flags;
@@ -1187,7 +1258,7 @@ xfs_ialloc(
* This is because we're setting fields here we need
* to prevent others from looking at until we're done.
*/
- error = xfs_iget(tp->t_mountp, tp, ino, XFS_IGET_CREATE,
+ error = xfs_iget(mp, tp, ino, XFS_IGET_CREATE,
XFS_ILOCK_EXCL, &ip);
if (error)
return error;
@@ -1208,7 +1279,7 @@ xfs_ialloc(
* the inode version number now. This way we only do the conversion
* here rather than here and in the flush/logging code.
*/
- if (xfs_sb_version_hasnlink(&tp->t_mountp->m_sb) &&
+ if (xfs_sb_version_hasnlink(&mp->m_sb) &&
ip->i_d.di_version == 1) {
ip->i_d.di_version = 2;
/*
@@ -1258,6 +1329,19 @@ xfs_ialloc(
ip->i_d.di_dmevmask = 0;
ip->i_d.di_dmstate = 0;
ip->i_d.di_flags = 0;
+
+ if (ip->i_d.di_version == 3) {
+ ASSERT(ip->i_d.di_ino == ino);
+ ASSERT(uuid_equal(&ip->i_d.di_uuid, &mp->m_sb.sb_uuid));
+ ip->i_d.di_crc = 0;
+ ip->i_d.di_changecount = 1;
+ ip->i_d.di_lsn = 0;
+ ip->i_d.di_flags2 = 0;
+ memset(&(ip->i_d.di_pad2[0]), 0, sizeof(ip->i_d.di_pad2));
+ ip->i_d.di_crtime = ip->i_d.di_mtime;
+ }
+
+
flags = XFS_ILOG_CORE;
switch (mode & S_IFMT) {
case S_IFIFO:
@@ -2716,20 +2800,18 @@ abort_out:
STATIC int
xfs_iflush_int(
- xfs_inode_t *ip,
- xfs_buf_t *bp)
+ struct xfs_inode *ip,
+ struct xfs_buf *bp)
{
- xfs_inode_log_item_t *iip;
- xfs_dinode_t *dip;
- xfs_mount_t *mp;
+ struct xfs_inode_log_item *iip = ip->i_itemp;
+ struct xfs_dinode *dip;
+ struct xfs_mount *mp = ip->i_mount;
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL|XFS_ILOCK_SHARED));
ASSERT(xfs_isiflocked(ip));
ASSERT(ip->i_d.di_format != XFS_DINODE_FMT_BTREE ||
ip->i_d.di_nextents > XFS_IFORK_MAXEXT(ip, XFS_DATA_FORK));
-
- iip = ip->i_itemp;
- mp = ip->i_mount;
+ ASSERT(iip != NULL && iip->ili_fields != 0);
/* set *dip = inode's place in the buffer */
dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset);
@@ -2790,9 +2872,9 @@ xfs_iflush_int(
}
/*
* bump the flush iteration count, used to detect flushes which
- * postdate a log record during recovery.
+ * postdate a log record during recovery. This is redundant as we now
+ * log every change and hence this can't happen. Still, it doesn't hurt.
*/
-
ip->i_d.di_flushiter++;
/*
@@ -2868,41 +2950,30 @@ xfs_iflush_int(
* need the AIL lock, because it is a 64 bit value that cannot be read
* atomically.
*/
- if (iip != NULL && iip->ili_fields != 0) {
- iip->ili_last_fields = iip->ili_fields;
- iip->ili_fields = 0;
- iip->ili_logged = 1;
+ iip->ili_last_fields = iip->ili_fields;
+ iip->ili_fields = 0;
+ iip->ili_logged = 1;
- xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn,
- &iip->ili_item.li_lsn);
+ xfs_trans_ail_copy_lsn(mp->m_ail, &iip->ili_flush_lsn,
+ &iip->ili_item.li_lsn);
- /*
- * Attach the function xfs_iflush_done to the inode's
- * buffer. This will remove the inode from the AIL
- * and unlock the inode's flush lock when the inode is
- * completely written to disk.
- */
- xfs_buf_attach_iodone(bp, xfs_iflush_done, &iip->ili_item);
+ /*
+ * Attach the function xfs_iflush_done to the inode's
+ * buffer. This will remove the inode from the AIL
+ * and unlock the inode's flush lock when the inode is
+ * completely written to disk.
+ */
+ xfs_buf_attach_iodone(bp, xfs_iflush_done, &iip->ili_item);
- ASSERT(bp->b_fspriv != NULL);
- ASSERT(bp->b_iodone != NULL);
- } else {
- /*
- * We're flushing an inode which is not in the AIL and has
- * not been logged. For this case we can immediately drop
- * the inode flush lock because we can avoid the whole
- * AIL state thing. It's OK to drop the flush lock now,
- * because we've already locked the buffer and to do anything
- * you really need both.
- */
- if (iip != NULL) {
- ASSERT(iip->ili_logged == 0);
- ASSERT(iip->ili_last_fields == 0);
- ASSERT((iip->ili_item.li_flags & XFS_LI_IN_AIL) == 0);
- }
- xfs_ifunlock(ip);
- }
+ /* update the lsn in the on disk inode if required */
+ if (ip->i_d.di_version == 3)
+ dip->di_lsn = cpu_to_be64(iip->ili_item.li_lsn);
+
+ /* generate the checksum. */
+ xfs_dinode_calc_crc(mp, dip);
+ ASSERT(bp->b_fspriv != NULL);
+ ASSERT(bp->b_iodone != NULL);
return 0;
corrupt_out:
diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h
index b8520b5..9112979 100644
--- a/fs/xfs/xfs_inode.h
+++ b/fs/xfs/xfs_inode.h
@@ -150,13 +150,38 @@ typedef struct xfs_icdinode {
__uint16_t di_dmstate; /* DMIG state info */
__uint16_t di_flags; /* random flags, XFS_DIFLAG_... */
__uint32_t di_gen; /* generation number */
+
+ /* di_next_unlinked is the only non-core field in the old dinode */
+ xfs_agino_t di_next_unlinked;/* agi unlinked list ptr */
+
+ /* start of the extended dinode, writable fields */
+ __uint32_t di_crc; /* CRC of the inode */
+ __uint64_t di_changecount; /* number of attribute changes */
+ xfs_lsn_t di_lsn; /* flush sequence */
+ __uint64_t di_flags2; /* more random flags */
+ __uint8_t di_pad2[16]; /* more padding for future expansion */
+
+ /* fields only written to during inode creation */
+ xfs_ictimestamp_t di_crtime; /* time created */
+ xfs_ino_t di_ino; /* inode number */
+ uuid_t di_uuid; /* UUID of the filesystem */
+
+ /* structure must be padded to 64 bit alignment */
} xfs_icdinode_t;
+static inline uint xfs_icdinode_size(int version)
+{
+ if (version == 3)
+ return sizeof(struct xfs_icdinode);
+ return offsetof(struct xfs_icdinode, di_next_unlinked);
+}
+
/*
* Flags for xfs_ichgtime().
*/
#define XFS_ICHGTIME_MOD 0x1 /* data fork modification timestamp */
#define XFS_ICHGTIME_CHG 0x2 /* inode field change timestamp */
+#define XFS_ICHGTIME_CREATE 0x4 /* inode create timestamp */
/*
* Per-fork incore inode flags.
@@ -556,6 +581,7 @@ int xfs_imap_to_bp(struct xfs_mount *, struct xfs_trans *,
struct xfs_buf **, uint, uint);
int xfs_iread(struct xfs_mount *, struct xfs_trans *,
struct xfs_inode *, uint);
+void xfs_dinode_calc_crc(struct xfs_mount *, struct xfs_dinode *);
void xfs_dinode_to_disk(struct xfs_dinode *,
struct xfs_icdinode *);
void xfs_idestroy_fork(struct xfs_inode *, int);
diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c
index f034bd1..f76ff52 100644
--- a/fs/xfs/xfs_inode_item.c
+++ b/fs/xfs/xfs_inode_item.c
@@ -179,7 +179,7 @@ xfs_inode_item_format(
nvecs = 1;
vecp->i_addr = &ip->i_d;
- vecp->i_len = sizeof(struct xfs_icdinode);
+ vecp->i_len = xfs_icdinode_size(ip->i_d.di_version);
vecp->i_type = XLOG_REG_TYPE_ICORE;
vecp++;
nvecs++;
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index e77c770..4096998 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1787,6 +1787,7 @@ xlog_recover_do_inode_buffer(
xfs_agino_t *buffer_nextp;
trace_xfs_log_recover_buf_inode_buf(mp->m_log, buf_f);
+ bp->b_ops = &xfs_inode_buf_ops;
inodes_per_buf = BBTOB(bp->b_io_length) >> mp->m_sb.sb_inodelog;
for (i = 0; i < inodes_per_buf; i++) {
@@ -1931,6 +1932,9 @@ xlog_recover_do_reg_buffer(
/* Shouldn't be any more regions */
ASSERT(i == item->ri_total);
+ /* Shouldn't be any more regions */
+ ASSERT(i == item->ri_total);
+
switch (buf_f->blf_flags & XFS_BLF_TYPE_MASK) {
case XFS_BLF_BTREE_BUF:
switch (be32_to_cpu(*(__be32 *)bp->b_addr)) {
@@ -1990,6 +1994,18 @@ xlog_recover_do_reg_buffer(
}
bp->b_ops = &xfs_dquot_buf_ops;
break;
+ case XFS_BLF_DINO_BUF:
+ /*
+ * we get here with inode allocation buffers, not buffers that
+ * track unlinked list changes.
+ */
+ if (*(__be16 *)bp->b_addr != cpu_to_be16(XFS_DINODE_MAGIC)) {
+ xfs_warn(mp, "Bad INODE block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_inode_buf_ops;
+ break;
default:
break;
}
@@ -2278,6 +2294,7 @@ xlog_recover_inode_pass2(
int attr_index;
uint fields;
xfs_icdinode_t *dicp;
+ uint isize;
int need_free = 0;
if (item->ri_buf[0].i_len == sizeof(xfs_inode_log_format_t)) {
@@ -2303,7 +2320,7 @@ xlog_recover_inode_pass2(
trace_xfs_log_recover_inode_recover(log, in_f);
bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len, 0,
- NULL);
+ &xfs_inode_buf_ops);
if (!bp) {
error = ENOMEM;
goto error;
@@ -2414,7 +2431,8 @@ xlog_recover_inode_pass2(
error = EFSCORRUPTED;
goto error;
}
- if (unlikely(item->ri_buf[1].i_len > sizeof(struct xfs_icdinode))) {
+ isize = xfs_icdinode_size(dicp->di_version);
+ if (unlikely(item->ri_buf[1].i_len > isize)) {
XFS_CORRUPTION_ERROR("xlog_recover_inode_pass2(7)",
XFS_ERRLEVEL_LOW, mp, dicp);
xfs_buf_relse(bp);
@@ -2426,13 +2444,13 @@ xlog_recover_inode_pass2(
}
/* The core is in in-core format */
- xfs_dinode_to_disk(dip, item->ri_buf[1].i_addr);
+ xfs_dinode_to_disk(dip, dicp);
/* the rest is in on-disk format */
- if (item->ri_buf[1].i_len > sizeof(struct xfs_icdinode)) {
- memcpy((xfs_caddr_t) dip + sizeof(struct xfs_icdinode),
- item->ri_buf[1].i_addr + sizeof(struct xfs_icdinode),
- item->ri_buf[1].i_len - sizeof(struct xfs_icdinode));
+ if (item->ri_buf[1].i_len > isize) {
+ memcpy((char *)dip + isize,
+ item->ri_buf[1].i_addr + isize,
+ item->ri_buf[1].i_len - isize);
}
fields = in_f->ilf_fields;
@@ -2516,6 +2534,9 @@ xlog_recover_inode_pass2(
}
write_inode_buffer:
+ /* re-generate the checksum. */
+ xfs_dinode_calc_crc(log->l_mp, dip);
+
ASSERT(bp->b_target->bt_mount == mp);
bp->b_iodone = xlog_recover_iodone;
xfs_buf_delwri_queue(bp, buffer_list);
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index f950edd..8a0f6af 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -704,12 +704,13 @@ xfs_trans_inode_buf(
ASSERT(atomic_read(&bip->bli_refcount) > 0);
bip->bli_flags |= XFS_BLI_INODE_BUF;
+ xfs_trans_buf_set_type(tp, bp, XFS_BLF_DINO_BUF);
}
/*
* This call is used to indicate that the buffer is going to
* be staled and was an inode buffer. This means it gets
- * special processing during unpin - where any inodes
+ * special processing during unpin - where any inodes
* associated with the buffer should be removed from ail.
* There is also special processing during recovery,
* any replay of the inodes in the buffer needs to be
@@ -728,6 +729,7 @@ xfs_trans_stale_inode_buf(
bip->bli_flags |= XFS_BLI_STALE_INODE;
bip->bli_item.li_cb = xfs_buf_iodone;
+ xfs_trans_buf_set_type(tp, bp, XFS_BLF_DINO_BUF);
}
/*
@@ -751,6 +753,7 @@ xfs_trans_inode_alloc_buf(
ASSERT(atomic_read(&bip->bli_refcount) > 0);
bip->bli_flags |= XFS_BLI_INODE_ALLOC_BUF;
+ xfs_trans_buf_set_type(tp, bp, XFS_BLF_DINO_BUF);
}
/*
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 09/16] xfs: add CRC checks to remote symlinks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (7 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 08/16] xfs: add version 3 inode format with CRCs Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-27 23:35 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 10/16] xfs: add CRC checks to block format directory blocks Dave Chinner
` (6 subsequent siblings)
15 siblings, 1 reply; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
Add a header to the remote symlink block, containing location and
owner information, as well as CRCs and LSN fields. This requires
verifiers to be added to the remote symlink buffers for CRC enabled
filesystems.
This also fixes a bug reading multiple block symlinks, where the second block
overwrites the first block when copying out the link name.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/Makefile | 5 +-
fs/xfs/xfs_bmap.c | 21 +-
fs/xfs/xfs_buf_item.h | 4 +-
fs/xfs/xfs_log_recover.c | 9 +
fs/xfs/xfs_vnodeops.c | 480 +---------------------------------------------
5 files changed, 25 insertions(+), 494 deletions(-)
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index d02201d..062f25c 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -45,11 +45,11 @@ xfs-y += xfs_aops.o \
xfs_itable.o \
xfs_message.o \
xfs_mru_cache.o \
- xfs_super.o \
- xfs_xattr.o \
xfs_rename.o \
+ xfs_super.o \
xfs_utils.o \
xfs_vnodeops.o \
+ xfs_xattr.o \
kmem.o \
uuid.o
@@ -73,6 +73,7 @@ xfs-y += xfs_alloc.o \
xfs_inode.o \
xfs_log_recover.o \
xfs_mount.o \
+ xfs_symlink.o \
xfs_trans.o
# low-level transaction/log code
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index 0531cd3..0fcb7f4 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -47,6 +47,7 @@
#include "xfs_filestream.h"
#include "xfs_vnodeops.h"
#include "xfs_trace.h"
+#include "xfs_symlink.h"
kmem_zone_t *xfs_bmap_free_item_zone;
@@ -1321,9 +1322,10 @@ xfs_bmap_add_attrfork_extents(
}
/*
- * Block initialisation functions for local to extent format conversion.
- * As these get more complex, they will be moved to the relevant files,
- * but for now they are too simple to worry about.
+ * Block initialisation function for local to extent format conversion.
+ *
+ * This shouldn't actually be called by anyone, so make sure debug kernels cause
+ * a noticable failure.
*/
STATIC void
xfs_bmap_local_to_extents_init_fn(
@@ -1332,23 +1334,12 @@ xfs_bmap_local_to_extents_init_fn(
struct xfs_inode *ip,
struct xfs_ifork *ifp)
{
+ ASSERT(0);
bp->b_ops = &xfs_bmbt_buf_ops;
memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
xfs_trans_buf_set_type(tp, bp, XFS_BLF_BTREE_BUF);
}
-STATIC void
-xfs_symlink_local_to_remote(
- struct xfs_trans *tp,
- struct xfs_buf *bp,
- struct xfs_inode *ip,
- struct xfs_ifork *ifp)
-{
- /* remote symlink blocks are not verifiable until CRCs come along */
- bp->b_ops = NULL;
- memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
-}
-
/*
* Called from xfs_bmap_add_attrfork to handle local format files. Each
* different data fork content type needs a different callout to do the
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index abae8c8..09cab4e 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -49,6 +49,7 @@ extern kmem_zone_t *xfs_buf_item_zone;
#define XFS_BLF_AGFL_BUF (1<<7)
#define XFS_BLF_AGI_BUF (1<<8)
#define XFS_BLF_DINO_BUF (1<<9)
+#define XFS_BLF_SYMLINK_BUF (1<<10)
#define XFS_BLF_TYPE_MASK \
(XFS_BLF_UDQUOT_BUF | \
@@ -58,7 +59,8 @@ extern kmem_zone_t *xfs_buf_item_zone;
XFS_BLF_AGF_BUF | \
XFS_BLF_AGFL_BUF | \
XFS_BLF_AGI_BUF | \
- XFS_BLF_DINO_BUF)
+ XFS_BLF_DINO_BUF | \
+ XFS_BLF_SYMLINK_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 4096998..e1e3dcf 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -45,6 +45,7 @@
#include "xfs_cksum.h"
#include "xfs_trace.h"
#include "xfs_icache.h"
+#include "xfs_symlink.h"
STATIC int
xlog_find_zeroed(
@@ -2006,6 +2007,14 @@ xlog_recover_do_reg_buffer(
}
bp->b_ops = &xfs_inode_buf_ops;
break;
+ case XFS_BLF_SYMLINK_BUF:
+ if (*(__be32 *)bp->b_addr != cpu_to_be32(XFS_SYMLINK_MAGIC)) {
+ xfs_warn(mp, "Bad symlink block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_symlink_buf_ops;
+ break;
default:
break;
}
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index aa0c066..1034440 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -1,5 +1,6 @@
/*
* Copyright (c) 2000-2006 Silicon Graphics, Inc.
+ * Copyright (c) 2012 Red Hat, Inc.
* All Rights Reserved.
*
* This program is free software; you can redistribute it and/or
@@ -48,103 +49,10 @@
#include "xfs_vnodeops.h"
#include "xfs_trace.h"
#include "xfs_icache.h"
+#include "xfs_symlink.h"
+#include "xfs_cksum.h"
+#include "xfs_buf_item.h"
-/*
- * The maximum pathlen is 1024 bytes. Since the minimum file system
- * blocksize is 512 bytes, we can get a max of 2 extents back from
- * bmapi.
- */
-#define SYMLINK_MAPS 2
-
-STATIC int
-xfs_readlink_bmap(
- xfs_inode_t *ip,
- char *link)
-{
- xfs_mount_t *mp = ip->i_mount;
- int pathlen = ip->i_d.di_size;
- int nmaps = SYMLINK_MAPS;
- xfs_bmbt_irec_t mval[SYMLINK_MAPS];
- xfs_daddr_t d;
- int byte_cnt;
- int n;
- xfs_buf_t *bp;
- int error = 0;
-
- error = xfs_bmapi_read(ip, 0, XFS_B_TO_FSB(mp, pathlen), mval, &nmaps,
- 0);
- if (error)
- goto out;
-
- for (n = 0; n < nmaps; n++) {
- d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
- byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
-
- bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), 0, NULL);
- if (!bp)
- return XFS_ERROR(ENOMEM);
- error = bp->b_error;
- if (error) {
- xfs_buf_ioerror_alert(bp, __func__);
- xfs_buf_relse(bp);
- goto out;
- }
- if (pathlen < byte_cnt)
- byte_cnt = pathlen;
- pathlen -= byte_cnt;
-
- memcpy(link, bp->b_addr, byte_cnt);
- xfs_buf_relse(bp);
- }
-
- link[ip->i_d.di_size] = '\0';
- error = 0;
-
- out:
- return error;
-}
-
-int
-xfs_readlink(
- xfs_inode_t *ip,
- char *link)
-{
- xfs_mount_t *mp = ip->i_mount;
- xfs_fsize_t pathlen;
- int error = 0;
-
- trace_xfs_readlink(ip);
-
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(EIO);
-
- xfs_ilock(ip, XFS_ILOCK_SHARED);
-
- pathlen = ip->i_d.di_size;
- if (!pathlen)
- goto out;
-
- if (pathlen < 0 || pathlen > MAXPATHLEN) {
- xfs_alert(mp, "%s: inode (%llu) bad symlink length (%lld)",
- __func__, (unsigned long long) ip->i_ino,
- (long long) pathlen);
- ASSERT(0);
- error = XFS_ERROR(EFSCORRUPTED);
- goto out;
- }
-
-
- if (ip->i_df.if_flags & XFS_IFINLINE) {
- memcpy(link, ip->i_df.if_u1.if_data, pathlen);
- link[pathlen] = '\0';
- } else {
- error = xfs_readlink_bmap(ip, link);
- }
-
- out:
- xfs_iunlock(ip, XFS_ILOCK_SHARED);
- return error;
-}
/*
* This is called by xfs_inactive to free any blocks beyond eof
@@ -249,145 +157,6 @@ xfs_free_eofblocks(
return error;
}
-/*
- * Free a symlink that has blocks associated with it.
- */
-STATIC int
-xfs_inactive_symlink_rmt(
- xfs_inode_t *ip,
- xfs_trans_t **tpp)
-{
- xfs_buf_t *bp;
- int committed;
- int done;
- int error;
- xfs_fsblock_t first_block;
- xfs_bmap_free_t free_list;
- int i;
- xfs_mount_t *mp;
- xfs_bmbt_irec_t mval[SYMLINK_MAPS];
- int nmaps;
- xfs_trans_t *ntp;
- int size;
- xfs_trans_t *tp;
-
- tp = *tpp;
- mp = ip->i_mount;
- ASSERT(ip->i_d.di_size > XFS_IFORK_DSIZE(ip));
- /*
- * We're freeing a symlink that has some
- * blocks allocated to it. Free the
- * blocks here. We know that we've got
- * either 1 or 2 extents and that we can
- * free them all in one bunmapi call.
- */
- ASSERT(ip->i_d.di_nextents > 0 && ip->i_d.di_nextents <= 2);
-
- /*
- * Lock the inode, fix the size, and join it to the transaction.
- * Hold it so in the normal path, we still have it locked for
- * the second transaction. In the error paths we need it
- * held so the cancel won't rele it, see below.
- */
- size = (int)ip->i_d.di_size;
- ip->i_d.di_size = 0;
- xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
- /*
- * Find the block(s) so we can inval and unmap them.
- */
- done = 0;
- xfs_bmap_init(&free_list, &first_block);
- nmaps = ARRAY_SIZE(mval);
- error = xfs_bmapi_read(ip, 0, XFS_B_TO_FSB(mp, size),
- mval, &nmaps, 0);
- if (error)
- goto error0;
- /*
- * Invalidate the block(s).
- */
- for (i = 0; i < nmaps; i++) {
- bp = xfs_trans_get_buf(tp, mp->m_ddev_targp,
- XFS_FSB_TO_DADDR(mp, mval[i].br_startblock),
- XFS_FSB_TO_BB(mp, mval[i].br_blockcount), 0);
- if (!bp) {
- error = ENOMEM;
- goto error1;
- }
- xfs_trans_binval(tp, bp);
- }
- /*
- * Unmap the dead block(s) to the free_list.
- */
- if ((error = xfs_bunmapi(tp, ip, 0, size, XFS_BMAPI_METADATA, nmaps,
- &first_block, &free_list, &done)))
- goto error1;
- ASSERT(done);
- /*
- * Commit the first transaction. This logs the EFI and the inode.
- */
- if ((error = xfs_bmap_finish(&tp, &free_list, &committed)))
- goto error1;
- /*
- * The transaction must have been committed, since there were
- * actually extents freed by xfs_bunmapi. See xfs_bmap_finish.
- * The new tp has the extent freeing and EFDs.
- */
- ASSERT(committed);
- /*
- * The first xact was committed, so add the inode to the new one.
- * Mark it dirty so it will be logged and moved forward in the log as
- * part of every commit.
- */
- xfs_trans_ijoin(tp, ip, 0);
- xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
- /*
- * Get a new, empty transaction to return to our caller.
- */
- ntp = xfs_trans_dup(tp);
- /*
- * Commit the transaction containing extent freeing and EFDs.
- * If we get an error on the commit here or on the reserve below,
- * we need to unlock the inode since the new transaction doesn't
- * have the inode attached.
- */
- error = xfs_trans_commit(tp, 0);
- tp = ntp;
- if (error) {
- ASSERT(XFS_FORCED_SHUTDOWN(mp));
- goto error0;
- }
- /*
- * transaction commit worked ok so we can drop the extra ticket
- * reference that we gained in xfs_trans_dup()
- */
- xfs_log_ticket_put(tp->t_ticket);
-
- /*
- * Remove the memory for extent descriptions (just bookkeeping).
- */
- if (ip->i_df.if_bytes)
- xfs_idata_realloc(ip, -ip->i_df.if_bytes, XFS_DATA_FORK);
- ASSERT(ip->i_df.if_bytes == 0);
- /*
- * Put an itruncate log reservation in the new transaction
- * for our caller.
- */
- if ((error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
- XFS_TRANS_PERM_LOG_RES, XFS_ITRUNCATE_LOG_COUNT))) {
- ASSERT(XFS_FORCED_SHUTDOWN(mp));
- goto error0;
- }
-
- xfs_trans_ijoin(tp, ip, 0);
- *tpp = tp;
- return 0;
-
- error1:
- xfs_bmap_cancel(&free_list);
- error0:
- return error;
-}
-
int
xfs_release(
xfs_inode_t *ip)
@@ -1353,247 +1122,6 @@ xfs_link(
}
int
-xfs_symlink(
- xfs_inode_t *dp,
- struct xfs_name *link_name,
- const char *target_path,
- umode_t mode,
- xfs_inode_t **ipp)
-{
- xfs_mount_t *mp = dp->i_mount;
- xfs_trans_t *tp;
- xfs_inode_t *ip;
- int error;
- int pathlen;
- xfs_bmap_free_t free_list;
- xfs_fsblock_t first_block;
- bool unlock_dp_on_error = false;
- uint cancel_flags;
- int committed;
- xfs_fileoff_t first_fsb;
- xfs_filblks_t fs_blocks;
- int nmaps;
- xfs_bmbt_irec_t mval[SYMLINK_MAPS];
- xfs_daddr_t d;
- const char *cur_chunk;
- int byte_cnt;
- int n;
- xfs_buf_t *bp;
- prid_t prid;
- struct xfs_dquot *udqp, *gdqp;
- uint resblks;
-
- *ipp = NULL;
- error = 0;
- ip = NULL;
- tp = NULL;
-
- trace_xfs_symlink(dp, link_name);
-
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(EIO);
-
- /*
- * Check component lengths of the target path name.
- */
- pathlen = strlen(target_path);
- if (pathlen >= MAXPATHLEN) /* total string too long */
- return XFS_ERROR(ENAMETOOLONG);
-
- udqp = gdqp = NULL;
- if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
- prid = xfs_get_projid(dp);
- else
- prid = XFS_PROJID_DEFAULT;
-
- /*
- * Make sure that we have allocated dquot(s) on disk.
- */
- error = xfs_qm_vop_dqalloc(dp, current_fsuid(), current_fsgid(), prid,
- XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT, &udqp, &gdqp);
- if (error)
- goto std_return;
-
- tp = xfs_trans_alloc(mp, XFS_TRANS_SYMLINK);
- cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
- /*
- * The symlink will fit into the inode data fork?
- * There can't be any attributes so we get the whole variable part.
- */
- if (pathlen <= XFS_LITINO(mp, dp->i_d.di_version))
- fs_blocks = 0;
- else
- fs_blocks = XFS_B_TO_FSB(mp, pathlen);
- resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
- error = xfs_trans_reserve(tp, resblks, XFS_SYMLINK_LOG_RES(mp), 0,
- XFS_TRANS_PERM_LOG_RES, XFS_SYMLINK_LOG_COUNT);
- if (error == ENOSPC && fs_blocks == 0) {
- resblks = 0;
- error = xfs_trans_reserve(tp, 0, XFS_SYMLINK_LOG_RES(mp), 0,
- XFS_TRANS_PERM_LOG_RES, XFS_SYMLINK_LOG_COUNT);
- }
- if (error) {
- cancel_flags = 0;
- goto error_return;
- }
-
- xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
- unlock_dp_on_error = true;
-
- /*
- * Check whether the directory allows new symlinks or not.
- */
- if (dp->i_d.di_flags & XFS_DIFLAG_NOSYMLINKS) {
- error = XFS_ERROR(EPERM);
- goto error_return;
- }
-
- /*
- * Reserve disk quota : blocks and inode.
- */
- error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp, resblks, 1, 0);
- if (error)
- goto error_return;
-
- /*
- * Check for ability to enter directory entry, if no space reserved.
- */
- error = xfs_dir_canenter(tp, dp, link_name, resblks);
- if (error)
- goto error_return;
- /*
- * Initialize the bmap freelist prior to calling either
- * bmapi or the directory create code.
- */
- xfs_bmap_init(&free_list, &first_block);
-
- /*
- * Allocate an inode for the symlink.
- */
- error = xfs_dir_ialloc(&tp, dp, S_IFLNK | (mode & ~S_IFMT), 1, 0,
- prid, resblks > 0, &ip, NULL);
- if (error) {
- if (error == ENOSPC)
- goto error_return;
- goto error1;
- }
-
- /*
- * An error after we've joined dp to the transaction will result in the
- * transaction cancel unlocking dp so don't do it explicitly in the
- * error path.
- */
- xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
- unlock_dp_on_error = false;
-
- /*
- * Also attach the dquot(s) to it, if applicable.
- */
- xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp);
-
- if (resblks)
- resblks -= XFS_IALLOC_SPACE_RES(mp);
- /*
- * If the symlink will fit into the inode, write it inline.
- */
- if (pathlen <= XFS_IFORK_DSIZE(ip)) {
- xfs_idata_realloc(ip, pathlen, XFS_DATA_FORK);
- memcpy(ip->i_df.if_u1.if_data, target_path, pathlen);
- ip->i_d.di_size = pathlen;
-
- /*
- * The inode was initially created in extent format.
- */
- ip->i_df.if_flags &= ~(XFS_IFEXTENTS | XFS_IFBROOT);
- ip->i_df.if_flags |= XFS_IFINLINE;
-
- ip->i_d.di_format = XFS_DINODE_FMT_LOCAL;
- xfs_trans_log_inode(tp, ip, XFS_ILOG_DDATA | XFS_ILOG_CORE);
-
- } else {
- first_fsb = 0;
- nmaps = SYMLINK_MAPS;
-
- error = xfs_bmapi_write(tp, ip, first_fsb, fs_blocks,
- XFS_BMAPI_METADATA, &first_block, resblks,
- mval, &nmaps, &free_list);
- if (error)
- goto error2;
-
- if (resblks)
- resblks -= fs_blocks;
- ip->i_d.di_size = pathlen;
- xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-
- cur_chunk = target_path;
- for (n = 0; n < nmaps; n++) {
- d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
- byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
- bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
- BTOBB(byte_cnt), 0);
- if (!bp) {
- error = ENOMEM;
- goto error2;
- }
- if (pathlen < byte_cnt) {
- byte_cnt = pathlen;
- }
- pathlen -= byte_cnt;
-
- memcpy(bp->b_addr, cur_chunk, byte_cnt);
- cur_chunk += byte_cnt;
-
- xfs_trans_log_buf(tp, bp, 0, byte_cnt - 1);
- }
- }
-
- /*
- * Create the directory entry for the symlink.
- */
- error = xfs_dir_createname(tp, dp, link_name, ip->i_ino,
- &first_block, &free_list, resblks);
- if (error)
- goto error2;
- xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
- xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
-
- /*
- * If this is a synchronous mount, make sure that the
- * symlink transaction goes to disk before returning to
- * the user.
- */
- if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
- xfs_trans_set_sync(tp);
- }
-
- error = xfs_bmap_finish(&tp, &free_list, &committed);
- if (error) {
- goto error2;
- }
- error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
- xfs_qm_dqrele(udqp);
- xfs_qm_dqrele(gdqp);
-
- *ipp = ip;
- return 0;
-
- error2:
- IRELE(ip);
- error1:
- xfs_bmap_cancel(&free_list);
- cancel_flags |= XFS_TRANS_ABORT;
- error_return:
- xfs_trans_cancel(tp, cancel_flags);
- xfs_qm_dqrele(udqp);
- xfs_qm_dqrele(gdqp);
-
- if (unlock_dp_on_error)
- xfs_iunlock(dp, XFS_ILOCK_EXCL);
- std_return:
- return error;
-}
-
-int
xfs_set_dmattrs(
xfs_inode_t *ip,
u_int evmask,
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH 09/16] xfs: add CRC checks to remote symlinks
2013-02-25 1:31 ` [PATCH 09/16] xfs: add CRC checks to remote symlinks Dave Chinner
@ 2013-02-27 23:35 ` Dave Chinner
0 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-27 23:35 UTC (permalink / raw)
To: xfs
On Mon, Feb 25, 2013 at 12:31:34PM +1100, Dave Chinner wrote:
> From: Dave Chinner <dchinner@redhat.com>
>
> Add a header to the remote symlink block, containing location and
> owner information, as well as CRCs and LSN fields. This requires
> verifiers to be added to the remote symlink buffers for CRC enabled
> filesystems.
>
> This also fixes a bug reading multiple block symlinks, where the second block
> overwrites the first block when copying out the link name.
>
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
> ---
> fs/xfs/Makefile | 5 +-
> fs/xfs/xfs_bmap.c | 21 +-
> fs/xfs/xfs_buf_item.h | 4 +-
> fs/xfs/xfs_log_recover.c | 9 +
> fs/xfs/xfs_vnodeops.c | 480 +---------------------------------------------
> 5 files changed, 25 insertions(+), 494 deletions(-)
I just noticed this is missing two new files (insert grumble about
guilt forgetting added files when a patch is force applied because
of conflicts).
The correct version is below (diffstat += 770 lines)...
-Dave.
--
Dave Chinner
david@fromorbit.com
xfs: add CRC checks to remote symlinks
From: Dave Chinner <dchinner@redhat.com>
Add a header to the remote symlink block, containing location and
owner information, as well as CRCs and LSN fields. This requires
verifiers to be added to the remote symlink buffers for CRC enabled
filesystems.
This also fixes a bug reading multiple block symlinks, where the second block
overwrites the first block when copying out the link name.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/Makefile | 5 +-
fs/xfs/xfs_bmap.c | 21 +-
fs/xfs/xfs_buf_item.h | 4 +-
fs/xfs/xfs_log_recover.c | 9 +
fs/xfs/xfs_symlink.c | 719 ++++++++++++++++++++++++++++++++++++++++++++++
fs/xfs/xfs_symlink.h | 53 ++++
fs/xfs/xfs_vnodeops.c | 480 +------------------------------
7 files changed, 797 insertions(+), 494 deletions(-)
diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
index d02201d..062f25c 100644
--- a/fs/xfs/Makefile
+++ b/fs/xfs/Makefile
@@ -45,11 +45,11 @@ xfs-y += xfs_aops.o \
xfs_itable.o \
xfs_message.o \
xfs_mru_cache.o \
- xfs_super.o \
- xfs_xattr.o \
xfs_rename.o \
+ xfs_super.o \
xfs_utils.o \
xfs_vnodeops.o \
+ xfs_xattr.o \
kmem.o \
uuid.o
@@ -73,6 +73,7 @@ xfs-y += xfs_alloc.o \
xfs_inode.o \
xfs_log_recover.o \
xfs_mount.o \
+ xfs_symlink.o \
xfs_trans.o
# low-level transaction/log code
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index 0531cd3..0fcb7f4 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -47,6 +47,7 @@
#include "xfs_filestream.h"
#include "xfs_vnodeops.h"
#include "xfs_trace.h"
+#include "xfs_symlink.h"
kmem_zone_t *xfs_bmap_free_item_zone;
@@ -1321,9 +1322,10 @@ xfs_bmap_add_attrfork_extents(
}
/*
- * Block initialisation functions for local to extent format conversion.
- * As these get more complex, they will be moved to the relevant files,
- * but for now they are too simple to worry about.
+ * Block initialisation function for local to extent format conversion.
+ *
+ * This shouldn't actually be called by anyone, so make sure debug kernels cause
+ * a noticable failure.
*/
STATIC void
xfs_bmap_local_to_extents_init_fn(
@@ -1332,23 +1334,12 @@ xfs_bmap_local_to_extents_init_fn(
struct xfs_inode *ip,
struct xfs_ifork *ifp)
{
+ ASSERT(0);
bp->b_ops = &xfs_bmbt_buf_ops;
memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
xfs_trans_buf_set_type(tp, bp, XFS_BLF_BTREE_BUF);
}
-STATIC void
-xfs_symlink_local_to_remote(
- struct xfs_trans *tp,
- struct xfs_buf *bp,
- struct xfs_inode *ip,
- struct xfs_ifork *ifp)
-{
- /* remote symlink blocks are not verifiable until CRCs come along */
- bp->b_ops = NULL;
- memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
-}
-
/*
* Called from xfs_bmap_add_attrfork to handle local format files. Each
* different data fork content type needs a different callout to do the
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index abae8c8..09cab4e 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -49,6 +49,7 @@ extern kmem_zone_t *xfs_buf_item_zone;
#define XFS_BLF_AGFL_BUF (1<<7)
#define XFS_BLF_AGI_BUF (1<<8)
#define XFS_BLF_DINO_BUF (1<<9)
+#define XFS_BLF_SYMLINK_BUF (1<<10)
#define XFS_BLF_TYPE_MASK \
(XFS_BLF_UDQUOT_BUF | \
@@ -58,7 +59,8 @@ extern kmem_zone_t *xfs_buf_item_zone;
XFS_BLF_AGF_BUF | \
XFS_BLF_AGFL_BUF | \
XFS_BLF_AGI_BUF | \
- XFS_BLF_DINO_BUF)
+ XFS_BLF_DINO_BUF | \
+ XFS_BLF_SYMLINK_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 4096998..e1e3dcf 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -45,6 +45,7 @@
#include "xfs_cksum.h"
#include "xfs_trace.h"
#include "xfs_icache.h"
+#include "xfs_symlink.h"
STATIC int
xlog_find_zeroed(
@@ -2006,6 +2007,14 @@ xlog_recover_do_reg_buffer(
}
bp->b_ops = &xfs_inode_buf_ops;
break;
+ case XFS_BLF_SYMLINK_BUF:
+ if (*(__be32 *)bp->b_addr != cpu_to_be32(XFS_SYMLINK_MAGIC)) {
+ xfs_warn(mp, "Bad symlink block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_symlink_buf_ops;
+ break;
default:
break;
}
diff --git a/fs/xfs/xfs_symlink.c b/fs/xfs/xfs_symlink.c
new file mode 100644
index 0000000..33e8f13
--- /dev/null
+++ b/fs/xfs/xfs_symlink.c
@@ -0,0 +1,719 @@
+/*
+ * Copyright (c) 2000-2006 Silicon Graphics, Inc.
+ * Copyright (c) 2012-2013 Red Hat, Inc.
+ * All rights reserved.
+ */
+
+#include "xfs.h"
+#include "xfs_fs.h"
+#include "xfs_types.h"
+#include "xfs_bit.h"
+#include "xfs_log.h"
+#include "xfs_trans.h"
+#include "xfs_sb.h"
+#include "xfs_ag.h"
+#include "xfs_dir2.h"
+#include "xfs_mount.h"
+#include "xfs_da_btree.h"
+#include "xfs_bmap_btree.h"
+#include "xfs_ialloc_btree.h"
+#include "xfs_dinode.h"
+#include "xfs_inode.h"
+#include "xfs_inode_item.h"
+#include "xfs_itable.h"
+#include "xfs_ialloc.h"
+#include "xfs_alloc.h"
+#include "xfs_bmap.h"
+#include "xfs_error.h"
+#include "xfs_quota.h"
+#include "xfs_utils.h"
+#include "xfs_trans_space.h"
+#include "xfs_log_priv.h"
+#include "xfs_trace.h"
+#include "xfs_symlink.h"
+#include "xfs_cksum.h"
+#include "xfs_buf_item.h"
+
+
+/*
+ * Each contiguous block has a header, so it is not just a simple pathlen
+ * to FSB conversion.
+ */
+int
+xfs_symlink_blocks(
+ struct xfs_mount *mp,
+ int pathlen)
+{
+ int fsblocks = 0;
+ int len = pathlen;
+
+ do {
+ fsblocks++;
+ len -= XFS_SYMLINK_BUF_SPACE(mp, mp->m_sb.sb_blocksize);
+ } while (len > 0);
+
+ ASSERT(fsblocks <= XFS_SYMLINK_MAPS);
+ return fsblocks;
+}
+
+static int
+xfs_symlink_hdr_set(
+ struct xfs_mount *mp,
+ xfs_ino_t ino,
+ uint32_t offset,
+ uint32_t size,
+ struct xfs_buf *bp)
+{
+ struct xfs_dsymlink_hdr *dsl = bp->b_addr;
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return 0;
+
+ dsl->sl_magic = cpu_to_be32(XFS_SYMLINK_MAGIC);
+ dsl->sl_offset = cpu_to_be32(offset);
+ dsl->sl_bytes = cpu_to_be32(size);
+ uuid_copy(&dsl->sl_uuid, &mp->m_sb.sb_uuid);
+ dsl->sl_owner = cpu_to_be64(ino);
+ dsl->sl_blkno = cpu_to_be64(bp->b_bn);
+ bp->b_ops = &xfs_symlink_buf_ops;
+
+ return sizeof(struct xfs_dsymlink_hdr);
+}
+
+/*
+ * Checking of the symlink header is split into two parts. the verifier does
+ * CRC, location and bounds checking, the unpacking function checks the path
+ * parameters and owner.
+ */
+bool
+xfs_symlink_hdr_ok(
+ struct xfs_mount *mp,
+ xfs_ino_t ino,
+ uint32_t offset,
+ uint32_t size,
+ struct xfs_buf *bp)
+{
+ struct xfs_dsymlink_hdr *dsl = bp->b_addr;
+
+ if (offset != be32_to_cpu(dsl->sl_offset))
+ return false;
+ if (size != be32_to_cpu(dsl->sl_bytes))
+ return false;
+ if (ino != be64_to_cpu(dsl->sl_owner))
+ return false;
+
+ /* ok */
+ return true;
+
+}
+
+static bool
+xfs_symlink_verify(
+ struct xfs_buf *bp)
+{
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_dsymlink_hdr *dsl = bp->b_addr;
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return false;
+ if (dsl->sl_magic != cpu_to_be32(XFS_SYMLINK_MAGIC))
+ return false;
+ if (!uuid_equal(&dsl->sl_uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (bp->b_bn != be64_to_cpu(dsl->sl_blkno))
+ return false;
+ if (be32_to_cpu(dsl->sl_offset) +
+ be32_to_cpu(dsl->sl_bytes) >= MAXPATHLEN)
+ return false;
+ if (dsl->sl_owner == 0)
+ return false;
+
+ return true;
+}
+
+static void
+xfs_symlink_read_verify(
+ struct xfs_buf *bp)
+{
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ /* no verification of non-crc buffers */
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (!xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_dsymlink_hdr, sl_crc)) ||
+ !xfs_symlink_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
+}
+
+static void
+xfs_symlink_write_verify(
+ struct xfs_buf *bp)
+{
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+
+ /* no verification of non-crc buffers */
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (!xfs_symlink_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (bip) {
+ struct xfs_dsymlink_hdr *dsl = bp->b_addr;
+ dsl->sl_lsn = cpu_to_be64(bip->bli_item.li_lsn);
+ }
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length),
+ offsetof(struct xfs_dsymlink_hdr, sl_crc));
+}
+
+const struct xfs_buf_ops xfs_symlink_buf_ops = {
+ .verify_read = xfs_symlink_read_verify,
+ .verify_write = xfs_symlink_write_verify,
+};
+
+void
+xfs_symlink_local_to_remote(
+ struct xfs_trans *tp,
+ struct xfs_buf *bp,
+ struct xfs_inode *ip,
+ struct xfs_ifork *ifp)
+{
+ struct xfs_mount *mp = ip->i_mount;
+ char *buf;
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb)) {
+ bp->b_ops = NULL;
+ memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes);
+ return;
+ }
+
+ /*
+ * As this symlink fits in an inode literal area, it must also fit in
+ * the smallest buffer the filesystem supports.
+ */
+ ASSERT(BBTOB(bp->b_length) >=
+ ifp->if_bytes + sizeof(struct xfs_dsymlink_hdr));
+
+ bp->b_ops = &xfs_symlink_buf_ops;
+
+ buf = bp->b_addr;
+ buf += xfs_symlink_hdr_set(mp, ip->i_ino, 0, ifp->if_bytes, bp);
+ memcpy(buf, ifp->if_u1.if_data, ifp->if_bytes);
+}
+
+/* ----- Kernel only functions below ----- */
+STATIC int
+xfs_readlink_bmap(
+ struct xfs_inode *ip,
+ char *link)
+{
+ struct xfs_mount *mp = ip->i_mount;
+ struct xfs_bmbt_irec mval[XFS_SYMLINK_MAPS];
+ struct xfs_buf *bp;
+ xfs_daddr_t d;
+ char *cur_chunk;
+ int pathlen = ip->i_d.di_size;
+ int nmaps = XFS_SYMLINK_MAPS;
+ int byte_cnt;
+ int n;
+ int error = 0;
+ int fsblocks = 0;
+ int offset;
+
+ fsblocks = xfs_symlink_blocks(mp, pathlen);
+ error = xfs_bmapi_read(ip, 0, fsblocks, mval, &nmaps, 0);
+ if (error)
+ goto out;
+
+ offset = 0;
+ for (n = 0; n < nmaps; n++) {
+ d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
+ byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
+
+ bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), 0,
+ &xfs_symlink_buf_ops);
+ if (!bp)
+ return XFS_ERROR(ENOMEM);
+ error = bp->b_error;
+ if (error) {
+ xfs_buf_ioerror_alert(bp, __func__);
+ xfs_buf_relse(bp);
+ goto out;
+ }
+ byte_cnt = XFS_SYMLINK_BUF_SPACE(mp, byte_cnt);
+ if (pathlen < byte_cnt)
+ byte_cnt = pathlen;
+
+ cur_chunk = bp->b_addr;
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ if (!xfs_symlink_hdr_ok(mp, ip->i_ino, offset,
+ byte_cnt, bp)) {
+ error = EFSCORRUPTED;
+ xfs_alert(mp,
+"symlink header does not match required off/len/owner (0x%x/Ox%x,0x%llx)",
+ offset, byte_cnt, ip->i_ino);
+ xfs_buf_relse(bp);
+ goto out;
+
+ }
+
+ cur_chunk += sizeof(struct xfs_dsymlink_hdr);
+ }
+
+ memcpy(link + offset, bp->b_addr, byte_cnt);
+
+ pathlen -= byte_cnt;
+ offset += byte_cnt;
+
+ xfs_buf_relse(bp);
+ }
+ ASSERT(pathlen == 0);
+
+ link[ip->i_d.di_size] = '\0';
+ error = 0;
+
+ out:
+ return error;
+}
+
+int
+xfs_readlink(
+ struct xfs_inode *ip,
+ char *link)
+{
+ struct xfs_mount *mp = ip->i_mount;
+ xfs_fsize_t pathlen;
+ int error = 0;
+
+ trace_xfs_readlink(ip);
+
+ if (XFS_FORCED_SHUTDOWN(mp))
+ return XFS_ERROR(EIO);
+
+ xfs_ilock(ip, XFS_ILOCK_SHARED);
+
+ pathlen = ip->i_d.di_size;
+ if (!pathlen)
+ goto out;
+
+ if (pathlen < 0 || pathlen > MAXPATHLEN) {
+ xfs_alert(mp, "%s: inode (%llu) bad symlink length (%lld)",
+ __func__, (unsigned long long) ip->i_ino,
+ (long long) pathlen);
+ ASSERT(0);
+ error = XFS_ERROR(EFSCORRUPTED);
+ goto out;
+ }
+
+
+ if (ip->i_df.if_flags & XFS_IFINLINE) {
+ memcpy(link, ip->i_df.if_u1.if_data, pathlen);
+ link[pathlen] = '\0';
+ } else {
+ error = xfs_readlink_bmap(ip, link);
+ }
+
+ out:
+ xfs_iunlock(ip, XFS_ILOCK_SHARED);
+ return error;
+}
+
+int
+xfs_symlink(
+ struct xfs_inode *dp,
+ struct xfs_name *link_name,
+ const char *target_path,
+ umode_t mode,
+ struct xfs_inode **ipp)
+{
+ struct xfs_mount *mp = dp->i_mount;
+ struct xfs_trans *tp = NULL;
+ struct xfs_inode *ip = NULL;
+ int error = 0;
+ int pathlen;
+ struct xfs_bmap_free free_list;
+ xfs_fsblock_t first_block;
+ bool unlock_dp_on_error = false;
+ uint cancel_flags;
+ int committed;
+ xfs_fileoff_t first_fsb;
+ xfs_filblks_t fs_blocks;
+ int nmaps;
+ struct xfs_bmbt_irec mval[XFS_SYMLINK_MAPS];
+ xfs_daddr_t d;
+ const char *cur_chunk;
+ int byte_cnt;
+ int n;
+ xfs_buf_t *bp;
+ prid_t prid;
+ struct xfs_dquot *udqp, *gdqp;
+ uint resblks;
+
+ *ipp = NULL;
+
+ trace_xfs_symlink(dp, link_name);
+
+ if (XFS_FORCED_SHUTDOWN(mp))
+ return XFS_ERROR(EIO);
+
+ /*
+ * Check component lengths of the target path name.
+ */
+ pathlen = strlen(target_path);
+ if (pathlen >= MAXPATHLEN) /* total string too long */
+ return XFS_ERROR(ENAMETOOLONG);
+
+ udqp = gdqp = NULL;
+ if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
+ prid = xfs_get_projid(dp);
+ else
+ prid = XFS_PROJID_DEFAULT;
+
+ /*
+ * Make sure that we have allocated dquot(s) on disk.
+ */
+ error = xfs_qm_vop_dqalloc(dp, current_fsuid(), current_fsgid(), prid,
+ XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT, &udqp, &gdqp);
+ if (error)
+ goto std_return;
+
+ tp = xfs_trans_alloc(mp, XFS_TRANS_SYMLINK);
+ cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
+ /*
+ * The symlink will fit into the inode data fork?
+ * There can't be any attributes so we get the whole variable part.
+ */
+ if (pathlen <= XFS_LITINO(mp, dp->i_d.di_version))
+ fs_blocks = 0;
+ else
+ fs_blocks = XFS_B_TO_FSB(mp, pathlen);
+ resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
+ error = xfs_trans_reserve(tp, resblks, XFS_SYMLINK_LOG_RES(mp), 0,
+ XFS_TRANS_PERM_LOG_RES, XFS_SYMLINK_LOG_COUNT);
+ if (error == ENOSPC && fs_blocks == 0) {
+ resblks = 0;
+ error = xfs_trans_reserve(tp, 0, XFS_SYMLINK_LOG_RES(mp), 0,
+ XFS_TRANS_PERM_LOG_RES, XFS_SYMLINK_LOG_COUNT);
+ }
+ if (error) {
+ cancel_flags = 0;
+ goto error_return;
+ }
+
+ xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
+ unlock_dp_on_error = true;
+
+ /*
+ * Check whether the directory allows new symlinks or not.
+ */
+ if (dp->i_d.di_flags & XFS_DIFLAG_NOSYMLINKS) {
+ error = XFS_ERROR(EPERM);
+ goto error_return;
+ }
+
+ /*
+ * Reserve disk quota : blocks and inode.
+ */
+ error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp, resblks, 1, 0);
+ if (error)
+ goto error_return;
+
+ /*
+ * Check for ability to enter directory entry, if no space reserved.
+ */
+ error = xfs_dir_canenter(tp, dp, link_name, resblks);
+ if (error)
+ goto error_return;
+ /*
+ * Initialize the bmap freelist prior to calling either
+ * bmapi or the directory create code.
+ */
+ xfs_bmap_init(&free_list, &first_block);
+
+ /*
+ * Allocate an inode for the symlink.
+ */
+ error = xfs_dir_ialloc(&tp, dp, S_IFLNK | (mode & ~S_IFMT), 1, 0,
+ prid, resblks > 0, &ip, NULL);
+ if (error) {
+ if (error == ENOSPC)
+ goto error_return;
+ goto error1;
+ }
+
+ /*
+ * An error after we've joined dp to the transaction will result in the
+ * transaction cancel unlocking dp so don't do it explicitly in the
+ * error path.
+ */
+ xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
+ unlock_dp_on_error = false;
+
+ /*
+ * Also attach the dquot(s) to it, if applicable.
+ */
+ xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp);
+
+ if (resblks)
+ resblks -= XFS_IALLOC_SPACE_RES(mp);
+ /*
+ * If the symlink will fit into the inode, write it inline.
+ */
+ if (pathlen <= XFS_IFORK_DSIZE(ip)) {
+ xfs_idata_realloc(ip, pathlen, XFS_DATA_FORK);
+ memcpy(ip->i_df.if_u1.if_data, target_path, pathlen);
+ ip->i_d.di_size = pathlen;
+
+ /*
+ * The inode was initially created in extent format.
+ */
+ ip->i_df.if_flags &= ~(XFS_IFEXTENTS | XFS_IFBROOT);
+ ip->i_df.if_flags |= XFS_IFINLINE;
+
+ ip->i_d.di_format = XFS_DINODE_FMT_LOCAL;
+ xfs_trans_log_inode(tp, ip, XFS_ILOG_DDATA | XFS_ILOG_CORE);
+
+ } else {
+ int offset;
+
+ first_fsb = 0;
+ nmaps = XFS_SYMLINK_MAPS;
+
+ error = xfs_bmapi_write(tp, ip, first_fsb, fs_blocks,
+ XFS_BMAPI_METADATA, &first_block, resblks,
+ mval, &nmaps, &free_list);
+ if (error)
+ goto error2;
+
+ if (resblks)
+ resblks -= fs_blocks;
+ ip->i_d.di_size = pathlen;
+ xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+
+ cur_chunk = target_path;
+ offset = 0;
+ for (n = 0; n < nmaps; n++) {
+ char *buf;
+
+ d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
+ byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
+ bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
+ BTOBB(byte_cnt), 0);
+ if (!bp) {
+ error = ENOMEM;
+ goto error2;
+ }
+ bp->b_ops = &xfs_symlink_buf_ops;
+
+ byte_cnt = XFS_SYMLINK_BUF_SPACE(mp, byte_cnt);
+ if (pathlen < byte_cnt) {
+ byte_cnt = pathlen;
+ }
+
+ buf = bp->b_addr;
+ buf += xfs_symlink_hdr_set(mp, ip->i_ino, offset,
+ byte_cnt, bp);
+
+ memcpy(buf, cur_chunk, byte_cnt);
+
+ cur_chunk += byte_cnt;
+ pathlen -= byte_cnt;
+ offset += byte_cnt;
+
+ xfs_trans_log_buf(tp, bp, 0, (buf + byte_cnt - 1) -
+ (char *)bp->b_addr);
+ }
+ }
+
+ /*
+ * Create the directory entry for the symlink.
+ */
+ error = xfs_dir_createname(tp, dp, link_name, ip->i_ino,
+ &first_block, &free_list, resblks);
+ if (error)
+ goto error2;
+ xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
+ xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
+
+ /*
+ * If this is a synchronous mount, make sure that the
+ * symlink transaction goes to disk before returning to
+ * the user.
+ */
+ if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
+ xfs_trans_set_sync(tp);
+ }
+
+ error = xfs_bmap_finish(&tp, &free_list, &committed);
+ if (error) {
+ goto error2;
+ }
+ error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
+ xfs_qm_dqrele(udqp);
+ xfs_qm_dqrele(gdqp);
+
+ *ipp = ip;
+ return 0;
+
+ error2:
+ IRELE(ip);
+ error1:
+ xfs_bmap_cancel(&free_list);
+ cancel_flags |= XFS_TRANS_ABORT;
+ error_return:
+ xfs_trans_cancel(tp, cancel_flags);
+ xfs_qm_dqrele(udqp);
+ xfs_qm_dqrele(gdqp);
+
+ if (unlock_dp_on_error)
+ xfs_iunlock(dp, XFS_ILOCK_EXCL);
+ std_return:
+ return error;
+}
+
+/*
+ * Free a symlink that has blocks associated with it.
+ */
+int
+xfs_inactive_symlink_rmt(
+ xfs_inode_t *ip,
+ xfs_trans_t **tpp)
+{
+ xfs_buf_t *bp;
+ int committed;
+ int done;
+ int error;
+ xfs_fsblock_t first_block;
+ xfs_bmap_free_t free_list;
+ int i;
+ xfs_mount_t *mp;
+ xfs_bmbt_irec_t mval[XFS_SYMLINK_MAPS];
+ int nmaps;
+ xfs_trans_t *ntp;
+ int size;
+ xfs_trans_t *tp;
+
+ tp = *tpp;
+ mp = ip->i_mount;
+ ASSERT(ip->i_d.di_size > XFS_IFORK_DSIZE(ip));
+ /*
+ * We're freeing a symlink that has some
+ * blocks allocated to it. Free the
+ * blocks here. We know that we've got
+ * either 1 or 2 extents and that we can
+ * free them all in one bunmapi call.
+ */
+ ASSERT(ip->i_d.di_nextents > 0 && ip->i_d.di_nextents <= 2);
+
+ /*
+ * Lock the inode, fix the size, and join it to the transaction.
+ * Hold it so in the normal path, we still have it locked for
+ * the second transaction. In the error paths we need it
+ * held so the cancel won't rele it, see below.
+ */
+ size = (int)ip->i_d.di_size;
+ ip->i_d.di_size = 0;
+ xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+ /*
+ * Find the block(s) so we can inval and unmap them.
+ */
+ done = 0;
+ xfs_bmap_init(&free_list, &first_block);
+ nmaps = ARRAY_SIZE(mval);
+ error = xfs_bmapi_read(ip, 0, xfs_symlink_blocks(mp, size),
+ mval, &nmaps, 0);
+ if (error)
+ goto error0;
+ /*
+ * Invalidate the block(s). No validation is done.
+ */
+ for (i = 0; i < nmaps; i++) {
+ bp = xfs_trans_get_buf(tp, mp->m_ddev_targp,
+ XFS_FSB_TO_DADDR(mp, mval[i].br_startblock),
+ XFS_FSB_TO_BB(mp, mval[i].br_blockcount), 0);
+ if (!bp) {
+ error = ENOMEM;
+ goto error1;
+ }
+ xfs_trans_binval(tp, bp);
+ }
+ /*
+ * Unmap the dead block(s) to the free_list.
+ */
+ if ((error = xfs_bunmapi(tp, ip, 0, size, XFS_BMAPI_METADATA, nmaps,
+ &first_block, &free_list, &done)))
+ goto error1;
+ ASSERT(done);
+ /*
+ * Commit the first transaction. This logs the EFI and the inode.
+ */
+ if ((error = xfs_bmap_finish(&tp, &free_list, &committed)))
+ goto error1;
+ /*
+ * The transaction must have been committed, since there were
+ * actually extents freed by xfs_bunmapi. See xfs_bmap_finish.
+ * The new tp has the extent freeing and EFDs.
+ */
+ ASSERT(committed);
+ /*
+ * The first xact was committed, so add the inode to the new one.
+ * Mark it dirty so it will be logged and moved forward in the log as
+ * part of every commit.
+ */
+ xfs_trans_ijoin(tp, ip, 0);
+ xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
+ /*
+ * Get a new, empty transaction to return to our caller.
+ */
+ ntp = xfs_trans_dup(tp);
+ /*
+ * Commit the transaction containing extent freeing and EFDs.
+ * If we get an error on the commit here or on the reserve below,
+ * we need to unlock the inode since the new transaction doesn't
+ * have the inode attached.
+ */
+ error = xfs_trans_commit(tp, 0);
+ tp = ntp;
+ if (error) {
+ ASSERT(XFS_FORCED_SHUTDOWN(mp));
+ goto error0;
+ }
+ /*
+ * transaction commit worked ok so we can drop the extra ticket
+ * reference that we gained in xfs_trans_dup()
+ */
+ xfs_log_ticket_put(tp->t_ticket);
+
+ /*
+ * Remove the memory for extent descriptions (just bookkeeping).
+ */
+ if (ip->i_df.if_bytes)
+ xfs_idata_realloc(ip, -ip->i_df.if_bytes, XFS_DATA_FORK);
+ ASSERT(ip->i_df.if_bytes == 0);
+ /*
+ * Put an itruncate log reservation in the new transaction
+ * for our caller.
+ */
+ if ((error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
+ XFS_TRANS_PERM_LOG_RES, XFS_ITRUNCATE_LOG_COUNT))) {
+ ASSERT(XFS_FORCED_SHUTDOWN(mp));
+ goto error0;
+ }
+
+ xfs_trans_ijoin(tp, ip, 0);
+ *tpp = tp;
+ return 0;
+
+ error1:
+ xfs_bmap_cancel(&free_list);
+ error0:
+ return error;
+}
diff --git a/fs/xfs/xfs_symlink.h b/fs/xfs/xfs_symlink.h
new file mode 100644
index 0000000..be20ae1
--- /dev/null
+++ b/fs/xfs/xfs_symlink.h
@@ -0,0 +1,53 @@
+/*
+ * Copyright (c) 2012 Red Hat, Inc. All rights reserved.
+ */
+#ifndef __XFS_SYMLINK_H
+#define __XFS_SYMLINK_H 1
+
+struct xfs_mount;
+struct xfs_trans;
+struct xfs_inode;
+struct xfs_buf;
+struct xfs_ifork;
+struct xfs_name;
+
+#define XFS_SYMLINK_MAGIC 0x58534c4d /* XSLM */
+
+struct xfs_dsymlink_hdr {
+ __be32 sl_magic;
+ __be32 sl_offset;
+ __be32 sl_bytes;
+ __be32 sl_crc;
+ uuid_t sl_uuid;
+ __be64 sl_owner;
+ __be64 sl_blkno;
+ __be64 sl_lsn;
+};
+
+/*
+ * The maximum pathlen is 1024 bytes. Since the minimum file system
+ * blocksize is 512 bytes, we can get a max of 3 extents back from
+ * bmapi when crc headers are taken into account.
+ */
+#define XFS_SYMLINK_MAPS 3
+
+#define XFS_SYMLINK_BUF_SPACE(mp, bufsize) \
+ ((bufsize) - (xfs_sb_version_hascrc(&(mp)->m_sb) ? \
+ sizeof(struct xfs_dsymlink_hdr) : 0))
+
+int xfs_symlink_blocks(struct xfs_mount *mp, int pathlen);
+
+void xfs_symlink_local_to_remote(struct xfs_trans *tp, struct xfs_buf *bp,
+ struct xfs_inode *ip, struct xfs_ifork *ifp);
+
+extern const struct xfs_buf_ops xfs_symlink_buf_ops;
+
+#ifdef __KERNEL__
+
+int xfs_symlink(struct xfs_inode *dp, struct xfs_name *link_name,
+ const char *target_path, umode_t mode, struct xfs_inode **ipp);
+int xfs_readlink(struct xfs_inode *ip, char *link);
+int xfs_inactive_symlink_rmt(struct xfs_inode *ip, struct xfs_trans **tpp);
+
+#endif /* __KERNEL__ */
+#endif /* __XFS_SYMLINK_H */
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index aa0c066..1034440 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -1,5 +1,6 @@
/*
* Copyright (c) 2000-2006 Silicon Graphics, Inc.
+ * Copyright (c) 2012 Red Hat, Inc.
* All Rights Reserved.
*
* This program is free software; you can redistribute it and/or
@@ -48,103 +49,10 @@
#include "xfs_vnodeops.h"
#include "xfs_trace.h"
#include "xfs_icache.h"
+#include "xfs_symlink.h"
+#include "xfs_cksum.h"
+#include "xfs_buf_item.h"
-/*
- * The maximum pathlen is 1024 bytes. Since the minimum file system
- * blocksize is 512 bytes, we can get a max of 2 extents back from
- * bmapi.
- */
-#define SYMLINK_MAPS 2
-
-STATIC int
-xfs_readlink_bmap(
- xfs_inode_t *ip,
- char *link)
-{
- xfs_mount_t *mp = ip->i_mount;
- int pathlen = ip->i_d.di_size;
- int nmaps = SYMLINK_MAPS;
- xfs_bmbt_irec_t mval[SYMLINK_MAPS];
- xfs_daddr_t d;
- int byte_cnt;
- int n;
- xfs_buf_t *bp;
- int error = 0;
-
- error = xfs_bmapi_read(ip, 0, XFS_B_TO_FSB(mp, pathlen), mval, &nmaps,
- 0);
- if (error)
- goto out;
-
- for (n = 0; n < nmaps; n++) {
- d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
- byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
-
- bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), 0, NULL);
- if (!bp)
- return XFS_ERROR(ENOMEM);
- error = bp->b_error;
- if (error) {
- xfs_buf_ioerror_alert(bp, __func__);
- xfs_buf_relse(bp);
- goto out;
- }
- if (pathlen < byte_cnt)
- byte_cnt = pathlen;
- pathlen -= byte_cnt;
-
- memcpy(link, bp->b_addr, byte_cnt);
- xfs_buf_relse(bp);
- }
-
- link[ip->i_d.di_size] = '\0';
- error = 0;
-
- out:
- return error;
-}
-
-int
-xfs_readlink(
- xfs_inode_t *ip,
- char *link)
-{
- xfs_mount_t *mp = ip->i_mount;
- xfs_fsize_t pathlen;
- int error = 0;
-
- trace_xfs_readlink(ip);
-
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(EIO);
-
- xfs_ilock(ip, XFS_ILOCK_SHARED);
-
- pathlen = ip->i_d.di_size;
- if (!pathlen)
- goto out;
-
- if (pathlen < 0 || pathlen > MAXPATHLEN) {
- xfs_alert(mp, "%s: inode (%llu) bad symlink length (%lld)",
- __func__, (unsigned long long) ip->i_ino,
- (long long) pathlen);
- ASSERT(0);
- error = XFS_ERROR(EFSCORRUPTED);
- goto out;
- }
-
-
- if (ip->i_df.if_flags & XFS_IFINLINE) {
- memcpy(link, ip->i_df.if_u1.if_data, pathlen);
- link[pathlen] = '\0';
- } else {
- error = xfs_readlink_bmap(ip, link);
- }
-
- out:
- xfs_iunlock(ip, XFS_ILOCK_SHARED);
- return error;
-}
/*
* This is called by xfs_inactive to free any blocks beyond eof
@@ -249,145 +157,6 @@ xfs_free_eofblocks(
return error;
}
-/*
- * Free a symlink that has blocks associated with it.
- */
-STATIC int
-xfs_inactive_symlink_rmt(
- xfs_inode_t *ip,
- xfs_trans_t **tpp)
-{
- xfs_buf_t *bp;
- int committed;
- int done;
- int error;
- xfs_fsblock_t first_block;
- xfs_bmap_free_t free_list;
- int i;
- xfs_mount_t *mp;
- xfs_bmbt_irec_t mval[SYMLINK_MAPS];
- int nmaps;
- xfs_trans_t *ntp;
- int size;
- xfs_trans_t *tp;
-
- tp = *tpp;
- mp = ip->i_mount;
- ASSERT(ip->i_d.di_size > XFS_IFORK_DSIZE(ip));
- /*
- * We're freeing a symlink that has some
- * blocks allocated to it. Free the
- * blocks here. We know that we've got
- * either 1 or 2 extents and that we can
- * free them all in one bunmapi call.
- */
- ASSERT(ip->i_d.di_nextents > 0 && ip->i_d.di_nextents <= 2);
-
- /*
- * Lock the inode, fix the size, and join it to the transaction.
- * Hold it so in the normal path, we still have it locked for
- * the second transaction. In the error paths we need it
- * held so the cancel won't rele it, see below.
- */
- size = (int)ip->i_d.di_size;
- ip->i_d.di_size = 0;
- xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
- /*
- * Find the block(s) so we can inval and unmap them.
- */
- done = 0;
- xfs_bmap_init(&free_list, &first_block);
- nmaps = ARRAY_SIZE(mval);
- error = xfs_bmapi_read(ip, 0, XFS_B_TO_FSB(mp, size),
- mval, &nmaps, 0);
- if (error)
- goto error0;
- /*
- * Invalidate the block(s).
- */
- for (i = 0; i < nmaps; i++) {
- bp = xfs_trans_get_buf(tp, mp->m_ddev_targp,
- XFS_FSB_TO_DADDR(mp, mval[i].br_startblock),
- XFS_FSB_TO_BB(mp, mval[i].br_blockcount), 0);
- if (!bp) {
- error = ENOMEM;
- goto error1;
- }
- xfs_trans_binval(tp, bp);
- }
- /*
- * Unmap the dead block(s) to the free_list.
- */
- if ((error = xfs_bunmapi(tp, ip, 0, size, XFS_BMAPI_METADATA, nmaps,
- &first_block, &free_list, &done)))
- goto error1;
- ASSERT(done);
- /*
- * Commit the first transaction. This logs the EFI and the inode.
- */
- if ((error = xfs_bmap_finish(&tp, &free_list, &committed)))
- goto error1;
- /*
- * The transaction must have been committed, since there were
- * actually extents freed by xfs_bunmapi. See xfs_bmap_finish.
- * The new tp has the extent freeing and EFDs.
- */
- ASSERT(committed);
- /*
- * The first xact was committed, so add the inode to the new one.
- * Mark it dirty so it will be logged and moved forward in the log as
- * part of every commit.
- */
- xfs_trans_ijoin(tp, ip, 0);
- xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
- /*
- * Get a new, empty transaction to return to our caller.
- */
- ntp = xfs_trans_dup(tp);
- /*
- * Commit the transaction containing extent freeing and EFDs.
- * If we get an error on the commit here or on the reserve below,
- * we need to unlock the inode since the new transaction doesn't
- * have the inode attached.
- */
- error = xfs_trans_commit(tp, 0);
- tp = ntp;
- if (error) {
- ASSERT(XFS_FORCED_SHUTDOWN(mp));
- goto error0;
- }
- /*
- * transaction commit worked ok so we can drop the extra ticket
- * reference that we gained in xfs_trans_dup()
- */
- xfs_log_ticket_put(tp->t_ticket);
-
- /*
- * Remove the memory for extent descriptions (just bookkeeping).
- */
- if (ip->i_df.if_bytes)
- xfs_idata_realloc(ip, -ip->i_df.if_bytes, XFS_DATA_FORK);
- ASSERT(ip->i_df.if_bytes == 0);
- /*
- * Put an itruncate log reservation in the new transaction
- * for our caller.
- */
- if ((error = xfs_trans_reserve(tp, 0, XFS_ITRUNCATE_LOG_RES(mp), 0,
- XFS_TRANS_PERM_LOG_RES, XFS_ITRUNCATE_LOG_COUNT))) {
- ASSERT(XFS_FORCED_SHUTDOWN(mp));
- goto error0;
- }
-
- xfs_trans_ijoin(tp, ip, 0);
- *tpp = tp;
- return 0;
-
- error1:
- xfs_bmap_cancel(&free_list);
- error0:
- return error;
-}
-
int
xfs_release(
xfs_inode_t *ip)
@@ -1353,247 +1122,6 @@ xfs_link(
}
int
-xfs_symlink(
- xfs_inode_t *dp,
- struct xfs_name *link_name,
- const char *target_path,
- umode_t mode,
- xfs_inode_t **ipp)
-{
- xfs_mount_t *mp = dp->i_mount;
- xfs_trans_t *tp;
- xfs_inode_t *ip;
- int error;
- int pathlen;
- xfs_bmap_free_t free_list;
- xfs_fsblock_t first_block;
- bool unlock_dp_on_error = false;
- uint cancel_flags;
- int committed;
- xfs_fileoff_t first_fsb;
- xfs_filblks_t fs_blocks;
- int nmaps;
- xfs_bmbt_irec_t mval[SYMLINK_MAPS];
- xfs_daddr_t d;
- const char *cur_chunk;
- int byte_cnt;
- int n;
- xfs_buf_t *bp;
- prid_t prid;
- struct xfs_dquot *udqp, *gdqp;
- uint resblks;
-
- *ipp = NULL;
- error = 0;
- ip = NULL;
- tp = NULL;
-
- trace_xfs_symlink(dp, link_name);
-
- if (XFS_FORCED_SHUTDOWN(mp))
- return XFS_ERROR(EIO);
-
- /*
- * Check component lengths of the target path name.
- */
- pathlen = strlen(target_path);
- if (pathlen >= MAXPATHLEN) /* total string too long */
- return XFS_ERROR(ENAMETOOLONG);
-
- udqp = gdqp = NULL;
- if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
- prid = xfs_get_projid(dp);
- else
- prid = XFS_PROJID_DEFAULT;
-
- /*
- * Make sure that we have allocated dquot(s) on disk.
- */
- error = xfs_qm_vop_dqalloc(dp, current_fsuid(), current_fsgid(), prid,
- XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT, &udqp, &gdqp);
- if (error)
- goto std_return;
-
- tp = xfs_trans_alloc(mp, XFS_TRANS_SYMLINK);
- cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
- /*
- * The symlink will fit into the inode data fork?
- * There can't be any attributes so we get the whole variable part.
- */
- if (pathlen <= XFS_LITINO(mp, dp->i_d.di_version))
- fs_blocks = 0;
- else
- fs_blocks = XFS_B_TO_FSB(mp, pathlen);
- resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
- error = xfs_trans_reserve(tp, resblks, XFS_SYMLINK_LOG_RES(mp), 0,
- XFS_TRANS_PERM_LOG_RES, XFS_SYMLINK_LOG_COUNT);
- if (error == ENOSPC && fs_blocks == 0) {
- resblks = 0;
- error = xfs_trans_reserve(tp, 0, XFS_SYMLINK_LOG_RES(mp), 0,
- XFS_TRANS_PERM_LOG_RES, XFS_SYMLINK_LOG_COUNT);
- }
- if (error) {
- cancel_flags = 0;
- goto error_return;
- }
-
- xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
- unlock_dp_on_error = true;
-
- /*
- * Check whether the directory allows new symlinks or not.
- */
- if (dp->i_d.di_flags & XFS_DIFLAG_NOSYMLINKS) {
- error = XFS_ERROR(EPERM);
- goto error_return;
- }
-
- /*
- * Reserve disk quota : blocks and inode.
- */
- error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp, resblks, 1, 0);
- if (error)
- goto error_return;
-
- /*
- * Check for ability to enter directory entry, if no space reserved.
- */
- error = xfs_dir_canenter(tp, dp, link_name, resblks);
- if (error)
- goto error_return;
- /*
- * Initialize the bmap freelist prior to calling either
- * bmapi or the directory create code.
- */
- xfs_bmap_init(&free_list, &first_block);
-
- /*
- * Allocate an inode for the symlink.
- */
- error = xfs_dir_ialloc(&tp, dp, S_IFLNK | (mode & ~S_IFMT), 1, 0,
- prid, resblks > 0, &ip, NULL);
- if (error) {
- if (error == ENOSPC)
- goto error_return;
- goto error1;
- }
-
- /*
- * An error after we've joined dp to the transaction will result in the
- * transaction cancel unlocking dp so don't do it explicitly in the
- * error path.
- */
- xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
- unlock_dp_on_error = false;
-
- /*
- * Also attach the dquot(s) to it, if applicable.
- */
- xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp);
-
- if (resblks)
- resblks -= XFS_IALLOC_SPACE_RES(mp);
- /*
- * If the symlink will fit into the inode, write it inline.
- */
- if (pathlen <= XFS_IFORK_DSIZE(ip)) {
- xfs_idata_realloc(ip, pathlen, XFS_DATA_FORK);
- memcpy(ip->i_df.if_u1.if_data, target_path, pathlen);
- ip->i_d.di_size = pathlen;
-
- /*
- * The inode was initially created in extent format.
- */
- ip->i_df.if_flags &= ~(XFS_IFEXTENTS | XFS_IFBROOT);
- ip->i_df.if_flags |= XFS_IFINLINE;
-
- ip->i_d.di_format = XFS_DINODE_FMT_LOCAL;
- xfs_trans_log_inode(tp, ip, XFS_ILOG_DDATA | XFS_ILOG_CORE);
-
- } else {
- first_fsb = 0;
- nmaps = SYMLINK_MAPS;
-
- error = xfs_bmapi_write(tp, ip, first_fsb, fs_blocks,
- XFS_BMAPI_METADATA, &first_block, resblks,
- mval, &nmaps, &free_list);
- if (error)
- goto error2;
-
- if (resblks)
- resblks -= fs_blocks;
- ip->i_d.di_size = pathlen;
- xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
-
- cur_chunk = target_path;
- for (n = 0; n < nmaps; n++) {
- d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
- byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
- bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
- BTOBB(byte_cnt), 0);
- if (!bp) {
- error = ENOMEM;
- goto error2;
- }
- if (pathlen < byte_cnt) {
- byte_cnt = pathlen;
- }
- pathlen -= byte_cnt;
-
- memcpy(bp->b_addr, cur_chunk, byte_cnt);
- cur_chunk += byte_cnt;
-
- xfs_trans_log_buf(tp, bp, 0, byte_cnt - 1);
- }
- }
-
- /*
- * Create the directory entry for the symlink.
- */
- error = xfs_dir_createname(tp, dp, link_name, ip->i_ino,
- &first_block, &free_list, resblks);
- if (error)
- goto error2;
- xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
- xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
-
- /*
- * If this is a synchronous mount, make sure that the
- * symlink transaction goes to disk before returning to
- * the user.
- */
- if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
- xfs_trans_set_sync(tp);
- }
-
- error = xfs_bmap_finish(&tp, &free_list, &committed);
- if (error) {
- goto error2;
- }
- error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
- xfs_qm_dqrele(udqp);
- xfs_qm_dqrele(gdqp);
-
- *ipp = ip;
- return 0;
-
- error2:
- IRELE(ip);
- error1:
- xfs_bmap_cancel(&free_list);
- cancel_flags |= XFS_TRANS_ABORT;
- error_return:
- xfs_trans_cancel(tp, cancel_flags);
- xfs_qm_dqrele(udqp);
- xfs_qm_dqrele(gdqp);
-
- if (unlock_dp_on_error)
- xfs_iunlock(dp, XFS_ILOCK_EXCL);
- std_return:
- return error;
-}
-
-int
xfs_set_dmattrs(
xfs_inode_t *ip,
u_int evmask,
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread
* [PATCH 10/16] xfs: add CRC checks to block format directory blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (8 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 09/16] xfs: add CRC checks to remote symlinks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 11/16] xfs: add CRC checking to dir2 free blocks Dave Chinner
` (5 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
Now that directory buffers are made from a single struct xfs_buf, we
can add CRC calculation and checking callbacks. While there, add all
the fields to the on disk structures for future functionality such
as d_type support, uuids, block numbers, owner inode, etc.
To distinguish between the different on disk formats, change the
magic numbers for the new format directory blocks.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_dir2_block.c | 124 +++++++++++++++++++++++++-----------
fs/xfs/xfs_dir2_data.c | 157 ++++++++++++++++++++++++++++------------------
fs/xfs/xfs_dir2_format.h | 134 +++++++++++++++++++++++++++++++++++++++
fs/xfs/xfs_dir2_leaf.c | 6 +-
fs/xfs/xfs_dir2_node.c | 2 +-
fs/xfs/xfs_dir2_priv.h | 4 +-
fs/xfs/xfs_log_recover.c | 2 +
7 files changed, 327 insertions(+), 102 deletions(-)
diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c
index 12afe07..f2b34b2 100644
--- a/fs/xfs/xfs_dir2_block.c
+++ b/fs/xfs/xfs_dir2_block.c
@@ -28,11 +28,13 @@
#include "xfs_dinode.h"
#include "xfs_inode.h"
#include "xfs_inode_item.h"
+#include "xfs_buf_item.h"
#include "xfs_dir2.h"
#include "xfs_dir2_format.h"
#include "xfs_dir2_priv.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_cksum.h"
/*
* Local function prototypes.
@@ -56,44 +58,74 @@ xfs_dir_startup(void)
xfs_dir_hash_dotdot = xfs_da_hashname((unsigned char *)"..", 2);
}
-static void
-xfs_dir2_block_verify(
+static bool
+xfs_dir3_block_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
- struct xfs_dir2_data_hdr *hdr = bp->b_addr;
- int block_ok = 0;
-
- block_ok = hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
- block_ok = block_ok && __xfs_dir2_data_check(NULL, bp) == 0;
-
- if (!block_ok) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ if (hdr3->magic != be32_to_cpu(XFS_DIR3_BLOCK_MAGIC))
+ return false;
+ if (!uuid_equal(&hdr3->uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be64_to_cpu(hdr3->blkno) != bp->b_bn)
+ return false;
+ } else {
+ if (hdr3->magic != be32_to_cpu(XFS_DIR2_BLOCK_MAGIC))
+ return false;
}
+ if (__xfs_dir2_data_check(NULL, bp))
+ return false;
+ return true;
}
static void
-xfs_dir2_block_read_verify(
+xfs_dir3_block_read_verify(
struct xfs_buf *bp)
{
- xfs_dir2_block_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ if ((xfs_sb_version_hascrc(&mp->m_sb) &&
+ !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_DIR3_DATA_CRC_OFF)) ||
+ !xfs_dir3_block_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
static void
-xfs_dir2_block_write_verify(
+xfs_dir3_block_write_verify(
struct xfs_buf *bp)
{
- xfs_dir2_block_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
+
+ if (!xfs_dir3_block_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ hdr3->lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_DATA_CRC_OFF);
}
-const struct xfs_buf_ops xfs_dir2_block_buf_ops = {
- .verify_read = xfs_dir2_block_read_verify,
- .verify_write = xfs_dir2_block_write_verify,
+const struct xfs_buf_ops xfs_dir3_block_buf_ops = {
+ .verify_read = xfs_dir3_block_read_verify,
+ .verify_write = xfs_dir3_block_write_verify,
};
static int
-xfs_dir2_block_read(
+xfs_dir3_block_read(
struct xfs_trans *tp,
struct xfs_inode *dp,
struct xfs_buf **bpp)
@@ -101,7 +133,29 @@ xfs_dir2_block_read(
struct xfs_mount *mp = dp->i_mount;
return xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, bpp,
- XFS_DATA_FORK, &xfs_dir2_block_buf_ops);
+ XFS_DATA_FORK, &xfs_dir3_block_buf_ops);
+}
+
+static void
+xfs_dir3_block_init(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp,
+ struct xfs_inode *dp)
+{
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
+
+ bp->b_ops = &xfs_dir3_block_buf_ops;
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ memset(hdr3, 0, sizeof(*hdr3));
+ hdr3->magic = cpu_to_be32(XFS_DIR3_BLOCK_MAGIC);
+ hdr3->blkno = cpu_to_be64(bp->b_bn);
+ hdr3->owner = cpu_to_be64(dp->i_ino);
+ uuid_copy(&hdr3->uuid, &mp->m_sb.sb_uuid);
+ return;
+
+ }
+ hdr3->magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
}
static void
@@ -121,7 +175,7 @@ xfs_dir2_block_need_space(
struct xfs_dir2_data_unused *enddup = NULL;
*compact = 0;
- bf = hdr->bestfree;
+ bf = xfs_dir3_data_bestfree_p(hdr);
/*
* If there are stale entries we'll use one for the leaf.
@@ -303,7 +357,7 @@ xfs_dir2_block_addname(
mp = dp->i_mount;
/* Read the (one and only) directory block into bp. */
- error = xfs_dir2_block_read(tp, dp, &bp);
+ error = xfs_dir3_block_read(tp, dp, &bp);
if (error)
return error;
@@ -531,7 +585,7 @@ xfs_dir2_block_getdents(
if (xfs_dir2_dataptr_to_db(mp, *offset) > mp->m_dirdatablk)
return 0;
- error = xfs_dir2_block_read(NULL, dp, &bp);
+ error = xfs_dir3_block_read(NULL, dp, &bp);
if (error)
return error;
@@ -711,7 +765,7 @@ xfs_dir2_block_lookup_int(
tp = args->trans;
mp = dp->i_mount;
- error = xfs_dir2_block_read(tp, dp, &bp);
+ error = xfs_dir3_block_read(tp, dp, &bp);
if (error)
return error;
@@ -1014,8 +1068,8 @@ xfs_dir2_leaf_to_block(
/*
* Start converting it to block form.
*/
- dbp->b_ops = &xfs_dir2_block_buf_ops;
- hdr->magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
+ xfs_dir3_block_init(mp, dbp, dp);
+
needlog = 1;
needscan = 0;
/*
@@ -1137,16 +1191,16 @@ xfs_dir2_sf_to_block(
return error;
}
/*
- * Initialize the data block.
+ * Initialize the data block, then convert it to block format.
*/
- error = xfs_dir2_data_init(args, blkno, &bp);
+ error = xfs_dir3_data_init(args, blkno, &bp);
if (error) {
kmem_free(sfp);
return error;
}
- bp->b_ops = &xfs_dir2_block_buf_ops;
+ xfs_dir3_block_init(mp, bp, dp);
hdr = bp->b_addr;
- hdr->magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC);
+
/*
* Compute size of block "tail" area.
*/
@@ -1156,7 +1210,7 @@ xfs_dir2_sf_to_block(
* The whole thing is initialized to free by the init routine.
* Say we're using the leaf and tail area.
*/
- dup = (xfs_dir2_data_unused_t *)(hdr + 1);
+ dup = xfs_dir3_data_unused_p(hdr);
needlog = needscan = 0;
xfs_dir2_data_use_free(tp, bp, dup, mp->m_dirblksize - i, i, &needlog,
&needscan);
@@ -1178,8 +1232,7 @@ xfs_dir2_sf_to_block(
/*
* Create entry for .
*/
- dep = (xfs_dir2_data_entry_t *)
- ((char *)hdr + XFS_DIR2_DATA_DOT_OFFSET);
+ dep = xfs_dir3_data_dot_entry_p(hdr);
dep->inumber = cpu_to_be64(dp->i_ino);
dep->namelen = 1;
dep->name[0] = '.';
@@ -1192,8 +1245,7 @@ xfs_dir2_sf_to_block(
/*
* Create entry for ..
*/
- dep = (xfs_dir2_data_entry_t *)
- ((char *)hdr + XFS_DIR2_DATA_DOTDOT_OFFSET);
+ dep = xfs_dir3_data_dotdot_entry_p(hdr);
dep->inumber = cpu_to_be64(xfs_dir2_sf_get_parent_ino(sfp));
dep->namelen = 2;
dep->name[0] = dep->name[1] = '.';
@@ -1203,7 +1255,7 @@ xfs_dir2_sf_to_block(
blp[1].hashval = cpu_to_be32(xfs_dir_hash_dotdot);
blp[1].address = cpu_to_be32(xfs_dir2_byte_to_dataptr(mp,
(char *)dep - (char *)hdr));
- offset = XFS_DIR2_DATA_FIRST_OFFSET;
+ offset = xfs_dir3_data_first_offset(hdr);
/*
* Loop over existing entries, stuff them in.
*/
diff --git a/fs/xfs/xfs_dir2_data.c b/fs/xfs/xfs_dir2_data.c
index ffcf177..93bc123 100644
--- a/fs/xfs/xfs_dir2_data.c
+++ b/fs/xfs/xfs_dir2_data.c
@@ -65,10 +65,11 @@ __xfs_dir2_data_check(
mp = bp->b_target->bt_mount;
hdr = bp->b_addr;
- bf = hdr->bestfree;
- p = (char *)(hdr + 1);
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ p = (char *)xfs_dir3_data_entry_p(hdr);
switch (hdr->magic) {
+ case cpu_to_be32(XFS_DIR3_BLOCK_MAGIC):
case cpu_to_be32(XFS_DIR2_BLOCK_MAGIC):
btp = xfs_dir2_block_tail_p(mp, hdr);
lep = xfs_dir2_block_leaf_p(btp);
@@ -148,7 +149,8 @@ __xfs_dir2_data_check(
(char *)dep - (char *)hdr);
count++;
lastfree = 0;
- if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) {
+ if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
addr = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk,
(xfs_dir2_data_aoff_t)
((char *)dep - (char *)hdr));
@@ -168,7 +170,8 @@ __xfs_dir2_data_check(
* Need to have seen all the entries and all the bestfree slots.
*/
XFS_WANT_CORRUPTED_RETURN(freeseen == 7);
- if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) {
+ if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
for (i = stale = 0; i < be32_to_cpu(btp->count); i++) {
if (lep[i].address ==
cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
@@ -216,7 +219,8 @@ xfs_dir2_data_reada_verify(
switch (hdr->magic) {
case cpu_to_be32(XFS_DIR2_BLOCK_MAGIC):
- bp->b_ops = &xfs_dir2_block_buf_ops;
+ case cpu_to_be32(XFS_DIR3_BLOCK_MAGIC):
+ bp->b_ops = &xfs_dir3_block_buf_ops;
bp->b_ops->verify_read(bp);
return;
case cpu_to_be32(XFS_DIR2_DATA_MAGIC):
@@ -288,12 +292,15 @@ xfs_dir2_data_freefind(
{
xfs_dir2_data_free_t *dfp; /* bestfree entry */
xfs_dir2_data_aoff_t off; /* offset value needed */
+ struct xfs_dir2_data_free *bf;
#if defined(DEBUG) && defined(__KERNEL__)
int matched; /* matched the value */
int seenzero; /* saw a 0 bestfree entry */
#endif
off = (xfs_dir2_data_aoff_t)((char *)dup - (char *)hdr);
+ bf = xfs_dir3_data_bestfree_p(hdr);
+
#if defined(DEBUG) && defined(__KERNEL__)
/*
* Validate some consistency in the bestfree table.
@@ -301,9 +308,10 @@ xfs_dir2_data_freefind(
* one we're looking for it has to be exact.
*/
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
- for (dfp = &hdr->bestfree[0], seenzero = matched = 0;
- dfp < &hdr->bestfree[XFS_DIR2_DATA_FD_COUNT];
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
+ for (dfp = &bf[0], seenzero = matched = 0;
+ dfp < &bf[XFS_DIR2_DATA_FD_COUNT];
dfp++) {
if (!dfp->offset) {
ASSERT(!dfp->length);
@@ -319,7 +327,7 @@ xfs_dir2_data_freefind(
else
ASSERT(be16_to_cpu(dfp->offset) + be16_to_cpu(dfp->length) <= off);
ASSERT(matched || be16_to_cpu(dfp->length) >= be16_to_cpu(dup->length));
- if (dfp > &hdr->bestfree[0])
+ if (dfp > &bf[0])
ASSERT(be16_to_cpu(dfp[-1].length) >= be16_to_cpu(dfp[0].length));
}
#endif
@@ -328,14 +336,12 @@ xfs_dir2_data_freefind(
* it can't be there since they're sorted.
*/
if (be16_to_cpu(dup->length) <
- be16_to_cpu(hdr->bestfree[XFS_DIR2_DATA_FD_COUNT - 1].length))
+ be16_to_cpu(bf[XFS_DIR2_DATA_FD_COUNT - 1].length))
return NULL;
/*
* Look at the three bestfree entries for our guy.
*/
- for (dfp = &hdr->bestfree[0];
- dfp < &hdr->bestfree[XFS_DIR2_DATA_FD_COUNT];
- dfp++) {
+ for (dfp = &bf[0]; dfp < &bf[XFS_DIR2_DATA_FD_COUNT]; dfp++) {
if (!dfp->offset)
return NULL;
if (be16_to_cpu(dfp->offset) == off)
@@ -359,11 +365,11 @@ xfs_dir2_data_freeinsert(
xfs_dir2_data_free_t *dfp; /* bestfree table pointer */
xfs_dir2_data_free_t new; /* new bestfree entry */
-#ifdef __KERNEL__
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
-#endif
- dfp = hdr->bestfree;
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
+
+ dfp = xfs_dir3_data_bestfree_p(hdr);
new.length = dup->length;
new.offset = cpu_to_be16((char *)dup - (char *)hdr);
@@ -400,32 +406,35 @@ xfs_dir2_data_freeremove(
xfs_dir2_data_free_t *dfp, /* bestfree entry pointer */
int *loghead) /* out: log data header */
{
-#ifdef __KERNEL__
+ struct xfs_dir2_data_free *bf;
+
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
-#endif
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
+
/*
* It's the first entry, slide the next 2 up.
*/
- if (dfp == &hdr->bestfree[0]) {
- hdr->bestfree[0] = hdr->bestfree[1];
- hdr->bestfree[1] = hdr->bestfree[2];
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ if (dfp == &bf[0]) {
+ bf[0] = bf[1];
+ bf[1] = bf[2];
}
/*
* It's the second entry, slide the 3rd entry up.
*/
- else if (dfp == &hdr->bestfree[1])
- hdr->bestfree[1] = hdr->bestfree[2];
+ else if (dfp == &bf[1])
+ bf[1] = bf[2];
/*
* Must be the last entry.
*/
else
- ASSERT(dfp == &hdr->bestfree[2]);
+ ASSERT(dfp == &bf[2]);
/*
* Clear the 3rd entry, must be zero now.
*/
- hdr->bestfree[2].length = 0;
- hdr->bestfree[2].offset = 0;
+ bf[2].length = 0;
+ bf[2].offset = 0;
*loghead = 1;
}
@@ -441,23 +450,26 @@ xfs_dir2_data_freescan(
xfs_dir2_block_tail_t *btp; /* block tail */
xfs_dir2_data_entry_t *dep; /* active data entry */
xfs_dir2_data_unused_t *dup; /* unused data entry */
+ struct xfs_dir2_data_free *bf;
char *endp; /* end of block's data */
char *p; /* current entry pointer */
-#ifdef __KERNEL__
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
-#endif
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
+
/*
* Start by clearing the table.
*/
- memset(hdr->bestfree, 0, sizeof(hdr->bestfree));
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ memset(bf, 0, sizeof(*bf) * XFS_DIR2_DATA_FD_COUNT);
*loghead = 1;
/*
* Set up pointers.
*/
- p = (char *)(hdr + 1);
- if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) {
+ p = (char *)xfs_dir3_data_entry_p(hdr);
+ if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
btp = xfs_dir2_block_tail_p(mp, hdr);
endp = (char *)xfs_dir2_block_leaf_p(btp);
} else
@@ -493,7 +505,7 @@ xfs_dir2_data_freescan(
* Give back the buffer for the created block.
*/
int /* error */
-xfs_dir2_data_init(
+xfs_dir3_data_init(
xfs_da_args_t *args, /* directory operation args */
xfs_dir2_db_t blkno, /* logical dir block number */
struct xfs_buf **bpp) /* output block buffer */
@@ -502,6 +514,7 @@ xfs_dir2_data_init(
xfs_dir2_data_hdr_t *hdr; /* data block header */
xfs_inode_t *dp; /* incore directory inode */
xfs_dir2_data_unused_t *dup; /* unused entry pointer */
+ struct xfs_dir2_data_free *bf;
int error; /* error return value */
int i; /* bestfree index */
xfs_mount_t *mp; /* filesystem mount point */
@@ -524,21 +537,34 @@ xfs_dir2_data_init(
* Initialize the header.
*/
hdr = bp->b_addr;
- hdr->magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
- hdr->bestfree[0].offset = cpu_to_be16(sizeof(*hdr));
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
+
+ memset(hdr3, 0, sizeof(*hdr3));
+ hdr3->magic = cpu_to_be32(XFS_DIR3_DATA_MAGIC);
+ hdr3->blkno = cpu_to_be64(bp->b_bn);
+ hdr3->owner = cpu_to_be64(dp->i_ino);
+ uuid_copy(&hdr3->uuid, &mp->m_sb.sb_uuid);
+
+ } else
+ hdr->magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
+
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ bf[0].offset = cpu_to_be16(xfs_dir3_data_entry_offset(hdr));
for (i = 1; i < XFS_DIR2_DATA_FD_COUNT; i++) {
- hdr->bestfree[i].length = 0;
- hdr->bestfree[i].offset = 0;
+ bf[i].length = 0;
+ bf[i].offset = 0;
}
/*
* Set up an unused entry for the block's body.
*/
- dup = (xfs_dir2_data_unused_t *)(hdr + 1);
+ dup = xfs_dir3_data_unused_p(hdr);
dup->freetag = cpu_to_be16(XFS_DIR2_DATA_FREE_TAG);
- t = mp->m_dirblksize - (uint)sizeof(*hdr);
- hdr->bestfree[0].length = cpu_to_be16(t);
+ t = mp->m_dirblksize - (uint)xfs_dir3_data_entry_offset(hdr);
+ bf[0].length = cpu_to_be16(t);
dup->length = cpu_to_be16(t);
*xfs_dir2_data_unused_tag_p(dup) = cpu_to_be16((char *)dup - (char *)hdr);
/*
@@ -562,7 +588,8 @@ xfs_dir2_data_log_entry(
xfs_dir2_data_hdr_t *hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
xfs_trans_log_buf(tp, bp, (uint)((char *)dep - (char *)hdr),
(uint)((char *)(xfs_dir2_data_entry_tag_p(dep) + 1) -
@@ -580,9 +607,10 @@ xfs_dir2_data_log_header(
xfs_dir2_data_hdr_t *hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
- xfs_trans_log_buf(tp, bp, 0, sizeof(*hdr) - 1);
+ xfs_trans_log_buf(tp, bp, 0, xfs_dir3_data_entry_offset(hdr) - 1);
}
/*
@@ -597,7 +625,8 @@ xfs_dir2_data_log_unused(
xfs_dir2_data_hdr_t *hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
/*
* Log the first part of the unused entry.
@@ -635,6 +664,7 @@ xfs_dir2_data_make_free(
xfs_dir2_data_unused_t *newdup; /* new unused entry */
xfs_dir2_data_unused_t *postdup; /* unused entry after us */
xfs_dir2_data_unused_t *prevdup; /* unused entry before us */
+ struct xfs_dir2_data_free *bf;
mp = tp->t_mountp;
hdr = bp->b_addr;
@@ -647,7 +677,8 @@ xfs_dir2_data_make_free(
else {
xfs_dir2_block_tail_t *btp; /* block tail */
- ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+ ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
btp = xfs_dir2_block_tail_p(mp, hdr);
endptr = (char *)xfs_dir2_block_leaf_p(btp);
}
@@ -655,7 +686,7 @@ xfs_dir2_data_make_free(
* If this isn't the start of the block, then back up to
* the previous entry and see if it's free.
*/
- if (offset > sizeof(*hdr)) {
+ if (offset > xfs_dir3_data_entry_offset(hdr)) {
__be16 *tagp; /* tag just before us */
tagp = (__be16 *)((char *)hdr + offset) - 1;
@@ -681,6 +712,7 @@ xfs_dir2_data_make_free(
* Previous and following entries are both free,
* merge everything into a single free entry.
*/
+ bf = xfs_dir3_data_bestfree_p(hdr);
if (prevdup && postdup) {
xfs_dir2_data_free_t *dfp2; /* another bestfree pointer */
@@ -695,7 +727,7 @@ xfs_dir2_data_make_free(
* since the third bestfree is there, there might be more
* entries.
*/
- needscan = (hdr->bestfree[2].length != 0);
+ needscan = (bf[2].length != 0);
/*
* Fix up the new big freespace.
*/
@@ -711,10 +743,10 @@ xfs_dir2_data_make_free(
* Remove entry 1 first then entry 0.
*/
ASSERT(dfp && dfp2);
- if (dfp == &hdr->bestfree[1]) {
- dfp = &hdr->bestfree[0];
+ if (dfp == &bf[1]) {
+ dfp = &bf[0];
ASSERT(dfp2 == dfp);
- dfp2 = &hdr->bestfree[1];
+ dfp2 = &bf[1];
}
xfs_dir2_data_freeremove(hdr, dfp2, needlogp);
xfs_dir2_data_freeremove(hdr, dfp, needlogp);
@@ -722,7 +754,7 @@ xfs_dir2_data_make_free(
* Now insert the new entry.
*/
dfp = xfs_dir2_data_freeinsert(hdr, prevdup, needlogp);
- ASSERT(dfp == &hdr->bestfree[0]);
+ ASSERT(dfp == &bf[0]);
ASSERT(dfp->length == prevdup->length);
ASSERT(!dfp[1].length);
ASSERT(!dfp[2].length);
@@ -751,7 +783,7 @@ xfs_dir2_data_make_free(
*/
else {
needscan = be16_to_cpu(prevdup->length) >
- be16_to_cpu(hdr->bestfree[2].length);
+ be16_to_cpu(bf[2].length);
}
}
/*
@@ -779,7 +811,7 @@ xfs_dir2_data_make_free(
*/
else {
needscan = be16_to_cpu(newdup->length) >
- be16_to_cpu(hdr->bestfree[2].length);
+ be16_to_cpu(bf[2].length);
}
}
/*
@@ -818,10 +850,12 @@ xfs_dir2_data_use_free(
xfs_dir2_data_unused_t *newdup; /* new unused entry */
xfs_dir2_data_unused_t *newdup2; /* another new unused entry */
int oldlen; /* old unused entry's length */
+ struct xfs_dir2_data_free *bf;
hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
- hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC));
+ hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
ASSERT(be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG);
ASSERT(offset >= (char *)dup - (char *)hdr);
ASSERT(offset + len <= (char *)dup + be16_to_cpu(dup->length) - (char *)hdr);
@@ -831,7 +865,8 @@ xfs_dir2_data_use_free(
*/
dfp = xfs_dir2_data_freefind(hdr, dup);
oldlen = be16_to_cpu(dup->length);
- ASSERT(dfp || oldlen <= be16_to_cpu(hdr->bestfree[2].length));
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ ASSERT(dfp || oldlen <= be16_to_cpu(bf[2].length));
/*
* Check for alignment with front and back of the entry.
*/
@@ -845,7 +880,7 @@ xfs_dir2_data_use_free(
*/
if (matchfront && matchback) {
if (dfp) {
- needscan = (hdr->bestfree[2].offset != 0);
+ needscan = (bf[2].offset != 0);
if (!needscan)
xfs_dir2_data_freeremove(hdr, dfp, needlogp);
}
@@ -875,7 +910,7 @@ xfs_dir2_data_use_free(
* that means we don't know if there was a better
* choice for the last slot, or not. Rescan.
*/
- needscan = dfp == &hdr->bestfree[2];
+ needscan = dfp == &bf[2];
}
}
/*
@@ -902,7 +937,7 @@ xfs_dir2_data_use_free(
* that means we don't know if there was a better
* choice for the last slot, or not. Rescan.
*/
- needscan = dfp == &hdr->bestfree[2];
+ needscan = dfp == &bf[2];
}
}
/*
@@ -930,7 +965,7 @@ xfs_dir2_data_use_free(
* the 2 new will work.
*/
if (dfp) {
- needscan = (hdr->bestfree[2].length != 0);
+ needscan = (bf[2].length != 0);
if (!needscan) {
xfs_dir2_data_freeremove(hdr, dfp, needlogp);
xfs_dir2_data_freeinsert(hdr, newdup, needlogp);
diff --git a/fs/xfs/xfs_dir2_format.h b/fs/xfs/xfs_dir2_format.h
index 0727098..ccd15a29 100644
--- a/fs/xfs/xfs_dir2_format.h
+++ b/fs/xfs/xfs_dir2_format.h
@@ -36,6 +36,37 @@
#define XFS_DIR2_FREE_MAGIC 0x58443246 /* XD2F: free index blocks */
/*
+ * Directory Version 3 With CRCs.
+ *
+ * The tree formats are the same as for version 2 directories. The difference
+ * is in the block header and dirent formats. In many cases the v3 structures
+ * use v2 definitions as they are no different and this makes code sharing much
+ * easier.
+ *
+ * Also, the xfs_dir3_*() functions handle both v2 and v3 formats - if the
+ * format is v2 then they switch to the existing v2 code, or the format is v3
+ * they implement the v3 functionality. This means the existing dir2 is a mix of
+ * xfs_dir2/xfs_dir3 calls and functions. The xfs_dir3 functions are called
+ * where there is a difference in the formats, otherwise the code is unchanged.
+ *
+ * Where it is possible, the code decides what to do based on the magic numbers
+ * in the blocks rather than feature bits in the superblock. This means the code
+ * is as independent of the external XFS code as possible as doesn't require
+ * passing struct xfs_mount pointers into places where it isn't really
+ * necessary.
+ *
+ * Version 3 includes:
+ *
+ * - a larger block header for CRC and identification purposes and so the
+ * offsets of all the structures inside the blocks are different.
+ *
+ * - new magic numbers to be able to detect the v2/v3 types on the fly.
+ */
+
+#define XFS_DIR3_BLOCK_MAGIC 0x58444233 /* XDB3: single block dirs */
+#define XFS_DIR3_DATA_MAGIC 0x58444433 /* XDD3: multiblock dirs */
+
+/*
* Byte offset in data block and shortform entry.
*/
typedef __uint16_t xfs_dir2_data_off_t;
@@ -226,6 +257,38 @@ typedef struct xfs_dir2_data_hdr {
} xfs_dir2_data_hdr_t;
/*
+ * define a structure for all the verification fields we are adding to the
+ * directory block structures. This will be used in several structures.
+ * The magic number must be the first entry to align with all the dir2
+ * structures so we determine how to decode them just by the magic number.
+ */
+struct xfs_dir3_blk_hdr {
+ __be32 magic; /* magic number */
+ __be32 crc; /* CRC of block */
+ __be64 blkno; /* first block of the buffer */
+ __be64 lsn; /* sequence number of last write */
+ uuid_t uuid; /* filesystem we belong to */
+ __be64 owner; /* inode that owns the block */
+};
+
+struct xfs_dir3_data_hdr {
+ struct xfs_dir3_blk_hdr hdr;
+ xfs_dir2_data_free_t best_free[XFS_DIR2_DATA_FD_COUNT];
+};
+
+#define XFS_DIR3_DATA_CRC_OFF offsetof(struct xfs_dir3_data_hdr, hdr.crc)
+
+static inline struct xfs_dir2_data_free *
+xfs_dir3_data_bestfree_p(struct xfs_dir2_data_hdr *hdr)
+{
+ if (hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
+ struct xfs_dir3_data_hdr *hdr3 = (struct xfs_dir3_data_hdr *)hdr;
+ return hdr3->best_free;
+ }
+ return hdr->bestfree;
+}
+
+/*
* Active entry in a data block.
*
* Aligned to 8 bytes. After the variable length name field there is a
@@ -280,6 +343,77 @@ xfs_dir2_data_unused_tag_p(struct xfs_dir2_data_unused *dup)
be16_to_cpu(dup->length) - sizeof(__be16));
}
+static inline struct xfs_dir2_data_unused *
+xfs_dir3_data_unused_p(struct xfs_dir2_data_hdr *hdr)
+{
+ if (hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
+ return (struct xfs_dir2_data_unused *)
+ ((char *)hdr + sizeof(struct xfs_dir3_data_hdr));
+ }
+ return (struct xfs_dir2_data_unused *)
+ ((char *)hdr + sizeof(struct xfs_dir2_data_hdr));
+}
+
+static inline size_t
+xfs_dir3_data_entry_offset(struct xfs_dir2_data_hdr *hdr)
+{
+ if (hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC))
+ return sizeof(struct xfs_dir3_data_hdr);
+ return sizeof(struct xfs_dir2_data_hdr);
+}
+
+static inline struct xfs_dir2_data_entry *
+xfs_dir3_data_entry_p(struct xfs_dir2_data_hdr *hdr)
+{
+ return (struct xfs_dir2_data_entry *)
+ ((char *)hdr + xfs_dir3_data_entry_offset(hdr));
+}
+
+/*
+ * Offsets of . and .. in data space (always block 0)
+ */
+static inline xfs_dir2_data_aoff_t
+xfs_dir3_data_dot_offset(struct xfs_dir2_data_hdr *hdr)
+{
+ return xfs_dir3_data_entry_offset(hdr);
+}
+
+static inline xfs_dir2_data_aoff_t
+xfs_dir3_data_dotdot_offset(struct xfs_dir2_data_hdr *hdr)
+{
+ return xfs_dir3_data_dot_offset(hdr) + xfs_dir2_data_entsize(1);
+}
+
+static inline xfs_dir2_data_aoff_t
+xfs_dir3_data_first_offset(struct xfs_dir2_data_hdr *hdr)
+{
+ return xfs_dir3_data_dotdot_offset(hdr) + xfs_dir2_data_entsize(2);
+}
+
+/*
+ * location of . and .. in data space (always block 0)
+ */
+static inline struct xfs_dir2_data_entry *
+xfs_dir3_data_dot_entry_p(struct xfs_dir2_data_hdr *hdr)
+{
+ return (struct xfs_dir2_data_entry *)
+ ((char *)hdr + xfs_dir3_data_dot_offset(hdr));
+}
+
+static inline struct xfs_dir2_data_entry *
+xfs_dir3_data_dotdot_entry_p(struct xfs_dir2_data_hdr *hdr)
+{
+ return (struct xfs_dir2_data_entry *)
+ ((char *)hdr + xfs_dir3_data_dotdot_offset(hdr));
+}
+
+static inline struct xfs_dir2_data_entry *
+xfs_dir3_data_first_entry_p(struct xfs_dir2_data_hdr *hdr)
+{
+ return (struct xfs_dir2_data_entry *)
+ ((char *)hdr + xfs_dir3_data_first_offset(hdr));
+}
+
/*
* Leaf block structures.
*
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index 60cd2fa..ef163f0 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -149,6 +149,7 @@ xfs_dir2_block_to_leaf(
int needlog; /* need to log block header */
int needscan; /* need to rescan bestfree */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir2_data_free *bf;
trace_xfs_dir2_block_to_leaf(args);
@@ -177,6 +178,7 @@ xfs_dir2_block_to_leaf(
xfs_dir2_data_check(dp, dbp);
btp = xfs_dir2_block_tail_p(mp, hdr);
blp = xfs_dir2_block_leaf_p(btp);
+ bf = xfs_dir3_data_bestfree_p(hdr);
/*
* Set the counts in the leaf header.
*/
@@ -212,7 +214,7 @@ xfs_dir2_block_to_leaf(
ltp = xfs_dir2_leaf_tail_p(mp, leaf);
ltp->bestcount = cpu_to_be32(1);
bestsp = xfs_dir2_leaf_bests_p(ltp);
- bestsp[0] = hdr->bestfree[0].length;
+ bestsp[0] = bf[0].length;
/*
* Log the data header and leaf bests table.
*/
@@ -544,7 +546,7 @@ xfs_dir2_leaf_addname(
/*
* Initialize the block.
*/
- if ((error = xfs_dir2_data_init(args, use_block, &dbp))) {
+ if ((error = xfs_dir3_data_init(args, use_block, &dbp))) {
xfs_trans_brelse(tp, lbp);
return error;
}
diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c
index 5980f9b..985d70b 100644
--- a/fs/xfs/xfs_dir2_node.c
+++ b/fs/xfs/xfs_dir2_node.c
@@ -1588,7 +1588,7 @@ xfs_dir2_node_addname_int(
if (unlikely((error = xfs_dir2_grow_inode(args,
XFS_DIR2_DATA_SPACE,
&dbno)) ||
- (error = xfs_dir2_data_init(args, dbno, &dbp))))
+ (error = xfs_dir3_data_init(args, dbno, &dbp))))
return error;
/*
diff --git a/fs/xfs/xfs_dir2_priv.h b/fs/xfs/xfs_dir2_priv.h
index 7da79f6..e6f2e0a 100644
--- a/fs/xfs/xfs_dir2_priv.h
+++ b/fs/xfs/xfs_dir2_priv.h
@@ -30,7 +30,7 @@ extern int xfs_dir_cilookup_result(struct xfs_da_args *args,
const unsigned char *name, int len);
/* xfs_dir2_block.c */
-extern const struct xfs_buf_ops xfs_dir2_block_buf_ops;
+extern const struct xfs_buf_ops xfs_dir3_block_buf_ops;
extern int xfs_dir2_block_addname(struct xfs_da_args *args);
extern int xfs_dir2_block_getdents(struct xfs_inode *dp, void *dirent,
@@ -61,7 +61,7 @@ xfs_dir2_data_freeinsert(struct xfs_dir2_data_hdr *hdr,
struct xfs_dir2_data_unused *dup, int *loghead);
extern void xfs_dir2_data_freescan(struct xfs_mount *mp,
struct xfs_dir2_data_hdr *hdr, int *loghead);
-extern int xfs_dir2_data_init(struct xfs_da_args *args, xfs_dir2_db_t blkno,
+extern int xfs_dir3_data_init(struct xfs_da_args *args, xfs_dir2_db_t blkno,
struct xfs_buf **bpp);
extern void xfs_dir2_data_log_entry(struct xfs_trans *tp, struct xfs_buf *bp,
struct xfs_dir2_data_entry *dep);
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index e1e3dcf..3216ea5 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -1877,6 +1877,7 @@ xlog_recover_do_reg_buffer(
int bit;
int nbits;
int error;
+ struct xfs_da_blkinfo *info;
trace_xfs_log_recover_buf_reg_buf(mp->m_log, buf_f);
@@ -1936,6 +1937,7 @@ xlog_recover_do_reg_buffer(
/* Shouldn't be any more regions */
ASSERT(i == item->ri_total);
+ info = bp->b_addr;
switch (buf_f->blf_flags & XFS_BLF_TYPE_MASK) {
case XFS_BLF_BTREE_BUF:
switch (be32_to_cpu(*(__be32 *)bp->b_addr)) {
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 11/16] xfs: add CRC checking to dir2 free blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (9 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 10/16] xfs: add CRC checks to block format directory blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 12/16] xfs: add CRC checking to dir2 data blocks Dave Chinner
` (4 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
This addition follows the same pattern as the dir2 block CRCs, but
with a few differences. The main difference is that the free block
header is different between the v2 and v3 formats, so an "in-core"
free block header has been added and _todisk/_from_disk functions
used to abstract the differences in structure format from the code.
This is similar to the on-disk superblock versus the in-core
superblock setup. The in-core strucutre is populated when the buffer
is read from disk, all the in memory checks and modifications are
done on the in-core version of the structure which is written back
to the buffer before the buffer is logged.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_dir2_format.h | 52 ++++-
fs/xfs/xfs_dir2_leaf.c | 2 +-
fs/xfs/xfs_dir2_node.c | 475 +++++++++++++++++++++++++++++++---------------
3 files changed, 373 insertions(+), 156 deletions(-)
diff --git a/fs/xfs/xfs_dir2_format.h b/fs/xfs/xfs_dir2_format.h
index ccd15a29..96390ef 100644
--- a/fs/xfs/xfs_dir2_format.h
+++ b/fs/xfs/xfs_dir2_format.h
@@ -65,6 +65,7 @@
#define XFS_DIR3_BLOCK_MAGIC 0x58444233 /* XDB3: single block dirs */
#define XFS_DIR3_DATA_MAGIC 0x58444433 /* XDD3: multiblock dirs */
+#define XFS_DIR3_FREE_MAGIC 0x58444633 /* XDF3: free index blocks */
/*
* Byte offset in data block and shortform entry.
@@ -654,19 +655,62 @@ typedef struct xfs_dir2_free {
/* unused entries are -1 */
} xfs_dir2_free_t;
-static inline int xfs_dir2_free_max_bests(struct xfs_mount *mp)
+struct xfs_dir3_free_hdr {
+ struct xfs_dir3_blk_hdr hdr;
+ __be32 firstdb; /* db of first entry */
+ __be32 nvalid; /* count of valid entries */
+ __be32 nused; /* count of used entries */
+};
+
+struct xfs_dir3_free {
+ struct xfs_dir3_free_hdr hdr;
+ __be16 bests[]; /* best free counts */
+ /* unused entries are -1 */
+};
+
+#define XFS_DIR3_FREE_CRC_OFF offsetof(struct xfs_dir3_free, hdr.hdr.crc)
+
+/*
+ * In core version of the free block header, abstracted away from on-disk format
+ * differences. Use this in the code, and convert to/from the disk version using
+ * xfs_dir3_free_hdr_from_disk/xfs_dir3_free_hdr_to_disk.
+ */
+struct xfs_dir3_icfree_hdr {
+ __uint32_t magic;
+ __uint32_t firstdb;
+ __uint32_t nvalid;
+ __uint32_t nused;
+
+};
+
+static inline int
+xfs_dir3_free_hdr_size(struct xfs_mount *mp)
+{
+ if (xfs_sb_version_hascrc(&mp->m_sb))
+ return sizeof(struct xfs_dir3_free_hdr);
+ return sizeof(struct xfs_dir2_free_hdr);
+}
+
+static inline int
+xfs_dir3_free_max_bests(struct xfs_mount *mp)
{
- return (mp->m_dirblksize - sizeof(struct xfs_dir2_free_hdr)) /
+ return (mp->m_dirblksize - xfs_dir3_free_hdr_size(mp)) /
sizeof(xfs_dir2_data_off_t);
}
+static inline __be16 *
+xfs_dir3_free_bests_p(struct xfs_mount *mp, struct xfs_dir2_free *free)
+{
+ return (__be16 *)((char *)free + xfs_dir3_free_hdr_size(mp));
+}
+
/*
* Convert data space db to the corresponding free db.
*/
static inline xfs_dir2_db_t
xfs_dir2_db_to_fdb(struct xfs_mount *mp, xfs_dir2_db_t db)
{
- return XFS_DIR2_FREE_FIRSTDB(mp) + db / xfs_dir2_free_max_bests(mp);
+ return XFS_DIR2_FREE_FIRSTDB(mp) + db / xfs_dir3_free_max_bests(mp);
}
/*
@@ -675,7 +719,7 @@ xfs_dir2_db_to_fdb(struct xfs_mount *mp, xfs_dir2_db_t db)
static inline int
xfs_dir2_db_to_fdindex(struct xfs_mount *mp, xfs_dir2_db_t db)
{
- return db % xfs_dir2_free_max_bests(mp);
+ return db % xfs_dir3_free_max_bests(mp);
}
/*
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index ef163f0..4adee8b 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -1971,7 +1971,7 @@ xfs_dir2_node_to_leaf(
/*
* Set up the leaf bests table.
*/
- memcpy(xfs_dir2_leaf_bests_p(ltp), free->bests,
+ memcpy(xfs_dir2_leaf_bests_p(ltp), xfs_dir3_free_bests_p(mp, free),
be32_to_cpu(ltp->bestcount) * sizeof(xfs_dir2_data_off_t));
xfs_dir2_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
xfs_dir2_leaf_log_tail(tp, lbp);
diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c
index 985d70b..480228a 100644
--- a/fs/xfs/xfs_dir2_node.c
+++ b/fs/xfs/xfs_dir2_node.c
@@ -32,6 +32,8 @@
#include "xfs_dir2_priv.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_buf_item.h"
+#include "xfs_cksum.h"
/*
* Function declarations.
@@ -55,44 +57,78 @@ static int xfs_dir2_leafn_remove(xfs_da_args_t *args, struct xfs_buf *bp,
static int xfs_dir2_node_addname_int(xfs_da_args_t *args,
xfs_da_state_blk_t *fblk);
-static void
-xfs_dir2_free_verify(
+static bool
+xfs_dir3_free_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_dir2_free_hdr *hdr = bp->b_addr;
- int block_ok = 0;
- block_ok = hdr->magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC);
- if (!block_ok) {
- XFS_CORRUPTION_ERROR("xfs_dir2_free_verify magic",
- XFS_ERRLEVEL_LOW, mp, hdr);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
+
+ if (hdr3->magic != be32_to_cpu(XFS_DIR3_FREE_MAGIC))
+ return false;
+ if (!uuid_equal(&hdr3->uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be64_to_cpu(hdr3->blkno) != bp->b_bn)
+ return false;
+ } else {
+ if (hdr->magic != be32_to_cpu(XFS_DIR2_FREE_MAGIC))
+ return false;
}
+
+ /* XXX: should bounds check the xfs_dir3_icfree_hdr here */
+
+ return true;
}
static void
-xfs_dir2_free_read_verify(
+xfs_dir3_free_read_verify(
struct xfs_buf *bp)
{
- xfs_dir2_free_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ if ((xfs_sb_version_hascrc(&mp->m_sb) &&
+ !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_DIR3_FREE_CRC_OFF)) ||
+ !xfs_dir3_free_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
static void
-xfs_dir2_free_write_verify(
+xfs_dir3_free_write_verify(
struct xfs_buf *bp)
{
- xfs_dir2_free_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
+
+ if (!xfs_dir3_free_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ hdr3->lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_FREE_CRC_OFF);
}
-static const struct xfs_buf_ops xfs_dir2_free_buf_ops = {
- .verify_read = xfs_dir2_free_read_verify,
- .verify_write = xfs_dir2_free_write_verify,
+static const struct xfs_buf_ops xfs_dir3_free_buf_ops = {
+ .verify_read = xfs_dir3_free_read_verify,
+ .verify_write = xfs_dir3_free_write_verify,
};
static int
-__xfs_dir2_free_read(
+__xfs_dir3_free_read(
struct xfs_trans *tp,
struct xfs_inode *dp,
xfs_dablk_t fbno,
@@ -100,7 +136,7 @@ __xfs_dir2_free_read(
struct xfs_buf **bpp)
{
return xfs_da_read_buf(tp, dp, fbno, mappedbno, bpp,
- XFS_DATA_FORK, &xfs_dir2_free_buf_ops);
+ XFS_DATA_FORK, &xfs_dir3_free_buf_ops);
}
int
@@ -110,7 +146,7 @@ xfs_dir2_free_read(
xfs_dablk_t fbno,
struct xfs_buf **bpp)
{
- return __xfs_dir2_free_read(tp, dp, fbno, -1, bpp);
+ return __xfs_dir3_free_read(tp, dp, fbno, -1, bpp);
}
static int
@@ -120,7 +156,95 @@ xfs_dir2_free_try_read(
xfs_dablk_t fbno,
struct xfs_buf **bpp)
{
- return __xfs_dir2_free_read(tp, dp, fbno, -2, bpp);
+ return __xfs_dir3_free_read(tp, dp, fbno, -2, bpp);
+}
+
+
+static void
+xfs_dir3_free_hdr_from_disk(
+ struct xfs_dir3_icfree_hdr *to,
+ struct xfs_dir2_free *from)
+{
+ if (from->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC)) {
+ to->magic = be32_to_cpu(from->hdr.magic);
+ to->firstdb = be32_to_cpu(from->hdr.firstdb);
+ to->nvalid = be32_to_cpu(from->hdr.nvalid);
+ to->nused = be32_to_cpu(from->hdr.nused);
+ } else {
+ struct xfs_dir3_free_hdr *hdr3 = (struct xfs_dir3_free_hdr *)from;
+
+ to->magic = be32_to_cpu(hdr3->hdr.magic);
+ to->firstdb = be32_to_cpu(hdr3->firstdb);
+ to->nvalid = be32_to_cpu(hdr3->nvalid);
+ to->nused = be32_to_cpu(hdr3->nused);
+ }
+
+ ASSERT(to->magic == XFS_DIR2_FREE_MAGIC ||
+ to->magic == XFS_DIR3_FREE_MAGIC);
+}
+
+static void
+xfs_dir3_free_hdr_to_disk(
+ struct xfs_dir2_free *to,
+ struct xfs_dir3_icfree_hdr *from)
+{
+ ASSERT(from->magic == XFS_DIR2_FREE_MAGIC ||
+ from->magic == XFS_DIR3_FREE_MAGIC);
+
+ if (from->magic == XFS_DIR2_FREE_MAGIC) {
+ to->hdr.magic = cpu_to_be32(from->magic);
+ to->hdr.firstdb = cpu_to_be32(from->firstdb);
+ to->hdr.nvalid = cpu_to_be32(from->nvalid);
+ to->hdr.nused = cpu_to_be32(from->nused);
+ } else {
+ struct xfs_dir3_free_hdr *hdr3 = (struct xfs_dir3_free_hdr *)to;
+
+ hdr3->hdr.magic = cpu_to_be32(from->magic);
+ hdr3->firstdb = cpu_to_be32(from->firstdb);
+ hdr3->nvalid = cpu_to_be32(from->nvalid);
+ hdr3->nused = cpu_to_be32(from->nused);
+ }
+}
+
+static int
+xfs_dir3_free_get_buf(
+ struct xfs_trans *tp,
+ struct xfs_inode *dp,
+ xfs_dir2_db_t fbno,
+ struct xfs_buf **bpp)
+{
+ struct xfs_mount *mp = dp->i_mount;
+ struct xfs_buf *bp;
+ int error;
+ struct xfs_dir3_icfree_hdr hdr;
+
+ error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, fbno),
+ -1, &bp, XFS_DATA_FORK);
+ if (error)
+ return error;
+
+ bp->b_ops = &xfs_dir3_free_buf_ops;;
+
+ /*
+ * Initialize the new block to be empty, and remember
+ * its first slot as our empty slot.
+ */
+ hdr.magic = XFS_DIR2_FREE_MAGIC;
+ hdr.firstdb = 0;
+ hdr.nused = 0;
+ hdr.nvalid = 0;
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_dir3_free_hdr *hdr3 = bp->b_addr;
+
+ hdr.magic = XFS_DIR3_FREE_MAGIC;
+ hdr3->hdr.blkno = cpu_to_be64(bp->b_bn);
+ hdr3->hdr.owner = cpu_to_be64(dp->i_ino);
+ uuid_copy(&hdr3->hdr.uuid, &mp->m_sb.sb_uuid);
+
+ }
+ xfs_dir3_free_hdr_to_disk(bp->b_addr, &hdr);
+ *bpp = bp;
+ return 0;
}
/*
@@ -134,13 +258,16 @@ xfs_dir2_free_log_bests(
int last) /* last entry to log */
{
xfs_dir2_free_t *free; /* freespace structure */
+ __be16 *bests;
free = bp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
+ bests = xfs_dir3_free_bests_p(tp->t_mountp, free);
+ ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC) ||
+ free->hdr.magic == cpu_to_be32(XFS_DIR3_FREE_MAGIC));
xfs_trans_log_buf(tp, bp,
- (uint)((char *)&free->bests[first] - (char *)free),
- (uint)((char *)&free->bests[last] - (char *)free +
- sizeof(free->bests[0]) - 1));
+ (uint)((char *)&bests[first] - (char *)free),
+ (uint)((char *)&bests[last] - (char *)free +
+ sizeof(bests[0]) - 1));
}
/*
@@ -154,9 +281,9 @@ xfs_dir2_free_log_header(
xfs_dir2_free_t *free; /* freespace structure */
free = bp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
- xfs_trans_log_buf(tp, bp, (uint)((char *)&free->hdr - (char *)free),
- (uint)(sizeof(xfs_dir2_free_hdr_t) - 1));
+ ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC) ||
+ free->hdr.magic == cpu_to_be32(XFS_DIR3_FREE_MAGIC));
+ xfs_trans_log_buf(tp, bp, 0, xfs_dir3_free_hdr_size(tp->t_mountp) - 1);
}
/*
@@ -183,6 +310,7 @@ xfs_dir2_leaf_to_node(
xfs_dir2_data_off_t off; /* freespace entry value */
__be16 *to; /* pointer to freespace entry */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir3_icfree_hdr freehdr;
trace_xfs_dir2_leaf_to_node(args);
@@ -199,43 +327,43 @@ xfs_dir2_leaf_to_node(
/*
* Get the buffer for the new freespace block.
*/
- error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, fdb), -1, &fbp,
- XFS_DATA_FORK);
+ error = xfs_dir3_free_get_buf(tp, dp, fdb, &fbp);
if (error)
return error;
- fbp->b_ops = &xfs_dir2_free_buf_ops;
free = fbp->b_addr;
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
leaf = lbp->b_addr;
ltp = xfs_dir2_leaf_tail_p(mp, leaf);
- /*
- * Initialize the freespace block header.
- */
- free->hdr.magic = cpu_to_be32(XFS_DIR2_FREE_MAGIC);
- free->hdr.firstdb = 0;
- ASSERT(be32_to_cpu(ltp->bestcount) <= (uint)dp->i_d.di_size / mp->m_dirblksize);
- free->hdr.nvalid = ltp->bestcount;
+ ASSERT(be32_to_cpu(ltp->bestcount) <=
+ (uint)dp->i_d.di_size / mp->m_dirblksize);
+
/*
* Copy freespace entries from the leaf block to the new block.
* Count active entries.
*/
- for (i = n = 0, from = xfs_dir2_leaf_bests_p(ltp), to = free->bests;
- i < be32_to_cpu(ltp->bestcount); i++, from++, to++) {
+ from = xfs_dir2_leaf_bests_p(ltp);
+ to = xfs_dir3_free_bests_p(mp, free);
+ for (i = n = 0; i < be32_to_cpu(ltp->bestcount); i++, from++, to++) {
if ((off = be16_to_cpu(*from)) != NULLDATAOFF)
n++;
*to = cpu_to_be16(off);
}
- free->hdr.nused = cpu_to_be32(n);
-
- lbp->b_ops = &xfs_dir2_leafn_buf_ops;
- leaf->hdr.info.magic = cpu_to_be16(XFS_DIR2_LEAFN_MAGIC);
/*
- * Log everything.
+ * Now initialize the freespace block header.
*/
- xfs_dir2_leaf_log_header(tp, lbp);
+ freehdr.nused = n;
+ freehdr.nvalid = be32_to_cpu(ltp->bestcount);
+
+ xfs_dir3_free_hdr_to_disk(fbp->b_addr, &freehdr);
+ xfs_dir2_free_log_bests(tp, fbp, 0, freehdr.nvalid - 1);
xfs_dir2_free_log_header(tp, fbp);
- xfs_dir2_free_log_bests(tp, fbp, 0, be32_to_cpu(free->hdr.nvalid) - 1);
+
+ /* convert the leaf to a leafnode */
+ leaf->hdr.info.magic = cpu_to_be16(XFS_DIR2_LEAFN_MAGIC);
+ lbp->b_ops = &xfs_dir2_leafn_buf_ops;
+ xfs_dir2_leaf_log_header(tp, lbp);
xfs_dir2_leafn_check(dp, lbp);
return 0;
}
@@ -354,6 +482,23 @@ xfs_dir2_leafn_check(
}
ASSERT(be16_to_cpu(leaf->hdr.stale) == stale);
}
+
+static void
+xfs_dir2_free_hdr_check(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp,
+ xfs_dir2_db_t db)
+{
+ struct xfs_dir3_icfree_hdr hdr;
+
+ xfs_dir3_free_hdr_from_disk(&hdr, bp->b_addr);
+
+ ASSERT((hdr.firstdb % xfs_dir3_free_max_bests(mp)) == 0);
+ ASSERT(hdr.firstdb <= db);
+ ASSERT(db < hdr.firstdb + hdr.nvalid);
+}
+#else
+#define xfs_dir2_free_hdr_check(mp, dp, db)
#endif /* DEBUG */
/*
@@ -424,7 +569,8 @@ xfs_dir2_leafn_lookup_for_addname(
curbp = state->extrablk.bp;
curfdb = state->extrablk.blkno;
free = curbp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
+ ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC) ||
+ free->hdr.magic == cpu_to_be32(XFS_DIR3_FREE_MAGIC));
}
length = xfs_dir2_data_entsize(args->namelen);
/*
@@ -451,6 +597,8 @@ xfs_dir2_leafn_lookup_for_addname(
* in hand, take a look at it.
*/
if (newdb != curdb) {
+ __be16 *bests;
+
curdb = newdb;
/*
* Convert the data block to the free block
@@ -473,13 +621,8 @@ xfs_dir2_leafn_lookup_for_addname(
if (error)
return error;
free = curbp->b_addr;
- ASSERT(be32_to_cpu(free->hdr.magic) ==
- XFS_DIR2_FREE_MAGIC);
- ASSERT((be32_to_cpu(free->hdr.firstdb) %
- xfs_dir2_free_max_bests(mp)) == 0);
- ASSERT(be32_to_cpu(free->hdr.firstdb) <= curdb);
- ASSERT(curdb < be32_to_cpu(free->hdr.firstdb) +
- be32_to_cpu(free->hdr.nvalid));
+
+ xfs_dir2_free_hdr_check(mp, curbp, curdb);
}
/*
* Get the index for our entry.
@@ -488,8 +631,8 @@ xfs_dir2_leafn_lookup_for_addname(
/*
* If it has room, return it.
*/
- if (unlikely(free->bests[fi] ==
- cpu_to_be16(NULLDATAOFF))) {
+ bests = xfs_dir3_free_bests_p(mp, free);
+ if (unlikely(bests[fi] == cpu_to_be16(NULLDATAOFF))) {
XFS_ERROR_REPORT("xfs_dir2_leafn_lookup_int",
XFS_ERRLEVEL_LOW, mp);
if (curfdb != newfdb)
@@ -497,7 +640,7 @@ xfs_dir2_leafn_lookup_for_addname(
return XFS_ERROR(EFSCORRUPTED);
}
curfdb = newfdb;
- if (be16_to_cpu(free->bests[fi]) >= length)
+ if (be16_to_cpu(bests[fi]) >= length)
goto out;
}
}
@@ -511,6 +654,12 @@ out:
state->extrablk.bp = curbp;
state->extrablk.index = fi;
state->extrablk.blkno = curfdb;
+
+ /*
+ * Important: this magic number is not in the buffer - it's for
+ * buffer type information and therefore only the free/data type
+ * matters here, not whether CRCs are enabled or not.
+ */
state->extrablk.magic = XFS_DIR2_FREE_MAGIC;
} else {
state->extravalid = 0;
@@ -898,7 +1047,7 @@ xfs_dir2_leafn_rebalance(
}
static int
-xfs_dir2_data_block_free(
+xfs_dir3_data_block_free(
xfs_da_args_t *args,
struct xfs_dir2_data_hdr *hdr,
struct xfs_dir2_free *free,
@@ -909,59 +1058,68 @@ xfs_dir2_data_block_free(
{
struct xfs_trans *tp = args->trans;
int logfree = 0;
+ __be16 *bests;
+ struct xfs_dir3_icfree_hdr freehdr;
- if (!hdr) {
- /* One less used entry in the free table. */
- be32_add_cpu(&free->hdr.nused, -1);
- xfs_dir2_free_log_header(tp, fbp);
- /*
- * If this was the last entry in the table, we can trim the
- * table size back. There might be other entries at the end
- * referring to non-existent data blocks, get those too.
- */
- if (findex == be32_to_cpu(free->hdr.nvalid) - 1) {
- int i; /* free entry index */
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
- for (i = findex - 1; i >= 0; i--) {
- if (free->bests[i] != cpu_to_be16(NULLDATAOFF))
- break;
- }
- free->hdr.nvalid = cpu_to_be32(i + 1);
- logfree = 0;
- } else {
- /* Not the last entry, just punch it out. */
- free->bests[findex] = cpu_to_be16(NULLDATAOFF);
- logfree = 1;
- }
+ bests = xfs_dir3_free_bests_p(tp->t_mountp, free);
+ if (hdr) {
/*
- * If there are no useful entries left in the block,
- * get rid of the block if we can.
+ * Data block is not empty, just set the free entry to the new
+ * value.
*/
- if (!free->hdr.nused) {
- int error;
+ bests[findex] = cpu_to_be16(longest);
+ xfs_dir2_free_log_bests(tp, fbp, findex, findex);
+ return 0;
+ }
- error = xfs_dir2_shrink_inode(args, fdb, fbp);
- if (error == 0) {
- fbp = NULL;
- logfree = 0;
- } else if (error != ENOSPC || args->total != 0)
- return error;
- /*
- * It's possible to get ENOSPC if there is no
- * space reservation. In this case some one
- * else will eventually get rid of this block.
- */
+ /*
+ * One less used entry in the free table. Unused is not converted
+ * because we only need to know if it zero
+ */
+ freehdr.nused--;
+
+ if (findex == freehdr.nvalid - 1) {
+ int i; /* free entry index */
+
+ for (i = findex - 1; i >= 0; i--) {
+ if (bests[i] != cpu_to_be16(NULLDATAOFF))
+ break;
}
+ freehdr.nvalid = i + 1;
+ logfree = 0;
} else {
+ /* Not the last entry, just punch it out. */
+ bests[findex] = cpu_to_be16(NULLDATAOFF);
+ logfree = 1;
+ }
+
+ xfs_dir3_free_hdr_to_disk(free, &freehdr);
+ xfs_dir2_free_log_header(tp, fbp);
+
+ /*
+ * If there are no useful entries left in the block, get rid of the
+ * block if we can.
+ */
+ if (!freehdr.nused) {
+ int error;
+
+ error = xfs_dir2_shrink_inode(args, fdb, fbp);
+ if (error == 0) {
+ fbp = NULL;
+ logfree = 0;
+ } else if (error != ENOSPC || args->total != 0)
+ return error;
/*
- * Data block is not empty, just set the free entry to the new
- * value.
+ * It's possible to get ENOSPC if there is no
+ * space reservation. In this case some one
+ * else will eventually get rid of this block.
*/
- free->bests[findex] = cpu_to_be16(longest);
- logfree = 1;
}
+
/* Log the free entry that changed, unless we got rid of it. */
if (logfree)
xfs_dir2_free_log_bests(tp, fbp, findex, findex);
@@ -1062,10 +1220,15 @@ xfs_dir2_leafn_remove(
if (error)
return error;
free = fbp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
- ASSERT(be32_to_cpu(free->hdr.firstdb) ==
- xfs_dir2_free_max_bests(mp) *
- (fdb - XFS_DIR2_FREE_FIRSTDB(mp)));
+#ifdef DEBUG
+ {
+ struct xfs_dir3_icfree_hdr freehdr;
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
+ ASSERT(be32_to_cpu(freehdr.firstdb) ==
+ xfs_dir3_free_max_bests(mp) *
+ (fdb - XFS_DIR2_FREE_FIRSTDB(mp)));
+ }
+#endif
/*
* Calculate which entry we need to fix.
*/
@@ -1096,7 +1259,7 @@ xfs_dir2_leafn_remove(
* If we got rid of the data block, we can eliminate that entry
* in the free block.
*/
- error = xfs_dir2_data_block_free(args, hdr, free,
+ error = xfs_dir3_data_block_free(args, hdr, free,
fdb, findex, fbp, longest);
if (error)
return error;
@@ -1447,6 +1610,8 @@ xfs_dir2_node_addname_int(
int needscan; /* need to rescan data frees */
__be16 *tagp; /* data entry tag pointer */
xfs_trans_t *tp; /* transaction pointer */
+ __be16 *bests;
+ struct xfs_dir3_icfree_hdr freehdr;
dp = args->dp;
mp = dp->i_mount;
@@ -1464,36 +1629,37 @@ xfs_dir2_node_addname_int(
*/
ifbno = fblk->blkno;
free = fbp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
findex = fblk->index;
+ bests = xfs_dir3_free_bests_p(mp, free);
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
+
/*
* This means the free entry showed that the data block had
* space for our entry, so we remembered it.
* Use that data block.
*/
if (findex >= 0) {
- ASSERT(findex < be32_to_cpu(free->hdr.nvalid));
- ASSERT(be16_to_cpu(free->bests[findex]) != NULLDATAOFF);
- ASSERT(be16_to_cpu(free->bests[findex]) >= length);
- dbno = be32_to_cpu(free->hdr.firstdb) + findex;
- }
- /*
- * The data block looked at didn't have enough room.
- * We'll start at the beginning of the freespace entries.
- */
- else {
+ ASSERT(findex < freehdr.nvalid);
+ ASSERT(be16_to_cpu(bests[findex]) != NULLDATAOFF);
+ ASSERT(be16_to_cpu(bests[findex]) >= length);
+ dbno = freehdr.firstdb + findex;
+ } else {
+ /*
+ * The data block looked at didn't have enough room.
+ * We'll start at the beginning of the freespace entries.
+ */
dbno = -1;
findex = 0;
}
- }
- /*
- * Didn't come in with a freespace block, so don't have a data block.
- */
- else {
+ } else {
+ /*
+ * Didn't come in with a freespace block, so no data block.
+ */
ifbno = dbno = -1;
fbp = NULL;
findex = 0;
}
+
/*
* If we don't have a data block yet, we're going to scan the
* freespace blocks looking for one. Figure out what the
@@ -1547,20 +1713,26 @@ xfs_dir2_node_addname_int(
if (!fbp)
continue;
free = fbp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
findex = 0;
}
/*
* Look at the current free entry. Is it good enough?
+ *
+ * The bests initialisation should be wher eteh bufer is read in
+ * the above branch. But gcc is too stupid to realise that bests
+ * iand the freehdr are actually initialised if they are placed
+ * there, so we have to do it here to avoid warnings. Blech.
*/
- if (be16_to_cpu(free->bests[findex]) != NULLDATAOFF &&
- be16_to_cpu(free->bests[findex]) >= length)
- dbno = be32_to_cpu(free->hdr.firstdb) + findex;
+ bests = xfs_dir3_free_bests_p(mp, free);
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
+ if (be16_to_cpu(bests[findex]) != NULLDATAOFF &&
+ be16_to_cpu(bests[findex]) >= length)
+ dbno = freehdr.firstdb + findex;
else {
/*
* Are we done with the freeblock?
*/
- if (++findex == be32_to_cpu(free->hdr.nvalid)) {
+ if (++findex == freehdr.nvalid) {
/*
* Drop the block.
*/
@@ -1614,11 +1786,11 @@ xfs_dir2_node_addname_int(
* If there wasn't a freespace block, the read will
* return a NULL fbp. Allocate and initialize a new one.
*/
- if( fbp == NULL ) {
- if ((error = xfs_dir2_grow_inode(args, XFS_DIR2_FREE_SPACE,
- &fbno))) {
+ if(!fbp) {
+ error = xfs_dir2_grow_inode(args, XFS_DIR2_FREE_SPACE,
+ &fbno);
+ if (error)
return error;
- }
if (unlikely(xfs_dir2_db_to_fdb(mp, dbno) != fbno)) {
xfs_alert(mp,
@@ -1646,27 +1818,24 @@ xfs_dir2_node_addname_int(
/*
* Get a buffer for the new block.
*/
- error = xfs_da_get_buf(tp, dp,
- xfs_dir2_db_to_da(mp, fbno),
- -1, &fbp, XFS_DATA_FORK);
+ error = xfs_dir3_free_get_buf(tp, dp, fbno, &fbp);
if (error)
return error;
- fbp->b_ops = &xfs_dir2_free_buf_ops;
+ free = fbp->b_addr;
+ bests = xfs_dir3_free_bests_p(mp, free);
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
/*
- * Initialize the new block to be empty, and remember
- * its first slot as our empty slot.
+ * Remember the first slot as our empty slot.
*/
- free = fbp->b_addr;
- free->hdr.magic = cpu_to_be32(XFS_DIR2_FREE_MAGIC);
- free->hdr.firstdb = cpu_to_be32(
- (fbno - XFS_DIR2_FREE_FIRSTDB(mp)) *
- xfs_dir2_free_max_bests(mp));
+ freehdr.firstdb = (fbno - XFS_DIR2_FREE_FIRSTDB(mp)) *
+ xfs_dir3_free_max_bests(mp);
free->hdr.nvalid = 0;
free->hdr.nused = 0;
} else {
free = fbp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
+ bests = xfs_dir3_free_bests_p(mp, free);
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
}
/*
@@ -1677,20 +1846,21 @@ xfs_dir2_node_addname_int(
* If it's after the end of the current entries in the
* freespace block, extend that table.
*/
- if (findex >= be32_to_cpu(free->hdr.nvalid)) {
- ASSERT(findex < xfs_dir2_free_max_bests(mp));
- free->hdr.nvalid = cpu_to_be32(findex + 1);
+ if (findex >= freehdr.nvalid) {
+ ASSERT(findex < xfs_dir3_free_max_bests(mp));
+ freehdr.nvalid = findex + 1;
/*
* Tag new entry so nused will go up.
*/
- free->bests[findex] = cpu_to_be16(NULLDATAOFF);
+ bests[findex] = cpu_to_be16(NULLDATAOFF);
}
/*
* If this entry was for an empty data block
* (this should always be true) then update the header.
*/
- if (free->bests[findex] == cpu_to_be16(NULLDATAOFF)) {
- be32_add_cpu(&free->hdr.nused, 1);
+ if (bests[findex] == cpu_to_be16(NULLDATAOFF)) {
+ freehdr.nused++;
+ xfs_dir3_free_hdr_to_disk(fbp->b_addr, &freehdr);
xfs_dir2_free_log_header(tp, fbp);
}
/*
@@ -1699,7 +1869,7 @@ xfs_dir2_node_addname_int(
* change again.
*/
hdr = dbp->b_addr;
- free->bests[findex] = hdr->bestfree[0].length;
+ bests[findex] = hdr->bestfree[0].length;
logfree = 1;
}
/*
@@ -1758,8 +1928,9 @@ xfs_dir2_node_addname_int(
/*
* If the freespace entry is now wrong, update it.
*/
- if (be16_to_cpu(free->bests[findex]) != be16_to_cpu(hdr->bestfree[0].length)) {
- free->bests[findex] = hdr->bestfree[0].length;
+ bests = xfs_dir3_free_bests_p(mp, free); /* gcc is so stupid */
+ if (be16_to_cpu(bests[findex]) != be16_to_cpu(hdr->bestfree[0].length)) {
+ bests[findex] = hdr->bestfree[0].length;
logfree = 1;
}
/*
@@ -1995,6 +2166,7 @@ xfs_dir2_node_trim_free(
xfs_dir2_free_t *free; /* freespace structure */
xfs_mount_t *mp; /* filesystem mount point */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir3_icfree_hdr freehdr;
dp = args->dp;
mp = dp->i_mount;
@@ -2012,11 +2184,12 @@ xfs_dir2_node_trim_free(
if (!bp)
return 0;
free = bp->b_addr;
- ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC));
+ xfs_dir3_free_hdr_from_disk(&freehdr, free);
+
/*
* If there are used entries, there's nothing to do.
*/
- if (be32_to_cpu(free->hdr.nused) > 0) {
+ if (freehdr.nused > 0) {
xfs_trans_brelse(tp, bp);
*rvalp = 0;
return 0;
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 12/16] xfs: add CRC checking to dir2 data blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (10 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 11/16] xfs: add CRC checking to dir2 free blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 13/16] xfs: add CRC checking to dir2 leaf blocks Dave Chinner
` (3 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
This addition follows the same pattern as the dir2 block CRCs.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_dir2_block.c | 22 +++++-----
fs/xfs/xfs_dir2_data.c | 100 ++++++++++++++++++++++++++++++++--------------
fs/xfs/xfs_dir2_format.h | 9 +++--
fs/xfs/xfs_dir2_leaf.c | 64 ++++++++++++++++-------------
fs/xfs/xfs_dir2_node.c | 36 ++++++++++-------
fs/xfs/xfs_dir2_priv.h | 12 +++---
fs/xfs/xfs_file.c | 2 +-
7 files changed, 153 insertions(+), 92 deletions(-)
diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c
index f2b34b2..175db20 100644
--- a/fs/xfs/xfs_dir2_block.c
+++ b/fs/xfs/xfs_dir2_block.c
@@ -76,7 +76,7 @@ xfs_dir3_block_verify(
if (hdr3->magic != be32_to_cpu(XFS_DIR2_BLOCK_MAGIC))
return false;
}
- if (__xfs_dir2_data_check(NULL, bp))
+ if (__xfs_dir3_data_check(NULL, bp))
return false;
return true;
}
@@ -552,7 +552,7 @@ xfs_dir2_block_addname(
xfs_dir2_data_log_header(tp, bp);
xfs_dir2_block_log_tail(tp, bp);
xfs_dir2_data_log_entry(tp, bp, dep);
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
return 0;
}
@@ -595,7 +595,7 @@ xfs_dir2_block_getdents(
*/
wantoff = xfs_dir2_dataptr_to_off(mp, *offset);
hdr = bp->b_addr;
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
/*
* Set up values for the loop.
*/
@@ -719,7 +719,7 @@ xfs_dir2_block_lookup(
dp = args->dp;
mp = dp->i_mount;
hdr = bp->b_addr;
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
btp = xfs_dir2_block_tail_p(mp, hdr);
blp = xfs_dir2_block_leaf_p(btp);
/*
@@ -770,7 +770,7 @@ xfs_dir2_block_lookup_int(
return error;
hdr = bp->b_addr;
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
btp = xfs_dir2_block_tail_p(mp, hdr);
blp = xfs_dir2_block_leaf_p(btp);
/*
@@ -907,7 +907,7 @@ xfs_dir2_block_removename(
xfs_dir2_data_freescan(mp, hdr, &needlog);
if (needlog)
xfs_dir2_data_log_header(tp, bp);
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
/*
* See if the size as a shortform is good enough.
*/
@@ -964,7 +964,7 @@ xfs_dir2_block_replace(
*/
dep->inumber = cpu_to_be64(args->inumber);
xfs_dir2_data_log_entry(args->trans, bp, dep);
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
return 0;
}
@@ -1042,12 +1042,14 @@ xfs_dir2_leaf_to_block(
* Read the data block if we don't already have it, give up if it fails.
*/
if (!dbp) {
- error = xfs_dir2_data_read(tp, dp, mp->m_dirdatablk, -1, &dbp);
+ error = xfs_dir3_data_read(tp, dp, mp->m_dirdatablk, -1, &dbp);
if (error)
return error;
}
hdr = dbp->b_addr;
- ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC));
+ ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC));
+
/*
* Size of the "leaf" area in the block.
*/
@@ -1325,6 +1327,6 @@ xfs_dir2_sf_to_block(
ASSERT(needscan == 0);
xfs_dir2_block_log_leaf(tp, bp, 0, be32_to_cpu(btp->count) - 1);
xfs_dir2_block_log_tail(tp, bp);
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
return 0;
}
diff --git a/fs/xfs/xfs_dir2_data.c b/fs/xfs/xfs_dir2_data.c
index 93bc123..f0284fa 100644
--- a/fs/xfs/xfs_dir2_data.c
+++ b/fs/xfs/xfs_dir2_data.c
@@ -30,6 +30,8 @@
#include "xfs_dir2_format.h"
#include "xfs_dir2_priv.h"
#include "xfs_error.h"
+#include "xfs_buf_item.h"
+#include "xfs_cksum.h"
STATIC xfs_dir2_data_free_t *
xfs_dir2_data_freefind(xfs_dir2_data_hdr_t *hdr, xfs_dir2_data_unused_t *dup);
@@ -40,7 +42,7 @@ xfs_dir2_data_freefind(xfs_dir2_data_hdr_t *hdr, xfs_dir2_data_unused_t *dup);
* Return 0 is the buffer is good, otherwise an error.
*/
int
-__xfs_dir2_data_check(
+__xfs_dir3_data_check(
struct xfs_inode *dp, /* incore inode pointer */
struct xfs_buf *bp) /* data block's buffer */
{
@@ -75,6 +77,7 @@ __xfs_dir2_data_check(
lep = xfs_dir2_block_leaf_p(btp);
endp = (char *)lep;
break;
+ case cpu_to_be32(XFS_DIR3_DATA_MAGIC):
case cpu_to_be32(XFS_DIR2_DATA_MAGIC):
endp = (char *)hdr + mp->m_dirblksize;
break;
@@ -188,21 +191,27 @@ __xfs_dir2_data_check(
return 0;
}
-static void
-xfs_dir2_data_verify(
+static bool
+xfs_dir3_data_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
- struct xfs_dir2_data_hdr *hdr = bp->b_addr;
- int block_ok = 0;
-
- block_ok = hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC);
- block_ok = block_ok && __xfs_dir2_data_check(NULL, bp) == 0;
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
- if (!block_ok) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ if (hdr3->magic != be32_to_cpu(XFS_DIR3_DATA_MAGIC))
+ return false;
+ if (!uuid_equal(&hdr3->uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be64_to_cpu(hdr3->blkno) != bp->b_bn)
+ return false;
+ } else {
+ if (hdr3->magic != be32_to_cpu(XFS_DIR2_DATA_MAGIC))
+ return false;
}
+ if (__xfs_dir3_data_check(NULL, bp))
+ return false;
+ return true;
}
/*
@@ -211,7 +220,7 @@ xfs_dir2_data_verify(
* format buffer or a data format buffer on readahead.
*/
static void
-xfs_dir2_data_reada_verify(
+xfs_dir3_data_reada_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
@@ -224,7 +233,8 @@ xfs_dir2_data_reada_verify(
bp->b_ops->verify_read(bp);
return;
case cpu_to_be32(XFS_DIR2_DATA_MAGIC):
- xfs_dir2_data_verify(bp);
+ case cpu_to_be32(XFS_DIR3_DATA_MAGIC):
+ xfs_dir3_data_verify(bp);
return;
default:
XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
@@ -234,32 +244,56 @@ xfs_dir2_data_reada_verify(
}
static void
-xfs_dir2_data_read_verify(
+xfs_dir3_data_read_verify(
struct xfs_buf *bp)
{
- xfs_dir2_data_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ if ((xfs_sb_version_hascrc(&mp->m_sb) &&
+ !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_DIR3_DATA_CRC_OFF)) ||
+ !xfs_dir3_data_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
static void
-xfs_dir2_data_write_verify(
+xfs_dir3_data_write_verify(
struct xfs_buf *bp)
{
- xfs_dir2_data_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+ struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
+
+ if (!xfs_dir3_data_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ hdr3->lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_DATA_CRC_OFF);
}
-const struct xfs_buf_ops xfs_dir2_data_buf_ops = {
- .verify_read = xfs_dir2_data_read_verify,
- .verify_write = xfs_dir2_data_write_verify,
+const struct xfs_buf_ops xfs_dir3_data_buf_ops = {
+ .verify_read = xfs_dir3_data_read_verify,
+ .verify_write = xfs_dir3_data_write_verify,
};
-static const struct xfs_buf_ops xfs_dir2_data_reada_buf_ops = {
- .verify_read = xfs_dir2_data_reada_verify,
- .verify_write = xfs_dir2_data_write_verify,
+static const struct xfs_buf_ops xfs_dir3_data_reada_buf_ops = {
+ .verify_read = xfs_dir3_data_reada_verify,
+ .verify_write = xfs_dir3_data_write_verify,
};
int
-xfs_dir2_data_read(
+xfs_dir3_data_read(
struct xfs_trans *tp,
struct xfs_inode *dp,
xfs_dablk_t bno,
@@ -267,18 +301,18 @@ xfs_dir2_data_read(
struct xfs_buf **bpp)
{
return xfs_da_read_buf(tp, dp, bno, mapped_bno, bpp,
- XFS_DATA_FORK, &xfs_dir2_data_buf_ops);
+ XFS_DATA_FORK, &xfs_dir3_data_buf_ops);
}
int
-xfs_dir2_data_readahead(
+xfs_dir3_data_readahead(
struct xfs_trans *tp,
struct xfs_inode *dp,
xfs_dablk_t bno,
xfs_daddr_t mapped_bno)
{
return xfs_da_reada_buf(tp, dp, bno, mapped_bno,
- XFS_DATA_FORK, &xfs_dir2_data_reada_buf_ops);
+ XFS_DATA_FORK, &xfs_dir3_data_reada_buf_ops);
}
/*
@@ -308,6 +342,7 @@ xfs_dir2_data_freefind(
* one we're looking for it has to be exact.
*/
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
for (dfp = &bf[0], seenzero = matched = 0;
@@ -455,6 +490,7 @@ xfs_dir2_data_freescan(
char *p; /* current entry pointer */
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
@@ -531,13 +567,12 @@ xfs_dir3_data_init(
XFS_DATA_FORK);
if (error)
return error;
- bp->b_ops = &xfs_dir2_data_buf_ops;
+ bp->b_ops = &xfs_dir3_data_buf_ops;
/*
* Initialize the header.
*/
hdr = bp->b_addr;
-
if (xfs_sb_version_hascrc(&mp->m_sb)) {
struct xfs_dir3_blk_hdr *hdr3 = bp->b_addr;
@@ -588,6 +623,7 @@ xfs_dir2_data_log_entry(
xfs_dir2_data_hdr_t *hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
@@ -607,6 +643,7 @@ xfs_dir2_data_log_header(
xfs_dir2_data_hdr_t *hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
@@ -625,6 +662,7 @@ xfs_dir2_data_log_unused(
xfs_dir2_data_hdr_t *hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
@@ -672,7 +710,8 @@ xfs_dir2_data_make_free(
/*
* Figure out where the end of the data area is.
*/
- if (hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC))
+ if (hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC))
endptr = (char *)hdr + mp->m_dirblksize;
else {
xfs_dir2_block_tail_t *btp; /* block tail */
@@ -854,6 +893,7 @@ xfs_dir2_data_use_free(
hdr = bp->b_addr;
ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC) ||
hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC));
ASSERT(be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG);
diff --git a/fs/xfs/xfs_dir2_format.h b/fs/xfs/xfs_dir2_format.h
index 96390ef..358efab 100644
--- a/fs/xfs/xfs_dir2_format.h
+++ b/fs/xfs/xfs_dir2_format.h
@@ -282,7 +282,8 @@ struct xfs_dir3_data_hdr {
static inline struct xfs_dir2_data_free *
xfs_dir3_data_bestfree_p(struct xfs_dir2_data_hdr *hdr)
{
- if (hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
+ if (hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
struct xfs_dir3_data_hdr *hdr3 = (struct xfs_dir3_data_hdr *)hdr;
return hdr3->best_free;
}
@@ -347,7 +348,8 @@ xfs_dir2_data_unused_tag_p(struct xfs_dir2_data_unused *dup)
static inline struct xfs_dir2_data_unused *
xfs_dir3_data_unused_p(struct xfs_dir2_data_hdr *hdr)
{
- if (hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
+ if (hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC)) {
return (struct xfs_dir2_data_unused *)
((char *)hdr + sizeof(struct xfs_dir3_data_hdr));
}
@@ -358,7 +360,8 @@ xfs_dir3_data_unused_p(struct xfs_dir2_data_hdr *hdr)
static inline size_t
xfs_dir3_data_entry_offset(struct xfs_dir2_data_hdr *hdr)
{
- if (hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC))
+ if (hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_BLOCK_MAGIC))
return sizeof(struct xfs_dir3_data_hdr);
return sizeof(struct xfs_dir2_data_hdr);
}
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index 4adee8b..78d3182 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -149,7 +149,7 @@ xfs_dir2_block_to_leaf(
int needlog; /* need to log block header */
int needscan; /* need to rescan bestfree */
xfs_trans_t *tp; /* transaction pointer */
- struct xfs_dir2_data_free *bf;
+ struct xfs_dir2_data_free *bf;
trace_xfs_dir2_block_to_leaf(args);
@@ -175,7 +175,7 @@ xfs_dir2_block_to_leaf(
ASSERT(lbp != NULL);
leaf = lbp->b_addr;
hdr = dbp->b_addr;
- xfs_dir2_data_check(dp, dbp);
+ xfs_dir3_data_check(dp, dbp);
btp = xfs_dir2_block_tail_p(mp, hdr);
blp = xfs_dir2_block_leaf_p(btp);
bf = xfs_dir3_data_bestfree_p(hdr);
@@ -204,8 +204,12 @@ xfs_dir2_block_to_leaf(
/*
* Fix up the block header, make it a data block.
*/
- dbp->b_ops = &xfs_dir2_data_buf_ops;
- hdr->magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
+ dbp->b_ops = &xfs_dir3_data_buf_ops;
+ if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC))
+ hdr->magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC);
+ else
+ hdr->magic = cpu_to_be32(XFS_DIR3_DATA_MAGIC);
+
if (needscan)
xfs_dir2_data_freescan(mp, hdr, &needlog);
/*
@@ -221,7 +225,7 @@ xfs_dir2_block_to_leaf(
if (needlog)
xfs_dir2_data_log_header(tp, dbp);
xfs_dir2_leaf_check(dp, lbp);
- xfs_dir2_data_check(dp, dbp);
+ xfs_dir3_data_check(dp, dbp);
xfs_dir2_leaf_log_bests(tp, lbp, 0, 0);
return 0;
}
@@ -385,6 +389,7 @@ xfs_dir2_leaf_addname(
__be16 *tagp; /* end of data entry */
xfs_trans_t *tp; /* transaction pointer */
xfs_dir2_db_t use_block; /* data block number */
+ struct xfs_dir2_data_free *bf; /* bestfree table */
trace_xfs_dir2_leaf_addname(args);
@@ -568,14 +573,15 @@ xfs_dir2_leaf_addname(
else
xfs_dir2_leaf_log_bests(tp, lbp, use_block, use_block);
hdr = dbp->b_addr;
- bestsp[use_block] = hdr->bestfree[0].length;
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ bestsp[use_block] = bf[0].length;
grown = 1;
} else {
/*
* Already had space in some data block.
* Just read that one in.
*/
- error = xfs_dir2_data_read(tp, dp,
+ error = xfs_dir3_data_read(tp, dp,
xfs_dir2_db_to_da(mp, use_block),
-1, &dbp);
if (error) {
@@ -583,13 +589,14 @@ xfs_dir2_leaf_addname(
return error;
}
hdr = dbp->b_addr;
+ bf = xfs_dir3_data_bestfree_p(hdr);
grown = 0;
}
/*
* Point to the biggest freespace in our data block.
*/
dup = (xfs_dir2_data_unused_t *)
- ((char *)hdr + be16_to_cpu(hdr->bestfree[0].offset));
+ ((char *)hdr + be16_to_cpu(bf[0].offset));
ASSERT(be16_to_cpu(dup->length) >= length);
needscan = needlog = 0;
/*
@@ -622,8 +629,8 @@ xfs_dir2_leaf_addname(
* If the bests table needs to be changed, do it.
* Log the change unless we've already done that.
*/
- if (be16_to_cpu(bestsp[use_block]) != be16_to_cpu(hdr->bestfree[0].length)) {
- bestsp[use_block] = hdr->bestfree[0].length;
+ if (be16_to_cpu(bestsp[use_block]) != be16_to_cpu(bf[0].length)) {
+ bestsp[use_block] = bf[0].length;
if (!grown)
xfs_dir2_leaf_log_bests(tp, lbp, use_block, use_block);
}
@@ -643,7 +650,7 @@ xfs_dir2_leaf_addname(
xfs_dir2_leaf_log_header(tp, lbp);
xfs_dir2_leaf_log_ents(tp, lbp, lfloglow, lfloghigh);
xfs_dir2_leaf_check(dp, lbp);
- xfs_dir2_data_check(dp, dbp);
+ xfs_dir3_data_check(dp, dbp);
return 0;
}
@@ -967,7 +974,7 @@ xfs_dir2_leaf_readbuf(
* Read the directory block starting at the first mapping.
*/
mip->curdb = xfs_dir2_da_to_db(mp, map->br_startoff);
- error = xfs_dir2_data_read(NULL, dp, map->br_startoff,
+ error = xfs_dir3_data_read(NULL, dp, map->br_startoff,
map->br_blockcount >= mp->m_dirblkfsbs ?
XFS_FSB_TO_DADDR(mp, map->br_startblock) : -1, &bp);
@@ -996,7 +1003,7 @@ xfs_dir2_leaf_readbuf(
*/
if (i > mip->ra_current &&
map[mip->ra_index].br_blockcount >= mp->m_dirblkfsbs) {
- xfs_dir2_data_readahead(NULL, dp,
+ xfs_dir3_data_readahead(NULL, dp,
map[mip->ra_index].br_startoff + mip->ra_offset,
XFS_FSB_TO_DADDR(mp,
map[mip->ra_index].br_startblock +
@@ -1009,7 +1016,7 @@ xfs_dir2_leaf_readbuf(
* use our mapping, but this is a very rare case.
*/
else if (i > mip->ra_current) {
- xfs_dir2_data_readahead(NULL, dp,
+ xfs_dir3_data_readahead(NULL, dp,
map[mip->ra_index].br_startoff +
mip->ra_offset, -1);
mip->ra_current = i;
@@ -1135,7 +1142,7 @@ xfs_dir2_leaf_getdents(
ASSERT(xfs_dir2_byte_to_db(mp, curoff) ==
map_info->curdb);
hdr = bp->b_addr;
- xfs_dir2_data_check(dp, bp);
+ xfs_dir3_data_check(dp, bp);
/*
* Find our position in the block.
*/
@@ -1481,7 +1488,7 @@ xfs_dir2_leaf_lookup_int(
if (newdb != curdb) {
if (dbp)
xfs_trans_brelse(tp, dbp);
- error = xfs_dir2_data_read(tp, dp,
+ error = xfs_dir3_data_read(tp, dp,
xfs_dir2_db_to_da(mp, newdb),
-1, &dbp);
if (error) {
@@ -1522,7 +1529,7 @@ xfs_dir2_leaf_lookup_int(
ASSERT(cidb != -1);
if (cidb != curdb) {
xfs_trans_brelse(tp, dbp);
- error = xfs_dir2_data_read(tp, dp,
+ error = xfs_dir3_data_read(tp, dp,
xfs_dir2_db_to_da(mp, cidb),
-1, &dbp);
if (error) {
@@ -1568,6 +1575,7 @@ xfs_dir2_leaf_removename(
int needscan; /* need to rescan data frees */
xfs_dir2_data_off_t oldbest; /* old value of best free */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir2_data_free *bf; /* bestfree table */
trace_xfs_dir2_leaf_removename(args);
@@ -1582,7 +1590,8 @@ xfs_dir2_leaf_removename(
mp = dp->i_mount;
leaf = lbp->b_addr;
hdr = dbp->b_addr;
- xfs_dir2_data_check(dp, dbp);
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ xfs_dir3_data_check(dp, dbp);
/*
* Point to the leaf entry, use that to point to the data entry.
*/
@@ -1591,7 +1600,7 @@ xfs_dir2_leaf_removename(
dep = (xfs_dir2_data_entry_t *)
((char *)hdr + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address)));
needscan = needlog = 0;
- oldbest = be16_to_cpu(hdr->bestfree[0].length);
+ oldbest = be16_to_cpu(bf[0].length);
ltp = xfs_dir2_leaf_tail_p(mp, leaf);
bestsp = xfs_dir2_leaf_bests_p(ltp);
ASSERT(be16_to_cpu(bestsp[db]) == oldbest);
@@ -1620,16 +1629,15 @@ xfs_dir2_leaf_removename(
* If the longest freespace in the data block has changed,
* put the new value in the bests table and log that.
*/
- if (be16_to_cpu(hdr->bestfree[0].length) != oldbest) {
- bestsp[db] = hdr->bestfree[0].length;
+ if (be16_to_cpu(bf[0].length) != oldbest) {
+ bestsp[db] = bf[0].length;
xfs_dir2_leaf_log_bests(tp, lbp, db, db);
}
- xfs_dir2_data_check(dp, dbp);
+ xfs_dir3_data_check(dp, dbp);
/*
* If the data block is now empty then get rid of the data block.
*/
- if (be16_to_cpu(hdr->bestfree[0].length) ==
- mp->m_dirblksize - (uint)sizeof(*hdr)) {
+ if (be16_to_cpu(bf[0].length) == mp->m_dirblksize - (uint)sizeof(*hdr)) {
ASSERT(db != mp->m_dirdatablk);
if ((error = xfs_dir2_shrink_inode(args, db, dbp))) {
/*
@@ -1809,7 +1817,7 @@ xfs_dir2_leaf_trim_data(
/*
* Read the offending data block. We need its buffer.
*/
- error = xfs_dir2_data_read(tp, dp, xfs_dir2_db_to_da(mp, db), -1, &dbp);
+ error = xfs_dir3_data_read(tp, dp, xfs_dir2_db_to_da(mp, db), -1, &dbp);
if (error)
return error;
@@ -1819,9 +1827,11 @@ xfs_dir2_leaf_trim_data(
#ifdef DEBUG
{
struct xfs_dir2_data_hdr *hdr = dbp->b_addr;
+ struct xfs_dir2_data_free *bf = xfs_dir3_data_bestfree_p(hdr);
- ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC));
- ASSERT(be16_to_cpu(hdr->bestfree[0].length) ==
+ ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC));
+ ASSERT(be16_to_cpu(bf[0].length) ==
mp->m_dirblksize - (uint)sizeof(*hdr));
ASSERT(db == be32_to_cpu(ltp->bestcount) - 1);
}
diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c
index 480228a..3b33d7d 100644
--- a/fs/xfs/xfs_dir2_node.c
+++ b/fs/xfs/xfs_dir2_node.c
@@ -753,13 +753,13 @@ xfs_dir2_leafn_lookup_for_entry(
ASSERT(state->extravalid);
curbp = state->extrablk.bp;
} else {
- error = xfs_dir2_data_read(tp, dp,
+ error = xfs_dir3_data_read(tp, dp,
xfs_dir2_db_to_da(mp, newdb),
-1, &curbp);
if (error)
return error;
}
- xfs_dir2_data_check(dp, curbp);
+ xfs_dir3_data_check(dp, curbp);
curdb = newdb;
}
/*
@@ -787,7 +787,7 @@ xfs_dir2_leafn_lookup_for_entry(
state->extrablk.index = (int)((char *)dep -
(char *)curbp->b_addr);
state->extrablk.magic = XFS_DIR2_DATA_MAGIC;
- curbp->b_ops = &xfs_dir2_data_buf_ops;
+ curbp->b_ops = &xfs_dir3_data_buf_ops;
if (cmp == XFS_CMP_EXACT)
return XFS_ERROR(EEXIST);
}
@@ -802,7 +802,7 @@ xfs_dir2_leafn_lookup_for_entry(
state->extrablk.index = -1;
state->extrablk.blkno = curdb;
state->extrablk.magic = XFS_DIR2_DATA_MAGIC;
- curbp->b_ops = &xfs_dir2_data_buf_ops;
+ curbp->b_ops = &xfs_dir3_data_buf_ops;
} else {
/* If the curbp is not the CI match block, drop it */
if (state->extrablk.bp != curbp)
@@ -1152,6 +1152,7 @@ xfs_dir2_leafn_remove(
int needlog; /* need to log data header */
int needscan; /* need to rescan data frees */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir2_data_free *bf; /* bestfree table */
trace_xfs_dir2_leafn_remove(args, index);
@@ -1186,7 +1187,8 @@ xfs_dir2_leafn_remove(
dbp = dblk->bp;
hdr = dbp->b_addr;
dep = (xfs_dir2_data_entry_t *)((char *)hdr + off);
- longest = be16_to_cpu(hdr->bestfree[0].length);
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ longest = be16_to_cpu(bf[0].length);
needlog = needscan = 0;
xfs_dir2_data_make_free(tp, dbp, off,
xfs_dir2_data_entsize(dep->namelen), &needlog, &needscan);
@@ -1198,12 +1200,12 @@ xfs_dir2_leafn_remove(
xfs_dir2_data_freescan(mp, hdr, &needlog);
if (needlog)
xfs_dir2_data_log_header(tp, dbp);
- xfs_dir2_data_check(dp, dbp);
+ xfs_dir3_data_check(dp, dbp);
/*
* If the longest data block freespace changes, need to update
* the corresponding freeblock entry.
*/
- if (longest < be16_to_cpu(hdr->bestfree[0].length)) {
+ if (longest < be16_to_cpu(bf[0].length)) {
int error; /* error return value */
struct xfs_buf *fbp; /* freeblock buffer */
xfs_dir2_db_t fdb; /* freeblock block number */
@@ -1233,7 +1235,7 @@ xfs_dir2_leafn_remove(
* Calculate which entry we need to fix.
*/
findex = xfs_dir2_db_to_fdindex(mp, db);
- longest = be16_to_cpu(hdr->bestfree[0].length);
+ longest = be16_to_cpu(bf[0].length);
/*
* If the data block is now empty we can get rid of it
* (usually).
@@ -1612,6 +1614,7 @@ xfs_dir2_node_addname_int(
xfs_trans_t *tp; /* transaction pointer */
__be16 *bests;
struct xfs_dir3_icfree_hdr freehdr;
+ struct xfs_dir2_data_free *bf;
dp = args->dp;
mp = dp->i_mount;
@@ -1869,7 +1872,8 @@ xfs_dir2_node_addname_int(
* change again.
*/
hdr = dbp->b_addr;
- bests[findex] = hdr->bestfree[0].length;
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ bests[findex] = bf[0].length;
logfree = 1;
}
/*
@@ -1885,19 +1889,20 @@ xfs_dir2_node_addname_int(
/*
* Read the data block in.
*/
- error = xfs_dir2_data_read(tp, dp, xfs_dir2_db_to_da(mp, dbno),
+ error = xfs_dir3_data_read(tp, dp, xfs_dir2_db_to_da(mp, dbno),
-1, &dbp);
if (error)
return error;
hdr = dbp->b_addr;
+ bf = xfs_dir3_data_bestfree_p(hdr);
logfree = 0;
}
- ASSERT(be16_to_cpu(hdr->bestfree[0].length) >= length);
+ ASSERT(be16_to_cpu(bf[0].length) >= length);
/*
* Point to the existing unused space.
*/
dup = (xfs_dir2_data_unused_t *)
- ((char *)hdr + be16_to_cpu(hdr->bestfree[0].offset));
+ ((char *)hdr + be16_to_cpu(bf[0].offset));
needscan = needlog = 0;
/*
* Mark the first part of the unused space, inuse for us.
@@ -1929,8 +1934,8 @@ xfs_dir2_node_addname_int(
* If the freespace entry is now wrong, update it.
*/
bests = xfs_dir3_free_bests_p(mp, free); /* gcc is so stupid */
- if (be16_to_cpu(bests[findex]) != be16_to_cpu(hdr->bestfree[0].length)) {
- bests[findex] = hdr->bestfree[0].length;
+ if (be16_to_cpu(bests[findex]) != be16_to_cpu(bf[0].length)) {
+ bests[findex] = bf[0].length;
logfree = 1;
}
/*
@@ -2120,7 +2125,8 @@ xfs_dir2_node_replace(
* Point to the data entry.
*/
hdr = state->extrablk.bp->b_addr;
- ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC));
+ ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC) ||
+ hdr->magic == cpu_to_be32(XFS_DIR3_DATA_MAGIC));
dep = (xfs_dir2_data_entry_t *)
((char *)hdr +
xfs_dir2_dataptr_to_off(state->mp, be32_to_cpu(lep->address)));
diff --git a/fs/xfs/xfs_dir2_priv.h b/fs/xfs/xfs_dir2_priv.h
index e6f2e0a..910e644 100644
--- a/fs/xfs/xfs_dir2_priv.h
+++ b/fs/xfs/xfs_dir2_priv.h
@@ -43,17 +43,17 @@ extern int xfs_dir2_leaf_to_block(struct xfs_da_args *args,
/* xfs_dir2_data.c */
#ifdef DEBUG
-#define xfs_dir2_data_check(dp,bp) __xfs_dir2_data_check(dp, bp);
+#define xfs_dir3_data_check(dp,bp) __xfs_dir3_data_check(dp, bp);
#else
-#define xfs_dir2_data_check(dp,bp)
+#define xfs_dir3_data_check(dp,bp)
#endif
-extern const struct xfs_buf_ops xfs_dir2_data_buf_ops;
+extern const struct xfs_buf_ops xfs_dir3_data_buf_ops;
-extern int __xfs_dir2_data_check(struct xfs_inode *dp, struct xfs_buf *bp);
-extern int xfs_dir2_data_read(struct xfs_trans *tp, struct xfs_inode *dp,
+extern int __xfs_dir3_data_check(struct xfs_inode *dp, struct xfs_buf *bp);
+extern int xfs_dir3_data_read(struct xfs_trans *tp, struct xfs_inode *dp,
xfs_dablk_t bno, xfs_daddr_t mapped_bno, struct xfs_buf **bpp);
-extern int xfs_dir2_data_readahead(struct xfs_trans *tp, struct xfs_inode *dp,
+extern int xfs_dir3_data_readahead(struct xfs_trans *tp, struct xfs_inode *dp,
xfs_dablk_t bno, xfs_daddr_t mapped_bno);
extern struct xfs_dir2_data_free *
diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 21b6f0b..967dad6 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -915,7 +915,7 @@ xfs_dir_open(
*/
mode = xfs_ilock_map_shared(ip);
if (ip->i_d.di_nextents > 0)
- xfs_dir2_data_readahead(NULL, ip, 0, -1);
+ xfs_dir3_data_readahead(NULL, ip, 0, -1);
xfs_iunlock(ip, mode);
return 0;
}
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 13/16] xfs: add CRC checking to dir2 leaf blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (11 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 12/16] xfs: add CRC checking to dir2 data blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 14/16] xfs: add CRCs to dir2/da node blocks Dave Chinner
` (2 subsequent siblings)
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
This addition follows the same pattern as the dir2 block CRCs.
Seeing as both LEAF1 and LEAFN types need to changed at the same
time, this is a pretty large amount of change. leaf block headers
neeed to be abstracted away from the on-disk structures (struct
xfs_dir3_icleaf_hdr), as do the base leaf entry locations.
This header abstract allows the in-core header and leaf entry
location to be passed around instead of the leaf block itself. This
saves a lot of converting individual variables from on-disk format
to host format where they are used, so there's a good chance that
the compiler will be able to produce much more optimal code as it's
not having to byteswap variables all over the place.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_da_btree.c | 42 ++-
fs/xfs/xfs_da_btree.h | 19 ++
fs/xfs/xfs_dir2_block.c | 17 +-
fs/xfs/xfs_dir2_format.h | 60 +++-
fs/xfs/xfs_dir2_leaf.c | 777 +++++++++++++++++++++++++++++-----------------
fs/xfs/xfs_dir2_node.c | 455 +++++++++++++++------------
fs/xfs/xfs_dir2_priv.h | 32 +-
7 files changed, 889 insertions(+), 513 deletions(-)
diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c
index 4d7696a..ebfb14a 100644
--- a/fs/xfs/xfs_da_btree.c
+++ b/fs/xfs/xfs_da_btree.c
@@ -139,7 +139,8 @@ xfs_da_node_read_verify(
bp->b_ops->verify_read(bp);
return;
case XFS_DIR2_LEAFN_MAGIC:
- bp->b_ops = &xfs_dir2_leafn_buf_ops;
+ case XFS_DIR3_LEAFN_MAGIC:
+ bp->b_ops = &xfs_dir3_leafn_buf_ops;
bp->b_ops->verify_read(bp);
return;
default:
@@ -396,11 +397,18 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
size = (int)((char *)&oldroot->btree[be16_to_cpu(oldroot->hdr.count)] -
(char *)oldroot);
} else {
- ASSERT(oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+ struct xfs_dir3_icleaf_hdr leafhdr;
+ struct xfs_dir2_leaf_entry *ents;
+
leaf = (xfs_dir2_leaf_t *)oldroot;
- size = (int)((char *)&leaf->ents[be16_to_cpu(leaf->hdr.count)] -
- (char *)leaf);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+
+ ASSERT(leafhdr.magic == XFS_DIR2_LEAFN_MAGIC ||
+ leafhdr.magic == XFS_DIR3_LEAFN_MAGIC);
+ size = (int)((char *)&ents[leafhdr.count] - (char *)leaf);
}
+ /* XXX: can't just copy CRC headers from one block to another */
memcpy(node, oldroot, size);
xfs_trans_log_buf(tp, bp, 0, size - 1);
@@ -424,7 +432,8 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
node->hdr.count = cpu_to_be16(2);
#ifdef DEBUG
- if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)) {
+ if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+ oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC)) {
ASSERT(blk1->blkno >= mp->m_dirleafblk &&
blk1->blkno < mp->m_dirfreeblk);
ASSERT(blk2->blkno >= mp->m_dirleafblk &&
@@ -782,6 +791,7 @@ xfs_da_blkinfo_onlychild_validate(struct xfs_da_blkinfo *blkinfo, __u16 level)
if (level == 1) {
ASSERT(magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+ magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC) ||
magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
} else
ASSERT(magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
@@ -1565,6 +1575,7 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
info = blk->bp->b_addr;
ASSERT(info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+ info->magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC) ||
info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
blk->magic = be16_to_cpu(info->magic);
if (blk->magic == XFS_DA_NODE_MAGIC) {
@@ -1588,8 +1599,7 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
NULL);
break;
default:
- ASSERT(blk->magic == XFS_ATTR_LEAF_MAGIC ||
- blk->magic == XFS_DIR2_LEAFN_MAGIC);
+ ASSERT(0);
break;
}
}
@@ -1833,10 +1843,16 @@ xfs_da_swap_lastblock(
/*
* Get values from the moved block.
*/
- if (dead_info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)) {
+ if (dead_info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+ dead_info->magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC)) {
+ struct xfs_dir3_icleaf_hdr leafhdr;
+ struct xfs_dir2_leaf_entry *ents;
+
dead_leaf2 = (xfs_dir2_leaf_t *)dead_info;
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, dead_leaf2);
+ ents = xfs_dir3_leaf_ents_p(dead_leaf2);
dead_level = 0;
- dead_hash = be32_to_cpu(dead_leaf2->ents[be16_to_cpu(dead_leaf2->hdr.count) - 1].hashval);
+ dead_hash = be32_to_cpu(ents[leafhdr.count - 1].hashval);
} else {
ASSERT(dead_info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
dead_node = (xfs_da_intnode_t *)dead_info;
@@ -2282,9 +2298,15 @@ xfs_da_read_buf(
(magic != XFS_ATTR_LEAF_MAGIC) &&
(magic != XFS_DIR2_LEAF1_MAGIC) &&
(magic != XFS_DIR2_LEAFN_MAGIC) &&
+ (magic != XFS_DIR3_LEAFN_MAGIC) &&
(magic1 != XFS_DIR2_BLOCK_MAGIC) &&
+ (magic1 != XFS_DIR3_BLOCK_MAGIC) &&
(magic1 != XFS_DIR2_DATA_MAGIC) &&
- (free->hdr.magic != cpu_to_be32(XFS_DIR2_FREE_MAGIC)),
+ (magic1 != XFS_DIR3_DATA_MAGIC) &&
+ (free->hdr.magic !=
+ cpu_to_be32(XFS_DIR2_FREE_MAGIC)) &&
+ (free->hdr.magic !=
+ cpu_to_be32(XFS_DIR3_FREE_MAGIC)),
mp, XFS_ERRTAG_DA_READ_BUF,
XFS_RANDOM_DA_READ_BUF))) {
trace_xfs_da_btree_corrupt(bp, _RET_IP_);
diff --git a/fs/xfs/xfs_da_btree.h b/fs/xfs/xfs_da_btree.h
index ee5170c..d408d9b 100644
--- a/fs/xfs/xfs_da_btree.h
+++ b/fs/xfs/xfs_da_btree.h
@@ -47,6 +47,25 @@ typedef struct xfs_da_blkinfo {
} xfs_da_blkinfo_t;
/*
+ * CRC enabled directory structure types
+ *
+ * The headers change size for the additional verification information, but
+ * otherwise the tree layouts and contents are unchanged.
+ */
+#define XFS_DIR3_LEAF1_MAGIC 0x3df1 /* magic number: v2 dirlf single blks */
+#define XFS_DIR3_LEAFN_MAGIC 0x3dff /* magic number: v2 dirlf multi blks */
+
+struct xfs_da3_blkinfo {
+/* XXX: check packing due to size of xfs_da_blkinfo */
+ struct xfs_da_blkinfo hdr;
+ __be32 crc; /* CRC of block */
+ __be64 blkno; /* first block of the buffer */
+ __be64 lsn; /* sequence number of last write */
+ uuid_t uuid; /* filesystem we belong to */
+ __be64 owner; /* inode that owns the block */
+};
+
+/*
* This is the structure of the root and intermediate nodes in the Btree.
* The leaf nodes are defined above.
*
diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c
index 175db20..af721a3 100644
--- a/fs/xfs/xfs_dir2_block.c
+++ b/fs/xfs/xfs_dir2_block.c
@@ -1012,6 +1012,8 @@ xfs_dir2_leaf_to_block(
__be16 *tagp; /* end of entry (tag) */
int to; /* block/leaf to index */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
trace_xfs_dir2_leaf_to_block(args);
@@ -1019,8 +1021,12 @@ xfs_dir2_leaf_to_block(
tp = args->trans;
mp = dp->i_mount;
leaf = lbp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
ltp = xfs_dir2_leaf_tail_p(mp, leaf);
+
+ ASSERT(leafhdr.magic == XFS_DIR2_LEAF1_MAGIC ||
+ leafhdr.magic == XFS_DIR3_LEAF1_MAGIC);
/*
* If there are data blocks other than the first one, take this
* opportunity to remove trailing empty data blocks that may have
@@ -1083,18 +1089,17 @@ xfs_dir2_leaf_to_block(
* Initialize the block tail.
*/
btp = xfs_dir2_block_tail_p(mp, hdr);
- btp->count = cpu_to_be32(be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale));
+ btp->count = cpu_to_be32(leafhdr.count - leafhdr.stale);
btp->stale = 0;
xfs_dir2_block_log_tail(tp, dbp);
/*
* Initialize the block leaf area. We compact out stale entries.
*/
lep = xfs_dir2_block_leaf_p(btp);
- for (from = to = 0; from < be16_to_cpu(leaf->hdr.count); from++) {
- if (leaf->ents[from].address ==
- cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
+ for (from = to = 0; from < leafhdr.count; from++) {
+ if (ents[from].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
continue;
- lep[to++] = leaf->ents[from];
+ lep[to++] = ents[from];
}
ASSERT(to == be32_to_cpu(btp->count));
xfs_dir2_block_log_leaf(tp, dbp, 0, be32_to_cpu(btp->count) - 1);
diff --git a/fs/xfs/xfs_dir2_format.h b/fs/xfs/xfs_dir2_format.h
index 358efab..2e3d66e 100644
--- a/fs/xfs/xfs_dir2_format.h
+++ b/fs/xfs/xfs_dir2_format.h
@@ -467,6 +467,21 @@ typedef struct xfs_dir2_leaf_hdr {
__be16 stale; /* count of stale entries */
} xfs_dir2_leaf_hdr_t;
+struct xfs_dir3_leaf_hdr {
+ struct xfs_da3_blkinfo info; /* header for da routines */
+ __be16 count; /* count of entries */
+ __be16 stale; /* count of stale entries */
+ __be32 pad;
+};
+
+struct xfs_dir3_icleaf_hdr {
+ __uint32_t forw;
+ __uint32_t back;
+ __uint16_t magic;
+ __uint16_t count;
+ __uint16_t stale;
+};
+
/*
* Leaf block entry.
*/
@@ -486,23 +501,50 @@ typedef struct xfs_dir2_leaf_tail {
* Leaf block.
*/
typedef struct xfs_dir2_leaf {
- xfs_dir2_leaf_hdr_t hdr; /* leaf header */
- xfs_dir2_leaf_entry_t ents[]; /* entries */
+ xfs_dir2_leaf_hdr_t hdr; /* leaf header */
+ xfs_dir2_leaf_entry_t __ents[]; /* entries */
} xfs_dir2_leaf_t;
-/*
- * DB blocks here are logical directory block numbers, not filesystem blocks.
- */
+struct xfs_dir3_leaf {
+ struct xfs_dir3_leaf_hdr hdr; /* leaf header */
+ struct xfs_dir2_leaf_entry __ents[]; /* entries */
+};
+
+#define XFS_DIR3_LEAF_CRC_OFF offsetof(struct xfs_dir3_leaf_hdr, info.crc)
+
+static inline int
+xfs_dir3_leaf_hdr_size(struct xfs_dir2_leaf *lp)
+{
+ if (lp->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAF1_MAGIC) ||
+ lp->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC))
+ return sizeof(struct xfs_dir3_leaf_hdr);
+ return sizeof(struct xfs_dir2_leaf_hdr);
+}
-static inline int xfs_dir2_max_leaf_ents(struct xfs_mount *mp)
+static inline int
+xfs_dir3_max_leaf_ents(struct xfs_mount *mp, struct xfs_dir2_leaf *lp)
{
- return (mp->m_dirblksize - (uint)sizeof(struct xfs_dir2_leaf_hdr)) /
+ return (mp->m_dirblksize - xfs_dir3_leaf_hdr_size(lp)) /
(uint)sizeof(struct xfs_dir2_leaf_entry);
}
/*
* Get address of the bestcount field in the single-leaf block.
*/
+static inline struct xfs_dir2_leaf_entry *
+xfs_dir3_leaf_ents_p(struct xfs_dir2_leaf *lp)
+{
+ if (lp->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAF1_MAGIC) ||
+ lp->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC)) {
+ struct xfs_dir3_leaf *lp3 = (struct xfs_dir3_leaf *)lp;
+ return lp3->__ents;
+ }
+ return lp->__ents;
+}
+
+/*
+ * Get address of the bestcount field in the single-leaf block.
+ */
static inline struct xfs_dir2_leaf_tail *
xfs_dir2_leaf_tail_p(struct xfs_mount *mp, struct xfs_dir2_leaf *lp)
{
@@ -521,6 +563,10 @@ xfs_dir2_leaf_bests_p(struct xfs_dir2_leaf_tail *ltp)
}
/*
+ * DB blocks here are logical directory block numbers, not filesystem blocks.
+ */
+
+/*
* Convert dataptr to byte in file space
*/
static inline xfs_dir2_off_t
diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c
index 78d3182..57d81ce 100644
--- a/fs/xfs/xfs_dir2_leaf.c
+++ b/fs/xfs/xfs_dir2_leaf.c
@@ -33,77 +33,258 @@
#include "xfs_dir2_priv.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_buf_item.h"
+#include "xfs_cksum.h"
/*
* Local function declarations.
*/
#ifdef DEBUG
-static void xfs_dir2_leaf_check(struct xfs_inode *dp, struct xfs_buf *bp);
-#else
-#define xfs_dir2_leaf_check(dp, bp)
+#define xfs_dir3_leaf_check(mp, bp) \
+do { \
+ if (!xfs_dir3_leaf1_check((mp), (bp))) \
+ ASSERT(0); \
+} while (0);
#endif
+
static int xfs_dir2_leaf_lookup_int(xfs_da_args_t *args, struct xfs_buf **lbpp,
int *indexp, struct xfs_buf **dbpp);
-static void xfs_dir2_leaf_log_bests(struct xfs_trans *tp, struct xfs_buf *bp,
+static void xfs_dir3_leaf_log_bests(struct xfs_trans *tp, struct xfs_buf *bp,
int first, int last);
-static void xfs_dir2_leaf_log_tail(struct xfs_trans *tp, struct xfs_buf *bp);
+static void xfs_dir3_leaf_log_tail(struct xfs_trans *tp, struct xfs_buf *bp);
-static void
-xfs_dir2_leaf_verify(
+void
+xfs_dir3_leaf_hdr_from_disk(
+ struct xfs_dir3_icleaf_hdr *to,
+ struct xfs_dir2_leaf *from)
+{
+ if (from->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC) ||
+ from->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)) {
+ to->forw = be32_to_cpu(from->hdr.info.forw);
+ to->back = be32_to_cpu(from->hdr.info.back);
+ to->magic = be16_to_cpu(from->hdr.info.magic);
+ to->count = be16_to_cpu(from->hdr.count);
+ to->stale = be16_to_cpu(from->hdr.stale);
+ } else {
+ struct xfs_dir3_leaf_hdr *hdr3 = (struct xfs_dir3_leaf_hdr *)from;
+
+ to->forw = be32_to_cpu(hdr3->info.hdr.forw);
+ to->back = be32_to_cpu(hdr3->info.hdr.back);
+ to->magic = be16_to_cpu(hdr3->info.hdr.magic);
+ to->count = be16_to_cpu(hdr3->count);
+ to->stale = be16_to_cpu(hdr3->stale);
+ }
+
+ ASSERT(to->magic == XFS_DIR2_LEAF1_MAGIC ||
+ to->magic == XFS_DIR3_LEAF1_MAGIC ||
+ to->magic == XFS_DIR2_LEAFN_MAGIC ||
+ to->magic == XFS_DIR3_LEAFN_MAGIC);
+}
+
+void
+xfs_dir3_leaf_hdr_to_disk(
+ struct xfs_dir2_leaf *to,
+ struct xfs_dir3_icleaf_hdr *from)
+{
+ ASSERT(from->magic == XFS_DIR2_LEAF1_MAGIC ||
+ from->magic == XFS_DIR3_LEAF1_MAGIC ||
+ from->magic == XFS_DIR2_LEAFN_MAGIC ||
+ from->magic == XFS_DIR3_LEAFN_MAGIC);
+
+ if (from->magic == XFS_DIR2_LEAF1_MAGIC ||
+ from->magic == XFS_DIR2_LEAFN_MAGIC) {
+ to->hdr.info.forw = cpu_to_be32(from->forw);
+ to->hdr.info.back = cpu_to_be32(from->back);
+ to->hdr.info.magic = cpu_to_be16(from->magic);
+ to->hdr.count = cpu_to_be16(from->count);
+ to->hdr.stale = cpu_to_be16(from->stale);
+ } else {
+ struct xfs_dir3_leaf_hdr *hdr3 = (struct xfs_dir3_leaf_hdr *)to;
+
+ hdr3->info.hdr.forw = cpu_to_be32(from->forw);
+ hdr3->info.hdr.back = cpu_to_be32(from->back);
+ hdr3->info.hdr.magic = cpu_to_be16(from->magic);
+ hdr3->count = cpu_to_be16(from->count);
+ hdr3->stale = cpu_to_be16(from->stale);
+ }
+}
+
+bool
+xfs_dir3_leaf_check_int(
+ struct xfs_mount *mp,
+ struct xfs_dir3_icleaf_hdr *hdr,
+ struct xfs_dir2_leaf *leaf)
+{
+ struct xfs_dir2_leaf_entry *ents;
+ xfs_dir2_leaf_tail_t *ltp;
+ int stale;
+ int i;
+
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ ltp = xfs_dir2_leaf_tail_p(mp, leaf);
+
+ /*
+ * This value is not restrictive enough.
+ * Should factor in the size of the bests table as well.
+ * We can deduce a value for that from di_size.
+ */
+ if (hdr->count > xfs_dir3_max_leaf_ents(mp, leaf))
+ return false;
+
+ /* Leaves and bests don't overlap in leaf format. */
+ if ((hdr->magic == XFS_DIR2_LEAF1_MAGIC ||
+ hdr->magic == XFS_DIR3_LEAF1_MAGIC) &&
+ (char *)&ents[hdr->count] > (char *)xfs_dir2_leaf_bests_p(ltp))
+ return false;
+
+ /* Check hash value order, count stale entries. */
+ for (i = stale = 0; i < hdr->count; i++) {
+ if (i + 1 < hdr->count) {
+ if (be32_to_cpu(ents[i].hashval) >
+ be32_to_cpu(ents[i + 1].hashval))
+ return false;
+ }
+ if (ents[i].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
+ stale++;
+ }
+ if (hdr->stale != stale)
+ return false;
+ return true;
+}
+
+/*
+ * Check the internal consistency of a leaf1 block.
+ * Pop an assert if something is wrong.
+ */
+STATIC bool
+xfs_dir3_leaf1_check(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp)
+{
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
+ struct xfs_dir3_icleaf_hdr leafhdr;
+
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+
+ if (leafhdr.magic != XFS_DIR2_LEAF1_MAGIC &&
+ leafhdr.magic != XFS_DIR3_LEAF1_MAGIC)
+ return false;
+
+ return xfs_dir3_leaf_check_int(mp, &leafhdr, leaf);
+}
+
+static bool
+xfs_dir3_leaf_verify(
struct xfs_buf *bp,
__be16 magic)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
- struct xfs_dir2_leaf_hdr *hdr = bp->b_addr;
- int block_ok = 0;
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
+ struct xfs_dir3_icleaf_hdr leafhdr;
+
+ ASSERT(magic == XFS_DIR2_LEAF1_MAGIC || magic == XFS_DIR2_LEAFN_MAGIC);
+
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_dir3_leaf_hdr *leaf3 = bp->b_addr;
+
+ if ((magic == XFS_DIR2_LEAF1_MAGIC &&
+ leafhdr.magic != XFS_DIR3_LEAF1_MAGIC) ||
+ (magic == XFS_DIR2_LEAFN_MAGIC &&
+ leafhdr.magic != XFS_DIR3_LEAFN_MAGIC))
+ return false;
+
+ if (!uuid_equal(&leaf3->info.uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be64_to_cpu(leaf3->info.blkno) != bp->b_bn)
+ return false;
+ } else {
+ if (leafhdr.magic != magic)
+ return false;
+ }
+ return xfs_dir3_leaf_check_int(mp, &leafhdr, leaf);
+}
+
+static void
+__read_verify(
+ struct xfs_buf *bp,
+ __uint16_t magic)
+{
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ if ((xfs_sb_version_hascrc(&mp->m_sb) &&
+ !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_DIR3_LEAF_CRC_OFF)) ||
+ !xfs_dir3_leaf_verify(bp, magic)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
+}
+
+static void
+__write_verify(
+ struct xfs_buf *bp,
+ __uint16_t magic)
+{
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+ struct xfs_dir3_leaf_hdr *hdr3 = bp->b_addr;
- block_ok = hdr->info.magic == magic;
- if (!block_ok) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
+ if (!xfs_dir3_leaf_verify(bp, magic)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
}
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ hdr3->info.lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DIR3_LEAF_CRC_OFF);
}
static void
-xfs_dir2_leaf1_read_verify(
+xfs_dir3_leaf1_read_verify(
struct xfs_buf *bp)
{
- xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
+ __read_verify(bp, XFS_DIR2_LEAF1_MAGIC);
}
static void
-xfs_dir2_leaf1_write_verify(
+xfs_dir3_leaf1_write_verify(
struct xfs_buf *bp)
{
- xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
+ __write_verify(bp, XFS_DIR2_LEAF1_MAGIC);
}
-void
-xfs_dir2_leafn_read_verify(
+static void
+xfs_dir3_leafn_read_verify(
struct xfs_buf *bp)
{
- xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+ __read_verify(bp, XFS_DIR2_LEAFN_MAGIC);
}
-void
-xfs_dir2_leafn_write_verify(
+static void
+xfs_dir3_leafn_write_verify(
struct xfs_buf *bp)
{
- xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+ __write_verify(bp, XFS_DIR2_LEAFN_MAGIC);
}
-static const struct xfs_buf_ops xfs_dir2_leaf1_buf_ops = {
- .verify_read = xfs_dir2_leaf1_read_verify,
- .verify_write = xfs_dir2_leaf1_write_verify,
+static const struct xfs_buf_ops xfs_dir3_leaf1_buf_ops = {
+ .verify_read = xfs_dir3_leaf1_read_verify,
+ .verify_write = xfs_dir3_leaf1_write_verify,
};
-const struct xfs_buf_ops xfs_dir2_leafn_buf_ops = {
- .verify_read = xfs_dir2_leafn_read_verify,
- .verify_write = xfs_dir2_leafn_write_verify,
+const struct xfs_buf_ops xfs_dir3_leafn_buf_ops = {
+ .verify_read = xfs_dir3_leafn_read_verify,
+ .verify_write = xfs_dir3_leafn_write_verify,
};
static int
-xfs_dir2_leaf_read(
+xfs_dir3_leaf_read(
struct xfs_trans *tp,
struct xfs_inode *dp,
xfs_dablk_t fbno,
@@ -111,11 +292,11 @@ xfs_dir2_leaf_read(
struct xfs_buf **bpp)
{
return xfs_da_read_buf(tp, dp, fbno, mappedbno, bpp,
- XFS_DATA_FORK, &xfs_dir2_leaf1_buf_ops);
+ XFS_DATA_FORK, &xfs_dir3_leaf1_buf_ops);
}
int
-xfs_dir2_leafn_read(
+xfs_dir3_leafn_read(
struct xfs_trans *tp,
struct xfs_inode *dp,
xfs_dablk_t fbno,
@@ -123,7 +304,81 @@ xfs_dir2_leafn_read(
struct xfs_buf **bpp)
{
return xfs_da_read_buf(tp, dp, fbno, mappedbno, bpp,
- XFS_DATA_FORK, &xfs_dir2_leafn_buf_ops);
+ XFS_DATA_FORK, &xfs_dir3_leafn_buf_ops);
+}
+
+/*
+ * Initialize a new leaf block, leaf1 or leafn magic accepted.
+ */
+static void
+xfs_dir3_leaf_init(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp,
+ xfs_ino_t owner,
+ __uint16_t type)
+{
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
+
+ ASSERT(type == XFS_DIR2_LEAF1_MAGIC || type == XFS_DIR2_LEAFN_MAGIC);
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_dir3_leaf_hdr *leaf3 = bp->b_addr;
+
+ memset(leaf3, 0, sizeof(*leaf3));
+
+ leaf3->info.hdr.magic = (type == XFS_DIR2_LEAF1_MAGIC)
+ ? cpu_to_be16(XFS_DIR3_LEAF1_MAGIC)
+ : cpu_to_be16(XFS_DIR3_LEAFN_MAGIC);
+ leaf3->info.blkno = cpu_to_be64(bp->b_bn);
+ leaf3->info.owner = cpu_to_be64(owner);
+ uuid_copy(&leaf3->info.uuid, &mp->m_sb.sb_uuid);
+ } else {
+ memset(leaf, 0, sizeof(*leaf));
+ leaf->hdr.info.magic = cpu_to_be16(type);
+ }
+
+ /*
+ * If it's a leaf-format directory initialize the tail.
+ * Caller is responsible for initialising the bests table.
+ */
+ if (type == XFS_DIR2_LEAF1_MAGIC) {
+ struct xfs_dir2_leaf_tail *ltp;
+
+ ltp = xfs_dir2_leaf_tail_p(mp, leaf);
+ ltp->bestcount = 0;
+ bp->b_ops = &xfs_dir3_leaf1_buf_ops;
+ } else
+ bp->b_ops = &xfs_dir3_leafn_buf_ops;
+}
+
+int
+xfs_dir3_leaf_get_buf(
+ xfs_da_args_t *args,
+ xfs_dir2_db_t bno,
+ struct xfs_buf **bpp,
+ __uint16_t magic)
+{
+ struct xfs_inode *dp = args->dp;
+ struct xfs_trans *tp = args->trans;
+ struct xfs_mount *mp = dp->i_mount;
+ struct xfs_buf *bp;
+ int error;
+
+ ASSERT(magic == XFS_DIR2_LEAF1_MAGIC || magic == XFS_DIR2_LEAFN_MAGIC);
+ ASSERT(bno >= XFS_DIR2_LEAF_FIRSTDB(mp) &&
+ bno < XFS_DIR2_FREE_FIRSTDB(mp));
+
+ error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, bno), -1, &bp,
+ XFS_DATA_FORK);
+ if (error)
+ return error;
+
+ xfs_dir3_leaf_init(mp, bp, dp->i_ino, magic);
+ xfs_dir3_leaf_log_header(tp, bp);
+ if (magic == XFS_DIR2_LEAF1_MAGIC)
+ xfs_dir3_leaf_log_tail(tp, bp);
+ *bpp = bp;
+ return 0;
}
/*
@@ -150,6 +405,8 @@ xfs_dir2_block_to_leaf(
int needscan; /* need to rescan bestfree */
xfs_trans_t *tp; /* transaction pointer */
struct xfs_dir2_data_free *bf;
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
trace_xfs_dir2_block_to_leaf(args);
@@ -169,27 +426,34 @@ xfs_dir2_block_to_leaf(
/*
* Initialize the leaf block, get a buffer for it.
*/
- if ((error = xfs_dir2_leaf_init(args, ldb, &lbp, XFS_DIR2_LEAF1_MAGIC))) {
+ error = xfs_dir3_leaf_get_buf(args, ldb, &lbp, XFS_DIR2_LEAF1_MAGIC);
+ if (error)
return error;
- }
- ASSERT(lbp != NULL);
+
leaf = lbp->b_addr;
hdr = dbp->b_addr;
xfs_dir3_data_check(dp, dbp);
btp = xfs_dir2_block_tail_p(mp, hdr);
blp = xfs_dir2_block_leaf_p(btp);
bf = xfs_dir3_data_bestfree_p(hdr);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+
/*
* Set the counts in the leaf header.
+ *
+ * XXX: are these actually logged, or just gathered by chance?
*/
- leaf->hdr.count = cpu_to_be16(be32_to_cpu(btp->count));
- leaf->hdr.stale = cpu_to_be16(be32_to_cpu(btp->stale));
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ leafhdr.count = be32_to_cpu(btp->count);
+ leafhdr.stale = be32_to_cpu(btp->stale);
+ xfs_dir3_leaf_hdr_to_disk(leaf, &leafhdr);
+
/*
* Could compact these but I think we always do the conversion
* after squeezing out stale entries.
*/
- memcpy(leaf->ents, blp, be32_to_cpu(btp->count) * sizeof(xfs_dir2_leaf_entry_t));
- xfs_dir2_leaf_log_ents(tp, lbp, 0, be16_to_cpu(leaf->hdr.count) - 1);
+ memcpy(ents, blp, be32_to_cpu(btp->count) * sizeof(xfs_dir2_leaf_entry_t));
+ xfs_dir3_leaf_log_ents(tp, lbp, 0, leafhdr.count - 1);
needscan = 0;
needlog = 1;
/*
@@ -224,15 +488,16 @@ xfs_dir2_block_to_leaf(
*/
if (needlog)
xfs_dir2_data_log_header(tp, dbp);
- xfs_dir2_leaf_check(dp, lbp);
+ xfs_dir3_leaf_check(mp, lbp);
xfs_dir3_data_check(dp, dbp);
- xfs_dir2_leaf_log_bests(tp, lbp, 0, 0);
+ xfs_dir3_leaf_log_bests(tp, lbp, 0, 0);
return 0;
}
STATIC void
-xfs_dir2_leaf_find_stale(
- struct xfs_dir2_leaf *leaf,
+xfs_dir3_leaf_find_stale(
+ struct xfs_dir3_icleaf_hdr *leafhdr,
+ struct xfs_dir2_leaf_entry *ents,
int index,
int *lowstale,
int *highstale)
@@ -241,7 +506,7 @@ xfs_dir2_leaf_find_stale(
* Find the first stale entry before our index, if any.
*/
for (*lowstale = index - 1; *lowstale >= 0; --*lowstale) {
- if (leaf->ents[*lowstale].address ==
+ if (ents[*lowstale].address ==
cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
break;
}
@@ -251,10 +516,8 @@ xfs_dir2_leaf_find_stale(
* Stop if the result would require moving more entries than using
* lowstale.
*/
- for (*highstale = index;
- *highstale < be16_to_cpu(leaf->hdr.count);
- ++*highstale) {
- if (leaf->ents[*highstale].address ==
+ for (*highstale = index; *highstale < leafhdr->count; ++*highstale) {
+ if (ents[*highstale].address ==
cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
break;
if (*lowstale >= 0 && index - *lowstale <= *highstale - index)
@@ -263,8 +526,9 @@ xfs_dir2_leaf_find_stale(
}
struct xfs_dir2_leaf_entry *
-xfs_dir2_leaf_find_entry(
- xfs_dir2_leaf_t *leaf, /* leaf structure */
+xfs_dir3_leaf_find_entry(
+ struct xfs_dir3_icleaf_hdr *leafhdr,
+ struct xfs_dir2_leaf_entry *ents,
int index, /* leaf table position */
int compact, /* need to compact leaves */
int lowstale, /* index of prev stale leaf */
@@ -272,7 +536,7 @@ xfs_dir2_leaf_find_entry(
int *lfloglow, /* low leaf logging index */
int *lfloghigh) /* high leaf logging index */
{
- if (!leaf->hdr.stale) {
+ if (!leafhdr->stale) {
xfs_dir2_leaf_entry_t *lep; /* leaf entry table pointer */
/*
@@ -280,18 +544,16 @@ xfs_dir2_leaf_find_entry(
*
* If there are no stale entries, just insert a hole at index.
*/
- lep = &leaf->ents[index];
- if (index < be16_to_cpu(leaf->hdr.count))
+ lep = &ents[index];
+ if (index < leafhdr->count)
memmove(lep + 1, lep,
- (be16_to_cpu(leaf->hdr.count) - index) *
- sizeof(*lep));
+ (leafhdr->count - index) * sizeof(*lep));
/*
* Record low and high logging indices for the leaf.
*/
*lfloglow = index;
- *lfloghigh = be16_to_cpu(leaf->hdr.count);
- be16_add_cpu(&leaf->hdr.count, 1);
+ *lfloghigh = leafhdr->count++;
return lep;
}
@@ -305,16 +567,17 @@ xfs_dir2_leaf_find_entry(
* entries before and after our insertion point.
*/
if (compact == 0)
- xfs_dir2_leaf_find_stale(leaf, index, &lowstale, &highstale);
+ xfs_dir3_leaf_find_stale(leafhdr, ents, index,
+ &lowstale, &highstale);
/*
* If the low one is better, use it.
*/
if (lowstale >= 0 &&
- (highstale == be16_to_cpu(leaf->hdr.count) ||
+ (highstale == leafhdr->count ||
index - lowstale - 1 < highstale - index)) {
ASSERT(index - lowstale - 1 >= 0);
- ASSERT(leaf->ents[lowstale].address ==
+ ASSERT(ents[lowstale].address ==
cpu_to_be32(XFS_DIR2_NULL_DATAPTR));
/*
@@ -322,37 +585,34 @@ xfs_dir2_leaf_find_entry(
* for the new entry.
*/
if (index - lowstale - 1 > 0) {
- memmove(&leaf->ents[lowstale],
- &leaf->ents[lowstale + 1],
+ memmove(&ents[lowstale], &ents[lowstale + 1],
(index - lowstale - 1) *
- sizeof(xfs_dir2_leaf_entry_t));
+ sizeof(xfs_dir2_leaf_entry_t));
}
*lfloglow = MIN(lowstale, *lfloglow);
*lfloghigh = MAX(index - 1, *lfloghigh);
- be16_add_cpu(&leaf->hdr.stale, -1);
- return &leaf->ents[index - 1];
+ leafhdr->stale--;
+ return &ents[index - 1];
}
/*
* The high one is better, so use that one.
*/
ASSERT(highstale - index >= 0);
- ASSERT(leaf->ents[highstale].address ==
- cpu_to_be32(XFS_DIR2_NULL_DATAPTR));
+ ASSERT(ents[highstale].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR));
/*
* Copy entries down to cover the stale entry and make room for the
* new entry.
*/
if (highstale - index > 0) {
- memmove(&leaf->ents[index + 1],
- &leaf->ents[index],
+ memmove(&ents[index + 1], &ents[index],
(highstale - index) * sizeof(xfs_dir2_leaf_entry_t));
}
*lfloglow = MIN(index, *lfloglow);
*lfloghigh = MAX(highstale, *lfloghigh);
- be16_add_cpu(&leaf->hdr.stale, -1);
- return &leaf->ents[index];
+ leafhdr->stale--;
+ return &ents[index];
}
/*
@@ -390,6 +650,8 @@ xfs_dir2_leaf_addname(
xfs_trans_t *tp; /* transaction pointer */
xfs_dir2_db_t use_block; /* data block number */
struct xfs_dir2_data_free *bf; /* bestfree table */
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
trace_xfs_dir2_leaf_addname(args);
@@ -397,7 +659,7 @@ xfs_dir2_leaf_addname(
tp = args->trans;
mp = dp->i_mount;
- error = xfs_dir2_leaf_read(tp, dp, mp->m_dirleafblk, -1, &lbp);
+ error = xfs_dir3_leaf_read(tp, dp, mp->m_dirleafblk, -1, &lbp);
if (error)
return error;
@@ -410,16 +672,19 @@ xfs_dir2_leaf_addname(
index = xfs_dir2_leaf_search_hash(args, lbp);
leaf = lbp->b_addr;
ltp = xfs_dir2_leaf_tail_p(mp, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
bestsp = xfs_dir2_leaf_bests_p(ltp);
length = xfs_dir2_data_entsize(args->namelen);
+
/*
* See if there are any entries with the same hash value
* and space in their block for the new entry.
* This is good because it puts multiple same-hash value entries
* in a data block, improving the lookup of those entries.
*/
- for (use_block = -1, lep = &leaf->ents[index];
- index < be16_to_cpu(leaf->hdr.count) && be32_to_cpu(lep->hashval) == args->hashval;
+ for (use_block = -1, lep = &ents[index];
+ index < leafhdr.count && be32_to_cpu(lep->hashval) == args->hashval;
index++, lep++) {
if (be32_to_cpu(lep->address) == XFS_DIR2_NULL_DATAPTR)
continue;
@@ -452,7 +717,7 @@ xfs_dir2_leaf_addname(
* How many bytes do we need in the leaf block?
*/
needbytes = 0;
- if (!leaf->hdr.stale)
+ if (!leafhdr.stale)
needbytes += sizeof(xfs_dir2_leaf_entry_t);
if (use_block == -1)
needbytes += sizeof(xfs_dir2_data_off_t);
@@ -467,16 +732,15 @@ xfs_dir2_leaf_addname(
* If we don't have enough free bytes but we can make enough
* by compacting out stale entries, we'll do that.
*/
- if ((char *)bestsp - (char *)&leaf->ents[be16_to_cpu(leaf->hdr.count)] <
- needbytes && be16_to_cpu(leaf->hdr.stale) > 1) {
+ if ((char *)bestsp - (char *)&ents[leafhdr.count] < needbytes &&
+ leafhdr.stale > 1)
compact = 1;
- }
+
/*
* Otherwise if we don't have enough free bytes we need to
* convert to node form.
*/
- else if ((char *)bestsp - (char *)&leaf->ents[be16_to_cpu(
- leaf->hdr.count)] < needbytes) {
+ else if ((char *)bestsp - (char *)&ents[leafhdr.count] < needbytes) {
/*
* Just checking or no space reservation, give up.
*/
@@ -524,15 +788,15 @@ xfs_dir2_leaf_addname(
* point later.
*/
if (compact) {
- xfs_dir2_leaf_compact_x1(lbp, &index, &lowstale, &highstale,
- &lfloglow, &lfloghigh);
+ xfs_dir3_leaf_compact_x1(&leafhdr, ents, &index, &lowstale,
+ &highstale, &lfloglow, &lfloghigh);
}
/*
* There are stale entries, so we'll need log-low and log-high
* impossibly bad values later.
*/
- else if (be16_to_cpu(leaf->hdr.stale)) {
- lfloglow = be16_to_cpu(leaf->hdr.count);
+ else if (leafhdr.stale) {
+ lfloglow = leafhdr.count;
lfloghigh = -1;
}
/*
@@ -564,14 +828,14 @@ xfs_dir2_leaf_addname(
memmove(&bestsp[0], &bestsp[1],
be32_to_cpu(ltp->bestcount) * sizeof(bestsp[0]));
be32_add_cpu(<p->bestcount, 1);
- xfs_dir2_leaf_log_tail(tp, lbp);
- xfs_dir2_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
+ xfs_dir3_leaf_log_tail(tp, lbp);
+ xfs_dir3_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
}
/*
* If we're filling in a previously empty block just log it.
*/
else
- xfs_dir2_leaf_log_bests(tp, lbp, use_block, use_block);
+ xfs_dir3_leaf_log_bests(tp, lbp, use_block, use_block);
hdr = dbp->b_addr;
bf = xfs_dir3_data_bestfree_p(hdr);
bestsp[use_block] = bf[0].length;
@@ -632,10 +896,10 @@ xfs_dir2_leaf_addname(
if (be16_to_cpu(bestsp[use_block]) != be16_to_cpu(bf[0].length)) {
bestsp[use_block] = bf[0].length;
if (!grown)
- xfs_dir2_leaf_log_bests(tp, lbp, use_block, use_block);
+ xfs_dir3_leaf_log_bests(tp, lbp, use_block, use_block);
}
- lep = xfs_dir2_leaf_find_entry(leaf, index, compact, lowstale,
+ lep = xfs_dir3_leaf_find_entry(&leafhdr, ents, index, compact, lowstale,
highstale, &lfloglow, &lfloghigh);
/*
@@ -647,82 +911,40 @@ xfs_dir2_leaf_addname(
/*
* Log the leaf fields and give up the buffers.
*/
- xfs_dir2_leaf_log_header(tp, lbp);
- xfs_dir2_leaf_log_ents(tp, lbp, lfloglow, lfloghigh);
- xfs_dir2_leaf_check(dp, lbp);
+ xfs_dir3_leaf_hdr_to_disk(leaf, &leafhdr);
+ xfs_dir3_leaf_log_header(tp, lbp);
+ xfs_dir3_leaf_log_ents(tp, lbp, lfloglow, lfloghigh);
+ xfs_dir3_leaf_check(mp, lbp);
xfs_dir3_data_check(dp, dbp);
return 0;
}
-#ifdef DEBUG
-/*
- * Check the internal consistency of a leaf1 block.
- * Pop an assert if something is wrong.
- */
-STATIC void
-xfs_dir2_leaf_check(
- struct xfs_inode *dp, /* incore directory inode */
- struct xfs_buf *bp) /* leaf's buffer */
-{
- int i; /* leaf index */
- xfs_dir2_leaf_t *leaf; /* leaf structure */
- xfs_dir2_leaf_tail_t *ltp; /* leaf tail pointer */
- xfs_mount_t *mp; /* filesystem mount point */
- int stale; /* count of stale leaves */
-
- leaf = bp->b_addr;
- mp = dp->i_mount;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
- /*
- * This value is not restrictive enough.
- * Should factor in the size of the bests table as well.
- * We can deduce a value for that from di_size.
- */
- ASSERT(be16_to_cpu(leaf->hdr.count) <= xfs_dir2_max_leaf_ents(mp));
- ltp = xfs_dir2_leaf_tail_p(mp, leaf);
- /*
- * Leaves and bests don't overlap.
- */
- ASSERT((char *)&leaf->ents[be16_to_cpu(leaf->hdr.count)] <=
- (char *)xfs_dir2_leaf_bests_p(ltp));
- /*
- * Check hash value order, count stale entries.
- */
- for (i = stale = 0; i < be16_to_cpu(leaf->hdr.count); i++) {
- if (i + 1 < be16_to_cpu(leaf->hdr.count))
- ASSERT(be32_to_cpu(leaf->ents[i].hashval) <=
- be32_to_cpu(leaf->ents[i + 1].hashval));
- if (leaf->ents[i].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
- stale++;
- }
- ASSERT(be16_to_cpu(leaf->hdr.stale) == stale);
-}
-#endif /* DEBUG */
-
/*
* Compact out any stale entries in the leaf.
* Log the header and changed leaf entries, if any.
*/
void
-xfs_dir2_leaf_compact(
+xfs_dir3_leaf_compact(
xfs_da_args_t *args, /* operation arguments */
+ struct xfs_dir3_icleaf_hdr *leafhdr,
struct xfs_buf *bp) /* leaf buffer */
{
int from; /* source leaf index */
xfs_dir2_leaf_t *leaf; /* leaf structure */
int loglow; /* first leaf entry to log */
int to; /* target leaf index */
+ struct xfs_dir2_leaf_entry *ents;
leaf = bp->b_addr;
- if (!leaf->hdr.stale) {
+ if (!leafhdr->stale)
return;
- }
+
/*
* Compress out the stale entries in place.
*/
- for (from = to = 0, loglow = -1; from < be16_to_cpu(leaf->hdr.count); from++) {
- if (leaf->ents[from].address ==
- cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ for (from = to = 0, loglow = -1; from < leafhdr->count; from++) {
+ if (ents[from].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
continue;
/*
* Only actually copy the entries that are different.
@@ -730,19 +952,21 @@ xfs_dir2_leaf_compact(
if (from > to) {
if (loglow == -1)
loglow = to;
- leaf->ents[to] = leaf->ents[from];
+ ents[to] = ents[from];
}
to++;
}
/*
* Update and log the header, log the leaf entries.
*/
- ASSERT(be16_to_cpu(leaf->hdr.stale) == from - to);
- be16_add_cpu(&leaf->hdr.count, -(be16_to_cpu(leaf->hdr.stale)));
- leaf->hdr.stale = 0;
- xfs_dir2_leaf_log_header(args->trans, bp);
+ ASSERT(leafhdr->stale == from - to);
+ leafhdr->count -= leafhdr->stale;
+ leafhdr->stale = 0;
+
+ xfs_dir3_leaf_hdr_to_disk(leaf, leafhdr);
+ xfs_dir3_leaf_log_header(args->trans, bp);
if (loglow != -1)
- xfs_dir2_leaf_log_ents(args->trans, bp, loglow, to - 1);
+ xfs_dir3_leaf_log_ents(args->trans, bp, loglow, to - 1);
}
/*
@@ -754,8 +978,9 @@ xfs_dir2_leaf_compact(
* and leaf logging indices.
*/
void
-xfs_dir2_leaf_compact_x1(
- struct xfs_buf *bp, /* leaf buffer */
+xfs_dir3_leaf_compact_x1(
+ struct xfs_dir3_icleaf_hdr *leafhdr,
+ struct xfs_dir2_leaf_entry *ents,
int *indexp, /* insertion index */
int *lowstalep, /* out: stale entry before us */
int *highstalep, /* out: stale entry after us */
@@ -766,22 +991,20 @@ xfs_dir2_leaf_compact_x1(
int highstale; /* stale entry at/after index */
int index; /* insertion index */
int keepstale; /* source index of kept stale */
- xfs_dir2_leaf_t *leaf; /* leaf structure */
int lowstale; /* stale entry before index */
int newindex=0; /* new insertion index */
int to; /* destination copy index */
- leaf = bp->b_addr;
- ASSERT(be16_to_cpu(leaf->hdr.stale) > 1);
+ ASSERT(leafhdr->stale > 1);
index = *indexp;
- xfs_dir2_leaf_find_stale(leaf, index, &lowstale, &highstale);
+ xfs_dir3_leaf_find_stale(leafhdr, ents, index, &lowstale, &highstale);
/*
* Pick the better of lowstale and highstale.
*/
if (lowstale >= 0 &&
- (highstale == be16_to_cpu(leaf->hdr.count) ||
+ (highstale == leafhdr->count ||
index - lowstale <= highstale - index))
keepstale = lowstale;
else
@@ -790,15 +1013,14 @@ xfs_dir2_leaf_compact_x1(
* Copy the entries in place, removing all the stale entries
* except keepstale.
*/
- for (from = to = 0; from < be16_to_cpu(leaf->hdr.count); from++) {
+ for (from = to = 0; from < leafhdr->count; from++) {
/*
* Notice the new value of index.
*/
if (index == from)
newindex = to;
if (from != keepstale &&
- leaf->ents[from].address ==
- cpu_to_be32(XFS_DIR2_NULL_DATAPTR)) {
+ ents[from].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR)) {
if (from == to)
*lowlogp = to;
continue;
@@ -812,7 +1034,7 @@ xfs_dir2_leaf_compact_x1(
* Copy only the entries that have moved.
*/
if (from > to)
- leaf->ents[to] = leaf->ents[from];
+ ents[to] = ents[from];
to++;
}
ASSERT(from > to);
@@ -826,8 +1048,8 @@ xfs_dir2_leaf_compact_x1(
/*
* Adjust the leaf header values.
*/
- be16_add_cpu(&leaf->hdr.count, -(from - to));
- leaf->hdr.stale = cpu_to_be16(1);
+ leafhdr->count -= from - to;
+ leafhdr->stale = 1;
/*
* Remember the low/high stale value only in the "right"
* direction.
@@ -835,8 +1057,8 @@ xfs_dir2_leaf_compact_x1(
if (lowstale >= newindex)
lowstale = -1;
else
- highstale = be16_to_cpu(leaf->hdr.count);
- *highlogp = be16_to_cpu(leaf->hdr.count) - 1;
+ highstale = leafhdr->count;
+ *highlogp = leafhdr->count - 1;
*lowstalep = lowstale;
*highstalep = highstale;
}
@@ -1229,69 +1451,12 @@ xfs_dir2_leaf_getdents(
return error;
}
-/*
- * Initialize a new leaf block, leaf1 or leafn magic accepted.
- */
-int
-xfs_dir2_leaf_init(
- xfs_da_args_t *args, /* operation arguments */
- xfs_dir2_db_t bno, /* directory block number */
- struct xfs_buf **bpp, /* out: leaf buffer */
- int magic) /* magic number for block */
-{
- struct xfs_buf *bp; /* leaf buffer */
- xfs_inode_t *dp; /* incore directory inode */
- int error; /* error return code */
- xfs_dir2_leaf_t *leaf; /* leaf structure */
- xfs_dir2_leaf_tail_t *ltp; /* leaf tail structure */
- xfs_mount_t *mp; /* filesystem mount point */
- xfs_trans_t *tp; /* transaction pointer */
-
- dp = args->dp;
- ASSERT(dp != NULL);
- tp = args->trans;
- mp = dp->i_mount;
- ASSERT(bno >= XFS_DIR2_LEAF_FIRSTDB(mp) &&
- bno < XFS_DIR2_FREE_FIRSTDB(mp));
- /*
- * Get the buffer for the block.
- */
- error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, bno), -1, &bp,
- XFS_DATA_FORK);
- if (error)
- return error;
-
- /*
- * Initialize the header.
- */
- leaf = bp->b_addr;
- leaf->hdr.info.magic = cpu_to_be16(magic);
- leaf->hdr.info.forw = 0;
- leaf->hdr.info.back = 0;
- leaf->hdr.count = 0;
- leaf->hdr.stale = 0;
- xfs_dir2_leaf_log_header(tp, bp);
- /*
- * If it's a leaf-format directory initialize the tail.
- * In this case our caller has the real bests table to copy into
- * the block.
- */
- if (magic == XFS_DIR2_LEAF1_MAGIC) {
- bp->b_ops = &xfs_dir2_leaf1_buf_ops;
- ltp = xfs_dir2_leaf_tail_p(mp, leaf);
- ltp->bestcount = 0;
- xfs_dir2_leaf_log_tail(tp, bp);
- } else
- bp->b_ops = &xfs_dir2_leafn_buf_ops;
- *bpp = bp;
- return 0;
-}
/*
* Log the bests entries indicated from a leaf1 block.
*/
static void
-xfs_dir2_leaf_log_bests(
+xfs_dir3_leaf_log_bests(
xfs_trans_t *tp, /* transaction pointer */
struct xfs_buf *bp, /* leaf buffer */
int first, /* first entry to log */
@@ -1299,11 +1464,12 @@ xfs_dir2_leaf_log_bests(
{
__be16 *firstb; /* pointer to first entry */
__be16 *lastb; /* pointer to last entry */
- xfs_dir2_leaf_t *leaf; /* leaf structure */
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
xfs_dir2_leaf_tail_t *ltp; /* leaf tail structure */
- leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
+ ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC) ||
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAF1_MAGIC));
+
ltp = xfs_dir2_leaf_tail_p(tp->t_mountp, leaf);
firstb = xfs_dir2_leaf_bests_p(ltp) + first;
lastb = xfs_dir2_leaf_bests_p(ltp) + last;
@@ -1315,7 +1481,7 @@ xfs_dir2_leaf_log_bests(
* Log the leaf entries indicated from a leaf1 or leafn block.
*/
void
-xfs_dir2_leaf_log_ents(
+xfs_dir3_leaf_log_ents(
xfs_trans_t *tp, /* transaction pointer */
struct xfs_buf *bp, /* leaf buffer */
int first, /* first entry to log */
@@ -1323,13 +1489,17 @@ xfs_dir2_leaf_log_ents(
{
xfs_dir2_leaf_entry_t *firstlep; /* pointer to first entry */
xfs_dir2_leaf_entry_t *lastlep; /* pointer to last entry */
- xfs_dir2_leaf_t *leaf; /* leaf structure */
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
+ struct xfs_dir2_leaf_entry *ents;
- leaf = bp->b_addr;
ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC) ||
- leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
- firstlep = &leaf->ents[first];
- lastlep = &leaf->ents[last];
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAF1_MAGIC) ||
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC));
+
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ firstlep = &ents[first];
+ lastlep = &ents[last];
xfs_trans_log_buf(tp, bp, (uint)((char *)firstlep - (char *)leaf),
(uint)((char *)lastlep - (char *)leaf + sizeof(*lastlep) - 1));
}
@@ -1338,34 +1508,36 @@ xfs_dir2_leaf_log_ents(
* Log the header of the leaf1 or leafn block.
*/
void
-xfs_dir2_leaf_log_header(
+xfs_dir3_leaf_log_header(
struct xfs_trans *tp,
struct xfs_buf *bp)
{
- xfs_dir2_leaf_t *leaf; /* leaf structure */
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
- leaf = bp->b_addr;
ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC) ||
- leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAF1_MAGIC) ||
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC));
+
xfs_trans_log_buf(tp, bp, (uint)((char *)&leaf->hdr - (char *)leaf),
- (uint)(sizeof(leaf->hdr) - 1));
+ xfs_dir3_leaf_hdr_size(leaf));
}
/*
* Log the tail of the leaf1 block.
*/
STATIC void
-xfs_dir2_leaf_log_tail(
+xfs_dir3_leaf_log_tail(
struct xfs_trans *tp,
struct xfs_buf *bp)
{
- xfs_dir2_leaf_t *leaf; /* leaf structure */
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
xfs_dir2_leaf_tail_t *ltp; /* leaf tail structure */
- xfs_mount_t *mp; /* filesystem mount point */
+ struct xfs_mount *mp = tp->t_mountp;
+
+ ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC) ||
+ leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
- mp = tp->t_mountp;
- leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC));
ltp = xfs_dir2_leaf_tail_p(mp, leaf);
xfs_trans_log_buf(tp, bp, (uint)((char *)ltp - (char *)leaf),
(uint)(mp->m_dirblksize - 1));
@@ -1389,6 +1561,7 @@ xfs_dir2_leaf_lookup(
xfs_dir2_leaf_t *leaf; /* leaf structure */
xfs_dir2_leaf_entry_t *lep; /* leaf entry */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir2_leaf_entry *ents;
trace_xfs_dir2_leaf_lookup(args);
@@ -1400,12 +1573,14 @@ xfs_dir2_leaf_lookup(
}
tp = args->trans;
dp = args->dp;
- xfs_dir2_leaf_check(dp, lbp);
+ xfs_dir3_leaf_check(dp->i_mount, lbp);
leaf = lbp->b_addr;
+ ents = xfs_dir3_leaf_ents_p(leaf);
/*
* Get to the leaf entry and contained data entry address.
*/
- lep = &leaf->ents[index];
+ lep = &ents[index];
+
/*
* Point to the data entry.
*/
@@ -1449,18 +1624,23 @@ xfs_dir2_leaf_lookup_int(
xfs_trans_t *tp; /* transaction pointer */
xfs_dir2_db_t cidb = -1; /* case match data block no. */
enum xfs_dacmp cmp; /* name compare result */
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
- error = xfs_dir2_leaf_read(tp, dp, mp->m_dirleafblk, -1, &lbp);
+ error = xfs_dir3_leaf_read(tp, dp, mp->m_dirleafblk, -1, &lbp);
if (error)
return error;
*lbpp = lbp;
leaf = lbp->b_addr;
- xfs_dir2_leaf_check(dp, lbp);
+ xfs_dir3_leaf_check(mp, lbp);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+
/*
* Look for the first leaf entry with our hash value.
*/
@@ -1469,9 +1649,9 @@ xfs_dir2_leaf_lookup_int(
* Loop over all the entries with the right hash value
* looking to match the name.
*/
- for (lep = &leaf->ents[index]; index < be16_to_cpu(leaf->hdr.count) &&
- be32_to_cpu(lep->hashval) == args->hashval;
- lep++, index++) {
+ for (lep = &ents[index];
+ index < leafhdr.count && be32_to_cpu(lep->hashval) == args->hashval;
+ lep++, index++) {
/*
* Skip over stale leaf entries.
*/
@@ -1576,6 +1756,8 @@ xfs_dir2_leaf_removename(
xfs_dir2_data_off_t oldbest; /* old value of best free */
xfs_trans_t *tp; /* transaction pointer */
struct xfs_dir2_data_free *bf; /* bestfree table */
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
trace_xfs_dir2_leaf_removename(args);
@@ -1590,12 +1772,14 @@ xfs_dir2_leaf_removename(
mp = dp->i_mount;
leaf = lbp->b_addr;
hdr = dbp->b_addr;
- bf = xfs_dir3_data_bestfree_p(hdr);
xfs_dir3_data_check(dp, dbp);
+ bf = xfs_dir3_data_bestfree_p(hdr);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
/*
* Point to the leaf entry, use that to point to the data entry.
*/
- lep = &leaf->ents[index];
+ lep = &ents[index];
db = xfs_dir2_dataptr_to_db(mp, be32_to_cpu(lep->address));
dep = (xfs_dir2_data_entry_t *)
((char *)hdr + xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address)));
@@ -1613,10 +1797,13 @@ xfs_dir2_leaf_removename(
/*
* We just mark the leaf entry stale by putting a null in it.
*/
- be16_add_cpu(&leaf->hdr.stale, 1);
- xfs_dir2_leaf_log_header(tp, lbp);
+ leafhdr.stale++;
+ xfs_dir3_leaf_hdr_to_disk(leaf, &leafhdr);
+ xfs_dir3_leaf_log_header(tp, lbp);
+
lep->address = cpu_to_be32(XFS_DIR2_NULL_DATAPTR);
- xfs_dir2_leaf_log_ents(tp, lbp, index, index);
+ xfs_dir3_leaf_log_ents(tp, lbp, index, index);
+
/*
* Scan the freespace in the data block again if necessary,
* log the data block header if necessary.
@@ -1631,7 +1818,7 @@ xfs_dir2_leaf_removename(
*/
if (be16_to_cpu(bf[0].length) != oldbest) {
bestsp[db] = bf[0].length;
- xfs_dir2_leaf_log_bests(tp, lbp, db, db);
+ xfs_dir3_leaf_log_bests(tp, lbp, db, db);
}
xfs_dir3_data_check(dp, dbp);
/*
@@ -1648,7 +1835,7 @@ xfs_dir2_leaf_removename(
*/
if (error == ENOSPC && args->total == 0)
error = 0;
- xfs_dir2_leaf_check(dp, lbp);
+ xfs_dir3_leaf_check(mp, lbp);
return error;
}
dbp = NULL;
@@ -1671,8 +1858,8 @@ xfs_dir2_leaf_removename(
memmove(&bestsp[db - i], bestsp,
(be32_to_cpu(ltp->bestcount) - (db - i)) * sizeof(*bestsp));
be32_add_cpu(<p->bestcount, -(db - i));
- xfs_dir2_leaf_log_tail(tp, lbp);
- xfs_dir2_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
+ xfs_dir3_leaf_log_tail(tp, lbp);
+ xfs_dir3_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
} else
bestsp[db] = cpu_to_be16(NULLDATAOFF);
}
@@ -1682,7 +1869,7 @@ xfs_dir2_leaf_removename(
else if (db != mp->m_dirdatablk)
dbp = NULL;
- xfs_dir2_leaf_check(dp, lbp);
+ xfs_dir3_leaf_check(mp, lbp);
/*
* See if we can convert to block form.
*/
@@ -1705,6 +1892,7 @@ xfs_dir2_leaf_replace(
xfs_dir2_leaf_t *leaf; /* leaf structure */
xfs_dir2_leaf_entry_t *lep; /* leaf entry */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir2_leaf_entry *ents;
trace_xfs_dir2_leaf_replace(args);
@@ -1716,10 +1904,11 @@ xfs_dir2_leaf_replace(
}
dp = args->dp;
leaf = lbp->b_addr;
+ ents = xfs_dir3_leaf_ents_p(leaf);
/*
* Point to the leaf entry, get data address from it.
*/
- lep = &leaf->ents[index];
+ lep = &ents[index];
/*
* Point to the data entry.
*/
@@ -1733,7 +1922,7 @@ xfs_dir2_leaf_replace(
dep->inumber = cpu_to_be64(args->inumber);
tp = args->trans;
xfs_dir2_data_log_entry(tp, dbp, dep);
- xfs_dir2_leaf_check(dp, lbp);
+ xfs_dir3_leaf_check(dp->i_mount, lbp);
xfs_trans_brelse(tp, lbp);
return 0;
}
@@ -1755,17 +1944,22 @@ xfs_dir2_leaf_search_hash(
xfs_dir2_leaf_t *leaf; /* leaf structure */
xfs_dir2_leaf_entry_t *lep; /* leaf entry */
int mid=0; /* current leaf index */
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
leaf = lbp->b_addr;
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+
#ifndef __KERNEL__
- if (!leaf->hdr.count)
+ if (!leafhdr.count)
return 0;
#endif
/*
* Note, the table cannot be empty, so we have to go through the loop.
* Binary search the leaf entries looking for our hash value.
*/
- for (lep = leaf->ents, low = 0, high = be16_to_cpu(leaf->hdr.count) - 1,
+ for (lep = ents, low = 0, high = leafhdr.count - 1,
hashwant = args->hashval;
low <= high; ) {
mid = (low + high) >> 1;
@@ -1851,23 +2045,29 @@ xfs_dir2_leaf_trim_data(
bestsp = xfs_dir2_leaf_bests_p(ltp);
be32_add_cpu(<p->bestcount, -1);
memmove(&bestsp[1], &bestsp[0], be32_to_cpu(ltp->bestcount) * sizeof(*bestsp));
- xfs_dir2_leaf_log_tail(tp, lbp);
- xfs_dir2_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
+ xfs_dir3_leaf_log_tail(tp, lbp);
+ xfs_dir3_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
return 0;
}
static inline size_t
-xfs_dir2_leaf_size(
- struct xfs_dir2_leaf_hdr *hdr,
+xfs_dir3_leaf_size(
+ struct xfs_dir3_icleaf_hdr *hdr,
int counts)
{
- int entries;
+ int entries;
+ int hdrsize;
- entries = be16_to_cpu(hdr->count) - be16_to_cpu(hdr->stale);
- return sizeof(xfs_dir2_leaf_hdr_t) +
- entries * sizeof(xfs_dir2_leaf_entry_t) +
- counts * sizeof(xfs_dir2_data_off_t) +
- sizeof(xfs_dir2_leaf_tail_t);
+ entries = hdr->count - hdr->stale;
+ if (hdr->magic == XFS_DIR2_LEAF1_MAGIC ||
+ hdr->magic == XFS_DIR2_LEAFN_MAGIC)
+ hdrsize = sizeof(struct xfs_dir2_leaf_hdr);
+ else
+ hdrsize = sizeof(struct xfs_dir3_leaf_hdr);
+
+ return hdrsize + entries * sizeof(xfs_dir2_leaf_entry_t)
+ + counts * sizeof(xfs_dir2_data_off_t)
+ + sizeof(xfs_dir2_leaf_tail_t);
}
/*
@@ -1891,6 +2091,7 @@ xfs_dir2_node_to_leaf(
xfs_mount_t *mp; /* filesystem mount point */
int rval; /* successful free trim? */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir3_icleaf_hdr leafhdr;
/*
* There's more than a leaf level in the btree, so there must
@@ -1940,7 +2141,11 @@ xfs_dir2_node_to_leaf(
return 0;
lbp = state->path.blk[0].bp;
leaf = lbp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+
+ ASSERT(leafhdr.magic == XFS_DIR2_LEAFN_MAGIC ||
+ leafhdr.magic == XFS_DIR3_LEAFN_MAGIC);
+
/*
* Read the freespace block.
*/
@@ -1955,7 +2160,7 @@ xfs_dir2_node_to_leaf(
* Now see if the leafn and free data will fit in a leaf1.
* If not, release the buffer and give up.
*/
- if (xfs_dir2_leaf_size(&leaf->hdr, be32_to_cpu(free->hdr.nvalid)) >
+ if (xfs_dir3_leaf_size(&leafhdr, be32_to_cpu(free->hdr.nvalid)) >
mp->m_dirblksize) {
xfs_trans_brelse(tp, fbp);
return 0;
@@ -1963,15 +2168,14 @@ xfs_dir2_node_to_leaf(
/*
* If the leaf has any stale entries in it, compress them out.
- * The compact routine will log the header.
*/
- if (be16_to_cpu(leaf->hdr.stale))
- xfs_dir2_leaf_compact(args, lbp);
- else
- xfs_dir2_leaf_log_header(tp, lbp);
+ if (leafhdr.stale)
+ xfs_dir3_leaf_compact(args, &leafhdr, lbp);
- lbp->b_ops = &xfs_dir2_leaf1_buf_ops;
- leaf->hdr.info.magic = cpu_to_be16(XFS_DIR2_LEAF1_MAGIC);
+ lbp->b_ops = &xfs_dir3_leaf1_buf_ops;
+ leafhdr.magic = (leafhdr.magic == XFS_DIR2_LEAFN_MAGIC)
+ ? XFS_DIR2_LEAF1_MAGIC
+ : XFS_DIR3_LEAFN_MAGIC;
/*
* Set up the leaf tail from the freespace block.
@@ -1983,9 +2187,12 @@ xfs_dir2_node_to_leaf(
*/
memcpy(xfs_dir2_leaf_bests_p(ltp), xfs_dir3_free_bests_p(mp, free),
be32_to_cpu(ltp->bestcount) * sizeof(xfs_dir2_data_off_t));
- xfs_dir2_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
- xfs_dir2_leaf_log_tail(tp, lbp);
- xfs_dir2_leaf_check(dp, lbp);
+
+ xfs_dir3_leaf_hdr_to_disk(leaf, &leafhdr);
+ xfs_dir3_leaf_log_header(tp, lbp);
+ xfs_dir3_leaf_log_bests(tp, lbp, 0, be32_to_cpu(ltp->bestcount) - 1);
+ xfs_dir3_leaf_log_tail(tp, lbp);
+ xfs_dir3_leaf_check(mp, lbp);
/*
* Get rid of the freespace block.
*/
diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c
index 3b33d7d..99d1df3 100644
--- a/fs/xfs/xfs_dir2_node.c
+++ b/fs/xfs/xfs_dir2_node.c
@@ -41,13 +41,12 @@
static int xfs_dir2_leafn_add(struct xfs_buf *bp, xfs_da_args_t *args,
int index);
#ifdef DEBUG
-static void xfs_dir2_leafn_check(struct xfs_inode *dp, struct xfs_buf *bp);
-#else
-#define xfs_dir2_leafn_check(dp, bp)
+#define xfs_dir3_leaf_check(mp, bp) \
+do { \
+ if (!xfs_dir3_leafn_check((mp), (bp))) \
+ ASSERT(0); \
+} while (0);
#endif
-static void xfs_dir2_leafn_moveents(xfs_da_args_t *args, struct xfs_buf *bp_s,
- int start_s, struct xfs_buf *bp_d,
- int start_d, int count);
static void xfs_dir2_leafn_rebalance(xfs_da_state_t *state,
xfs_da_state_blk_t *blk1,
xfs_da_state_blk_t *blk2);
@@ -57,6 +56,26 @@ static int xfs_dir2_leafn_remove(xfs_da_args_t *args, struct xfs_buf *bp,
static int xfs_dir2_node_addname_int(xfs_da_args_t *args,
xfs_da_state_blk_t *fblk);
+/*
+ * Check internal consistency of a leafn block.
+ */
+static bool
+xfs_dir3_leafn_check(
+ struct xfs_mount *mp,
+ struct xfs_buf *bp)
+{
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
+ struct xfs_dir3_icleaf_hdr leafhdr;
+
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+
+ if (leafhdr.magic != XFS_DIR2_LEAFN_MAGIC &&
+ leafhdr.magic != XFS_DIR3_LEAFN_MAGIC)
+ return false;
+
+ return xfs_dir3_leaf_check_int(mp, &leafhdr, leaf);
+}
+
static bool
xfs_dir3_free_verify(
struct xfs_buf *bp)
@@ -360,11 +379,19 @@ xfs_dir2_leaf_to_node(
xfs_dir2_free_log_bests(tp, fbp, 0, freehdr.nvalid - 1);
xfs_dir2_free_log_header(tp, fbp);
- /* convert the leaf to a leafnode */
- leaf->hdr.info.magic = cpu_to_be16(XFS_DIR2_LEAFN_MAGIC);
- lbp->b_ops = &xfs_dir2_leafn_buf_ops;
- xfs_dir2_leaf_log_header(tp, lbp);
- xfs_dir2_leafn_check(dp, lbp);
+ /*
+ * Converting the leaf to a leafnode is just a matter of changing the
+ * magic number and the ops. Do the change directly to the buffer as
+ * it's less work (and less code) than decoding the header to host
+ * format and back again.
+ */
+ if (leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAF1_MAGIC))
+ leaf->hdr.info.magic = cpu_to_be16(XFS_DIR2_LEAFN_MAGIC);
+ else
+ leaf->hdr.info.magic = cpu_to_be16(XFS_DIR3_LEAFN_MAGIC);
+ lbp->b_ops = &xfs_dir3_leafn_buf_ops;
+ xfs_dir3_leaf_log_header(tp, lbp);
+ xfs_dir3_leaf_check(mp, lbp);
return 0;
}
@@ -388,6 +415,8 @@ xfs_dir2_leafn_add(
int lowstale; /* previous stale entry */
xfs_mount_t *mp; /* filesystem mount point */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir3_icleaf_hdr leafhdr;
+ struct xfs_dir2_leaf_entry *ents;
trace_xfs_dir2_leafn_add(args, index);
@@ -395,6 +424,8 @@ xfs_dir2_leafn_add(
mp = dp->i_mount;
tp = args->trans;
leaf = bp->b_addr;
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
/*
* Quick check just to make sure we are not going to index
@@ -410,15 +441,15 @@ xfs_dir2_leafn_add(
* a compact.
*/
- if (be16_to_cpu(leaf->hdr.count) == xfs_dir2_max_leaf_ents(mp)) {
- if (!leaf->hdr.stale)
+ if (leafhdr.count == xfs_dir3_max_leaf_ents(mp, leaf)) {
+ if (!leafhdr.stale)
return XFS_ERROR(ENOSPC);
- compact = be16_to_cpu(leaf->hdr.stale) > 1;
+ compact = leafhdr.stale > 1;
} else
compact = 0;
- ASSERT(index == 0 || be32_to_cpu(leaf->ents[index - 1].hashval) <= args->hashval);
- ASSERT(index == be16_to_cpu(leaf->hdr.count) ||
- be32_to_cpu(leaf->ents[index].hashval) >= args->hashval);
+ ASSERT(index == 0 || be32_to_cpu(ents[index - 1].hashval) <= args->hashval);
+ ASSERT(index == leafhdr.count ||
+ be32_to_cpu(ents[index].hashval) >= args->hashval);
if (args->op_flags & XFS_DA_OP_JUSTCHECK)
return 0;
@@ -427,62 +458,35 @@ xfs_dir2_leafn_add(
* Compact out all but one stale leaf entry. Leaves behind
* the entry closest to index.
*/
- if (compact) {
- xfs_dir2_leaf_compact_x1(bp, &index, &lowstale, &highstale,
- &lfloglow, &lfloghigh);
- }
- /*
- * Set impossible logging indices for this case.
- */
- else if (leaf->hdr.stale) {
- lfloglow = be16_to_cpu(leaf->hdr.count);
+ if (compact)
+ xfs_dir3_leaf_compact_x1(&leafhdr, ents, &index, &lowstale,
+ &highstale, &lfloglow, &lfloghigh);
+ else if (leafhdr.stale) {
+ /*
+ * Set impossible logging indices for this case.
+ */
+ lfloglow = leafhdr.count;
lfloghigh = -1;
}
/*
* Insert the new entry, log everything.
*/
- lep = xfs_dir2_leaf_find_entry(leaf, index, compact, lowstale,
+ lep = xfs_dir3_leaf_find_entry(&leafhdr, ents, index, compact, lowstale,
highstale, &lfloglow, &lfloghigh);
lep->hashval = cpu_to_be32(args->hashval);
lep->address = cpu_to_be32(xfs_dir2_db_off_to_dataptr(mp,
args->blkno, args->index));
- xfs_dir2_leaf_log_header(tp, bp);
- xfs_dir2_leaf_log_ents(tp, bp, lfloglow, lfloghigh);
- xfs_dir2_leafn_check(dp, bp);
+
+ xfs_dir3_leaf_hdr_to_disk(leaf, &leafhdr);
+ xfs_dir3_leaf_log_header(tp, bp);
+ xfs_dir3_leaf_log_ents(tp, bp, lfloglow, lfloghigh);
+ xfs_dir3_leaf_check(mp, bp);
return 0;
}
#ifdef DEBUG
-/*
- * Check internal consistency of a leafn block.
- */
-void
-xfs_dir2_leafn_check(
- struct xfs_inode *dp,
- struct xfs_buf *bp)
-{
- int i; /* leaf index */
- xfs_dir2_leaf_t *leaf; /* leaf structure */
- xfs_mount_t *mp; /* filesystem mount point */
- int stale; /* count of stale leaves */
-
- leaf = bp->b_addr;
- mp = dp->i_mount;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
- ASSERT(be16_to_cpu(leaf->hdr.count) <= xfs_dir2_max_leaf_ents(mp));
- for (i = stale = 0; i < be16_to_cpu(leaf->hdr.count); i++) {
- if (i + 1 < be16_to_cpu(leaf->hdr.count)) {
- ASSERT(be32_to_cpu(leaf->ents[i].hashval) <=
- be32_to_cpu(leaf->ents[i + 1].hashval));
- }
- if (leaf->ents[i].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
- stale++;
- }
- ASSERT(be16_to_cpu(leaf->hdr.stale) == stale);
-}
-
static void
xfs_dir2_free_hdr_check(
struct xfs_mount *mp,
@@ -510,15 +514,18 @@ xfs_dir2_leafn_lasthash(
struct xfs_buf *bp, /* leaf buffer */
int *count) /* count of entries in leaf */
{
- xfs_dir2_leaf_t *leaf; /* leaf structure */
+ struct xfs_dir2_leaf *leaf = bp->b_addr;
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
+
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
- leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
if (count)
- *count = be16_to_cpu(leaf->hdr.count);
- if (!leaf->hdr.count)
+ *count = leafhdr.count;
+ if (!leafhdr.count)
return 0;
- return be32_to_cpu(leaf->ents[be16_to_cpu(leaf->hdr.count) - 1].hashval);
+ return be32_to_cpu(ents[leafhdr.count - 1].hashval);
}
/*
@@ -547,16 +554,19 @@ xfs_dir2_leafn_lookup_for_addname(
xfs_dir2_db_t newdb; /* new data block number */
xfs_dir2_db_t newfdb; /* new free block number */
xfs_trans_t *tp; /* transaction pointer */
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
-#ifdef __KERNEL__
- ASSERT(be16_to_cpu(leaf->hdr.count) > 0);
-#endif
- xfs_dir2_leafn_check(dp, bp);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+
+ xfs_dir3_leaf_check(mp, bp);
+ ASSERT(leafhdr.count > 0);
+
/*
* Look up the hash value in the leaf entries.
*/
@@ -576,9 +586,9 @@ xfs_dir2_leafn_lookup_for_addname(
/*
* Loop over leaf entries with the right hash value.
*/
- for (lep = &leaf->ents[index]; index < be16_to_cpu(leaf->hdr.count) &&
- be32_to_cpu(lep->hashval) == args->hashval;
- lep++, index++) {
+ for (lep = &ents[index];
+ index < leafhdr.count && be32_to_cpu(lep->hashval) == args->hashval;
+ lep++, index++) {
/*
* Skip stale leaf entries.
*/
@@ -694,16 +704,19 @@ xfs_dir2_leafn_lookup_for_entry(
xfs_dir2_db_t newdb; /* new data block number */
xfs_trans_t *tp; /* transaction pointer */
enum xfs_dacmp cmp; /* comparison result */
+ struct xfs_dir2_leaf_entry *ents;
+ struct xfs_dir3_icleaf_hdr leafhdr;
dp = args->dp;
tp = args->trans;
mp = dp->i_mount;
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
-#ifdef __KERNEL__
- ASSERT(be16_to_cpu(leaf->hdr.count) > 0);
-#endif
- xfs_dir2_leafn_check(dp, bp);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+
+ xfs_dir3_leaf_check(mp, bp);
+ ASSERT(leafhdr.count > 0);
+
/*
* Look up the hash value in the leaf entries.
*/
@@ -718,9 +731,9 @@ xfs_dir2_leafn_lookup_for_entry(
/*
* Loop over leaf entries with the right hash value.
*/
- for (lep = &leaf->ents[index]; index < be16_to_cpu(leaf->hdr.count) &&
- be32_to_cpu(lep->hashval) == args->hashval;
- lep++, index++) {
+ for (lep = &ents[index];
+ index < leafhdr.count && be32_to_cpu(lep->hashval) == args->hashval;
+ lep++, index++) {
/*
* Skip stale leaf entries.
*/
@@ -792,8 +805,7 @@ xfs_dir2_leafn_lookup_for_entry(
return XFS_ERROR(EEXIST);
}
}
- ASSERT(index == be16_to_cpu(leaf->hdr.count) ||
- (args->op_flags & XFS_DA_OP_OKNOENT));
+ ASSERT(index == leafhdr.count || (args->op_flags & XFS_DA_OP_OKNOENT));
if (curbp) {
if (args->cmpresult == XFS_CMP_DIFFERENT) {
/* Giving back last used data block. */
@@ -838,52 +850,50 @@ xfs_dir2_leafn_lookup_int(
* Log entries and headers. Stale entries are preserved.
*/
static void
-xfs_dir2_leafn_moveents(
- xfs_da_args_t *args, /* operation arguments */
- struct xfs_buf *bp_s, /* source leaf buffer */
- int start_s, /* source leaf index */
- struct xfs_buf *bp_d, /* destination leaf buffer */
- int start_d, /* destination leaf index */
- int count) /* count of leaves to copy */
+xfs_dir3_leafn_moveents(
+ xfs_da_args_t *args, /* operation arguments */
+ struct xfs_buf *bp_s, /* source */
+ struct xfs_dir3_icleaf_hdr *shdr,
+ struct xfs_dir2_leaf_entry *sents,
+ int start_s,/* source leaf index */
+ struct xfs_buf *bp_d, /* destination */
+ struct xfs_dir3_icleaf_hdr *dhdr,
+ struct xfs_dir2_leaf_entry *dents,
+ int start_d,/* destination leaf index */
+ int count) /* count of leaves to copy */
{
- xfs_dir2_leaf_t *leaf_d; /* destination leaf structure */
- xfs_dir2_leaf_t *leaf_s; /* source leaf structure */
- int stale; /* count stale leaves copied */
- xfs_trans_t *tp; /* transaction pointer */
+ int stale;
+ xfs_trans_t *tp = args->trans;
trace_xfs_dir2_leafn_moveents(args, start_s, start_d, count);
/*
* Silently return if nothing to do.
*/
- if (count == 0) {
+ if (count == 0)
return;
- }
- tp = args->trans;
- leaf_s = bp_s->b_addr;
- leaf_d = bp_d->b_addr;
+
/*
* If the destination index is not the end of the current
* destination leaf entries, open up a hole in the destination
* to hold the new entries.
*/
- if (start_d < be16_to_cpu(leaf_d->hdr.count)) {
- memmove(&leaf_d->ents[start_d + count], &leaf_d->ents[start_d],
- (be16_to_cpu(leaf_d->hdr.count) - start_d) *
- sizeof(xfs_dir2_leaf_entry_t));
- xfs_dir2_leaf_log_ents(tp, bp_d, start_d + count,
- count + be16_to_cpu(leaf_d->hdr.count) - 1);
+ if (start_d < dhdr->count) {
+ memmove(&dents[start_d + count], &dents[start_d],
+ (dhdr->count - start_d) * sizeof(xfs_dir2_leaf_entry_t));
+ xfs_dir3_leaf_log_ents(tp, bp_d, start_d + count,
+ count + dhdr->count - 1);
}
/*
* If the source has stale leaves, count the ones in the copy range
* so we can update the header correctly.
*/
- if (leaf_s->hdr.stale) {
+ if (shdr->stale) {
int i; /* temp leaf index */
for (i = start_s, stale = 0; i < start_s + count; i++) {
- if (leaf_s->ents[i].address ==
- cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
+ if (sents[i].address ==
+ cpu_to_be32(XFS_DIR2_NULL_DATAPTR))
stale++;
}
} else
@@ -891,29 +901,27 @@ xfs_dir2_leafn_moveents(
/*
* Copy the leaf entries from source to destination.
*/
- memcpy(&leaf_d->ents[start_d], &leaf_s->ents[start_s],
+ memcpy(&dents[start_d], &sents[start_s],
count * sizeof(xfs_dir2_leaf_entry_t));
- xfs_dir2_leaf_log_ents(tp, bp_d, start_d, start_d + count - 1);
+ xfs_dir3_leaf_log_ents(tp, bp_d, start_d, start_d + count - 1);
+
/*
* If there are source entries after the ones we copied,
* delete the ones we copied by sliding the next ones down.
*/
- if (start_s + count < be16_to_cpu(leaf_s->hdr.count)) {
- memmove(&leaf_s->ents[start_s], &leaf_s->ents[start_s + count],
+ if (start_s + count < shdr->count) {
+ memmove(&sents[start_s], &sents[start_s + count],
count * sizeof(xfs_dir2_leaf_entry_t));
- xfs_dir2_leaf_log_ents(tp, bp_s, start_s, start_s + count - 1);
+ xfs_dir3_leaf_log_ents(tp, bp_s, start_s, start_s + count - 1);
}
+
/*
* Update the headers and log them.
*/
- be16_add_cpu(&leaf_s->hdr.count, -(count));
- be16_add_cpu(&leaf_s->hdr.stale, -(stale));
- be16_add_cpu(&leaf_d->hdr.count, count);
- be16_add_cpu(&leaf_d->hdr.stale, stale);
- xfs_dir2_leaf_log_header(tp, bp_s);
- xfs_dir2_leaf_log_header(tp, bp_d);
- xfs_dir2_leafn_check(args->dp, bp_s);
- xfs_dir2_leafn_check(args->dp, bp_d);
+ shdr->count -= count;
+ shdr->stale -= stale;
+ dhdr->count += count;
+ dhdr->stale += stale;
}
/*
@@ -922,21 +930,25 @@ xfs_dir2_leafn_moveents(
*/
int /* sort order */
xfs_dir2_leafn_order(
- struct xfs_buf *leaf1_bp, /* leaf1 buffer */
- struct xfs_buf *leaf2_bp) /* leaf2 buffer */
+ struct xfs_buf *leaf1_bp, /* leaf1 buffer */
+ struct xfs_buf *leaf2_bp) /* leaf2 buffer */
{
- xfs_dir2_leaf_t *leaf1; /* leaf1 structure */
- xfs_dir2_leaf_t *leaf2; /* leaf2 structure */
-
- leaf1 = leaf1_bp->b_addr;
- leaf2 = leaf2_bp->b_addr;
- ASSERT(leaf1->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
- ASSERT(leaf2->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
- if (be16_to_cpu(leaf1->hdr.count) > 0 &&
- be16_to_cpu(leaf2->hdr.count) > 0 &&
- (be32_to_cpu(leaf2->ents[0].hashval) < be32_to_cpu(leaf1->ents[0].hashval) ||
- be32_to_cpu(leaf2->ents[be16_to_cpu(leaf2->hdr.count) - 1].hashval) <
- be32_to_cpu(leaf1->ents[be16_to_cpu(leaf1->hdr.count) - 1].hashval)))
+ struct xfs_dir2_leaf *leaf1 = leaf1_bp->b_addr;
+ struct xfs_dir2_leaf *leaf2 = leaf2_bp->b_addr;
+ struct xfs_dir2_leaf_entry *ents1;
+ struct xfs_dir2_leaf_entry *ents2;
+ struct xfs_dir3_icleaf_hdr hdr1;
+ struct xfs_dir3_icleaf_hdr hdr2;
+
+ xfs_dir3_leaf_hdr_from_disk(&hdr1, leaf1);
+ xfs_dir3_leaf_hdr_from_disk(&hdr2, leaf2);
+ ents1 = xfs_dir3_leaf_ents_p(leaf1);
+ ents2 = xfs_dir3_leaf_ents_p(leaf2);
+
+ if (hdr1.count > 0 && hdr2.count > 0 &&
+ (be32_to_cpu(ents2[0].hashval) < be32_to_cpu(ents1[0].hashval) ||
+ be32_to_cpu(ents2[hdr2.count - 1].hashval) <
+ be32_to_cpu(ents1[hdr1.count - 1].hashval)))
return 1;
return 0;
}
@@ -965,6 +977,10 @@ xfs_dir2_leafn_rebalance(
#endif
int oldsum; /* old total leaf count */
int swap; /* swapped leaf blocks */
+ struct xfs_dir2_leaf_entry *ents1;
+ struct xfs_dir2_leaf_entry *ents2;
+ struct xfs_dir3_icleaf_hdr hdr1;
+ struct xfs_dir3_icleaf_hdr hdr2;
args = state->args;
/*
@@ -979,11 +995,17 @@ xfs_dir2_leafn_rebalance(
}
leaf1 = blk1->bp->b_addr;
leaf2 = blk2->bp->b_addr;
- oldsum = be16_to_cpu(leaf1->hdr.count) + be16_to_cpu(leaf2->hdr.count);
+ xfs_dir3_leaf_hdr_from_disk(&hdr1, leaf1);
+ xfs_dir3_leaf_hdr_from_disk(&hdr2, leaf2);
+ ents1 = xfs_dir3_leaf_ents_p(leaf1);
+ ents2 = xfs_dir3_leaf_ents_p(leaf2);
+
+ oldsum = hdr1.count + hdr2.count;
#ifdef DEBUG
- oldstale = be16_to_cpu(leaf1->hdr.stale) + be16_to_cpu(leaf2->hdr.stale);
+ oldstale = hdr1.stale + hdr2.stale;
#endif
mid = oldsum >> 1;
+
/*
* If the old leaf count was odd then the new one will be even,
* so we need to divide the new count evenly.
@@ -991,10 +1013,10 @@ xfs_dir2_leafn_rebalance(
if (oldsum & 1) {
xfs_dahash_t midhash; /* middle entry hash value */
- if (mid >= be16_to_cpu(leaf1->hdr.count))
- midhash = be32_to_cpu(leaf2->ents[mid - be16_to_cpu(leaf1->hdr.count)].hashval);
+ if (mid >= hdr1.count)
+ midhash = be32_to_cpu(ents2[mid - hdr1.count].hashval);
else
- midhash = be32_to_cpu(leaf1->ents[mid].hashval);
+ midhash = be32_to_cpu(ents1[mid].hashval);
isleft = args->hashval <= midhash;
}
/*
@@ -1008,30 +1030,42 @@ xfs_dir2_leafn_rebalance(
* Calculate moved entry count. Positive means left-to-right,
* negative means right-to-left. Then move the entries.
*/
- count = be16_to_cpu(leaf1->hdr.count) - mid + (isleft == 0);
+ count = hdr1.count - mid + (isleft == 0);
if (count > 0)
- xfs_dir2_leafn_moveents(args, blk1->bp,
- be16_to_cpu(leaf1->hdr.count) - count, blk2->bp, 0, count);
+ xfs_dir3_leafn_moveents(args, blk1->bp, &hdr1, ents1,
+ hdr1.count - count, blk2->bp,
+ &hdr2, ents2, 0, count);
else if (count < 0)
- xfs_dir2_leafn_moveents(args, blk2->bp, 0, blk1->bp,
- be16_to_cpu(leaf1->hdr.count), count);
- ASSERT(be16_to_cpu(leaf1->hdr.count) + be16_to_cpu(leaf2->hdr.count) == oldsum);
- ASSERT(be16_to_cpu(leaf1->hdr.stale) + be16_to_cpu(leaf2->hdr.stale) == oldstale);
+ xfs_dir3_leafn_moveents(args, blk2->bp, &hdr2, ents2, 0,
+ blk1->bp, &hdr1, ents1,
+ hdr1.count, count);
+
+ ASSERT(hdr1.count + hdr2.count == oldsum);
+ ASSERT(hdr1.stale + hdr2.stale == oldstale);
+
+ /* log the changes made when moving the entries */
+ xfs_dir3_leaf_hdr_to_disk(leaf1, &hdr1);
+ xfs_dir3_leaf_hdr_to_disk(leaf2, &hdr2);
+ xfs_dir3_leaf_log_header(args->trans, blk1->bp);
+ xfs_dir3_leaf_log_header(args->trans, blk2->bp);
+
+ xfs_dir3_leaf_check(args->dp->i_mount, blk1->bp);
+ xfs_dir3_leaf_check(args->dp->i_mount, blk2->bp);
+
/*
* Mark whether we're inserting into the old or new leaf.
*/
- if (be16_to_cpu(leaf1->hdr.count) < be16_to_cpu(leaf2->hdr.count))
+ if (hdr1.count < hdr2.count)
state->inleaf = swap;
- else if (be16_to_cpu(leaf1->hdr.count) > be16_to_cpu(leaf2->hdr.count))
+ else if (hdr1.count > hdr2.count)
state->inleaf = !swap;
else
- state->inleaf =
- swap ^ (blk1->index <= be16_to_cpu(leaf1->hdr.count));
+ state->inleaf = swap ^ (blk1->index <= hdr1.count);
/*
* Adjust the expected index for insertion.
*/
if (!state->inleaf)
- blk2->index = blk1->index - be16_to_cpu(leaf1->hdr.count);
+ blk2->index = blk1->index - hdr1.count;
/*
* Finally sanity check just to make sure we are not returning a
@@ -1153,6 +1187,8 @@ xfs_dir2_leafn_remove(
int needscan; /* need to rescan data frees */
xfs_trans_t *tp; /* transaction pointer */
struct xfs_dir2_data_free *bf; /* bestfree table */
+ struct xfs_dir3_icleaf_hdr leafhdr;
+ struct xfs_dir2_leaf_entry *ents;
trace_xfs_dir2_leafn_remove(args, index);
@@ -1160,11 +1196,14 @@ xfs_dir2_leafn_remove(
tp = args->trans;
mp = dp->i_mount;
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+
/*
* Point to the entry we're removing.
*/
- lep = &leaf->ents[index];
+ lep = &ents[index];
+
/*
* Extract the data block and offset from the entry.
*/
@@ -1172,14 +1211,18 @@ xfs_dir2_leafn_remove(
ASSERT(dblk->blkno == db);
off = xfs_dir2_dataptr_to_off(mp, be32_to_cpu(lep->address));
ASSERT(dblk->index == off);
+
/*
* Kill the leaf entry by marking it stale.
* Log the leaf block changes.
*/
- be16_add_cpu(&leaf->hdr.stale, 1);
- xfs_dir2_leaf_log_header(tp, bp);
+ leafhdr.stale++;
+ xfs_dir3_leaf_hdr_to_disk(leaf, &leafhdr);
+ xfs_dir3_leaf_log_header(tp, bp);
+
lep->address = cpu_to_be32(XFS_DIR2_NULL_DATAPTR);
- xfs_dir2_leaf_log_ents(tp, bp, index, index);
+ xfs_dir3_leaf_log_ents(tp, bp, index, index);
+
/*
* Make the data entry free. Keep track of the longest freespace
* in the data block in case it changes.
@@ -1267,15 +1310,13 @@ xfs_dir2_leafn_remove(
return error;
}
- xfs_dir2_leafn_check(dp, bp);
+ xfs_dir3_leaf_check(mp, bp);
/*
* Return indication of whether this leaf block is empty enough
* to justify trying to join it with a neighbor.
*/
- *rval =
- ((uint)sizeof(leaf->hdr) +
- (uint)sizeof(leaf->ents[0]) *
- (be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale))) <
+ *rval = (xfs_dir3_leaf_hdr_size(leaf) +
+ (uint)sizeof(ents[0]) * (leafhdr.count - leafhdr.stale)) <
mp->m_dir_magicpct;
return 0;
}
@@ -1308,11 +1349,11 @@ xfs_dir2_leafn_split(
/*
* Initialize the new leaf block.
*/
- error = xfs_dir2_leaf_init(args, xfs_dir2_da_to_db(mp, blkno),
- &newblk->bp, XFS_DIR2_LEAFN_MAGIC);
- if (error) {
+ error = xfs_dir3_leaf_get_buf(args, xfs_dir2_da_to_db(mp, blkno),
+ &newblk->bp, XFS_DIR2_LEAFN_MAGIC);
+ if (error)
return error;
- }
+
newblk->blkno = blkno;
newblk->magic = XFS_DIR2_LEAFN_MAGIC;
/*
@@ -1336,8 +1377,8 @@ xfs_dir2_leafn_split(
*/
oldblk->hashval = xfs_dir2_leafn_lasthash(oldblk->bp, NULL);
newblk->hashval = xfs_dir2_leafn_lasthash(newblk->bp, NULL);
- xfs_dir2_leafn_check(args->dp, oldblk->bp);
- xfs_dir2_leafn_check(args->dp, newblk->bp);
+ xfs_dir3_leaf_check(mp, oldblk->bp);
+ xfs_dir3_leaf_check(mp, newblk->bp);
return error;
}
@@ -1366,6 +1407,8 @@ xfs_dir2_leafn_toosmall(
xfs_da_blkinfo_t *info; /* leaf block header */
xfs_dir2_leaf_t *leaf; /* leaf structure */
int rval; /* result from path_shift */
+ struct xfs_dir3_icleaf_hdr leafhdr;
+ struct xfs_dir2_leaf_entry *ents;
/*
* Check for the degenerate case of the block being over 50% full.
@@ -1374,10 +1417,12 @@ xfs_dir2_leafn_toosmall(
*/
blk = &state->path.blk[state->path.active - 1];
info = blk->bp->b_addr;
- ASSERT(info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
leaf = (xfs_dir2_leaf_t *)info;
- count = be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale);
- bytes = (uint)sizeof(leaf->hdr) + count * (uint)sizeof(leaf->ents[0]);
+ xfs_dir3_leaf_hdr_from_disk(&leafhdr, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+
+ count = leafhdr.count - leafhdr.stale;
+ bytes = xfs_dir3_leaf_hdr_size(leaf) + count * sizeof(ents[0]);
if (bytes > (state->blocksize >> 1)) {
/*
* Blk over 50%, don't try to join.
@@ -1396,7 +1441,7 @@ xfs_dir2_leafn_toosmall(
* Make altpath point to the block we want to keep and
* path point to the block we want to drop (this one).
*/
- forward = (info->forw != 0);
+ forward = (leafhdr.forw != 0);
memcpy(&state->altpath, &state->path, sizeof(state->path));
error = xfs_da_path_shift(state, &state->altpath, forward, 0,
&rval);
@@ -1412,15 +1457,17 @@ xfs_dir2_leafn_toosmall(
* We prefer coalescing with the lower numbered sibling so as
* to shrink a directory over time.
*/
- forward = be32_to_cpu(info->forw) < be32_to_cpu(info->back);
+ forward = leafhdr.forw < leafhdr.back;
for (i = 0, bp = NULL; i < 2; forward = !forward, i++) {
- blkno = forward ? be32_to_cpu(info->forw) : be32_to_cpu(info->back);
+ struct xfs_dir3_icleaf_hdr hdr2;
+
+ blkno = forward ? leafhdr.forw : leafhdr.back;
if (blkno == 0)
continue;
/*
* Read the sibling leaf block.
*/
- error = xfs_dir2_leafn_read(state->args->trans, state->args->dp,
+ error = xfs_dir3_leafn_read(state->args->trans, state->args->dp,
blkno, -1, &bp);
if (error)
return error;
@@ -1428,13 +1475,15 @@ xfs_dir2_leafn_toosmall(
/*
* Count bytes in the two blocks combined.
*/
- leaf = (xfs_dir2_leaf_t *)info;
- count = be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale);
+ count = leafhdr.count - leafhdr.stale;
bytes = state->blocksize - (state->blocksize >> 2);
+
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
- count += be16_to_cpu(leaf->hdr.count) - be16_to_cpu(leaf->hdr.stale);
- bytes -= count * (uint)sizeof(leaf->ents[0]);
+ xfs_dir3_leaf_hdr_from_disk(&hdr2, leaf);
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ count += hdr2.count - hdr2.stale;
+ bytes -= count * sizeof(ents[0]);
+
/*
* Fits with at least 25% to spare.
*/
@@ -1481,34 +1530,52 @@ xfs_dir2_leafn_unbalance(
xfs_da_args_t *args; /* operation arguments */
xfs_dir2_leaf_t *drop_leaf; /* dead leaf structure */
xfs_dir2_leaf_t *save_leaf; /* surviving leaf structure */
+ struct xfs_dir3_icleaf_hdr shdr;
+ struct xfs_dir3_icleaf_hdr dhdr;
+ struct xfs_dir2_leaf_entry *sents;
+ struct xfs_dir2_leaf_entry *dents;
args = state->args;
ASSERT(drop_blk->magic == XFS_DIR2_LEAFN_MAGIC);
ASSERT(save_blk->magic == XFS_DIR2_LEAFN_MAGIC);
drop_leaf = drop_blk->bp->b_addr;
save_leaf = save_blk->bp->b_addr;
- ASSERT(drop_leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
- ASSERT(save_leaf->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC));
+
+ xfs_dir3_leaf_hdr_from_disk(&shdr, save_leaf);
+ xfs_dir3_leaf_hdr_from_disk(&dhdr, drop_leaf);
+ sents = xfs_dir3_leaf_ents_p(save_leaf);
+ dents = xfs_dir3_leaf_ents_p(drop_leaf);
+
/*
* If there are any stale leaf entries, take this opportunity
* to purge them.
*/
- if (drop_leaf->hdr.stale)
- xfs_dir2_leaf_compact(args, drop_blk->bp);
- if (save_leaf->hdr.stale)
- xfs_dir2_leaf_compact(args, save_blk->bp);
+ if (dhdr.stale)
+ xfs_dir3_leaf_compact(args, &dhdr, drop_blk->bp);
+ if (shdr.stale)
+ xfs_dir3_leaf_compact(args, &shdr, save_blk->bp);
+
/*
* Move the entries from drop to the appropriate end of save.
*/
- drop_blk->hashval = be32_to_cpu(drop_leaf->ents[be16_to_cpu(drop_leaf->hdr.count) - 1].hashval);
+ drop_blk->hashval = be32_to_cpu(dents[dhdr.count - 1].hashval);
if (xfs_dir2_leafn_order(save_blk->bp, drop_blk->bp))
- xfs_dir2_leafn_moveents(args, drop_blk->bp, 0, save_blk->bp, 0,
- be16_to_cpu(drop_leaf->hdr.count));
+ xfs_dir3_leafn_moveents(args, drop_blk->bp, &dhdr, dents, 0,
+ save_blk->bp, &shdr, sents, 0,
+ dhdr.count);
else
- xfs_dir2_leafn_moveents(args, drop_blk->bp, 0, save_blk->bp,
- be16_to_cpu(save_leaf->hdr.count), be16_to_cpu(drop_leaf->hdr.count));
- save_blk->hashval = be32_to_cpu(save_leaf->ents[be16_to_cpu(save_leaf->hdr.count) - 1].hashval);
- xfs_dir2_leafn_check(args->dp, save_blk->bp);
+ xfs_dir3_leafn_moveents(args, drop_blk->bp, &dhdr, dents, 0,
+ save_blk->bp, &shdr, sents,
+ shdr.count, dhdr.count);
+ save_blk->hashval = be32_to_cpu(sents[shdr.count - 1].hashval);
+
+ /* log the changes made when moving the entries */
+ xfs_dir3_leaf_hdr_to_disk(save_leaf, &shdr);
+ xfs_dir3_leaf_hdr_to_disk(drop_leaf, &dhdr);
+ xfs_dir3_leaf_log_header(args->trans, save_blk->bp);
+ xfs_dir3_leaf_log_header(args->trans, drop_blk->bp);
+
+ xfs_dir3_leaf_check(args->dp->i_mount, save_blk->bp);
}
/*
@@ -2113,13 +2180,15 @@ xfs_dir2_node_replace(
* and locked it. But paranoia is good.
*/
if (rval == EEXIST) {
+ struct xfs_dir2_leaf_entry *ents;
/*
* Find the leaf entry.
*/
blk = &state->path.blk[state->path.active - 1];
ASSERT(blk->magic == XFS_DIR2_LEAFN_MAGIC);
leaf = blk->bp->b_addr;
- lep = &leaf->ents[blk->index];
+ ents = xfs_dir3_leaf_ents_p(leaf);
+ lep = &ents[blk->index];
ASSERT(state->extravalid);
/*
* Point to the data entry.
diff --git a/fs/xfs/xfs_dir2_priv.h b/fs/xfs/xfs_dir2_priv.h
index 910e644..932565d 100644
--- a/fs/xfs/xfs_dir2_priv.h
+++ b/fs/xfs/xfs_dir2_priv.h
@@ -77,24 +77,25 @@ extern void xfs_dir2_data_use_free(struct xfs_trans *tp, struct xfs_buf *bp,
xfs_dir2_data_aoff_t len, int *needlogp, int *needscanp);
/* xfs_dir2_leaf.c */
-extern const struct xfs_buf_ops xfs_dir2_leafn_buf_ops;
+extern const struct xfs_buf_ops xfs_dir3_leafn_buf_ops;
-extern int xfs_dir2_leafn_read(struct xfs_trans *tp, struct xfs_inode *dp,
+extern int xfs_dir3_leafn_read(struct xfs_trans *tp, struct xfs_inode *dp,
xfs_dablk_t fbno, xfs_daddr_t mappedbno, struct xfs_buf **bpp);
extern int xfs_dir2_block_to_leaf(struct xfs_da_args *args,
struct xfs_buf *dbp);
extern int xfs_dir2_leaf_addname(struct xfs_da_args *args);
-extern void xfs_dir2_leaf_compact(struct xfs_da_args *args,
- struct xfs_buf *bp);
-extern void xfs_dir2_leaf_compact_x1(struct xfs_buf *bp, int *indexp,
+extern void xfs_dir3_leaf_compact(struct xfs_da_args *args,
+ struct xfs_dir3_icleaf_hdr *leafhdr, struct xfs_buf *bp);
+extern void xfs_dir3_leaf_compact_x1(struct xfs_dir3_icleaf_hdr *leafhdr,
+ struct xfs_dir2_leaf_entry *ents, int *indexp,
int *lowstalep, int *highstalep, int *lowlogp, int *highlogp);
extern int xfs_dir2_leaf_getdents(struct xfs_inode *dp, void *dirent,
size_t bufsize, xfs_off_t *offset, filldir_t filldir);
-extern int xfs_dir2_leaf_init(struct xfs_da_args *args, xfs_dir2_db_t bno,
- struct xfs_buf **bpp, int magic);
-extern void xfs_dir2_leaf_log_ents(struct xfs_trans *tp, struct xfs_buf *bp,
+extern int xfs_dir3_leaf_get_buf(struct xfs_da_args *args, xfs_dir2_db_t bno,
+ struct xfs_buf **bpp, __uint16_t magic);
+extern void xfs_dir3_leaf_log_ents(struct xfs_trans *tp, struct xfs_buf *bp,
int first, int last);
-extern void xfs_dir2_leaf_log_header(struct xfs_trans *tp,
+extern void xfs_dir3_leaf_log_header(struct xfs_trans *tp,
struct xfs_buf *bp);
extern int xfs_dir2_leaf_lookup(struct xfs_da_args *args);
extern int xfs_dir2_leaf_removename(struct xfs_da_args *args);
@@ -104,11 +105,18 @@ extern int xfs_dir2_leaf_search_hash(struct xfs_da_args *args,
extern int xfs_dir2_leaf_trim_data(struct xfs_da_args *args,
struct xfs_buf *lbp, xfs_dir2_db_t db);
extern struct xfs_dir2_leaf_entry *
-xfs_dir2_leaf_find_entry(struct xfs_dir2_leaf *leaf, int index, int compact,
- int lowstale, int highstale,
- int *lfloglow, int *lfloghigh);
+xfs_dir3_leaf_find_entry(struct xfs_dir3_icleaf_hdr *leafhdr,
+ struct xfs_dir2_leaf_entry *ents, int index, int compact,
+ int lowstale, int highstale, int *lfloglow, int *lfloghigh);
extern int xfs_dir2_node_to_leaf(struct xfs_da_state *state);
+extern void xfs_dir3_leaf_hdr_from_disk(struct xfs_dir3_icleaf_hdr *to,
+ struct xfs_dir2_leaf *from);
+extern void xfs_dir3_leaf_hdr_to_disk(struct xfs_dir2_leaf *to,
+ struct xfs_dir3_icleaf_hdr *from);
+extern bool xfs_dir3_leaf_check_int(struct xfs_mount *mp,
+ struct xfs_dir3_icleaf_hdr *hdr, struct xfs_dir2_leaf *leaf);
+
/* xfs_dir2_node.c */
extern int xfs_dir2_leaf_to_node(struct xfs_da_args *args,
struct xfs_buf *lbp);
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 14/16] xfs: add CRCs to dir2/da node blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (12 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 13/16] xfs: add CRC checking to dir2 leaf blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 15/16] xfs: add CRCs to attr leaf blocks Dave Chinner
2013-02-25 1:31 ` [PATCH 16/16] xfs: add CRC checks to the superblock Dave Chinner
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_attr.c | 33 +-
fs/xfs/xfs_attr_leaf.c | 29 +-
fs/xfs/xfs_bmap.c | 1 +
fs/xfs/xfs_da_btree.c | 1372 +++++++++++++++++++++++++++++-------------------
fs/xfs/xfs_da_btree.h | 105 +++-
fs/xfs/xfs_dir2_node.c | 26 +-
fs/xfs/xfs_trace.c | 2 +-
7 files changed, 955 insertions(+), 613 deletions(-)
diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c
index 8886838..e03128c 100644
--- a/fs/xfs/xfs_attr.c
+++ b/fs/xfs/xfs_attr.c
@@ -15,7 +15,6 @@
* 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_types.h"
@@ -25,6 +24,7 @@
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_mount.h"
+#include "xfs_error.h"
#include "xfs_da_btree.h"
#include "xfs_bmap_btree.h"
#include "xfs_attr_sf.h"
@@ -35,7 +35,6 @@
#include "xfs_bmap.h"
#include "xfs_attr.h"
#include "xfs_attr_leaf.h"
-#include "xfs_error.h"
#include "xfs_quota.h"
#include "xfs_trans_space.h"
#include "xfs_vnodeops.h"
@@ -1236,7 +1235,7 @@ restart:
* Search to see if name already exists, and get back a pointer
* to where it should go.
*/
- error = xfs_da_node_lookup_int(state, &retval);
+ error = xfs_da3_node_lookup_int(state, &retval);
if (error)
goto out;
blk = &state->path.blk[ state->path.active-1 ];
@@ -1307,7 +1306,7 @@ restart:
* in the index2/blkno2/rmtblkno2/rmtblkcnt2 fields.
*/
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_da_split(state);
+ error = xfs_da3_split(state);
if (!error) {
error = xfs_bmap_finish(&args->trans, args->flist,
&committed);
@@ -1329,7 +1328,7 @@ restart:
/*
* Addition succeeded, update Btree hashvals.
*/
- xfs_da_fixhashpath(state, &state->path);
+ xfs_da3_fixhashpath(state, &state->path);
}
/*
@@ -1400,7 +1399,7 @@ restart:
state->blocksize = state->mp->m_sb.sb_blocksize;
state->node_ents = state->mp->m_attr_node_ents;
state->inleaf = 0;
- error = xfs_da_node_lookup_int(state, &retval);
+ error = xfs_da3_node_lookup_int(state, &retval);
if (error)
goto out;
@@ -1410,14 +1409,14 @@ restart:
blk = &state->path.blk[ state->path.active-1 ];
ASSERT(blk->magic == XFS_ATTR_LEAF_MAGIC);
error = xfs_attr_leaf_remove(blk->bp, args);
- xfs_da_fixhashpath(state, &state->path);
+ xfs_da3_fixhashpath(state, &state->path);
/*
* Check to see if the tree needs to be collapsed.
*/
if (retval && (state->path.active > 1)) {
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_da_join(state);
+ error = xfs_da3_join(state);
if (!error) {
error = xfs_bmap_finish(&args->trans,
args->flist,
@@ -1495,7 +1494,7 @@ xfs_attr_node_removename(xfs_da_args_t *args)
/*
* Search to see if name exists, and get back a pointer to it.
*/
- error = xfs_da_node_lookup_int(state, &retval);
+ error = xfs_da3_node_lookup_int(state, &retval);
if (error || (retval != EEXIST)) {
if (error == 0)
error = retval;
@@ -1546,14 +1545,14 @@ xfs_attr_node_removename(xfs_da_args_t *args)
blk = &state->path.blk[ state->path.active-1 ];
ASSERT(blk->magic == XFS_ATTR_LEAF_MAGIC);
retval = xfs_attr_leaf_remove(blk->bp, args);
- xfs_da_fixhashpath(state, &state->path);
+ xfs_da3_fixhashpath(state, &state->path);
/*
* Check to see if the tree needs to be collapsed.
*/
if (retval && (state->path.active > 1)) {
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_da_join(state);
+ error = xfs_da3_join(state);
if (!error) {
error = xfs_bmap_finish(&args->trans, args->flist,
&committed);
@@ -1699,7 +1698,7 @@ xfs_attr_refillstate(xfs_da_state_t *state)
ASSERT((path->active >= 0) && (path->active < XFS_DA_NODE_MAXDEPTH));
for (blk = path->blk, level = 0; level < path->active; blk++, level++) {
if (blk->disk_blkno) {
- error = xfs_da_node_read(state->args->trans,
+ error = xfs_da3_node_read(state->args->trans,
state->args->dp,
blk->blkno, blk->disk_blkno,
&blk->bp, XFS_ATTR_FORK);
@@ -1718,7 +1717,7 @@ xfs_attr_refillstate(xfs_da_state_t *state)
ASSERT((path->active >= 0) && (path->active < XFS_DA_NODE_MAXDEPTH));
for (blk = path->blk, level = 0; level < path->active; blk++, level++) {
if (blk->disk_blkno) {
- error = xfs_da_node_read(state->args->trans,
+ error = xfs_da3_node_read(state->args->trans,
state->args->dp,
blk->blkno, blk->disk_blkno,
&blk->bp, XFS_ATTR_FORK);
@@ -1758,7 +1757,7 @@ xfs_attr_node_get(xfs_da_args_t *args)
/*
* Search to see if name exists, and get back a pointer to it.
*/
- error = xfs_da_node_lookup_int(state, &retval);
+ error = xfs_da3_node_lookup_int(state, &retval);
if (error) {
retval = error;
} else if (retval == EEXIST) {
@@ -1810,7 +1809,7 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
*/
bp = NULL;
if (cursor->blkno > 0) {
- error = xfs_da_node_read(NULL, context->dp, cursor->blkno, -1,
+ error = xfs_da3_node_read(NULL, context->dp, cursor->blkno, -1,
&bp, XFS_ATTR_FORK);
if ((error != 0) && (error != EFSCORRUPTED))
return(error);
@@ -1852,7 +1851,7 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
if (bp == NULL) {
cursor->blkno = 0;
for (;;) {
- error = xfs_da_node_read(NULL, context->dp,
+ error = xfs_da3_node_read(NULL, context->dp,
cursor->blkno, -1, &bp,
XFS_ATTR_FORK);
if (error)
@@ -1870,7 +1869,7 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
xfs_trans_brelse(NULL, bp);
return(XFS_ERROR(EFSCORRUPTED));
}
- btree = node->btree;
+ btree = xfs_da3_node_tree_p(node);
for (i = 0; i < be16_to_cpu(node->hdr.count);
btree++, i++) {
if (cursor->hashval
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index aa4765f..54b0dc8 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -910,6 +910,7 @@ xfs_attr_leaf_to_node(xfs_da_args_t *args)
struct xfs_buf *bp1, *bp2;
xfs_dablk_t blkno;
int error;
+ struct xfs_da_node_entry *btree;
trace_xfs_attr_leaf_to_node(args);
@@ -935,16 +936,16 @@ xfs_attr_leaf_to_node(xfs_da_args_t *args)
/*
* Set up the new root node.
*/
- error = xfs_da_node_create(args, 0, 1, &bp1, XFS_ATTR_FORK);
+ error = xfs_da3_node_create(args, 0, 1, &bp1, XFS_ATTR_FORK);
if (error)
goto out;
node = bp1->b_addr;
leaf = bp2->b_addr;
ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
/* both on-disk, don't endian-flip twice */
- node->btree[0].hashval =
- leaf->entries[be16_to_cpu(leaf->hdr.count)-1 ].hashval;
- node->btree[0].before = cpu_to_be32(blkno);
+ btree = xfs_da3_node_tree_p(node);
+ btree[0].hashval = leaf->entries[be16_to_cpu(leaf->hdr.count)-1 ].hashval;
+ btree[0].before = cpu_to_be32(blkno);
node->hdr.count = cpu_to_be16(1);
xfs_trans_log_buf(args->trans, bp1, 0, XFS_LBSIZE(dp->i_mount) - 1);
error = 0;
@@ -1032,7 +1033,7 @@ xfs_attr_leaf_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* NOTE: rebalance() currently depends on the 2nd block being empty.
*/
xfs_attr_leaf_rebalance(state, oldblk, newblk);
- error = xfs_da_blk_link(state, oldblk, newblk);
+ error = xfs_da3_blk_link(state, oldblk, newblk);
if (error)
return(error);
@@ -1660,7 +1661,7 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
*/
forward = (info->forw != 0);
memcpy(&state->altpath, &state->path, sizeof(state->path));
- error = xfs_da_path_shift(state, &state->altpath, forward,
+ error = xfs_da3_path_shift(state, &state->altpath, forward,
0, &retval);
if (error)
return(error);
@@ -1717,10 +1718,10 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
*/
memcpy(&state->altpath, &state->path, sizeof(state->path));
if (blkno < blk->blkno) {
- error = xfs_da_path_shift(state, &state->altpath, forward,
+ error = xfs_da3_path_shift(state, &state->altpath, forward,
0, &retval);
} else {
- error = xfs_da_path_shift(state, &state->path, forward,
+ error = xfs_da3_path_shift(state, &state->path, forward,
0, &retval);
}
if (error)
@@ -2783,7 +2784,7 @@ xfs_attr_root_inactive(xfs_trans_t **trans, xfs_inode_t *dp)
* the extents in reverse order the extent containing
* block 0 must still be there.
*/
- error = xfs_da_node_read(*trans, dp, 0, -1, &bp, XFS_ATTR_FORK);
+ error = xfs_da3_node_read(*trans, dp, 0, -1, &bp, XFS_ATTR_FORK);
if (error)
return(error);
blkno = XFS_BUF_ADDR(bp);
@@ -2836,6 +2837,7 @@ xfs_attr_node_inactive(
xfs_daddr_t parent_blkno, child_blkno;
int error, count, i;
struct xfs_buf *child_bp;
+ struct xfs_da_node_entry *btree;
/*
* Since this code is recursive (gasp!) we must protect ourselves.
@@ -2853,7 +2855,8 @@ xfs_attr_node_inactive(
xfs_trans_brelse(*trans, bp);
return(0);
}
- child_fsb = be32_to_cpu(node->btree[0].before);
+ btree = xfs_da3_node_tree_p(node);
+ child_fsb = be32_to_cpu(btree[0].before);
xfs_trans_brelse(*trans, bp); /* no locks for later trans */
/*
@@ -2868,7 +2871,7 @@ xfs_attr_node_inactive(
* traversal of the tree so we may deal with many blocks
* before we come back to this one.
*/
- error = xfs_da_node_read(*trans, dp, child_fsb, -2, &child_bp,
+ error = xfs_da3_node_read(*trans, dp, child_fsb, -2, &child_bp,
XFS_ATTR_FORK);
if (error)
return(error);
@@ -2909,11 +2912,11 @@ xfs_attr_node_inactive(
* child block number.
*/
if ((i+1) < count) {
- error = xfs_da_node_read(*trans, dp, 0, parent_blkno,
+ error = xfs_da3_node_read(*trans, dp, 0, parent_blkno,
&bp, XFS_ATTR_FORK);
if (error)
return(error);
- child_fsb = be32_to_cpu(node->btree[i+1].before);
+ child_fsb = be32_to_cpu(btree[i+1].before);
xfs_trans_brelse(*trans, bp);
}
/*
diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index 0fcb7f4..2844724 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -25,6 +25,7 @@
#include "xfs_sb.h"
#include "xfs_ag.h"
#include "xfs_dir2.h"
+#include "xfs_mount.h"
#include "xfs_da_btree.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c
index ebfb14a..f0ffc7c 100644
--- a/fs/xfs/xfs_da_btree.c
+++ b/fs/xfs/xfs_da_btree.c
@@ -38,6 +38,8 @@
#include "xfs_attr_leaf.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_cksum.h"
+#include "xfs_buf_item.h"
/*
* xfs_da_btree.c
@@ -52,69 +54,195 @@
/*
* Routines used for growing the Btree.
*/
-STATIC int xfs_da_root_split(xfs_da_state_t *state,
+STATIC int xfs_da3_root_split(xfs_da_state_t *state,
xfs_da_state_blk_t *existing_root,
xfs_da_state_blk_t *new_child);
-STATIC int xfs_da_node_split(xfs_da_state_t *state,
+STATIC int xfs_da3_node_split(xfs_da_state_t *state,
xfs_da_state_blk_t *existing_blk,
xfs_da_state_blk_t *split_blk,
xfs_da_state_blk_t *blk_to_add,
int treelevel,
int *result);
-STATIC void xfs_da_node_rebalance(xfs_da_state_t *state,
+STATIC void xfs_da3_node_rebalance(xfs_da_state_t *state,
xfs_da_state_blk_t *node_blk_1,
xfs_da_state_blk_t *node_blk_2);
-STATIC void xfs_da_node_add(xfs_da_state_t *state,
+STATIC void xfs_da3_node_add(xfs_da_state_t *state,
xfs_da_state_blk_t *old_node_blk,
xfs_da_state_blk_t *new_node_blk);
/*
* Routines used for shrinking the Btree.
*/
-STATIC int xfs_da_root_join(xfs_da_state_t *state,
+STATIC int xfs_da3_root_join(xfs_da_state_t *state,
xfs_da_state_blk_t *root_blk);
-STATIC int xfs_da_node_toosmall(xfs_da_state_t *state, int *retval);
-STATIC void xfs_da_node_remove(xfs_da_state_t *state,
+STATIC int xfs_da3_node_toosmall(xfs_da_state_t *state, int *retval);
+STATIC void xfs_da3_node_remove(xfs_da_state_t *state,
xfs_da_state_blk_t *drop_blk);
-STATIC void xfs_da_node_unbalance(xfs_da_state_t *state,
+STATIC void xfs_da3_node_unbalance(xfs_da_state_t *state,
xfs_da_state_blk_t *src_node_blk,
xfs_da_state_blk_t *dst_node_blk);
/*
* Utility routines.
*/
-STATIC uint xfs_da_node_lasthash(struct xfs_buf *bp, int *count);
-STATIC int xfs_da_node_order(struct xfs_buf *node1_bp,
- struct xfs_buf *node2_bp);
-STATIC int xfs_da_blk_unlink(xfs_da_state_t *state,
+STATIC int xfs_da3_blk_unlink(xfs_da_state_t *state,
xfs_da_state_blk_t *drop_blk,
xfs_da_state_blk_t *save_blk);
-STATIC void xfs_da_state_kill_altpath(xfs_da_state_t *state);
-static void
-xfs_da_node_verify(
+
+kmem_zone_t *xfs_da_state_zone; /* anchor for state struct zone */
+
+/*
+ * Allocate a dir-state structure.
+ * We don't put them on the stack since they're large.
+ */
+xfs_da_state_t *
+xfs_da_state_alloc(void)
+{
+ return kmem_zone_zalloc(xfs_da_state_zone, KM_NOFS);
+}
+
+/*
+ * Kill the altpath contents of a da-state structure.
+ */
+STATIC void
+xfs_da_state_kill_altpath(xfs_da_state_t *state)
+{
+ int i;
+
+ for (i = 0; i < state->altpath.active; i++)
+ state->altpath.blk[i].bp = NULL;
+ state->altpath.active = 0;
+}
+
+/*
+ * Free a da-state structure.
+ */
+void
+xfs_da_state_free(xfs_da_state_t *state)
+{
+ xfs_da_state_kill_altpath(state);
+#ifdef DEBUG
+ memset((char *)state, 0, sizeof(*state));
+#endif /* DEBUG */
+ kmem_zone_free(xfs_da_state_zone, state);
+}
+
+void
+xfs_da3_node_hdr_from_disk(
+ struct xfs_da3_icnode_hdr *to,
+ struct xfs_da_intnode *from)
+{
+ ASSERT(from->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
+ from->hdr.info.magic == cpu_to_be16(XFS_DA3_NODE_MAGIC));
+
+ if (from->hdr.info.magic == cpu_to_be16(XFS_DA3_NODE_MAGIC)) {
+ struct xfs_da3_node_hdr *hdr3 = (struct xfs_da3_node_hdr *)from;
+
+ to->forw = be32_to_cpu(hdr3->info.hdr.forw);
+ to->back = be32_to_cpu(hdr3->info.hdr.back);
+ to->magic = be16_to_cpu(hdr3->info.hdr.magic);
+ to->count = be16_to_cpu(hdr3->count);
+ to->level = be16_to_cpu(hdr3->level);
+ return;
+ }
+ to->forw = be32_to_cpu(from->hdr.info.forw);
+ to->back = be32_to_cpu(from->hdr.info.back);
+ to->magic = be16_to_cpu(from->hdr.info.magic);
+ to->count = be16_to_cpu(from->hdr.count);
+ to->level = be16_to_cpu(from->hdr.level);
+}
+
+void
+xfs_da3_node_hdr_to_disk(
+ struct xfs_da_intnode *to,
+ struct xfs_da3_icnode_hdr *from)
+{
+ ASSERT(from->magic == XFS_DA_NODE_MAGIC ||
+ from->magic == XFS_DA3_NODE_MAGIC);
+
+ if (from->magic == XFS_DA3_NODE_MAGIC) {
+ struct xfs_da3_node_hdr *hdr3 = (struct xfs_da3_node_hdr *)to;
+
+ hdr3->info.hdr.forw = cpu_to_be32(from->forw);
+ hdr3->info.hdr.back = cpu_to_be32(from->back);
+ hdr3->info.hdr.magic = cpu_to_be16(from->magic);
+ hdr3->count = cpu_to_be16(from->count);
+ hdr3->level = cpu_to_be16(from->level);
+ return;
+ }
+ to->hdr.info.forw = cpu_to_be32(from->forw);
+ to->hdr.info.back = cpu_to_be32(from->back);
+ to->hdr.info.magic = cpu_to_be16(from->magic);
+ to->hdr.count = cpu_to_be16(from->count);
+ to->hdr.level = cpu_to_be16(from->level);
+}
+
+static bool
+xfs_da3_node_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
- struct xfs_da_node_hdr *hdr = bp->b_addr;
- int block_ok = 0;
-
- block_ok = hdr->info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC);
- block_ok = block_ok &&
- be16_to_cpu(hdr->level) > 0 &&
- be16_to_cpu(hdr->count) > 0 ;
- if (!block_ok) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
+ struct xfs_da_intnode *hdr = bp->b_addr;
+ struct xfs_da3_icnode_hdr ichdr;
+
+ xfs_da3_node_hdr_from_disk(&ichdr, hdr);
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_da3_node_hdr *hdr3 = bp->b_addr;
+
+ if (ichdr.magic != XFS_DA3_NODE_MAGIC)
+ return false;
+
+ if (!uuid_equal(&hdr3->info.uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be64_to_cpu(hdr3->info.blkno) != bp->b_bn)
+ return false;
+ } else {
+ if (ichdr.magic != XFS_DA_NODE_MAGIC)
+ return false;
}
+ if (ichdr.level == 0)
+ return false;
+ if (ichdr.level > XFS_DA_NODE_MAXDEPTH)
+ return false;
+ if (ichdr.count == 0)
+ return false;
+
+ /*
+ * we don't know if the node is for and attribute or directory tree,
+ * so only fail if the count is outside both bounds
+ */
+ if (ichdr.count > mp->m_dir_node_ents &&
+ ichdr.count > mp->m_attr_node_ents)
+ return false;
+
+ /* XXX: hash order check? */
+ return true;
}
static void
-xfs_da_node_write_verify(
+xfs_da3_node_write_verify(
struct xfs_buf *bp)
{
- xfs_da_node_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+ struct xfs_da3_node_hdr *hdr3 = bp->b_addr;
+
+ if (!xfs_da3_node_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ hdr3->info.lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_DA3_NODE_CRC_OFF);
}
/*
@@ -124,16 +252,22 @@ xfs_da_node_write_verify(
* format of the block being read.
*/
static void
-xfs_da_node_read_verify(
+xfs_da3_node_read_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_da_blkinfo *info = bp->b_addr;
switch (be16_to_cpu(info->magic)) {
+ case XFS_DA3_NODE_MAGIC:
+ if (!xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_DA3_NODE_CRC_OFF))
+ break;
+ /* fall through */
case XFS_DA_NODE_MAGIC:
- xfs_da_node_verify(bp);
- break;
+ if (!xfs_da3_node_verify(bp))
+ break;
+ return;
case XFS_ATTR_LEAF_MAGIC:
bp->b_ops = &xfs_attr_leaf_buf_ops;
bp->b_ops->verify_read(bp);
@@ -144,21 +278,22 @@ xfs_da_node_read_verify(
bp->b_ops->verify_read(bp);
return;
default:
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW,
- mp, info);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
break;
}
+
+ /* corrupt block */
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
}
-const struct xfs_buf_ops xfs_da_node_buf_ops = {
- .verify_read = xfs_da_node_read_verify,
- .verify_write = xfs_da_node_write_verify,
+const struct xfs_buf_ops xfs_da3_node_buf_ops = {
+ .verify_read = xfs_da3_node_read_verify,
+ .verify_write = xfs_da3_node_write_verify,
};
int
-xfs_da_node_read(
+xfs_da3_node_read(
struct xfs_trans *tp,
struct xfs_inode *dp,
xfs_dablk_t bno,
@@ -167,7 +302,7 @@ xfs_da_node_read(
int which_fork)
{
return xfs_da_read_buf(tp, dp, bno, mappedbno, bpp,
- which_fork, &xfs_da_node_buf_ops);
+ which_fork, &xfs_da3_node_buf_ops);
}
/*========================================================================
@@ -178,33 +313,44 @@ xfs_da_node_read(
* Create the initial contents of an intermediate node.
*/
int
-xfs_da_node_create(xfs_da_args_t *args, xfs_dablk_t blkno, int level,
- struct xfs_buf **bpp, int whichfork)
+xfs_da3_node_create(
+ struct xfs_da_args *args,
+ xfs_dablk_t blkno,
+ int level,
+ struct xfs_buf **bpp,
+ int whichfork)
{
- xfs_da_intnode_t *node;
- struct xfs_buf *bp;
- int error;
- xfs_trans_t *tp;
+ struct xfs_da_intnode *node;
+ struct xfs_trans *tp = args->trans;
+ struct xfs_mount *mp = tp->t_mountp;
+ struct xfs_da3_icnode_hdr ichdr = {0};
+ struct xfs_buf *bp;
+ int error;
trace_xfs_da_node_create(args);
- tp = args->trans;
error = xfs_da_get_buf(tp, args->dp, blkno, -1, &bp, whichfork);
if (error)
return(error);
- ASSERT(bp != NULL);
node = bp->b_addr;
- node->hdr.info.forw = 0;
- node->hdr.info.back = 0;
- node->hdr.info.magic = cpu_to_be16(XFS_DA_NODE_MAGIC);
- node->hdr.info.pad = 0;
- node->hdr.count = 0;
- node->hdr.level = cpu_to_be16(level);
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_da3_node_hdr *hdr3 = bp->b_addr;
+
+ ichdr.magic = XFS_DA3_NODE_MAGIC;
+ hdr3->info.blkno = cpu_to_be64(bp->b_bn);
+ hdr3->info.owner = cpu_to_be64(args->dp->i_ino);
+ uuid_copy(&hdr3->info.uuid, &mp->m_sb.sb_uuid);
+ } else {
+ ichdr.magic = XFS_DA_NODE_MAGIC;
+ }
+ ichdr.level = level;
+
+ xfs_da3_node_hdr_to_disk(node, &ichdr);
xfs_trans_log_buf(tp, bp,
XFS_DA_LOGRANGE(node, &node->hdr, sizeof(node->hdr)));
- bp->b_ops = &xfs_da_node_buf_ops;
+ bp->b_ops = &xfs_da3_node_buf_ops;
*bpp = bp;
return(0);
}
@@ -214,12 +360,18 @@ xfs_da_node_create(xfs_da_args_t *args, xfs_dablk_t blkno, int level,
* intermediate nodes, rebalance, etc.
*/
int /* error */
-xfs_da_split(xfs_da_state_t *state)
+xfs_da3_split(
+ struct xfs_da_state *state)
{
- xfs_da_state_blk_t *oldblk, *newblk, *addblk;
- xfs_da_intnode_t *node;
- struct xfs_buf *bp;
- int max, action, error, i;
+ struct xfs_da_state_blk *oldblk;
+ struct xfs_da_state_blk *newblk;
+ struct xfs_da_state_blk *addblk;
+ struct xfs_da_intnode *node;
+ struct xfs_buf *bp;
+ int max;
+ int action;
+ int error;
+ int i;
trace_xfs_da_split(state->args);
@@ -281,7 +433,7 @@ xfs_da_split(xfs_da_state_t *state)
addblk = newblk;
break;
case XFS_DA_NODE_MAGIC:
- error = xfs_da_node_split(state, oldblk, newblk, addblk,
+ error = xfs_da3_node_split(state, oldblk, newblk, addblk,
max - i, &action);
addblk->bp = NULL;
if (error)
@@ -299,7 +451,7 @@ xfs_da_split(xfs_da_state_t *state)
/*
* Update the btree to show the new hashval for this child.
*/
- xfs_da_fixhashpath(state, &state->path);
+ xfs_da3_fixhashpath(state, &state->path);
}
if (!addblk)
return(0);
@@ -309,7 +461,7 @@ xfs_da_split(xfs_da_state_t *state)
*/
ASSERT(state->path.active == 0);
oldblk = &state->path.blk[0];
- error = xfs_da_root_split(state, oldblk, addblk);
+ error = xfs_da3_root_split(state, oldblk, addblk);
if (error) {
addblk->bp = NULL;
return(error); /* GROT: dir is inconsistent */
@@ -320,8 +472,10 @@ xfs_da_split(xfs_da_state_t *state)
* just got bumped because of the addition of a new root node.
* There might be three blocks involved if a double split occurred,
* and the original block 0 could be at any position in the list.
+ *
+ * Note: the info structures being modified here for both v2 and v3 da
+ * headers, so we can do this linkage just using the v2 structures.
*/
-
node = oldblk->bp->b_addr;
if (node->hdr.info.forw) {
if (be32_to_cpu(node->hdr.info.forw) == addblk->blkno) {
@@ -360,18 +514,25 @@ xfs_da_split(xfs_da_state_t *state)
* the EOF, extending the inode in process.
*/
STATIC int /* error */
-xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
- xfs_da_state_blk_t *blk2)
+xfs_da3_root_split(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *blk1,
+ struct xfs_da_state_blk *blk2)
{
- xfs_da_intnode_t *node, *oldroot;
- xfs_da_args_t *args;
- xfs_dablk_t blkno;
- struct xfs_buf *bp;
- int error, size;
- xfs_inode_t *dp;
- xfs_trans_t *tp;
- xfs_mount_t *mp;
- xfs_dir2_leaf_t *leaf;
+ struct xfs_da_intnode *node;
+ struct xfs_da_intnode *oldroot;
+ struct xfs_da_node_entry *btree;
+ struct xfs_da3_icnode_hdr nodehdr;
+ struct xfs_da_args *args;
+ struct xfs_buf *bp;
+ struct xfs_inode *dp;
+ struct xfs_trans *tp;
+ struct xfs_mount *mp;
+ struct xfs_dir2_leaf *leaf;
+ xfs_dablk_t blkno;
+ int level;
+ int error;
+ int size;
trace_xfs_da_root_split(state->args);
@@ -380,22 +541,26 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* to a free space somewhere.
*/
args = state->args;
- ASSERT(args != NULL);
error = xfs_da_grow_inode(args, &blkno);
if (error)
- return(error);
+ return error;
+
dp = args->dp;
tp = args->trans;
mp = state->mp;
error = xfs_da_get_buf(tp, dp, blkno, -1, &bp, args->whichfork);
if (error)
- return(error);
- ASSERT(bp != NULL);
+ return error;
node = bp->b_addr;
oldroot = blk1->bp->b_addr;
- if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC)) {
- size = (int)((char *)&oldroot->btree[be16_to_cpu(oldroot->hdr.count)] -
- (char *)oldroot);
+ if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
+ oldroot->hdr.info.magic == cpu_to_be16(XFS_DA3_NODE_MAGIC)) {
+ struct xfs_da3_icnode_hdr nodehdr;
+
+ xfs_da3_node_hdr_from_disk(&nodehdr, oldroot);
+ btree = xfs_da3_node_tree_p(oldroot);
+ size = (int)((char *)&btree[nodehdr.count] - (char *)oldroot);
+ level = nodehdr.level;
} else {
struct xfs_dir3_icleaf_hdr leafhdr;
struct xfs_dir2_leaf_entry *ents;
@@ -407,9 +572,22 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
ASSERT(leafhdr.magic == XFS_DIR2_LEAFN_MAGIC ||
leafhdr.magic == XFS_DIR3_LEAFN_MAGIC);
size = (int)((char *)&ents[leafhdr.count] - (char *)leaf);
+ level = 0;
}
- /* XXX: can't just copy CRC headers from one block to another */
+
+ /*
+ * we can copy most of the information in the node from one block to
+ * another, but for CRC enabled headers we have to make sure that the
+ * block specific identifiers are kept intact. We update the buffer
+ * directly for this.
+ */
memcpy(node, oldroot, size);
+ if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DA3_NODE_MAGIC) ||
+ oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC)) {
+ struct xfs_da3_intnode *node3 = (struct xfs_da3_intnode *)node;
+
+ node3->hdr.info.blkno = cpu_to_be64(bp->b_bn);
+ }
xfs_trans_log_buf(tp, bp, 0, size - 1);
bp->b_ops = blk1->bp->b_ops;
@@ -419,17 +597,21 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
/*
* Set up the new root node.
*/
- error = xfs_da_node_create(args,
+ error = xfs_da3_node_create(args,
(args->whichfork == XFS_DATA_FORK) ? mp->m_dirleafblk : 0,
- be16_to_cpu(node->hdr.level) + 1, &bp, args->whichfork);
+ level + 1, &bp, args->whichfork);
if (error)
- return(error);
+ return error;
+
node = bp->b_addr;
- node->btree[0].hashval = cpu_to_be32(blk1->hashval);
- node->btree[0].before = cpu_to_be32(blk1->blkno);
- node->btree[1].hashval = cpu_to_be32(blk2->hashval);
- node->btree[1].before = cpu_to_be32(blk2->blkno);
- node->hdr.count = cpu_to_be16(2);
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ btree = xfs_da3_node_tree_p(node);
+ btree[0].hashval = cpu_to_be32(blk1->hashval);
+ btree[0].before = cpu_to_be32(blk1->blkno);
+ btree[1].hashval = cpu_to_be32(blk2->hashval);
+ btree[1].before = cpu_to_be32(blk2->blkno);
+ nodehdr.count = 2;
+ xfs_da3_node_hdr_to_disk(node, &nodehdr);
#ifdef DEBUG
if (oldroot->hdr.info.magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
@@ -443,30 +625,34 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
/* Header is already logged by xfs_da_node_create */
xfs_trans_log_buf(tp, bp,
- XFS_DA_LOGRANGE(node, node->btree,
- sizeof(xfs_da_node_entry_t) * 2));
+ XFS_DA_LOGRANGE(node, btree, sizeof(xfs_da_node_entry_t) * 2));
- return(0);
+ return 0;
}
/*
* Split the node, rebalance, then add the new entry.
*/
STATIC int /* error */
-xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
- xfs_da_state_blk_t *newblk,
- xfs_da_state_blk_t *addblk,
- int treelevel, int *result)
+xfs_da3_node_split(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *oldblk,
+ struct xfs_da_state_blk *newblk,
+ struct xfs_da_state_blk *addblk,
+ int treelevel,
+ int *result)
{
- xfs_da_intnode_t *node;
- xfs_dablk_t blkno;
- int newcount, error;
- int useextra;
+ struct xfs_da_intnode *node;
+ struct xfs_da3_icnode_hdr nodehdr;
+ xfs_dablk_t blkno;
+ int newcount;
+ int error;
+ int useextra;
trace_xfs_da_node_split(state->args);
node = oldblk->bp->b_addr;
- ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
/*
* With V2 dirs the extra block is data or freespace.
@@ -476,7 +662,7 @@ xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
/*
* Do we have to split the node?
*/
- if ((be16_to_cpu(node->hdr.count) + newcount) > state->node_ents) {
+ if (nodehdr.count + newcount > state->node_ents) {
/*
* Allocate a new node, add to the doubly linked chain of
* nodes, then move some of our excess entries into it.
@@ -485,14 +671,14 @@ xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
if (error)
return(error); /* GROT: dir is inconsistent */
- error = xfs_da_node_create(state->args, blkno, treelevel,
+ error = xfs_da3_node_create(state->args, blkno, treelevel,
&newblk->bp, state->args->whichfork);
if (error)
return(error); /* GROT: dir is inconsistent */
newblk->blkno = blkno;
newblk->magic = XFS_DA_NODE_MAGIC;
- xfs_da_node_rebalance(state, oldblk, newblk);
- error = xfs_da_blk_link(state, oldblk, newblk);
+ xfs_da3_node_rebalance(state, oldblk, newblk);
+ error = xfs_da3_blk_link(state, oldblk, newblk);
if (error)
return(error);
*result = 1;
@@ -504,7 +690,7 @@ xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* Insert the new entry(s) into the correct block
* (updating last hashval in the process).
*
- * xfs_da_node_add() inserts BEFORE the given index,
+ * xfs_da3_node_add() inserts BEFORE the given index,
* and as a result of using node_lookup_int() we always
* point to a valid entry (not after one), but a split
* operation always results in a new block whose hashvals
@@ -513,22 +699,23 @@ xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* If we had double-split op below us, then add the extra block too.
*/
node = oldblk->bp->b_addr;
- if (oldblk->index <= be16_to_cpu(node->hdr.count)) {
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ if (oldblk->index <= nodehdr.count) {
oldblk->index++;
- xfs_da_node_add(state, oldblk, addblk);
+ xfs_da3_node_add(state, oldblk, addblk);
if (useextra) {
if (state->extraafter)
oldblk->index++;
- xfs_da_node_add(state, oldblk, &state->extrablk);
+ xfs_da3_node_add(state, oldblk, &state->extrablk);
state->extravalid = 0;
}
} else {
newblk->index++;
- xfs_da_node_add(state, newblk, addblk);
+ xfs_da3_node_add(state, newblk, addblk);
if (useextra) {
if (state->extraafter)
newblk->index++;
- xfs_da_node_add(state, newblk, &state->extrablk);
+ xfs_da3_node_add(state, newblk, &state->extrablk);
state->extravalid = 0;
}
}
@@ -543,33 +730,53 @@ xfs_da_node_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* NOTE: if blk2 is empty, then it will get the upper half of blk1.
*/
STATIC void
-xfs_da_node_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
- xfs_da_state_blk_t *blk2)
+xfs_da3_node_rebalance(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *blk1,
+ struct xfs_da_state_blk *blk2)
{
- xfs_da_intnode_t *node1, *node2, *tmpnode;
- xfs_da_node_entry_t *btree_s, *btree_d;
- int count, tmp;
- xfs_trans_t *tp;
+ struct xfs_da_intnode *node1;
+ struct xfs_da_intnode *node2;
+ struct xfs_da_intnode *tmpnode;
+ struct xfs_da_node_entry *btree1;
+ struct xfs_da_node_entry *btree2;
+ struct xfs_da_node_entry *btree_s;
+ struct xfs_da_node_entry *btree_d;
+ struct xfs_da3_icnode_hdr nodehdr1;
+ struct xfs_da3_icnode_hdr nodehdr2;
+ struct xfs_trans *tp;
+ int count;
+ int tmp;
+ int swap = 0;
trace_xfs_da_node_rebalance(state->args);
node1 = blk1->bp->b_addr;
node2 = blk2->bp->b_addr;
+ xfs_da3_node_hdr_from_disk(&nodehdr1, node1);
+ xfs_da3_node_hdr_from_disk(&nodehdr2, node2);
+ btree1 = xfs_da3_node_tree_p(node1);
+ btree2 = xfs_da3_node_tree_p(node2);
+
/*
* Figure out how many entries need to move, and in which direction.
* Swap the nodes around if that makes it simpler.
*/
- if ((be16_to_cpu(node1->hdr.count) > 0) && (be16_to_cpu(node2->hdr.count) > 0) &&
- ((be32_to_cpu(node2->btree[0].hashval) < be32_to_cpu(node1->btree[0].hashval)) ||
- (be32_to_cpu(node2->btree[be16_to_cpu(node2->hdr.count)-1].hashval) <
- be32_to_cpu(node1->btree[be16_to_cpu(node1->hdr.count)-1].hashval)))) {
+ if (nodehdr1.count > 0 && nodehdr2.count > 0 &&
+ ((be32_to_cpu(btree2[0].hashval) < be32_to_cpu(btree1[0].hashval)) ||
+ (be32_to_cpu(btree2[nodehdr2.count - 1].hashval) <
+ be32_to_cpu(btree1[nodehdr1.count - 1].hashval)))) {
tmpnode = node1;
node1 = node2;
node2 = tmpnode;
+ xfs_da3_node_hdr_from_disk(&nodehdr1, node1);
+ xfs_da3_node_hdr_from_disk(&nodehdr2, node2);
+ btree1 = xfs_da3_node_tree_p(node1);
+ btree2 = xfs_da3_node_tree_p(node2);
+ swap = 1;
}
- ASSERT(node1->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- ASSERT(node2->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- count = (be16_to_cpu(node1->hdr.count) - be16_to_cpu(node2->hdr.count)) / 2;
+
+ count = (nodehdr1.count - nodehdr2.count) / 2;
if (count == 0)
return;
tp = state->args->trans;
@@ -580,10 +787,11 @@ xfs_da_node_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
/*
* Move elements in node2 up to make a hole.
*/
- if ((tmp = be16_to_cpu(node2->hdr.count)) > 0) {
+ tmp = nodehdr2.count;
+ if (tmp > 0) {
tmp *= (uint)sizeof(xfs_da_node_entry_t);
- btree_s = &node2->btree[0];
- btree_d = &node2->btree[count];
+ btree_s = &btree2[0];
+ btree_d = &btree2[count];
memmove(btree_d, btree_s, tmp);
}
@@ -591,12 +799,12 @@ xfs_da_node_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* Move the req'd B-tree elements from high in node1 to
* low in node2.
*/
- be16_add_cpu(&node2->hdr.count, count);
+ nodehdr2.count += count;
tmp = count * (uint)sizeof(xfs_da_node_entry_t);
- btree_s = &node1->btree[be16_to_cpu(node1->hdr.count) - count];
- btree_d = &node2->btree[0];
+ btree_s = &btree1[nodehdr1.count- count];
+ btree_d = &btree2[0];
memcpy(btree_d, btree_s, tmp);
- be16_add_cpu(&node1->hdr.count, -count);
+ nodehdr1.count -= count;
} else {
/*
* Move the req'd B-tree elements from low in node2 to
@@ -604,49 +812,60 @@ xfs_da_node_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
*/
count = -count;
tmp = count * (uint)sizeof(xfs_da_node_entry_t);
- btree_s = &node2->btree[0];
- btree_d = &node1->btree[be16_to_cpu(node1->hdr.count)];
+ btree_s = &btree2[0];
+ btree_d = &btree1[nodehdr1.count];
memcpy(btree_d, btree_s, tmp);
- be16_add_cpu(&node1->hdr.count, count);
+ nodehdr1.count += count;
+
xfs_trans_log_buf(tp, blk1->bp,
XFS_DA_LOGRANGE(node1, btree_d, tmp));
/*
* Move elements in node2 down to fill the hole.
*/
- tmp = be16_to_cpu(node2->hdr.count) - count;
+ tmp = nodehdr2.count - count;
tmp *= (uint)sizeof(xfs_da_node_entry_t);
- btree_s = &node2->btree[count];
- btree_d = &node2->btree[0];
+ btree_s = &btree2[count];
+ btree_d = &btree2[0];
memmove(btree_d, btree_s, tmp);
- be16_add_cpu(&node2->hdr.count, -count);
+ nodehdr2.count -= count;
}
/*
* Log header of node 1 and all current bits of node 2.
*/
+ xfs_da3_node_hdr_to_disk(node1, &nodehdr1);
xfs_trans_log_buf(tp, blk1->bp,
- XFS_DA_LOGRANGE(node1, &node1->hdr, sizeof(node1->hdr)));
+ XFS_DA_LOGRANGE(node1, &node1->hdr,
+ xfs_da3_node_hdr_size(node1)));
+
+ xfs_da3_node_hdr_to_disk(node2, &nodehdr2);
xfs_trans_log_buf(tp, blk2->bp,
XFS_DA_LOGRANGE(node2, &node2->hdr,
- sizeof(node2->hdr) +
- sizeof(node2->btree[0]) * be16_to_cpu(node2->hdr.count)));
+ xfs_da3_node_hdr_size(node2) +
+ (sizeof(btree2[0]) * nodehdr2.count)));
/*
* Record the last hashval from each block for upward propagation.
* (note: don't use the swapped node pointers)
*/
- node1 = blk1->bp->b_addr;
- node2 = blk2->bp->b_addr;
- blk1->hashval = be32_to_cpu(node1->btree[be16_to_cpu(node1->hdr.count)-1].hashval);
- blk2->hashval = be32_to_cpu(node2->btree[be16_to_cpu(node2->hdr.count)-1].hashval);
+ if (swap) {
+ node1 = blk1->bp->b_addr;
+ node2 = blk2->bp->b_addr;
+ xfs_da3_node_hdr_from_disk(&nodehdr1, node1);
+ xfs_da3_node_hdr_from_disk(&nodehdr2, node2);
+ btree1 = xfs_da3_node_tree_p(node1);
+ btree2 = xfs_da3_node_tree_p(node2);
+ }
+ blk1->hashval = be32_to_cpu(btree1[nodehdr1.count - 1].hashval);
+ blk2->hashval = be32_to_cpu(btree2[nodehdr2.count - 1].hashval);
/*
* Adjust the expected index for insertion.
*/
- if (blk1->index >= be16_to_cpu(node1->hdr.count)) {
- blk2->index = blk1->index - be16_to_cpu(node1->hdr.count);
- blk1->index = be16_to_cpu(node1->hdr.count) + 1; /* make it invalid */
+ if (blk1->index >= nodehdr1.count) {
+ blk2->index = blk1->index - nodehdr1.count;
+ blk1->index = nodehdr1.count + 1; /* make it invalid */
}
}
@@ -654,18 +873,23 @@ xfs_da_node_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* Add a new entry to an intermediate node.
*/
STATIC void
-xfs_da_node_add(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
- xfs_da_state_blk_t *newblk)
+xfs_da3_node_add(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *oldblk,
+ struct xfs_da_state_blk *newblk)
{
- xfs_da_intnode_t *node;
- xfs_da_node_entry_t *btree;
- int tmp;
+ struct xfs_da_intnode *node;
+ struct xfs_da3_icnode_hdr nodehdr;
+ struct xfs_da_node_entry *btree;
+ int tmp;
trace_xfs_da_node_add(state->args);
node = oldblk->bp->b_addr;
- ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- ASSERT((oldblk->index >= 0) && (oldblk->index <= be16_to_cpu(node->hdr.count)));
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ btree = xfs_da3_node_tree_p(node);
+
+ ASSERT(oldblk->index >= 0 && oldblk->index <= nodehdr.count);
ASSERT(newblk->blkno != 0);
if (state->args->whichfork == XFS_DATA_FORK)
ASSERT(newblk->blkno >= state->mp->m_dirleafblk &&
@@ -675,23 +899,25 @@ xfs_da_node_add(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* We may need to make some room before we insert the new node.
*/
tmp = 0;
- btree = &node->btree[ oldblk->index ];
- if (oldblk->index < be16_to_cpu(node->hdr.count)) {
- tmp = (be16_to_cpu(node->hdr.count) - oldblk->index) * (uint)sizeof(*btree);
- memmove(btree + 1, btree, tmp);
+ if (oldblk->index < nodehdr.count) {
+ tmp = (nodehdr.count - oldblk->index) * (uint)sizeof(*btree);
+ memmove(&btree[oldblk->index + 1], &btree[oldblk->index], tmp);
}
- btree->hashval = cpu_to_be32(newblk->hashval);
- btree->before = cpu_to_be32(newblk->blkno);
+ btree[oldblk->index].hashval = cpu_to_be32(newblk->hashval);
+ btree[oldblk->index].before = cpu_to_be32(newblk->blkno);
xfs_trans_log_buf(state->args->trans, oldblk->bp,
- XFS_DA_LOGRANGE(node, btree, tmp + sizeof(*btree)));
- be16_add_cpu(&node->hdr.count, 1);
+ XFS_DA_LOGRANGE(node, &btree[oldblk->index],
+ tmp + sizeof(*btree)));
+
+ nodehdr.count += 1;
+ xfs_da3_node_hdr_to_disk(node, &nodehdr);
xfs_trans_log_buf(state->args->trans, oldblk->bp,
XFS_DA_LOGRANGE(node, &node->hdr, sizeof(node->hdr)));
/*
* Copy the last hash value from the oldblk to propagate upwards.
*/
- oldblk->hashval = be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1 ].hashval);
+ oldblk->hashval = be32_to_cpu(btree[nodehdr.count - 1].hashval);
}
/*========================================================================
@@ -703,14 +929,16 @@ xfs_da_node_add(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* possibly deallocating that block, etc...
*/
int
-xfs_da_join(xfs_da_state_t *state)
+xfs_da3_join(
+ struct xfs_da_state *state)
{
- xfs_da_state_blk_t *drop_blk, *save_blk;
- int action, error;
+ struct xfs_da_state_blk *drop_blk;
+ struct xfs_da_state_blk *save_blk;
+ int action = 0;
+ int error;
trace_xfs_da_join(state->args);
- action = 0;
drop_blk = &state->path.blk[ state->path.active-1 ];
save_blk = &state->altpath.blk[ state->path.active-1 ];
ASSERT(state->path.blk[0].magic == XFS_DA_NODE_MAGIC);
@@ -751,18 +979,18 @@ xfs_da_join(xfs_da_state_t *state)
* Remove the offending node, fixup hashvals,
* check for a toosmall neighbor.
*/
- xfs_da_node_remove(state, drop_blk);
- xfs_da_fixhashpath(state, &state->path);
- error = xfs_da_node_toosmall(state, &action);
+ xfs_da3_node_remove(state, drop_blk);
+ xfs_da3_fixhashpath(state, &state->path);
+ error = xfs_da3_node_toosmall(state, &action);
if (error)
return(error);
if (action == 0)
return 0;
- xfs_da_node_unbalance(state, drop_blk, save_blk);
+ xfs_da3_node_unbalance(state, drop_blk, save_blk);
break;
}
- xfs_da_fixhashpath(state, &state->altpath);
- error = xfs_da_blk_unlink(state, drop_blk, save_blk);
+ xfs_da3_fixhashpath(state, &state->altpath);
+ error = xfs_da3_blk_unlink(state, drop_blk, save_blk);
xfs_da_state_kill_altpath(state);
if (error)
return(error);
@@ -777,9 +1005,9 @@ xfs_da_join(xfs_da_state_t *state)
* we only have one entry in the root, make the child block
* the new root.
*/
- xfs_da_node_remove(state, drop_blk);
- xfs_da_fixhashpath(state, &state->path);
- error = xfs_da_root_join(state, &state->path.blk[0]);
+ xfs_da3_node_remove(state, drop_blk);
+ xfs_da3_fixhashpath(state, &state->path);
+ error = xfs_da3_root_join(state, &state->path.blk[0]);
return(error);
}
@@ -793,8 +1021,10 @@ xfs_da_blkinfo_onlychild_validate(struct xfs_da_blkinfo *blkinfo, __u16 level)
ASSERT(magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC) ||
magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- } else
- ASSERT(magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
+ } else {
+ ASSERT(magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
+ magic == cpu_to_be16(XFS_DA3_NODE_MAGIC));
+ }
ASSERT(!blkinfo->forw);
ASSERT(!blkinfo->back);
}
@@ -807,43 +1037,46 @@ xfs_da_blkinfo_onlychild_validate(struct xfs_da_blkinfo *blkinfo, __u16 level)
* the old root to block 0 as the new root node.
*/
STATIC int
-xfs_da_root_join(xfs_da_state_t *state, xfs_da_state_blk_t *root_blk)
+xfs_da3_root_join(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *root_blk)
{
- xfs_da_intnode_t *oldroot;
- xfs_da_args_t *args;
- xfs_dablk_t child;
- struct xfs_buf *bp;
- int error;
+ struct xfs_da_intnode *oldroot;
+ struct xfs_da_args *args;
+ xfs_dablk_t child;
+ struct xfs_buf *bp;
+ struct xfs_da3_icnode_hdr oldroothdr;
+ struct xfs_da_node_entry *btree;
+ int error;
trace_xfs_da_root_join(state->args);
- args = state->args;
- ASSERT(args != NULL);
ASSERT(root_blk->magic == XFS_DA_NODE_MAGIC);
+
+ args = state->args;
oldroot = root_blk->bp->b_addr;
- ASSERT(oldroot->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- ASSERT(!oldroot->hdr.info.forw);
- ASSERT(!oldroot->hdr.info.back);
+ xfs_da3_node_hdr_from_disk(&oldroothdr, oldroot);
+ ASSERT(oldroothdr.forw == 0);
+ ASSERT(oldroothdr.back == 0);
/*
* If the root has more than one child, then don't do anything.
*/
- if (be16_to_cpu(oldroot->hdr.count) > 1)
- return(0);
+ if (oldroothdr.count > 1)
+ return 0;
/*
* Read in the (only) child block, then copy those bytes into
* the root block's buffer and free the original child block.
*/
- child = be32_to_cpu(oldroot->btree[0].before);
+ btree = xfs_da3_node_tree_p(oldroot);
+ child = be32_to_cpu(btree[0].before);
ASSERT(child != 0);
- error = xfs_da_node_read(args->trans, args->dp, child, -1, &bp,
+ error = xfs_da3_node_read(args->trans, args->dp, child, -1, &bp,
args->whichfork);
if (error)
- return(error);
- ASSERT(bp != NULL);
- xfs_da_blkinfo_onlychild_validate(bp->b_addr,
- be16_to_cpu(oldroot->hdr.level));
+ return error;
+ xfs_da_blkinfo_onlychild_validate(bp->b_addr, oldroothdr.level);
/*
* This could be copying a leaf back into the root block in the case of
@@ -868,14 +1101,21 @@ xfs_da_root_join(xfs_da_state_t *state, xfs_da_state_blk_t *root_blk)
* If nothing can be done, return 0.
*/
STATIC int
-xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
+xfs_da3_node_toosmall(
+ struct xfs_da_state *state,
+ int *action)
{
- xfs_da_intnode_t *node;
- xfs_da_state_blk_t *blk;
- xfs_da_blkinfo_t *info;
- int count, forward, error, retval, i;
- xfs_dablk_t blkno;
- struct xfs_buf *bp;
+ struct xfs_da_intnode *node;
+ struct xfs_da_state_blk *blk;
+ struct xfs_da_blkinfo *info;
+ xfs_dablk_t blkno;
+ struct xfs_buf *bp;
+ struct xfs_da3_icnode_hdr nodehdr;
+ int count;
+ int forward;
+ int error;
+ int retval;
+ int i;
trace_xfs_da_node_toosmall(state->args);
@@ -888,8 +1128,8 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
info = blk->bp->b_addr;
ASSERT(info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
node = (xfs_da_intnode_t *)info;
- count = be16_to_cpu(node->hdr.count);
- if (count > (state->node_ents >> 1)) {
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ if (nodehdr.count > (state->node_ents >> 1)) {
*action = 0; /* blk over 50%, don't try to join */
return(0); /* blk over 50%, don't try to join */
}
@@ -900,14 +1140,14 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
* coalesce it with a sibling block. We choose (arbitrarily)
* to merge with the forward block unless it is NULL.
*/
- if (count == 0) {
+ if (nodehdr.count == 0) {
/*
* Make altpath point to the block we want to keep and
* path point to the block we want to drop (this one).
*/
forward = (info->forw != 0);
memcpy(&state->altpath, &state->path, sizeof(state->path));
- error = xfs_da_path_shift(state, &state->altpath, forward,
+ error = xfs_da3_path_shift(state, &state->altpath, forward,
0, &retval);
if (error)
return(error);
@@ -926,35 +1166,34 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
* We prefer coalescing with the lower numbered sibling so as
* to shrink a directory over time.
*/
+ count = state->node_ents;
+ count -= state->node_ents >> 2;
+ count -= nodehdr.count;
+
/* start with smaller blk num */
- forward = (be32_to_cpu(info->forw) < be32_to_cpu(info->back));
+ forward = nodehdr.forw < nodehdr.back;
for (i = 0; i < 2; forward = !forward, i++) {
if (forward)
- blkno = be32_to_cpu(info->forw);
+ blkno = nodehdr.forw;
else
- blkno = be32_to_cpu(info->back);
+ blkno = nodehdr.back;
if (blkno == 0)
continue;
- error = xfs_da_node_read(state->args->trans, state->args->dp,
+ error = xfs_da3_node_read(state->args->trans, state->args->dp,
blkno, -1, &bp, state->args->whichfork);
if (error)
return(error);
- ASSERT(bp != NULL);
- node = (xfs_da_intnode_t *)info;
- count = state->node_ents;
- count -= state->node_ents >> 2;
- count -= be16_to_cpu(node->hdr.count);
node = bp->b_addr;
- ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- count -= be16_to_cpu(node->hdr.count);
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
xfs_trans_brelse(state->args->trans, bp);
- if (count >= 0)
+
+ if (count - nodehdr.count >= 0)
break; /* fits with at least 25% to spare */
}
if (i >= 2) {
*action = 0;
- return(0);
+ return 0;
}
/*
@@ -963,28 +1202,42 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
*/
memcpy(&state->altpath, &state->path, sizeof(state->path));
if (blkno < blk->blkno) {
- error = xfs_da_path_shift(state, &state->altpath, forward,
+ error = xfs_da3_path_shift(state, &state->altpath, forward,
0, &retval);
- if (error) {
- return(error);
- }
- if (retval) {
- *action = 0;
- return(0);
- }
} else {
- error = xfs_da_path_shift(state, &state->path, forward,
+ error = xfs_da3_path_shift(state, &state->path, forward,
0, &retval);
- if (error) {
- return(error);
- }
- if (retval) {
- *action = 0;
- return(0);
- }
+ }
+ if (error)
+ return error;
+ if (retval) {
+ *action = 0;
+ return 0;
}
*action = 1;
- return(0);
+ return 0;
+}
+
+/*
+ * Pick up the last hashvalue from an intermediate node.
+ */
+STATIC uint
+xfs_da3_node_lasthash(
+ struct xfs_buf *bp,
+ int *count)
+{
+ struct xfs_da_intnode *node;
+ struct xfs_da_node_entry *btree;
+ struct xfs_da3_icnode_hdr nodehdr;
+
+ node = bp->b_addr;
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ if (count)
+ *count = nodehdr.count;
+ if (!nodehdr.count)
+ return 0;
+ btree = xfs_da3_node_tree_p(node);
+ return be32_to_cpu(btree[nodehdr.count - 1].hashval);
}
/*
@@ -992,13 +1245,16 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action)
* when we stop making changes, return.
*/
void
-xfs_da_fixhashpath(xfs_da_state_t *state, xfs_da_state_path_t *path)
+xfs_da3_fixhashpath(
+ struct xfs_da_state *state,
+ struct xfs_da_state_path *path)
{
- xfs_da_state_blk_t *blk;
- xfs_da_intnode_t *node;
- xfs_da_node_entry_t *btree;
- xfs_dahash_t lasthash=0;
- int level, count;
+ struct xfs_da_state_blk *blk;
+ struct xfs_da_intnode *node;
+ struct xfs_da_node_entry *btree;
+ xfs_dahash_t lasthash=0;
+ int level;
+ int count;
trace_xfs_da_fixhashpath(state->args);
@@ -1016,23 +1272,26 @@ xfs_da_fixhashpath(xfs_da_state_t *state, xfs_da_state_path_t *path)
return;
break;
case XFS_DA_NODE_MAGIC:
- lasthash = xfs_da_node_lasthash(blk->bp, &count);
+ lasthash = xfs_da3_node_lasthash(blk->bp, &count);
if (count == 0)
return;
break;
}
for (blk--, level--; level >= 0; blk--, level--) {
+ struct xfs_da3_icnode_hdr nodehdr;
+
node = blk->bp->b_addr;
- ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- btree = &node->btree[ blk->index ];
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ btree = xfs_da3_node_tree_p(node);
if (be32_to_cpu(btree->hashval) == lasthash)
break;
blk->hashval = lasthash;
- btree->hashval = cpu_to_be32(lasthash);
+ btree[blk->index].hashval = cpu_to_be32(lasthash);
xfs_trans_log_buf(state->args->trans, blk->bp,
- XFS_DA_LOGRANGE(node, btree, sizeof(*btree)));
+ XFS_DA_LOGRANGE(node, &btree[blk->index],
+ sizeof(*btree)));
- lasthash = be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1].hashval);
+ lasthash = be32_to_cpu(btree[nodehdr.count - 1].hashval);
}
}
@@ -1040,104 +1299,119 @@ xfs_da_fixhashpath(xfs_da_state_t *state, xfs_da_state_path_t *path)
* Remove an entry from an intermediate node.
*/
STATIC void
-xfs_da_node_remove(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk)
+xfs_da3_node_remove(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *drop_blk)
{
- xfs_da_intnode_t *node;
- xfs_da_node_entry_t *btree;
- int tmp;
+ struct xfs_da_intnode *node;
+ struct xfs_da3_icnode_hdr nodehdr;
+ struct xfs_da_node_entry *btree;
+ int index;
+ int tmp;
trace_xfs_da_node_remove(state->args);
node = drop_blk->bp->b_addr;
- ASSERT(drop_blk->index < be16_to_cpu(node->hdr.count));
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ ASSERT(drop_blk->index < nodehdr.count);
ASSERT(drop_blk->index >= 0);
/*
* Copy over the offending entry, or just zero it out.
*/
- btree = &node->btree[drop_blk->index];
- if (drop_blk->index < (be16_to_cpu(node->hdr.count)-1)) {
- tmp = be16_to_cpu(node->hdr.count) - drop_blk->index - 1;
+ index = drop_blk->index;
+ btree = xfs_da3_node_tree_p(node);
+ if (index < nodehdr.count - 1) {
+ tmp = nodehdr.count - index - 1;
tmp *= (uint)sizeof(xfs_da_node_entry_t);
- memmove(btree, btree + 1, tmp);
+ memmove(&btree[index], &btree[index + 1], tmp);
xfs_trans_log_buf(state->args->trans, drop_blk->bp,
- XFS_DA_LOGRANGE(node, btree, tmp));
- btree = &node->btree[be16_to_cpu(node->hdr.count)-1];
+ XFS_DA_LOGRANGE(node, &btree[index], tmp));
+ index = nodehdr.count - 1;
}
- memset((char *)btree, 0, sizeof(xfs_da_node_entry_t));
+ memset(&btree[index], 0, sizeof(xfs_da_node_entry_t));
xfs_trans_log_buf(state->args->trans, drop_blk->bp,
- XFS_DA_LOGRANGE(node, btree, sizeof(*btree)));
- be16_add_cpu(&node->hdr.count, -1);
+ XFS_DA_LOGRANGE(node, &btree[index], sizeof(btree[index])));
+ nodehdr.count -= 1;
+ xfs_da3_node_hdr_to_disk(node, &nodehdr);
xfs_trans_log_buf(state->args->trans, drop_blk->bp,
- XFS_DA_LOGRANGE(node, &node->hdr, sizeof(node->hdr)));
+ XFS_DA_LOGRANGE(node, &node->hdr, xfs_da3_node_hdr_size(node)));
/*
* Copy the last hash value from the block to propagate upwards.
*/
- btree--;
- drop_blk->hashval = be32_to_cpu(btree->hashval);
+ drop_blk->hashval = be32_to_cpu(btree[index - 1].hashval);
}
/*
- * Unbalance the btree elements between two intermediate nodes,
+ * Unbalance the elements between two intermediate nodes,
* move all Btree elements from one node into another.
*/
STATIC void
-xfs_da_node_unbalance(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
- xfs_da_state_blk_t *save_blk)
+xfs_da3_node_unbalance(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *drop_blk,
+ struct xfs_da_state_blk *save_blk)
{
- xfs_da_intnode_t *drop_node, *save_node;
- xfs_da_node_entry_t *btree;
- int tmp;
- xfs_trans_t *tp;
+ struct xfs_da_intnode *drop_node;
+ struct xfs_da_intnode *save_node;
+ struct xfs_da_node_entry *dbtree;
+ struct xfs_da_node_entry *sbtree;
+ struct xfs_da3_icnode_hdr dhdr;
+ struct xfs_da3_icnode_hdr shdr;
+ struct xfs_trans *tp;
+ int sindex;
+ int tmp;
trace_xfs_da_node_unbalance(state->args);
drop_node = drop_blk->bp->b_addr;
save_node = save_blk->bp->b_addr;
- ASSERT(drop_node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- ASSERT(save_node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
+ xfs_da3_node_hdr_from_disk(&dhdr, drop_node);
+ xfs_da3_node_hdr_from_disk(&shdr, save_node);
+ dbtree = xfs_da3_node_tree_p(drop_node);
+ sbtree = xfs_da3_node_tree_p(save_node);
tp = state->args->trans;
/*
* If the dying block has lower hashvals, then move all the
* elements in the remaining block up to make a hole.
*/
- if ((be32_to_cpu(drop_node->btree[0].hashval) < be32_to_cpu(save_node->btree[ 0 ].hashval)) ||
- (be32_to_cpu(drop_node->btree[be16_to_cpu(drop_node->hdr.count)-1].hashval) <
- be32_to_cpu(save_node->btree[be16_to_cpu(save_node->hdr.count)-1].hashval)))
- {
- btree = &save_node->btree[be16_to_cpu(drop_node->hdr.count)];
- tmp = be16_to_cpu(save_node->hdr.count) * (uint)sizeof(xfs_da_node_entry_t);
- memmove(btree, &save_node->btree[0], tmp);
- btree = &save_node->btree[0];
+ if ((be32_to_cpu(dbtree[0].hashval) < be32_to_cpu(sbtree[ 0 ].hashval)) ||
+ (be32_to_cpu(dbtree[dhdr.count - 1].hashval) <
+ be32_to_cpu(sbtree[shdr.count - 1].hashval))) {
+ /* XXX: check this - is memmove dst correct? */
+ tmp = shdr.count * (uint)sizeof(xfs_da_node_entry_t);
+ memmove(&sbtree[dhdr.count], &sbtree[0], tmp);
+
+ sindex = 0;
xfs_trans_log_buf(tp, save_blk->bp,
- XFS_DA_LOGRANGE(save_node, btree,
- (be16_to_cpu(save_node->hdr.count) + be16_to_cpu(drop_node->hdr.count)) *
- sizeof(xfs_da_node_entry_t)));
+ XFS_DA_LOGRANGE(save_node, &sbtree[0],
+ (shdr.count + dhdr.count) *
+ sizeof(xfs_da_node_entry_t)));
} else {
- btree = &save_node->btree[be16_to_cpu(save_node->hdr.count)];
+ sindex = shdr.count;
xfs_trans_log_buf(tp, save_blk->bp,
- XFS_DA_LOGRANGE(save_node, btree,
- be16_to_cpu(drop_node->hdr.count) *
- sizeof(xfs_da_node_entry_t)));
+ XFS_DA_LOGRANGE(save_node, &sbtree[sindex],
+ dhdr.count * sizeof(xfs_da_node_entry_t)));
}
/*
* Move all the B-tree elements from drop_blk to save_blk.
*/
- tmp = be16_to_cpu(drop_node->hdr.count) * (uint)sizeof(xfs_da_node_entry_t);
- memcpy(btree, &drop_node->btree[0], tmp);
- be16_add_cpu(&save_node->hdr.count, be16_to_cpu(drop_node->hdr.count));
+ tmp = dhdr.count * (uint)sizeof(xfs_da_node_entry_t);
+ memcpy(&sbtree[sindex], &dbtree[0], tmp);
+ shdr.count += dhdr.count;
+ xfs_da3_node_hdr_to_disk(save_node, &shdr);
xfs_trans_log_buf(tp, save_blk->bp,
XFS_DA_LOGRANGE(save_node, &save_node->hdr,
- sizeof(save_node->hdr)));
+ xfs_da3_node_hdr_size(save_node)));
/*
* Save the last hashval in the remaining block for upward propagation.
*/
- save_blk->hashval = be32_to_cpu(save_node->btree[be16_to_cpu(save_node->hdr.count)-1].hashval);
+ save_blk->hashval = be32_to_cpu(sbtree[be16_to_cpu(save_node->hdr.count)-1].hashval);
}
/*========================================================================
@@ -1156,16 +1430,24 @@ xfs_da_node_unbalance(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
* pruned depth-first tree search.
*/
int /* error */
-xfs_da_node_lookup_int(xfs_da_state_t *state, int *result)
+xfs_da3_node_lookup_int(
+ struct xfs_da_state *state,
+ int *result)
{
- xfs_da_state_blk_t *blk;
- xfs_da_blkinfo_t *curr;
- xfs_da_intnode_t *node;
- xfs_da_node_entry_t *btree;
- xfs_dablk_t blkno;
- int probe, span, max, error, retval;
- xfs_dahash_t hashval, btreehashval;
- xfs_da_args_t *args;
+ struct xfs_da_state_blk *blk;
+ struct xfs_da_blkinfo *curr;
+ struct xfs_da_intnode *node;
+ struct xfs_da_node_entry *btree;
+ struct xfs_da3_icnode_hdr nodehdr;
+ struct xfs_da_args *args;
+ xfs_dablk_t blkno;
+ xfs_dahash_t hashval;
+ xfs_dahash_t btreehashval;
+ int probe;
+ int span;
+ int max;
+ int error;
+ int retval;
args = state->args;
@@ -1181,7 +1463,7 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result)
* Read the next node down in the tree.
*/
blk->blkno = blkno;
- error = xfs_da_node_read(args->trans, args->dp, blkno,
+ error = xfs_da3_node_read(args->trans, args->dp, blkno,
-1, &blk->bp, args->whichfork);
if (error) {
blk->blkno = 0;
@@ -1190,66 +1472,70 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result)
}
curr = blk->bp->b_addr;
blk->magic = be16_to_cpu(curr->magic);
- ASSERT(blk->magic == XFS_DA_NODE_MAGIC ||
- blk->magic == XFS_DIR2_LEAFN_MAGIC ||
- blk->magic == XFS_ATTR_LEAF_MAGIC);
+
+ if (blk->magic == XFS_ATTR_LEAF_MAGIC) {
+ blk->hashval = xfs_attr_leaf_lasthash(blk->bp, NULL);
+ break;
+ }
+
+ if (blk->magic == XFS_DIR2_LEAFN_MAGIC ||
+ blk->magic == XFS_DIR3_LEAFN_MAGIC) {
+ blk->magic = XFS_DIR2_LEAFN_MAGIC;
+ blk->hashval = xfs_dir2_leafn_lasthash(blk->bp, NULL);
+ break;
+ }
/*
* Search an intermediate node for a match.
*/
- if (blk->magic == XFS_DA_NODE_MAGIC) {
- node = blk->bp->b_addr;
- max = be16_to_cpu(node->hdr.count);
- blk->hashval = be32_to_cpu(node->btree[max-1].hashval);
+ node = blk->bp->b_addr;
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ btree = xfs_da3_node_tree_p(node);
- /*
- * Binary search. (note: small blocks will skip loop)
- */
- probe = span = max / 2;
- hashval = args->hashval;
- for (btree = &node->btree[probe]; span > 4;
- btree = &node->btree[probe]) {
- span /= 2;
- btreehashval = be32_to_cpu(btree->hashval);
- if (btreehashval < hashval)
- probe += span;
- else if (btreehashval > hashval)
- probe -= span;
- else
- break;
- }
- ASSERT((probe >= 0) && (probe < max));
- ASSERT((span <= 4) || (be32_to_cpu(btree->hashval) == hashval));
+ max = nodehdr.count;
+ blk->hashval = be32_to_cpu(btree[max - 1].hashval);
- /*
- * Since we may have duplicate hashval's, find the first
- * matching hashval in the node.
- */
- while ((probe > 0) && (be32_to_cpu(btree->hashval) >= hashval)) {
- btree--;
- probe--;
- }
- while ((probe < max) && (be32_to_cpu(btree->hashval) < hashval)) {
- btree++;
- probe++;
- }
+ /*
+ * Binary search. (note: small blocks will skip loop)
+ */
+ probe = span = max / 2;
+ hashval = args->hashval;
+ while (span > 4) {
+ span /= 2;
+ btreehashval = be32_to_cpu(btree[probe].hashval);
+ if (btreehashval < hashval)
+ probe += span;
+ else if (btreehashval > hashval)
+ probe -= span;
+ else
+ break;
+ }
+ ASSERT((probe >= 0) && (probe < max));
+ ASSERT((span <= 4) ||
+ (be32_to_cpu(btree[probe].hashval) == hashval));
- /*
- * Pick the right block to descend on.
- */
- if (probe == max) {
- blk->index = max-1;
- blkno = be32_to_cpu(node->btree[max-1].before);
- } else {
- blk->index = probe;
- blkno = be32_to_cpu(btree->before);
- }
- } else if (blk->magic == XFS_ATTR_LEAF_MAGIC) {
- blk->hashval = xfs_attr_leaf_lasthash(blk->bp, NULL);
- break;
- } else if (blk->magic == XFS_DIR2_LEAFN_MAGIC) {
- blk->hashval = xfs_dir2_leafn_lasthash(blk->bp, NULL);
- break;
+ /*
+ * Since we may have duplicate hashval's, find the first
+ * matching hashval in the node.
+ */
+ while (probe > 0 &&
+ be32_to_cpu(btree[probe].hashval) >= hashval) {
+ probe--;
+ }
+ while (probe < max &&
+ be32_to_cpu(btree[probe].hashval) < hashval) {
+ probe++;
+ }
+
+ /*
+ * Pick the right block to descend on.
+ */
+ if (probe == max) {
+ blk->index = max - 1;
+ blkno = be32_to_cpu(btree[max - 1].before);
+ } else {
+ blk->index = probe;
+ blkno = be32_to_cpu(btree[probe].before);
}
}
@@ -1273,7 +1559,7 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result)
}
if (((retval == ENOENT) || (retval == ENOATTR)) &&
(blk->hashval == args->hashval)) {
- error = xfs_da_path_shift(state, &state->path, 1, 1,
+ error = xfs_da3_path_shift(state, &state->path, 1, 1,
&retval);
if (error)
return(error);
@@ -1295,16 +1581,52 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result)
*========================================================================*/
/*
+ * Compare two intermediate nodes for "order".
+ */
+STATIC int
+xfs_da3_node_order(
+ struct xfs_buf *node1_bp,
+ struct xfs_buf *node2_bp)
+{
+ struct xfs_da_intnode *node1;
+ struct xfs_da_intnode *node2;
+ struct xfs_da_node_entry *btree1;
+ struct xfs_da_node_entry *btree2;
+ struct xfs_da3_icnode_hdr node1hdr;
+ struct xfs_da3_icnode_hdr node2hdr;
+
+ node1 = node1_bp->b_addr;
+ node2 = node2_bp->b_addr;
+ xfs_da3_node_hdr_from_disk(&node1hdr, node1);
+ xfs_da3_node_hdr_from_disk(&node2hdr, node2);
+ btree1 = xfs_da3_node_tree_p(node1);
+ btree2 = xfs_da3_node_tree_p(node2);
+
+ if (node1hdr.count > 0 && node2hdr.count > 0 &&
+ ((be32_to_cpu(btree2[0].hashval) < be32_to_cpu(btree1[0].hashval)) ||
+ (be32_to_cpu(btree2[node2hdr.count - 1].hashval) <
+ be32_to_cpu(btree1[node1hdr.count - 1].hashval)))) {
+ return 1;
+ }
+ return 0;
+}
+
+/*
* Link a new block into a doubly linked list of blocks (of whatever type).
*/
int /* error */
-xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
- xfs_da_state_blk_t *new_blk)
+xfs_da3_blk_link(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *old_blk,
+ struct xfs_da_state_blk *new_blk)
{
- xfs_da_blkinfo_t *old_info, *new_info, *tmp_info;
- xfs_da_args_t *args;
- int before=0, error;
- struct xfs_buf *bp;
+ struct xfs_da_blkinfo *old_info;
+ struct xfs_da_blkinfo *new_info;
+ struct xfs_da_blkinfo *tmp_info;
+ struct xfs_da_args *args;
+ struct xfs_buf *bp;
+ int before = 0;
+ int error;
/*
* Set up environment.
@@ -1316,9 +1638,6 @@ xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
ASSERT(old_blk->magic == XFS_DA_NODE_MAGIC ||
old_blk->magic == XFS_DIR2_LEAFN_MAGIC ||
old_blk->magic == XFS_ATTR_LEAF_MAGIC);
- ASSERT(old_blk->magic == be16_to_cpu(old_info->magic));
- ASSERT(new_blk->magic == be16_to_cpu(new_info->magic));
- ASSERT(old_blk->magic == new_blk->magic);
switch (old_blk->magic) {
case XFS_ATTR_LEAF_MAGIC:
@@ -1328,7 +1647,7 @@ xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
before = xfs_dir2_leafn_order(old_blk->bp, new_blk->bp);
break;
case XFS_DA_NODE_MAGIC:
- before = xfs_da_node_order(old_blk->bp, new_blk->bp);
+ before = xfs_da3_node_order(old_blk->bp, new_blk->bp);
break;
}
@@ -1343,14 +1662,14 @@ xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
new_info->forw = cpu_to_be32(old_blk->blkno);
new_info->back = old_info->back;
if (old_info->back) {
- error = xfs_da_node_read(args->trans, args->dp,
+ error = xfs_da3_node_read(args->trans, args->dp,
be32_to_cpu(old_info->back),
-1, &bp, args->whichfork);
if (error)
return(error);
ASSERT(bp != NULL);
tmp_info = bp->b_addr;
- ASSERT(be16_to_cpu(tmp_info->magic) == be16_to_cpu(old_info->magic));
+ ASSERT(tmp_info->magic == old_info->magic);
ASSERT(be32_to_cpu(tmp_info->forw) == old_blk->blkno);
tmp_info->forw = cpu_to_be32(new_blk->blkno);
xfs_trans_log_buf(args->trans, bp, 0, sizeof(*tmp_info)-1);
@@ -1364,7 +1683,7 @@ xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
new_info->forw = old_info->forw;
new_info->back = cpu_to_be32(old_blk->blkno);
if (old_info->forw) {
- error = xfs_da_node_read(args->trans, args->dp,
+ error = xfs_da3_node_read(args->trans, args->dp,
be32_to_cpu(old_info->forw),
-1, &bp, args->whichfork);
if (error)
@@ -1385,59 +1704,20 @@ xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
}
/*
- * Compare two intermediate nodes for "order".
- */
-STATIC int
-xfs_da_node_order(
- struct xfs_buf *node1_bp,
- struct xfs_buf *node2_bp)
-{
- xfs_da_intnode_t *node1, *node2;
-
- node1 = node1_bp->b_addr;
- node2 = node2_bp->b_addr;
- ASSERT(node1->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC) &&
- node2->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- if ((be16_to_cpu(node1->hdr.count) > 0) && (be16_to_cpu(node2->hdr.count) > 0) &&
- ((be32_to_cpu(node2->btree[0].hashval) <
- be32_to_cpu(node1->btree[0].hashval)) ||
- (be32_to_cpu(node2->btree[be16_to_cpu(node2->hdr.count)-1].hashval) <
- be32_to_cpu(node1->btree[be16_to_cpu(node1->hdr.count)-1].hashval)))) {
- return(1);
- }
- return(0);
-}
-
-/*
- * Pick up the last hashvalue from an intermediate node.
- */
-STATIC uint
-xfs_da_node_lasthash(
- struct xfs_buf *bp,
- int *count)
-{
- xfs_da_intnode_t *node;
-
- node = bp->b_addr;
- ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- if (count)
- *count = be16_to_cpu(node->hdr.count);
- if (!node->hdr.count)
- return(0);
- return be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1].hashval);
-}
-
-/*
* Unlink a block from a doubly linked list of blocks.
*/
STATIC int /* error */
-xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
- xfs_da_state_blk_t *save_blk)
+xfs_da3_blk_unlink(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *drop_blk,
+ struct xfs_da_state_blk *save_blk)
{
- xfs_da_blkinfo_t *drop_info, *save_info, *tmp_info;
- xfs_da_args_t *args;
- struct xfs_buf *bp;
- int error;
+ struct xfs_da_blkinfo *drop_info;
+ struct xfs_da_blkinfo *save_info;
+ struct xfs_da_blkinfo *tmp_info;
+ struct xfs_da_args *args;
+ struct xfs_buf *bp;
+ int error;
/*
* Set up environment.
@@ -1449,8 +1729,6 @@ xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
ASSERT(save_blk->magic == XFS_DA_NODE_MAGIC ||
save_blk->magic == XFS_DIR2_LEAFN_MAGIC ||
save_blk->magic == XFS_ATTR_LEAF_MAGIC);
- ASSERT(save_blk->magic == be16_to_cpu(save_info->magic));
- ASSERT(drop_blk->magic == be16_to_cpu(drop_info->magic));
ASSERT(save_blk->magic == drop_blk->magic);
ASSERT((be32_to_cpu(save_info->forw) == drop_blk->blkno) ||
(be32_to_cpu(save_info->back) == drop_blk->blkno));
@@ -1464,7 +1742,7 @@ xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
trace_xfs_da_unlink_back(args);
save_info->back = drop_info->back;
if (drop_info->back) {
- error = xfs_da_node_read(args->trans, args->dp,
+ error = xfs_da3_node_read(args->trans, args->dp,
be32_to_cpu(drop_info->back),
-1, &bp, args->whichfork);
if (error)
@@ -1481,7 +1759,7 @@ xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
trace_xfs_da_unlink_forward(args);
save_info->forw = drop_info->forw;
if (drop_info->forw) {
- error = xfs_da_node_read(args->trans, args->dp,
+ error = xfs_da3_node_read(args->trans, args->dp,
be32_to_cpu(drop_info->forw),
-1, &bp, args->whichfork);
if (error)
@@ -1509,15 +1787,22 @@ xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
* the new bottom and the root.
*/
int /* error */
-xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
- int forward, int release, int *result)
+xfs_da3_path_shift(
+ struct xfs_da_state *state,
+ struct xfs_da_state_path *path,
+ int forward,
+ int release,
+ int *result)
{
- xfs_da_state_blk_t *blk;
- xfs_da_blkinfo_t *info;
- xfs_da_intnode_t *node;
- xfs_da_args_t *args;
- xfs_dablk_t blkno=0;
- int level, error;
+ struct xfs_da_state_blk *blk;
+ struct xfs_da_blkinfo *info;
+ struct xfs_da_intnode *node;
+ struct xfs_da_args *args;
+ struct xfs_da_node_entry *btree;
+ struct xfs_da3_icnode_hdr nodehdr;
+ xfs_dablk_t blkno = 0;
+ int level;
+ int error;
trace_xfs_da_path_shift(state->args);
@@ -1532,16 +1817,17 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
ASSERT((path->active > 0) && (path->active < XFS_DA_NODE_MAXDEPTH));
level = (path->active-1) - 1; /* skip bottom layer in path */
for (blk = &path->blk[level]; level >= 0; blk--, level--) {
- ASSERT(blk->bp != NULL);
node = blk->bp->b_addr;
- ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- if (forward && (blk->index < be16_to_cpu(node->hdr.count)-1)) {
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ btree = xfs_da3_node_tree_p(node);
+
+ if (forward && (blk->index < nodehdr.count - 1)) {
blk->index++;
- blkno = be32_to_cpu(node->btree[blk->index].before);
+ blkno = be32_to_cpu(btree[blk->index].before);
break;
} else if (!forward && (blk->index > 0)) {
blk->index--;
- blkno = be32_to_cpu(node->btree[blk->index].before);
+ blkno = be32_to_cpu(btree[blk->index].before);
break;
}
}
@@ -1567,45 +1853,59 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
* Read the next child block.
*/
blk->blkno = blkno;
- error = xfs_da_node_read(args->trans, args->dp, blkno, -1,
+ error = xfs_da3_node_read(args->trans, args->dp, blkno, -1,
&blk->bp, args->whichfork);
if (error)
return(error);
ASSERT(blk->bp != NULL);
info = blk->bp->b_addr;
ASSERT(info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
+ info->magic == cpu_to_be16(XFS_DA3_NODE_MAGIC) ||
info->magic == cpu_to_be16(XFS_DIR2_LEAFN_MAGIC) ||
info->magic == cpu_to_be16(XFS_DIR3_LEAFN_MAGIC) ||
info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
blk->magic = be16_to_cpu(info->magic);
- if (blk->magic == XFS_DA_NODE_MAGIC) {
+
+ /*
+ * Note: we flatten the magic number to a single type so we
+ * don't have to compare against crc/non-crc types elsewhere.
+ */
+ switch (blk->magic) {
+ case XFS_DA_NODE_MAGIC:
+ case XFS_DA3_NODE_MAGIC:
node = (xfs_da_intnode_t *)info;
- blk->hashval = be32_to_cpu(node->btree[be16_to_cpu(node->hdr.count)-1].hashval);
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
+ btree = xfs_da3_node_tree_p(node);
+ blk->hashval = be32_to_cpu(btree[nodehdr.count - 1].hashval);
if (forward)
blk->index = 0;
else
- blk->index = be16_to_cpu(node->hdr.count)-1;
- blkno = be32_to_cpu(node->btree[blk->index].before);
- } else {
+ blk->index = nodehdr.count - 1;
+ blkno = be32_to_cpu(btree[blk->index].before);
+ blk->magic = XFS_DA_NODE_MAGIC;
+ break;
+ case XFS_ATTR_LEAF_MAGIC:
ASSERT(level == path->active-1);
blk->index = 0;
- switch(blk->magic) {
- case XFS_ATTR_LEAF_MAGIC:
- blk->hashval = xfs_attr_leaf_lasthash(blk->bp,
- NULL);
- break;
- case XFS_DIR2_LEAFN_MAGIC:
- blk->hashval = xfs_dir2_leafn_lasthash(blk->bp,
- NULL);
- break;
- default:
- ASSERT(0);
- break;
- }
+ blk->hashval = xfs_attr_leaf_lasthash(blk->bp,
+ NULL);
+ blk->magic = XFS_ATTR_LEAF_MAGIC;
+ break;
+ case XFS_DIR2_LEAFN_MAGIC:
+ case XFS_DIR3_LEAFN_MAGIC:
+ ASSERT(level == path->active-1);
+ blk->index = 0;
+ blk->hashval = xfs_dir2_leafn_lasthash(blk->bp,
+ NULL);
+ blk->magic = XFS_DIR2_LEAFN_MAGIC;
+ break;
+ default:
+ ASSERT(0);
+ break;
}
}
*result = 0;
- return(0);
+ return 0;
}
@@ -1792,22 +2092,36 @@ xfs_da_grow_inode(
* a bmap btree split to do that.
*/
STATIC int
-xfs_da_swap_lastblock(
- xfs_da_args_t *args,
- xfs_dablk_t *dead_blknop,
- struct xfs_buf **dead_bufp)
+xfs_da3_swap_lastblock(
+ struct xfs_da_args *args,
+ xfs_dablk_t *dead_blknop,
+ struct xfs_buf **dead_bufp)
{
- xfs_dablk_t dead_blkno, last_blkno, sib_blkno, par_blkno;
- struct xfs_buf *dead_buf, *last_buf, *sib_buf, *par_buf;
- xfs_fileoff_t lastoff;
- xfs_inode_t *ip;
- xfs_trans_t *tp;
- xfs_mount_t *mp;
- int error, w, entno, level, dead_level;
- xfs_da_blkinfo_t *dead_info, *sib_info;
- xfs_da_intnode_t *par_node, *dead_node;
- xfs_dir2_leaf_t *dead_leaf2;
- xfs_dahash_t dead_hash;
+ struct xfs_da_blkinfo *dead_info;
+ struct xfs_da_blkinfo *sib_info;
+ struct xfs_da_intnode *par_node;
+ struct xfs_da_intnode *dead_node;
+ struct xfs_dir2_leaf *dead_leaf2;
+ struct xfs_da_node_entry *btree;
+ struct xfs_da3_icnode_hdr par_hdr;
+ struct xfs_inode *ip;
+ struct xfs_trans *tp;
+ struct xfs_mount *mp;
+ struct xfs_buf *dead_buf;
+ struct xfs_buf *last_buf;
+ struct xfs_buf *sib_buf;
+ struct xfs_buf *par_buf;
+ xfs_dahash_t dead_hash;
+ xfs_fileoff_t lastoff;
+ xfs_dablk_t dead_blkno;
+ xfs_dablk_t last_blkno;
+ xfs_dablk_t sib_blkno;
+ xfs_dablk_t par_blkno;
+ int error;
+ int w;
+ int entno;
+ int level;
+ int dead_level;
trace_xfs_da_swap_lastblock(args);
@@ -1831,7 +2145,7 @@ xfs_da_swap_lastblock(
* Read the last block in the btree space.
*/
last_blkno = (xfs_dablk_t)lastoff - mp->m_dirblkfsbs;
- error = xfs_da_node_read(tp, ip, last_blkno, -1, &last_buf, w);
+ error = xfs_da3_node_read(tp, ip, last_blkno, -1, &last_buf, w);
if (error)
return error;
/*
@@ -1854,17 +2168,22 @@ xfs_da_swap_lastblock(
dead_level = 0;
dead_hash = be32_to_cpu(ents[leafhdr.count - 1].hashval);
} else {
- ASSERT(dead_info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
+ struct xfs_da3_icnode_hdr deadhdr;
+
+ ASSERT(dead_info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC) ||
+ dead_info->magic == cpu_to_be16(XFS_DA3_NODE_MAGIC));
dead_node = (xfs_da_intnode_t *)dead_info;
- dead_level = be16_to_cpu(dead_node->hdr.level);
- dead_hash = be32_to_cpu(dead_node->btree[be16_to_cpu(dead_node->hdr.count) - 1].hashval);
+ xfs_da3_node_hdr_from_disk(&deadhdr, dead_node);
+ btree = xfs_da3_node_tree_p(dead_node);
+ dead_level = deadhdr.level;
+ dead_hash = be32_to_cpu(btree[deadhdr.count - 1].hashval);
}
sib_buf = par_buf = NULL;
/*
* If the moved block has a left sibling, fix up the pointers.
*/
if ((sib_blkno = be32_to_cpu(dead_info->back))) {
- error = xfs_da_node_read(tp, ip, sib_blkno, -1, &sib_buf, w);
+ error = xfs_da3_node_read(tp, ip, sib_blkno, -1, &sib_buf, w);
if (error)
goto done;
sib_info = sib_buf->b_addr;
@@ -1886,7 +2205,7 @@ xfs_da_swap_lastblock(
* If the moved block has a right sibling, fix up the pointers.
*/
if ((sib_blkno = be32_to_cpu(dead_info->forw))) {
- error = xfs_da_node_read(tp, ip, sib_blkno, -1, &sib_buf, w);
+ error = xfs_da3_node_read(tp, ip, sib_blkno, -1, &sib_buf, w);
if (error)
goto done;
sib_info = sib_buf->b_addr;
@@ -1910,31 +2229,31 @@ xfs_da_swap_lastblock(
* Walk down the tree looking for the parent of the moved block.
*/
for (;;) {
- error = xfs_da_node_read(tp, ip, par_blkno, -1, &par_buf, w);
+ error = xfs_da3_node_read(tp, ip, par_blkno, -1, &par_buf, w);
if (error)
goto done;
par_node = par_buf->b_addr;
- if (unlikely(par_node->hdr.info.magic !=
- cpu_to_be16(XFS_DA_NODE_MAGIC) ||
- (level >= 0 && level != be16_to_cpu(par_node->hdr.level) + 1))) {
+ xfs_da3_node_hdr_from_disk(&par_hdr, par_node);
+ if (level >= 0 && level != be16_to_cpu(par_hdr.level) + 1) {
XFS_ERROR_REPORT("xfs_da_swap_lastblock(4)",
XFS_ERRLEVEL_LOW, mp);
error = XFS_ERROR(EFSCORRUPTED);
goto done;
}
- level = be16_to_cpu(par_node->hdr.level);
+ level = par_hdr.level;
+ btree = xfs_da3_node_tree_p(par_node);
for (entno = 0;
- entno < be16_to_cpu(par_node->hdr.count) &&
- be32_to_cpu(par_node->btree[entno].hashval) < dead_hash;
+ entno < par_hdr.count &&
+ be32_to_cpu(btree[entno].hashval) < dead_hash;
entno++)
continue;
- if (unlikely(entno == be16_to_cpu(par_node->hdr.count))) {
+ if (entno == par_hdr.count) {
XFS_ERROR_REPORT("xfs_da_swap_lastblock(5)",
XFS_ERRLEVEL_LOW, mp);
error = XFS_ERROR(EFSCORRUPTED);
goto done;
}
- par_blkno = be32_to_cpu(par_node->btree[entno].before);
+ par_blkno = be32_to_cpu(btree[entno].before);
if (level == dead_level + 1)
break;
xfs_trans_brelse(tp, par_buf);
@@ -1946,13 +2265,13 @@ xfs_da_swap_lastblock(
*/
for (;;) {
for (;
- entno < be16_to_cpu(par_node->hdr.count) &&
- be32_to_cpu(par_node->btree[entno].before) != last_blkno;
+ entno < par_hdr.count &&
+ be32_to_cpu(btree[entno].before) != last_blkno;
entno++)
continue;
- if (entno < be16_to_cpu(par_node->hdr.count))
+ if (entno < par_hdr.count)
break;
- par_blkno = be32_to_cpu(par_node->hdr.info.forw);
+ par_blkno = par_hdr.forw;
xfs_trans_brelse(tp, par_buf);
par_buf = NULL;
if (unlikely(par_blkno == 0)) {
@@ -1961,27 +2280,27 @@ xfs_da_swap_lastblock(
error = XFS_ERROR(EFSCORRUPTED);
goto done;
}
- error = xfs_da_node_read(tp, ip, par_blkno, -1, &par_buf, w);
+ error = xfs_da3_node_read(tp, ip, par_blkno, -1, &par_buf, w);
if (error)
goto done;
par_node = par_buf->b_addr;
- if (unlikely(
- be16_to_cpu(par_node->hdr.level) != level ||
- par_node->hdr.info.magic != cpu_to_be16(XFS_DA_NODE_MAGIC))) {
+ xfs_da3_node_hdr_from_disk(&par_hdr, par_node);
+ if (par_hdr.level != level) {
XFS_ERROR_REPORT("xfs_da_swap_lastblock(7)",
XFS_ERRLEVEL_LOW, mp);
error = XFS_ERROR(EFSCORRUPTED);
goto done;
}
+ btree = xfs_da3_node_tree_p(par_node);
entno = 0;
}
/*
* Update the parent entry pointing to the moved block.
*/
- par_node->btree[entno].before = cpu_to_be32(dead_blkno);
+ btree[entno].before = cpu_to_be32(dead_blkno);
xfs_trans_log_buf(tp, par_buf,
- XFS_DA_LOGRANGE(par_node, &par_node->btree[entno].before,
- sizeof(par_node->btree[entno].before)));
+ XFS_DA_LOGRANGE(par_node, &btree[entno].before,
+ sizeof(btree[entno].before)));
*dead_blknop = last_blkno;
*dead_bufp = last_buf;
return 0;
@@ -2023,14 +2342,15 @@ xfs_da_shrink_inode(
* Remove extents. If we get ENOSPC for a dir we have to move
* the last block to the place we want to kill.
*/
- if ((error = xfs_bunmapi(tp, dp, dead_blkno, count,
- xfs_bmapi_aflag(w)|XFS_BMAPI_METADATA,
- 0, args->firstblock, args->flist,
- &done)) == ENOSPC) {
+ error = xfs_bunmapi(tp, dp, dead_blkno, count,
+ xfs_bmapi_aflag(w)|XFS_BMAPI_METADATA,
+ 0, args->firstblock, args->flist, &done);
+ if (error == ENOSPC) {
if (w != XFS_DATA_FORK)
break;
- if ((error = xfs_da_swap_lastblock(args, &dead_blkno,
- &dead_buf)))
+ error = xfs_da3_swap_lastblock(args, &dead_blkno,
+ &dead_buf);
+ if (error)
break;
} else {
break;
@@ -2364,41 +2684,3 @@ out_free:
return -1;
return mappedbno;
}
-
-kmem_zone_t *xfs_da_state_zone; /* anchor for state struct zone */
-
-/*
- * Allocate a dir-state structure.
- * We don't put them on the stack since they're large.
- */
-xfs_da_state_t *
-xfs_da_state_alloc(void)
-{
- return kmem_zone_zalloc(xfs_da_state_zone, KM_NOFS);
-}
-
-/*
- * Kill the altpath contents of a da-state structure.
- */
-STATIC void
-xfs_da_state_kill_altpath(xfs_da_state_t *state)
-{
- int i;
-
- for (i = 0; i < state->altpath.active; i++)
- state->altpath.blk[i].bp = NULL;
- state->altpath.active = 0;
-}
-
-/*
- * Free a da-state structure.
- */
-void
-xfs_da_state_free(xfs_da_state_t *state)
-{
- xfs_da_state_kill_altpath(state);
-#ifdef DEBUG
- memset((char *)state, 0, sizeof(*state));
-#endif /* DEBUG */
- kmem_zone_free(xfs_da_state_zone, state);
-}
diff --git a/fs/xfs/xfs_da_btree.h b/fs/xfs/xfs_da_btree.h
index d408d9b..f20741e 100644
--- a/fs/xfs/xfs_da_btree.h
+++ b/fs/xfs/xfs_da_btree.h
@@ -20,7 +20,6 @@
struct xfs_bmap_free;
struct xfs_inode;
-struct xfs_mount;
struct xfs_trans;
struct zone;
@@ -50,13 +49,15 @@ typedef struct xfs_da_blkinfo {
* CRC enabled directory structure types
*
* The headers change size for the additional verification information, but
- * otherwise the tree layouts and contents are unchanged.
+ * otherwise the tree layouts and contents are unchanged. Hence the da btree
+ * code can use the struct xfs_da_blkinfo for manipulating the tree links and
+ * magic numbers without modification for both v2 and v3 nodes.
*/
+#define XFS_DA3_NODE_MAGIC 0x3ebe /* magic number: non-leaf blocks */
#define XFS_DIR3_LEAF1_MAGIC 0x3df1 /* magic number: v2 dirlf single blks */
#define XFS_DIR3_LEAFN_MAGIC 0x3dff /* magic number: v2 dirlf multi blks */
struct xfs_da3_blkinfo {
-/* XXX: check packing due to size of xfs_da_blkinfo */
struct xfs_da_blkinfo hdr;
__be32 crc; /* CRC of block */
__be64 blkno; /* first block of the buffer */
@@ -76,19 +77,75 @@ struct xfs_da3_blkinfo {
*/
#define XFS_DA_NODE_MAXDEPTH 5 /* max depth of Btree */
+typedef struct xfs_da_node_hdr {
+ struct xfs_da_blkinfo info; /* block type, links, etc. */
+ __be16 count; /* count of active entries */
+ __be16 level; /* level above leaves (leaf == 0) */
+} xfs_da_node_hdr_t;
+
+struct xfs_da3_node_hdr {
+ struct xfs_da3_blkinfo info; /* block type, links, etc. */
+ __be16 count; /* count of active entries */
+ __be16 level; /* level above leaves (leaf == 0) */
+};
+
+#define XFS_DA3_NODE_CRC_OFF (offsetof(struct xfs_da3_node_hdr, info.crc))
+
+typedef struct xfs_da_node_entry {
+ __be32 hashval; /* hash value for this descendant */
+ __be32 before; /* Btree block before this key */
+} xfs_da_node_entry_t;
+
typedef struct xfs_da_intnode {
- struct xfs_da_node_hdr { /* constant-structure header block */
- xfs_da_blkinfo_t info; /* block type, links, etc. */
- __be16 count; /* count of active entries */
- __be16 level; /* level above leaves (leaf == 0) */
- } hdr;
- struct xfs_da_node_entry {
- __be32 hashval; /* hash value for this descendant */
- __be32 before; /* Btree block before this key */
- } btree[1]; /* variable sized array of keys */
+ struct xfs_da_node_hdr hdr;
+ struct xfs_da_node_entry __btree[];
} xfs_da_intnode_t;
-typedef struct xfs_da_node_hdr xfs_da_node_hdr_t;
-typedef struct xfs_da_node_entry xfs_da_node_entry_t;
+
+struct xfs_da3_intnode {
+ struct xfs_da3_node_hdr hdr;
+ struct xfs_da_node_entry __btree[];
+};
+
+/*
+ * In-core version of the node header to abstract the differences in the v2 and
+ * v3 disk format of the headers. Callers need to convert to/from disk format as
+ * appropriate.
+ */
+struct xfs_da3_icnode_hdr {
+ __uint32_t forw;
+ __uint32_t back;
+ __uint16_t magic;
+ __uint16_t count;
+ __uint16_t level;
+};
+
+extern void xfs_da3_node_hdr_from_disk(struct xfs_da3_icnode_hdr *to,
+ struct xfs_da_intnode *from);
+extern void xfs_da3_node_hdr_to_disk(struct xfs_da_intnode *to,
+ struct xfs_da3_icnode_hdr *from);
+
+static inline int
+xfs_da3_node_hdr_size(struct xfs_da_intnode *dap)
+{
+ if (dap->hdr.info.magic == cpu_to_be16(XFS_DA3_NODE_MAGIC))
+ return sizeof(struct xfs_da3_node_hdr);
+ return sizeof(struct xfs_da_node_hdr);
+}
+
+static inline struct xfs_da_node_entry *
+xfs_da3_node_tree_p(struct xfs_da_intnode *dap)
+{
+ if (dap->hdr.info.magic == cpu_to_be16(XFS_DA3_NODE_MAGIC)) {
+ struct xfs_da3_intnode *dap3 = (struct xfs_da3_intnode *)dap;
+ return dap3->__btree;
+ }
+ return dap->__btree;
+}
+
+extern void xfs_da3_intnode_from_disk(struct xfs_da3_icnode_hdr *to,
+ struct xfs_da_intnode *from);
+extern void xfs_da3_intnode_to_disk(struct xfs_da_intnode *to,
+ struct xfs_da3_icnode_hdr *from);
#define XFS_LBSIZE(mp) (mp)->m_sb.sb_blocksize
@@ -210,29 +267,29 @@ struct xfs_nameops {
/*
* Routines used for growing the Btree.
*/
-int xfs_da_node_create(xfs_da_args_t *args, xfs_dablk_t blkno, int level,
- struct xfs_buf **bpp, int whichfork);
-int xfs_da_split(xfs_da_state_t *state);
+int xfs_da3_node_create(struct xfs_da_args *args, xfs_dablk_t blkno,
+ int level, struct xfs_buf **bpp, int whichfork);
+int xfs_da3_split(xfs_da_state_t *state);
/*
* Routines used for shrinking the Btree.
*/
-int xfs_da_join(xfs_da_state_t *state);
-void xfs_da_fixhashpath(xfs_da_state_t *state,
- xfs_da_state_path_t *path_to_to_fix);
+int xfs_da3_join(xfs_da_state_t *state);
+void xfs_da3_fixhashpath(struct xfs_da_state *state,
+ struct xfs_da_state_path *path_to_to_fix);
/*
* Routines used for finding things in the Btree.
*/
-int xfs_da_node_lookup_int(xfs_da_state_t *state, int *result);
-int xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
+int xfs_da3_node_lookup_int(xfs_da_state_t *state, int *result);
+int xfs_da3_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path,
int forward, int release, int *result);
/*
* Utility routines.
*/
-int xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
+int xfs_da3_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk,
xfs_da_state_blk_t *new_blk);
-int xfs_da_node_read(struct xfs_trans *tp, struct xfs_inode *dp,
+int xfs_da3_node_read(struct xfs_trans *tp, struct xfs_inode *dp,
xfs_dablk_t bno, xfs_daddr_t mappedbno,
struct xfs_buf **bpp, int which_fork);
diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c
index 99d1df3..5dcafcd 100644
--- a/fs/xfs/xfs_dir2_node.c
+++ b/fs/xfs/xfs_dir2_node.c
@@ -1361,7 +1361,7 @@ xfs_dir2_leafn_split(
* block into the leaves.
*/
xfs_dir2_leafn_rebalance(state, oldblk, newblk);
- error = xfs_da_blk_link(state, oldblk, newblk);
+ error = xfs_da3_blk_link(state, oldblk, newblk);
if (error) {
return error;
}
@@ -1443,7 +1443,7 @@ xfs_dir2_leafn_toosmall(
*/
forward = (leafhdr.forw != 0);
memcpy(&state->altpath, &state->path, sizeof(state->path));
- error = xfs_da_path_shift(state, &state->altpath, forward, 0,
+ error = xfs_da3_path_shift(state, &state->altpath, forward, 0,
&rval);
if (error)
return error;
@@ -1505,10 +1505,10 @@ xfs_dir2_leafn_toosmall(
*/
memcpy(&state->altpath, &state->path, sizeof(state->path));
if (blkno < blk->blkno)
- error = xfs_da_path_shift(state, &state->altpath, forward, 0,
+ error = xfs_da3_path_shift(state, &state->altpath, forward, 0,
&rval);
else
- error = xfs_da_path_shift(state, &state->path, forward, 0,
+ error = xfs_da3_path_shift(state, &state->path, forward, 0,
&rval);
if (error) {
return error;
@@ -1604,7 +1604,7 @@ xfs_dir2_node_addname(
* Look up the name. We're not supposed to find it, but
* this gives us the insertion point.
*/
- error = xfs_da_node_lookup_int(state, &rval);
+ error = xfs_da3_node_lookup_int(state, &rval);
if (error)
rval = error;
if (rval != ENOENT) {
@@ -1630,7 +1630,7 @@ xfs_dir2_node_addname(
* It worked, fix the hash values up the btree.
*/
if (!(args->op_flags & XFS_DA_OP_JUSTCHECK))
- xfs_da_fixhashpath(state, &state->path);
+ xfs_da3_fixhashpath(state, &state->path);
} else {
/*
* It didn't work, we need to split the leaf block.
@@ -1642,7 +1642,7 @@ xfs_dir2_node_addname(
/*
* Split the leaf block and insert the new entry.
*/
- rval = xfs_da_split(state);
+ rval = xfs_da3_split(state);
}
done:
xfs_da_state_free(state);
@@ -2020,7 +2020,7 @@ xfs_dir2_node_addname_int(
/*
* Lookup an entry in a node-format directory.
- * All the real work happens in xfs_da_node_lookup_int.
+ * All the real work happens in xfs_da3_node_lookup_int.
* The only real output is the inode number of the entry.
*/
int /* error */
@@ -2045,7 +2045,7 @@ xfs_dir2_node_lookup(
/*
* Fill in the path to the entry in the cursor.
*/
- error = xfs_da_node_lookup_int(state, &rval);
+ error = xfs_da3_node_lookup_int(state, &rval);
if (error)
rval = error;
else if (rval == ENOENT && args->cmpresult == XFS_CMP_CASE) {
@@ -2100,7 +2100,7 @@ xfs_dir2_node_removename(
/*
* Look up the entry we're deleting, set up the cursor.
*/
- error = xfs_da_node_lookup_int(state, &rval);
+ error = xfs_da3_node_lookup_int(state, &rval);
if (error)
rval = error;
/*
@@ -2124,12 +2124,12 @@ xfs_dir2_node_removename(
/*
* Fix the hash values up the btree.
*/
- xfs_da_fixhashpath(state, &state->path);
+ xfs_da3_fixhashpath(state, &state->path);
/*
* If we need to join leaf blocks, do it.
*/
if (rval && state->path.active > 1)
- error = xfs_da_join(state);
+ error = xfs_da3_join(state);
/*
* If no errors so far, try conversion to leaf format.
*/
@@ -2171,7 +2171,7 @@ xfs_dir2_node_replace(
/*
* Lookup the entry to change in the btree.
*/
- error = xfs_da_node_lookup_int(state, &rval);
+ error = xfs_da3_node_lookup_int(state, &rval);
if (error) {
rval = error;
}
diff --git a/fs/xfs/xfs_trace.c b/fs/xfs/xfs_trace.c
index 624bedd..b6e3897 100644
--- a/fs/xfs/xfs_trace.c
+++ b/fs/xfs/xfs_trace.c
@@ -22,7 +22,6 @@
#include "xfs_trans.h"
#include "xfs_sb.h"
#include "xfs_ag.h"
-#include "xfs_da_btree.h"
#include "xfs_bmap_btree.h"
#include "xfs_alloc_btree.h"
#include "xfs_ialloc_btree.h"
@@ -30,6 +29,7 @@
#include "xfs_inode.h"
#include "xfs_btree.h"
#include "xfs_mount.h"
+#include "xfs_da_btree.h"
#include "xfs_ialloc.h"
#include "xfs_itable.h"
#include "xfs_alloc.h"
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 15/16] xfs: add CRCs to attr leaf blocks
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (13 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 14/16] xfs: add CRCs to dir2/da node blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
2013-02-25 1:31 ` [PATCH 16/16] xfs: add CRC checks to the superblock Dave Chinner
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
fs/xfs/xfs_attr.c | 125 ++--
fs/xfs/xfs_attr_leaf.c | 1738 +++++++++++++++++++++++++++---------------------
fs/xfs/xfs_attr_leaf.h | 121 +++-
fs/xfs/xfs_da_btree.c | 18 +-
fs/xfs/xfs_da_btree.h | 1 +
5 files changed, 1163 insertions(+), 840 deletions(-)
diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c
index e03128c..de4093f 100644
--- a/fs/xfs/xfs_attr.c
+++ b/fs/xfs/xfs_attr.c
@@ -819,7 +819,7 @@ xfs_attr_inactive(xfs_inode_t *dp)
error = 0;
goto out;
}
- error = xfs_attr_root_inactive(&trans, dp);
+ error = xfs_attr3_root_inactive(&trans, dp);
if (error)
goto out;
@@ -905,7 +905,7 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
*/
dp = args->dp;
args->blkno = 0;
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
if (error)
return error;
@@ -913,14 +913,14 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
* Look up the given attribute in the leaf block. Figure out if
* the given flags produce an error or call for an atomic rename.
*/
- retval = xfs_attr_leaf_lookup_int(bp, args);
+ retval = xfs_attr3_leaf_lookup_int(bp, args);
if ((args->flags & ATTR_REPLACE) && (retval == ENOATTR)) {
xfs_trans_brelse(args->trans, bp);
- return(retval);
+ return retval;
} else if (retval == EEXIST) {
if (args->flags & ATTR_CREATE) { /* pure create op */
xfs_trans_brelse(args->trans, bp);
- return(retval);
+ return retval;
}
trace_xfs_attr_leaf_replace(args);
@@ -936,7 +936,7 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
* Add the attribute to the leaf block, transitioning to a Btree
* if required.
*/
- retval = xfs_attr_leaf_add(bp, args);
+ retval = xfs_attr3_leaf_add(bp, args);
if (retval == ENOSPC) {
/*
* Promote the attribute list to the Btree format, then
@@ -944,7 +944,7 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
* can manage its own transactions.
*/
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_attr_leaf_to_node(args);
+ error = xfs_attr3_leaf_to_node(args);
if (!error) {
error = xfs_bmap_finish(&args->trans, args->flist,
&committed);
@@ -1009,7 +1009,7 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
* In a separate transaction, set the incomplete flag on the
* "old" attr and clear the incomplete flag on the "new" attr.
*/
- error = xfs_attr_leaf_flipflags(args);
+ error = xfs_attr3_leaf_flipflags(args);
if (error)
return(error);
@@ -1031,19 +1031,19 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
* Read in the block containing the "old" attr, then
* remove the "old" attr from that block (neat, huh!)
*/
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno,
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno,
-1, &bp);
if (error)
return error;
- xfs_attr_leaf_remove(bp, args);
+ xfs_attr3_leaf_remove(bp, args);
/*
* If the result is small enough, shrink it all into the inode.
*/
if ((forkoff = xfs_attr_shortform_allfit(bp, dp))) {
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_attr_leaf_to_shortform(bp, args, forkoff);
+ error = xfs_attr3_leaf_to_shortform(bp, args, forkoff);
/* bp is gone due to xfs_da_shrink_inode */
if (!error) {
error = xfs_bmap_finish(&args->trans,
@@ -1075,9 +1075,9 @@ xfs_attr_leaf_addname(xfs_da_args_t *args)
/*
* Added a "remote" value, just clear the incomplete flag.
*/
- error = xfs_attr_leaf_clearflag(args);
+ error = xfs_attr3_leaf_clearflag(args);
}
- return(error);
+ return error;
}
/*
@@ -1100,24 +1100,24 @@ xfs_attr_leaf_removename(xfs_da_args_t *args)
*/
dp = args->dp;
args->blkno = 0;
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
if (error)
return error;
- error = xfs_attr_leaf_lookup_int(bp, args);
+ error = xfs_attr3_leaf_lookup_int(bp, args);
if (error == ENOATTR) {
xfs_trans_brelse(args->trans, bp);
- return(error);
+ return error;
}
- xfs_attr_leaf_remove(bp, args);
+ xfs_attr3_leaf_remove(bp, args);
/*
* If the result is small enough, shrink it all into the inode.
*/
if ((forkoff = xfs_attr_shortform_allfit(bp, dp))) {
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_attr_leaf_to_shortform(bp, args, forkoff);
+ error = xfs_attr3_leaf_to_shortform(bp, args, forkoff);
/* bp is gone due to xfs_da_shrink_inode */
if (!error) {
error = xfs_bmap_finish(&args->trans, args->flist,
@@ -1127,7 +1127,7 @@ xfs_attr_leaf_removename(xfs_da_args_t *args)
ASSERT(committed);
args->trans = NULL;
xfs_bmap_cancel(args->flist);
- return(error);
+ return error;
}
/*
@@ -1137,7 +1137,7 @@ xfs_attr_leaf_removename(xfs_da_args_t *args)
if (committed)
xfs_trans_ijoin(args->trans, dp, 0);
}
- return(0);
+ return 0;
}
/*
@@ -1155,21 +1155,21 @@ xfs_attr_leaf_get(xfs_da_args_t *args)
trace_xfs_attr_leaf_get(args);
args->blkno = 0;
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
if (error)
return error;
- error = xfs_attr_leaf_lookup_int(bp, args);
+ error = xfs_attr3_leaf_lookup_int(bp, args);
if (error != EEXIST) {
xfs_trans_brelse(args->trans, bp);
- return(error);
+ return error;
}
- error = xfs_attr_leaf_getvalue(bp, args);
+ error = xfs_attr3_leaf_getvalue(bp, args);
xfs_trans_brelse(args->trans, bp);
if (!error && (args->rmtblkno > 0) && !(args->flags & ATTR_KERNOVAL)) {
error = xfs_attr_rmtval_get(args);
}
- return(error);
+ return error;
}
/*
@@ -1184,11 +1184,11 @@ xfs_attr_leaf_list(xfs_attr_list_context_t *context)
trace_xfs_attr_leaf_list(context);
context->cursor->blkno = 0;
- error = xfs_attr_leaf_read(NULL, context->dp, 0, -1, &bp);
+ error = xfs_attr3_leaf_read(NULL, context->dp, 0, -1, &bp);
if (error)
return XFS_ERROR(error);
- error = xfs_attr_leaf_list_int(bp, context);
+ error = xfs_attr3_leaf_list_int(bp, context);
xfs_trans_brelse(NULL, bp);
return XFS_ERROR(error);
}
@@ -1257,7 +1257,7 @@ restart:
args->rmtblkcnt = 0;
}
- retval = xfs_attr_leaf_add(blk->bp, state->args);
+ retval = xfs_attr3_leaf_add(blk->bp, state->args);
if (retval == ENOSPC) {
if (state->path.active == 1) {
/*
@@ -1267,7 +1267,7 @@ restart:
*/
xfs_da_state_free(state);
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_attr_leaf_to_node(args);
+ error = xfs_attr3_leaf_to_node(args);
if (!error) {
error = xfs_bmap_finish(&args->trans,
args->flist,
@@ -1369,7 +1369,7 @@ restart:
* In a separate transaction, set the incomplete flag on the
* "old" attr and clear the incomplete flag on the "new" attr.
*/
- error = xfs_attr_leaf_flipflags(args);
+ error = xfs_attr3_leaf_flipflags(args);
if (error)
goto out;
@@ -1408,7 +1408,7 @@ restart:
*/
blk = &state->path.blk[ state->path.active-1 ];
ASSERT(blk->magic == XFS_ATTR_LEAF_MAGIC);
- error = xfs_attr_leaf_remove(blk->bp, args);
+ error = xfs_attr3_leaf_remove(blk->bp, args);
xfs_da3_fixhashpath(state, &state->path);
/*
@@ -1449,7 +1449,7 @@ restart:
/*
* Added a "remote" value, just clear the incomplete flag.
*/
- error = xfs_attr_leaf_clearflag(args);
+ error = xfs_attr3_leaf_clearflag(args);
if (error)
goto out;
}
@@ -1523,7 +1523,7 @@ xfs_attr_node_removename(xfs_da_args_t *args)
* Mark the attribute as INCOMPLETE, then bunmapi() the
* remote value.
*/
- error = xfs_attr_leaf_setflag(args);
+ error = xfs_attr3_leaf_setflag(args);
if (error)
goto out;
error = xfs_attr_rmtval_remove(args);
@@ -1544,7 +1544,7 @@ xfs_attr_node_removename(xfs_da_args_t *args)
*/
blk = &state->path.blk[ state->path.active-1 ];
ASSERT(blk->magic == XFS_ATTR_LEAF_MAGIC);
- retval = xfs_attr_leaf_remove(blk->bp, args);
+ retval = xfs_attr3_leaf_remove(blk->bp, args);
xfs_da3_fixhashpath(state, &state->path);
/*
@@ -1590,13 +1590,13 @@ xfs_attr_node_removename(xfs_da_args_t *args)
ASSERT(state->path.blk[0].bp);
state->path.blk[0].bp = NULL;
- error = xfs_attr_leaf_read(args->trans, args->dp, 0, -1, &bp);
+ error = xfs_attr3_leaf_read(args->trans, args->dp, 0, -1, &bp);
if (error)
goto out;
if ((forkoff = xfs_attr_shortform_allfit(bp, dp))) {
xfs_bmap_init(args->flist, args->firstblock);
- error = xfs_attr_leaf_to_shortform(bp, args, forkoff);
+ error = xfs_attr3_leaf_to_shortform(bp, args, forkoff);
/* bp is gone due to xfs_da_shrink_inode */
if (!error) {
error = xfs_bmap_finish(&args->trans,
@@ -1768,7 +1768,7 @@ xfs_attr_node_get(xfs_da_args_t *args)
/*
* Get the value, local or "remote"
*/
- retval = xfs_attr_leaf_getvalue(blk->bp, args);
+ retval = xfs_attr3_leaf_getvalue(blk->bp, args);
if (!retval && (args->rmtblkno > 0)
&& !(args->flags & ATTR_KERNOVAL)) {
retval = xfs_attr_rmtval_get(args);
@@ -1793,7 +1793,9 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
attrlist_cursor_kern_t *cursor;
xfs_attr_leafblock_t *leaf;
xfs_da_intnode_t *node;
- xfs_da_node_entry_t *btree;
+ struct xfs_attr3_icleaf_hdr leafhdr;
+ struct xfs_da3_icnode_hdr nodehdr;
+ struct xfs_da_node_entry *btree;
int error, i;
struct xfs_buf *bp;
@@ -1814,22 +1816,28 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
if ((error != 0) && (error != EFSCORRUPTED))
return(error);
if (bp) {
+ struct xfs_attr_leaf_entry *entries;
+
node = bp->b_addr;
switch (be16_to_cpu(node->hdr.info.magic)) {
case XFS_DA_NODE_MAGIC:
+ case XFS_DA3_NODE_MAGIC:
trace_xfs_attr_list_wrong_blk(context);
xfs_trans_brelse(NULL, bp);
bp = NULL;
break;
case XFS_ATTR_LEAF_MAGIC:
+ case XFS_ATTR3_LEAF_MAGIC:
leaf = bp->b_addr;
- if (cursor->hashval > be32_to_cpu(leaf->entries[
- be16_to_cpu(leaf->hdr.count)-1].hashval)) {
+ xfs_attr3_leaf_hdr_from_disk(&leafhdr, leaf);
+ entries = xfs_attr3_leaf_entryp(leaf);
+ if (cursor->hashval > be32_to_cpu(
+ entries[leafhdr.count - 1].hashval)) {
trace_xfs_attr_list_wrong_blk(context);
xfs_trans_brelse(NULL, bp);
bp = NULL;
- } else if (cursor->hashval <=
- be32_to_cpu(leaf->entries[0].hashval)) {
+ } else if (cursor->hashval <= be32_to_cpu(
+ entries[0].hashval)) {
trace_xfs_attr_list_wrong_blk(context);
xfs_trans_brelse(NULL, bp);
bp = NULL;
@@ -1851,27 +1859,31 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
if (bp == NULL) {
cursor->blkno = 0;
for (;;) {
+ __uint16_t magic;
+
error = xfs_da3_node_read(NULL, context->dp,
cursor->blkno, -1, &bp,
XFS_ATTR_FORK);
if (error)
return(error);
node = bp->b_addr;
- if (node->hdr.info.magic ==
- cpu_to_be16(XFS_ATTR_LEAF_MAGIC))
+ magic = be16_to_cpu(node->hdr.info.magic);
+ if (magic == XFS_ATTR_LEAF_MAGIC ||
+ magic == XFS_ATTR3_LEAF_MAGIC)
break;
- if (unlikely(node->hdr.info.magic !=
- cpu_to_be16(XFS_DA_NODE_MAGIC))) {
+ if (magic != XFS_DA_NODE_MAGIC &&
+ magic != XFS_DA3_NODE_MAGIC) {
XFS_CORRUPTION_ERROR("xfs_attr_node_list(3)",
XFS_ERRLEVEL_LOW,
context->dp->i_mount,
node);
xfs_trans_brelse(NULL, bp);
- return(XFS_ERROR(EFSCORRUPTED));
+ return XFS_ERROR(EFSCORRUPTED);
}
+
+ xfs_da3_node_hdr_from_disk(&nodehdr, node);
btree = xfs_da3_node_tree_p(node);
- for (i = 0; i < be16_to_cpu(node->hdr.count);
- btree++, i++) {
+ for (i = 0; i < nodehdr.count; btree++, i++) {
if (cursor->hashval
<= be32_to_cpu(btree->hashval)) {
cursor->blkno = be32_to_cpu(btree->before);
@@ -1880,9 +1892,9 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
break;
}
}
- if (i == be16_to_cpu(node->hdr.count)) {
+ if (i == nodehdr.count) {
xfs_trans_brelse(NULL, bp);
- return(0);
+ return 0;
}
xfs_trans_brelse(NULL, bp);
}
@@ -1896,22 +1908,23 @@ xfs_attr_node_list(xfs_attr_list_context_t *context)
*/
for (;;) {
leaf = bp->b_addr;
- error = xfs_attr_leaf_list_int(bp, context);
+ error = xfs_attr3_leaf_list_int(bp, context);
if (error) {
xfs_trans_brelse(NULL, bp);
return error;
}
- if (context->seen_enough || leaf->hdr.info.forw == 0)
+ xfs_attr3_leaf_hdr_from_disk(&leafhdr, leaf);
+ if (context->seen_enough || leafhdr.forw == 0)
break;
- cursor->blkno = be32_to_cpu(leaf->hdr.info.forw);
+ cursor->blkno = leafhdr.forw;
xfs_trans_brelse(NULL, bp);
- error = xfs_attr_leaf_read(NULL, context->dp, cursor->blkno, -1,
+ error = xfs_attr3_leaf_read(NULL, context->dp, cursor->blkno, -1,
&bp);
if (error)
return error;
}
xfs_trans_brelse(NULL, bp);
- return(0);
+ return 0;
}
diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c
index 54b0dc8..25bc98c 100644
--- a/fs/xfs/xfs_attr_leaf.c
+++ b/fs/xfs/xfs_attr_leaf.c
@@ -39,6 +39,9 @@
#include "xfs_attr_leaf.h"
#include "xfs_error.h"
#include "xfs_trace.h"
+#include "xfs_buf_item.h"
+#include "xfs_cksum.h"
+
/*
* xfs_attr_leaf.c
@@ -53,77 +56,213 @@
/*
* Routines used for growing the Btree.
*/
-STATIC int xfs_attr_leaf_create(xfs_da_args_t *args, xfs_dablk_t which_block,
- struct xfs_buf **bpp);
-STATIC int xfs_attr_leaf_add_work(struct xfs_buf *leaf_buffer,
- xfs_da_args_t *args, int freemap_index);
-STATIC void xfs_attr_leaf_compact(struct xfs_da_args *args,
- struct xfs_buf *leaf_buffer);
-STATIC void xfs_attr_leaf_rebalance(xfs_da_state_t *state,
+STATIC int xfs_attr3_leaf_create(struct xfs_da_args *args,
+ xfs_dablk_t which_block, struct xfs_buf **bpp);
+STATIC int xfs_attr3_leaf_add_work(struct xfs_buf *leaf_buffer,
+ struct xfs_attr3_icleaf_hdr *ichdr,
+ struct xfs_da_args *args, int freemap_index);
+STATIC void xfs_attr3_leaf_compact(struct xfs_da_args *args,
+ struct xfs_attr3_icleaf_hdr *ichdr,
+ struct xfs_buf *leaf_buffer);
+STATIC void xfs_attr3_leaf_rebalance(xfs_da_state_t *state,
xfs_da_state_blk_t *blk1,
xfs_da_state_blk_t *blk2);
-STATIC int xfs_attr_leaf_figure_balance(xfs_da_state_t *state,
- xfs_da_state_blk_t *leaf_blk_1,
- xfs_da_state_blk_t *leaf_blk_2,
- int *number_entries_in_blk1,
- int *number_usedbytes_in_blk1);
+STATIC int xfs_attr3_leaf_figure_balance(xfs_da_state_t *state,
+ xfs_da_state_blk_t *leaf_blk_1,
+ struct xfs_attr3_icleaf_hdr *ichdr1,
+ xfs_da_state_blk_t *leaf_blk_2,
+ struct xfs_attr3_icleaf_hdr *ichdr2,
+ int *number_entries_in_blk1,
+ int *number_usedbytes_in_blk1);
/*
* Routines used for shrinking the Btree.
*/
-STATIC int xfs_attr_node_inactive(xfs_trans_t **trans, xfs_inode_t *dp,
+STATIC int xfs_attr3_node_inactive(xfs_trans_t **trans, xfs_inode_t *dp,
struct xfs_buf *bp, int level);
-STATIC int xfs_attr_leaf_inactive(xfs_trans_t **trans, xfs_inode_t *dp,
+STATIC int xfs_attr3_leaf_inactive(xfs_trans_t **trans, xfs_inode_t *dp,
struct xfs_buf *bp);
-STATIC int xfs_attr_leaf_freextent(xfs_trans_t **trans, xfs_inode_t *dp,
+STATIC int xfs_attr3_leaf_freextent(xfs_trans_t **trans, xfs_inode_t *dp,
xfs_dablk_t blkno, int blkcnt);
/*
* Utility routines.
*/
-STATIC void xfs_attr_leaf_moveents(xfs_attr_leafblock_t *src_leaf,
- int src_start,
- xfs_attr_leafblock_t *dst_leaf,
- int dst_start, int move_count,
- xfs_mount_t *mp);
+STATIC void xfs_attr3_leaf_moveents(struct xfs_attr_leafblock *src_leaf,
+ struct xfs_attr3_icleaf_hdr *src_ichdr, int src_start,
+ struct xfs_attr_leafblock *dst_leaf,
+ struct xfs_attr3_icleaf_hdr *dst_ichdr, int dst_start,
+ int move_count, struct xfs_mount *mp);
STATIC int xfs_attr_leaf_entsize(xfs_attr_leafblock_t *leaf, int index);
-static void
-xfs_attr_leaf_verify(
+void
+xfs_attr3_leaf_hdr_from_disk(
+ struct xfs_attr3_icleaf_hdr *to,
+ struct xfs_attr_leafblock *from)
+{
+ int i;
+
+ ASSERT(from->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC) ||
+ from->hdr.info.magic == cpu_to_be16(XFS_ATTR3_LEAF_MAGIC));
+
+ if (from->hdr.info.magic == cpu_to_be16(XFS_ATTR3_LEAF_MAGIC)) {
+ struct xfs_attr3_leaf_hdr *hdr3 = (struct xfs_attr3_leaf_hdr *)from;
+
+ to->forw = be32_to_cpu(hdr3->info.hdr.forw);
+ to->back = be32_to_cpu(hdr3->info.hdr.back);
+ to->magic = be16_to_cpu(hdr3->info.hdr.magic);
+ to->count = be16_to_cpu(hdr3->count);
+ to->usedbytes = be16_to_cpu(hdr3->usedbytes);
+ to->firstused = be16_to_cpu(hdr3->firstused);
+ to->holes = hdr3->holes;
+
+ for (i = 0; i < XFS_ATTR_LEAF_MAPSIZE; i++) {
+ to->freemap[i].base = be16_to_cpu(hdr3->freemap[i].base);
+ to->freemap[i].size = be16_to_cpu(hdr3->freemap[i].size);
+ }
+ return;
+ }
+ to->forw = be32_to_cpu(from->hdr.info.forw);
+ to->back = be32_to_cpu(from->hdr.info.back);
+ to->magic = be16_to_cpu(from->hdr.info.magic);
+ to->count = be16_to_cpu(from->hdr.count);
+ to->usedbytes = be16_to_cpu(from->hdr.usedbytes);
+ to->firstused = be16_to_cpu(from->hdr.firstused);
+ to->holes = from->hdr.holes;
+
+ for (i = 0; i < XFS_ATTR_LEAF_MAPSIZE; i++) {
+ to->freemap[i].base = be16_to_cpu(from->hdr.freemap[i].base);
+ to->freemap[i].size = be16_to_cpu(from->hdr.freemap[i].size);
+ }
+}
+
+void
+xfs_attr3_leaf_hdr_to_disk(
+ struct xfs_attr_leafblock *to,
+ struct xfs_attr3_icleaf_hdr *from)
+{
+ int i;
+
+ ASSERT(from->magic == XFS_ATTR_LEAF_MAGIC ||
+ from->magic == XFS_ATTR3_LEAF_MAGIC);
+
+ if (from->magic == XFS_ATTR3_LEAF_MAGIC) {
+ struct xfs_attr3_leaf_hdr *hdr3 = (struct xfs_attr3_leaf_hdr *)to;
+
+ hdr3->info.hdr.forw = cpu_to_be32(from->forw);
+ hdr3->info.hdr.back = cpu_to_be32(from->back);
+ hdr3->info.hdr.magic = cpu_to_be16(from->magic);
+ hdr3->count = cpu_to_be16(from->count);
+ hdr3->usedbytes = cpu_to_be16(from->usedbytes);
+ hdr3->firstused = cpu_to_be16(from->firstused);
+ hdr3->holes = from->holes;
+ hdr3->pad1 = 0;
+
+ for (i = 0; i < XFS_ATTR_LEAF_MAPSIZE; i++) {
+ hdr3->freemap[i].base = cpu_to_be16(from->freemap[i].base);
+ hdr3->freemap[i].size = cpu_to_be16(from->freemap[i].size);
+ }
+ return;
+ }
+ to->hdr.info.forw = cpu_to_be32(from->forw);
+ to->hdr.info.back = cpu_to_be32(from->back);
+ to->hdr.info.magic = cpu_to_be16(from->magic);
+ to->hdr.count = cpu_to_be16(from->count);
+ to->hdr.usedbytes = cpu_to_be16(from->usedbytes);
+ to->hdr.firstused = cpu_to_be16(from->firstused);
+ to->hdr.holes = from->holes;
+ to->hdr.pad1 = 0;
+
+ for (i = 0; i < XFS_ATTR_LEAF_MAPSIZE; i++) {
+ to->hdr.freemap[i].base = cpu_to_be16(from->freemap[i].base);
+ to->hdr.freemap[i].size = cpu_to_be16(from->freemap[i].size);
+ }
+}
+
+static bool
+xfs_attr3_leaf_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
- struct xfs_attr_leaf_hdr *hdr = bp->b_addr;
- int block_ok = 0;
+ struct xfs_attr_leafblock *leaf = bp->b_addr;
+ struct xfs_attr3_icleaf_hdr ichdr;
- block_ok = hdr->info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC);
- if (!block_ok) {
- XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr);
- xfs_buf_ioerror(bp, EFSCORRUPTED);
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_da3_node_hdr *hdr3 = bp->b_addr;
+
+ if (ichdr.magic != XFS_ATTR3_LEAF_MAGIC)
+ return false;
+
+ if (!uuid_equal(&hdr3->info.uuid, &mp->m_sb.sb_uuid))
+ return false;
+ if (be64_to_cpu(hdr3->info.blkno) != bp->b_bn)
+ return false;
+ } else {
+ if (ichdr.magic != XFS_ATTR_LEAF_MAGIC)
+ return false;
}
+ if (ichdr.count == 0)
+ return false;
+
+ /* XXX: need to range check rest of attr header values */
+ /* XXX: hash order check? */
+
+ return true;
}
static void
-xfs_attr_leaf_read_verify(
+xfs_attr3_leaf_write_verify(
struct xfs_buf *bp)
{
- xfs_attr_leaf_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+ struct xfs_buf_log_item *bip = bp->b_fspriv;
+ struct xfs_attr3_leaf_hdr *hdr3 = bp->b_addr;
+
+ if (!xfs_attr3_leaf_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ return;
+ }
+
+ if (!xfs_sb_version_hascrc(&mp->m_sb))
+ return;
+
+ if (bip)
+ hdr3->info.lsn = cpu_to_be64(bip->bli_item.li_lsn);
+
+ xfs_update_cksum(bp->b_addr, BBTOB(bp->b_length), XFS_ATTR3_LEAF_CRC_OFF);
}
+/*
+ * leaf/node format detection on trees is sketchy, so a node read can be done on
+ * leaf level blocks when detection identifies the tree as a node format tree
+ * incorrectly. In this case, we need to swap the verifier to match the correct
+ * format of the block being read.
+ */
static void
-xfs_attr_leaf_write_verify(
- struct xfs_buf *bp)
+xfs_attr3_leaf_read_verify(
+ struct xfs_buf *bp)
{
- xfs_attr_leaf_verify(bp);
+ struct xfs_mount *mp = bp->b_target->bt_mount;
+
+ if ((xfs_sb_version_hascrc(&mp->m_sb) &&
+ !xfs_verify_cksum(bp->b_addr, BBTOB(bp->b_length),
+ XFS_ATTR3_LEAF_CRC_OFF)) ||
+ !xfs_attr3_leaf_verify(bp)) {
+ XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr);
+ xfs_buf_ioerror(bp, EFSCORRUPTED);
+ }
}
-const struct xfs_buf_ops xfs_attr_leaf_buf_ops = {
- .verify_read = xfs_attr_leaf_read_verify,
- .verify_write = xfs_attr_leaf_write_verify,
+const struct xfs_buf_ops xfs_attr3_leaf_buf_ops = {
+ .verify_read = xfs_attr3_leaf_read_verify,
+ .verify_write = xfs_attr3_leaf_write_verify,
};
int
-xfs_attr_leaf_read(
+xfs_attr3_leaf_read(
struct xfs_trans *tp,
struct xfs_inode *dp,
xfs_dablk_t bno,
@@ -131,7 +270,7 @@ xfs_attr_leaf_read(
struct xfs_buf **bpp)
{
return xfs_da_read_buf(tp, dp, bno, mappedbno, bpp,
- XFS_ATTR_FORK, &xfs_attr_leaf_buf_ops);
+ XFS_ATTR_FORK, &xfs_attr3_leaf_buf_ops);
}
/*========================================================================
@@ -559,7 +698,7 @@ xfs_attr_shortform_to_leaf(xfs_da_args_t *args)
}
ASSERT(blkno == 0);
- error = xfs_attr_leaf_create(args, blkno, &bp);
+ error = xfs_attr3_leaf_create(args, blkno, &bp);
if (error) {
error = xfs_da_shrink_inode(args, 0, bp);
bp = NULL;
@@ -588,9 +727,9 @@ xfs_attr_shortform_to_leaf(xfs_da_args_t *args)
nargs.hashval = xfs_da_hashname(sfe->nameval,
sfe->namelen);
nargs.flags = XFS_ATTR_NSP_ONDISK_TO_ARGS(sfe->flags);
- error = xfs_attr_leaf_lookup_int(bp, &nargs); /* set a->index */
+ error = xfs_attr3_leaf_lookup_int(bp, &nargs); /* set a->index */
ASSERT(error == ENOATTR);
- error = xfs_attr_leaf_add(bp, &nargs);
+ error = xfs_attr3_leaf_add(bp, &nargs);
ASSERT(error != ENOSPC);
if (error)
goto out;
@@ -803,7 +942,7 @@ xfs_attr_shortform_allfit(
continue; /* don't copy partial entries */
if (!(entry->flags & XFS_ATTR_LOCAL))
return(0);
- name_loc = xfs_attr_leaf_name_local(leaf, i);
+ name_loc = xfs_attr3_leaf_name_local(leaf, i);
if (name_loc->namelen >= XFS_ATTR_SF_ENTSIZE_MAX)
return(0);
if (be16_to_cpu(name_loc->valuelen) >= XFS_ATTR_SF_ENTSIZE_MAX)
@@ -823,29 +962,34 @@ xfs_attr_shortform_allfit(
* Convert a leaf attribute list to shortform attribute list
*/
int
-xfs_attr_leaf_to_shortform(
- struct xfs_buf *bp,
- xfs_da_args_t *args,
- int forkoff)
+xfs_attr3_leaf_to_shortform(
+ struct xfs_buf *bp,
+ struct xfs_da_args *args,
+ int forkoff)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_entry_t *entry;
- xfs_attr_leaf_name_local_t *name_loc;
- xfs_da_args_t nargs;
- xfs_inode_t *dp;
- char *tmpbuffer;
- int error, i;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_attr_leaf_name_local *name_loc;
+ struct xfs_da_args nargs;
+ struct xfs_inode *dp = args->dp;
+ char *tmpbuffer;
+ int error;
+ int i;
trace_xfs_attr_leaf_to_sf(args);
- dp = args->dp;
tmpbuffer = kmem_alloc(XFS_LBSIZE(dp->i_mount), KM_SLEEP);
- ASSERT(tmpbuffer != NULL);
+ if (!tmpbuffer)
+ return ENOMEM;
- ASSERT(bp != NULL);
memcpy(tmpbuffer, bp->b_addr, XFS_LBSIZE(dp->i_mount));
+
leaf = (xfs_attr_leafblock_t *)tmpbuffer;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ entry = xfs_attr3_leaf_entryp(leaf);
+
+ /* XXX (dgc): buffer is about to be marked stale - why zero it? */
memset(bp->b_addr, 0, XFS_LBSIZE(dp->i_mount));
/*
@@ -875,14 +1019,14 @@ xfs_attr_leaf_to_shortform(
nargs.whichfork = XFS_ATTR_FORK;
nargs.trans = args->trans;
nargs.op_flags = XFS_DA_OP_OKNOENT;
- entry = &leaf->entries[0];
- for (i = 0; i < be16_to_cpu(leaf->hdr.count); entry++, i++) {
+
+ for (i = 0; i < ichdr.count; entry++, i++) {
if (entry->flags & XFS_ATTR_INCOMPLETE)
continue; /* don't copy partial entries */
if (!entry->nameidx)
continue;
ASSERT(entry->flags & XFS_ATTR_LOCAL);
- name_loc = xfs_attr_leaf_name_local(leaf, i);
+ name_loc = xfs_attr3_leaf_name_local(leaf, i);
nargs.name = name_loc->nameval;
nargs.namelen = name_loc->namelen;
nargs.value = &name_loc->nameval[nargs.namelen];
@@ -895,43 +1039,50 @@ xfs_attr_leaf_to_shortform(
out:
kmem_free(tmpbuffer);
- return(error);
+ return error;
}
/*
* Convert from using a single leaf to a root node and a leaf.
*/
int
-xfs_attr_leaf_to_node(xfs_da_args_t *args)
+xfs_attr3_leaf_to_node(
+ struct xfs_da_args *args)
{
- xfs_attr_leafblock_t *leaf;
- xfs_da_intnode_t *node;
- xfs_inode_t *dp;
- struct xfs_buf *bp1, *bp2;
- xfs_dablk_t blkno;
- int error;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr icleafhdr;
+ struct xfs_attr_leaf_entry *entries;
struct xfs_da_node_entry *btree;
+ struct xfs_da3_icnode_hdr icnodehdr;
+ struct xfs_da_intnode *node;
+ struct xfs_inode *dp = args->dp;
+ struct xfs_mount *mp = dp->i_mount;
+ struct xfs_buf *bp1 = NULL;
+ struct xfs_buf *bp2 = NULL;
+ xfs_dablk_t blkno;
+ int error;
trace_xfs_attr_leaf_to_node(args);
- dp = args->dp;
- bp1 = bp2 = NULL;
error = xfs_da_grow_inode(args, &blkno);
if (error)
goto out;
- error = xfs_attr_leaf_read(args->trans, args->dp, 0, -1, &bp1);
+ error = xfs_attr3_leaf_read(args->trans, dp, 0, -1, &bp1);
if (error)
goto out;
- bp2 = NULL;
- error = xfs_da_get_buf(args->trans, args->dp, blkno, -1, &bp2,
- XFS_ATTR_FORK);
+ error = xfs_da_get_buf(args->trans, dp, blkno, -1, &bp2, XFS_ATTR_FORK);
if (error)
goto out;
+
+ /* copy leaf to new buffer, update identifiers */
bp2->b_ops = bp1->b_ops;
- memcpy(bp2->b_addr, bp1->b_addr, XFS_LBSIZE(dp->i_mount));
- bp1 = NULL;
- xfs_trans_log_buf(args->trans, bp2, 0, XFS_LBSIZE(dp->i_mount) - 1);
+ memcpy(bp2->b_addr, bp1->b_addr, XFS_LBSIZE(mp));
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_da3_blkinfo *hdr3 = bp2->b_addr;
+ hdr3->blkno = cpu_to_be64(bp2->b_bn);
+ }
+ xfs_trans_log_buf(args->trans, bp2, 0, XFS_LBSIZE(mp) - 1);
/*
* Set up the new root node.
@@ -940,17 +1091,22 @@ xfs_attr_leaf_to_node(xfs_da_args_t *args)
if (error)
goto out;
node = bp1->b_addr;
+ xfs_da3_node_hdr_from_disk(&icnodehdr, node);
+ btree = xfs_da3_node_tree_p(node);
+
leaf = bp2->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
+ xfs_attr3_leaf_hdr_from_disk(&icleafhdr, leaf);
+ entries = xfs_attr3_leaf_entryp(leaf);
+
/* both on-disk, don't endian-flip twice */
- btree = xfs_da3_node_tree_p(node);
- btree[0].hashval = leaf->entries[be16_to_cpu(leaf->hdr.count)-1 ].hashval;
+ btree[0].hashval = entries[icleafhdr.count - 1].hashval;
btree[0].before = cpu_to_be32(blkno);
- node->hdr.count = cpu_to_be16(1);
- xfs_trans_log_buf(args->trans, bp1, 0, XFS_LBSIZE(dp->i_mount) - 1);
+ icnodehdr.count = 1;
+ xfs_da3_node_hdr_to_disk(node, &icnodehdr);
+ xfs_trans_log_buf(args->trans, bp1, 0, XFS_LBSIZE(mp) - 1);
error = 0;
out:
- return(error);
+ return error;
}
@@ -963,52 +1119,62 @@ out:
* or a leaf in a node attribute list.
*/
STATIC int
-xfs_attr_leaf_create(
- xfs_da_args_t *args,
- xfs_dablk_t blkno,
- struct xfs_buf **bpp)
+xfs_attr3_leaf_create(
+ struct xfs_da_args *args,
+ xfs_dablk_t blkno,
+ struct xfs_buf **bpp)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_hdr_t *hdr;
- xfs_inode_t *dp;
- struct xfs_buf *bp;
- int error;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_inode *dp = args->dp;
+ struct xfs_mount *mp = dp->i_mount;
+ struct xfs_buf *bp;
+ int error;
trace_xfs_attr_leaf_create(args);
- dp = args->dp;
- ASSERT(dp != NULL);
error = xfs_da_get_buf(args->trans, args->dp, blkno, -1, &bp,
XFS_ATTR_FORK);
if (error)
- return(error);
- bp->b_ops = &xfs_attr_leaf_buf_ops;
+ return error;
+ bp->b_ops = &xfs_attr3_leaf_buf_ops;
leaf = bp->b_addr;
- memset((char *)leaf, 0, XFS_LBSIZE(dp->i_mount));
- hdr = &leaf->hdr;
- hdr->info.magic = cpu_to_be16(XFS_ATTR_LEAF_MAGIC);
- hdr->firstused = cpu_to_be16(XFS_LBSIZE(dp->i_mount));
- if (!hdr->firstused) {
- hdr->firstused = cpu_to_be16(
- XFS_LBSIZE(dp->i_mount) - XFS_ATTR_LEAF_NAME_ALIGN);
- }
+ memset(leaf, 0, XFS_LBSIZE(mp));
- hdr->freemap[0].base = cpu_to_be16(sizeof(xfs_attr_leaf_hdr_t));
- hdr->freemap[0].size = cpu_to_be16(be16_to_cpu(hdr->firstused) -
- sizeof(xfs_attr_leaf_hdr_t));
+ memset(&ichdr, 0, sizeof(ichdr));
+ ichdr.firstused = XFS_LBSIZE(mp);
+
+ if (xfs_sb_version_hascrc(&mp->m_sb)) {
+ struct xfs_da3_blkinfo *hdr3 = bp->b_addr;
+
+ ichdr.magic = XFS_ATTR3_LEAF_MAGIC;
+
+ hdr3->blkno = cpu_to_be64(bp->b_bn);
+ hdr3->owner = cpu_to_be64(dp->i_ino);
+ uuid_copy(&hdr3->uuid, &mp->m_sb.sb_uuid);
+
+ ichdr.freemap[0].base = sizeof(struct xfs_attr3_leaf_hdr);
+ } else {
+ ichdr.magic = XFS_ATTR_LEAF_MAGIC;
+ ichdr.freemap[0].base = sizeof(struct xfs_attr_leaf_hdr);
+ }
+ ichdr.freemap[0].size = ichdr.firstused - ichdr.freemap[0].base;
- xfs_trans_log_buf(args->trans, bp, 0, XFS_LBSIZE(dp->i_mount) - 1);
+ xfs_attr3_leaf_hdr_to_disk(leaf, &ichdr);
+ xfs_trans_log_buf(args->trans, bp, 0, XFS_LBSIZE(mp) - 1);
*bpp = bp;
- return(0);
+ return 0;
}
/*
* Split the leaf node, rebalance, then add the new entry.
*/
int
-xfs_attr_leaf_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
- xfs_da_state_blk_t *newblk)
+xfs_attr3_leaf_split(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *oldblk,
+ struct xfs_da_state_blk *newblk)
{
xfs_dablk_t blkno;
int error;
@@ -1022,7 +1188,7 @@ xfs_attr_leaf_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
error = xfs_da_grow_inode(state->args, &blkno);
if (error)
return(error);
- error = xfs_attr_leaf_create(state->args, blkno, &newblk->bp);
+ error = xfs_attr3_leaf_create(state->args, blkno, &newblk->bp);
if (error)
return(error);
newblk->blkno = blkno;
@@ -1032,7 +1198,7 @@ xfs_attr_leaf_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* Rebalance the entries across the two leaves.
* NOTE: rebalance() currently depends on the 2nd block being empty.
*/
- xfs_attr_leaf_rebalance(state, oldblk, newblk);
+ xfs_attr3_leaf_rebalance(state, oldblk, newblk);
error = xfs_da3_blk_link(state, oldblk, newblk);
if (error)
return(error);
@@ -1046,10 +1212,10 @@ xfs_attr_leaf_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
*/
if (state->inleaf) {
trace_xfs_attr_leaf_add_old(state->args);
- error = xfs_attr_leaf_add(oldblk->bp, state->args);
+ error = xfs_attr3_leaf_add(oldblk->bp, state->args);
} else {
trace_xfs_attr_leaf_add_new(state->args);
- error = xfs_attr_leaf_add(newblk->bp, state->args);
+ error = xfs_attr3_leaf_add(newblk->bp, state->args);
}
/*
@@ -1064,22 +1230,23 @@ xfs_attr_leaf_split(xfs_da_state_t *state, xfs_da_state_blk_t *oldblk,
* Add a name to the leaf attribute list structure.
*/
int
-xfs_attr_leaf_add(
+xfs_attr3_leaf_add(
struct xfs_buf *bp,
struct xfs_da_args *args)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_hdr_t *hdr;
- xfs_attr_leaf_map_t *map;
- int tablesize, entsize, sum, tmp, i;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ int tablesize;
+ int entsize;
+ int sum;
+ int tmp;
+ int i;
trace_xfs_attr_leaf_add(args);
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- ASSERT((args->index >= 0)
- && (args->index <= be16_to_cpu(leaf->hdr.count)));
- hdr = &leaf->hdr;
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ ASSERT(args->index >= 0 && args->index <= ichdr.count);
entsize = xfs_attr_leaf_newentsize(args->namelen, args->valuelen,
args->trans->t_mountp->m_sb.sb_blocksize, NULL);
@@ -1087,25 +1254,23 @@ xfs_attr_leaf_add(
* Search through freemap for first-fit on new name length.
* (may need to figure in size of entry struct too)
*/
- tablesize = (be16_to_cpu(hdr->count) + 1)
- * sizeof(xfs_attr_leaf_entry_t)
- + sizeof(xfs_attr_leaf_hdr_t);
- map = &hdr->freemap[XFS_ATTR_LEAF_MAPSIZE-1];
- for (sum = 0, i = XFS_ATTR_LEAF_MAPSIZE-1; i >= 0; map--, i--) {
- if (tablesize > be16_to_cpu(hdr->firstused)) {
- sum += be16_to_cpu(map->size);
+ tablesize = (ichdr.count + 1) * sizeof(xfs_attr_leaf_entry_t)
+ + xfs_attr3_leaf_hdr_size(leaf);
+ for (sum = 0, i = XFS_ATTR_LEAF_MAPSIZE - 1; i >= 0; i--) {
+ if (tablesize > ichdr.firstused) {
+ sum += ichdr.freemap[i].size;
continue;
}
- if (!map->size)
+ if (!ichdr.freemap[i].size)
continue; /* no space in this map */
tmp = entsize;
- if (be16_to_cpu(map->base) < be16_to_cpu(hdr->firstused))
+ if (ichdr.freemap[i].base < ichdr.firstused)
tmp += sizeof(xfs_attr_leaf_entry_t);
- if (be16_to_cpu(map->size) >= tmp) {
- tmp = xfs_attr_leaf_add_work(bp, args, i);
- return(tmp);
+ if (ichdr.freemap[i].size >= tmp) {
+ tmp = xfs_attr3_leaf_add_work(bp, &ichdr, args, i);
+ goto out_log_hdr;
}
- sum += be16_to_cpu(map->size);
+ sum += ichdr.freemap[i].size;
}
/*
@@ -1113,82 +1278,90 @@ xfs_attr_leaf_add(
* and we don't have enough freespace, then compaction will do us
* no good and we should just give up.
*/
- if (!hdr->holes && (sum < entsize))
- return(XFS_ERROR(ENOSPC));
+ if (!ichdr.holes && sum < entsize)
+ return XFS_ERROR(ENOSPC);
/*
* Compact the entries to coalesce free space.
* This may change the hdr->count via dropping INCOMPLETE entries.
*/
- xfs_attr_leaf_compact(args, bp);
+ xfs_attr3_leaf_compact(args, &ichdr, bp);
/*
* After compaction, the block is guaranteed to have only one
* free region, in freemap[0]. If it is not big enough, give up.
*/
- if (be16_to_cpu(hdr->freemap[0].size)
- < (entsize + sizeof(xfs_attr_leaf_entry_t)))
- return(XFS_ERROR(ENOSPC));
+ if (ichdr.freemap[0].size < (entsize + sizeof(xfs_attr_leaf_entry_t))) {
+ tmp = ENOSPC;
+ goto out_log_hdr;
+ }
+
+ tmp = xfs_attr3_leaf_add_work(bp, &ichdr, args, 0);
- return(xfs_attr_leaf_add_work(bp, args, 0));
+out_log_hdr:
+ xfs_attr3_leaf_hdr_to_disk(leaf, &ichdr);
+ xfs_trans_log_buf(args->trans, bp,
+ XFS_DA_LOGRANGE(leaf, &leaf->hdr,
+ xfs_attr3_leaf_hdr_size(leaf)));
+ return tmp;
}
/*
* Add a name to a leaf attribute list structure.
*/
STATIC int
-xfs_attr_leaf_add_work(
- struct xfs_buf *bp,
- xfs_da_args_t *args,
- int mapindex)
+xfs_attr3_leaf_add_work(
+ struct xfs_buf *bp,
+ struct xfs_attr3_icleaf_hdr *ichdr,
+ struct xfs_da_args *args,
+ int mapindex)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_hdr_t *hdr;
- xfs_attr_leaf_entry_t *entry;
- xfs_attr_leaf_name_local_t *name_loc;
- xfs_attr_leaf_name_remote_t *name_rmt;
- xfs_attr_leaf_map_t *map;
- xfs_mount_t *mp;
- int tmp, i;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_attr_leaf_name_local *name_loc;
+ struct xfs_attr_leaf_name_remote *name_rmt;
+ struct xfs_attr_leaf_map *map;
+ struct xfs_mount *mp;
+ int tmp;
+ int i;
trace_xfs_attr_leaf_add_work(args);
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- hdr = &leaf->hdr;
- ASSERT((mapindex >= 0) && (mapindex < XFS_ATTR_LEAF_MAPSIZE));
- ASSERT((args->index >= 0) && (args->index <= be16_to_cpu(hdr->count)));
+ ASSERT(mapindex >= 0 && mapindex < XFS_ATTR_LEAF_MAPSIZE);
+ ASSERT(args->index >= 0 && args->index <= ichdr->count);
/*
* Force open some space in the entry array and fill it in.
*/
- entry = &leaf->entries[args->index];
- if (args->index < be16_to_cpu(hdr->count)) {
- tmp = be16_to_cpu(hdr->count) - args->index;
+ entry = &xfs_attr3_leaf_entryp(leaf)[args->index];
+ if (args->index < ichdr->count) {
+ tmp = ichdr->count - args->index;
tmp *= sizeof(xfs_attr_leaf_entry_t);
- memmove((char *)(entry+1), (char *)entry, tmp);
+ memmove(entry + 1, entry, tmp);
xfs_trans_log_buf(args->trans, bp,
XFS_DA_LOGRANGE(leaf, entry, tmp + sizeof(*entry)));
}
- be16_add_cpu(&hdr->count, 1);
+ ichdr->count++;
/*
* Allocate space for the new string (at the end of the run).
*/
- map = &hdr->freemap[mapindex];
mp = args->trans->t_mountp;
- ASSERT(be16_to_cpu(map->base) < XFS_LBSIZE(mp));
- ASSERT((be16_to_cpu(map->base) & 0x3) == 0);
- ASSERT(be16_to_cpu(map->size) >=
+ ASSERT(ichdr->freemap[mapindex].base < XFS_LBSIZE(mp));
+ ASSERT((ichdr->freemap[mapindex].base & 0x3) == 0);
+ ASSERT(ichdr->freemap[mapindex].size >=
xfs_attr_leaf_newentsize(args->namelen, args->valuelen,
mp->m_sb.sb_blocksize, NULL));
- ASSERT(be16_to_cpu(map->size) < XFS_LBSIZE(mp));
- ASSERT((be16_to_cpu(map->size) & 0x3) == 0);
- be16_add_cpu(&map->size,
- -xfs_attr_leaf_newentsize(args->namelen, args->valuelen,
- mp->m_sb.sb_blocksize, &tmp));
- entry->nameidx = cpu_to_be16(be16_to_cpu(map->base) +
- be16_to_cpu(map->size));
+ ASSERT(ichdr->freemap[mapindex].size < XFS_LBSIZE(mp));
+ ASSERT((ichdr->freemap[mapindex].size & 0x3) == 0);
+
+ ichdr->freemap[mapindex].size -=
+ xfs_attr_leaf_newentsize(args->namelen, args->valuelen,
+ mp->m_sb.sb_blocksize, &tmp);
+
+ entry->nameidx = cpu_to_be16(ichdr->freemap[mapindex].base +
+ ichdr->freemap[mapindex].size);
entry->hashval = cpu_to_be32(args->hashval);
entry->flags = tmp ? XFS_ATTR_LOCAL : 0;
entry->flags |= XFS_ATTR_NSP_ARGS_TO_ONDISK(args->flags);
@@ -1203,7 +1376,7 @@ xfs_attr_leaf_add_work(
XFS_DA_LOGRANGE(leaf, entry, sizeof(*entry)));
ASSERT((args->index == 0) ||
(be32_to_cpu(entry->hashval) >= be32_to_cpu((entry-1)->hashval)));
- ASSERT((args->index == be16_to_cpu(hdr->count)-1) ||
+ ASSERT((args->index == ichdr->count - 1) ||
(be32_to_cpu(entry->hashval) <= be32_to_cpu((entry+1)->hashval)));
/*
@@ -1214,14 +1387,14 @@ xfs_attr_leaf_add_work(
* as part of this transaction (a split operation for example).
*/
if (entry->flags & XFS_ATTR_LOCAL) {
- name_loc = xfs_attr_leaf_name_local(leaf, args->index);
+ name_loc = xfs_attr3_leaf_name_local(leaf, args->index);
name_loc->namelen = args->namelen;
name_loc->valuelen = cpu_to_be16(args->valuelen);
memcpy((char *)name_loc->nameval, args->name, args->namelen);
memcpy((char *)&name_loc->nameval[args->namelen], args->value,
be16_to_cpu(name_loc->valuelen));
} else {
- name_rmt = xfs_attr_leaf_name_remote(leaf, args->index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, args->index);
name_rmt->namelen = args->namelen;
memcpy((char *)name_rmt->name, args->name, args->namelen);
entry->flags |= XFS_ATTR_INCOMPLETE;
@@ -1232,44 +1405,41 @@ xfs_attr_leaf_add_work(
args->rmtblkcnt = XFS_B_TO_FSB(mp, args->valuelen);
}
xfs_trans_log_buf(args->trans, bp,
- XFS_DA_LOGRANGE(leaf, xfs_attr_leaf_name(leaf, args->index),
+ XFS_DA_LOGRANGE(leaf, xfs_attr3_leaf_name(leaf, args->index),
xfs_attr_leaf_entsize(leaf, args->index)));
/*
* Update the control info for this leaf node
*/
- if (be16_to_cpu(entry->nameidx) < be16_to_cpu(hdr->firstused)) {
- /* both on-disk, don't endian-flip twice */
- hdr->firstused = entry->nameidx;
- }
- ASSERT(be16_to_cpu(hdr->firstused) >=
- ((be16_to_cpu(hdr->count) * sizeof(*entry)) + sizeof(*hdr)));
- tmp = (be16_to_cpu(hdr->count)-1) * sizeof(xfs_attr_leaf_entry_t)
- + sizeof(xfs_attr_leaf_hdr_t);
- map = &hdr->freemap[0];
+ if (be16_to_cpu(entry->nameidx) < ichdr->firstused)
+ ichdr->firstused = be16_to_cpu(entry->nameidx);
+
+ ASSERT(ichdr->firstused >= ichdr->count * sizeof(xfs_attr_leaf_entry_t)
+ + xfs_attr3_leaf_hdr_size(leaf));
+ tmp = (ichdr->count - 1) * sizeof(xfs_attr_leaf_entry_t)
+ + xfs_attr3_leaf_hdr_size(leaf);
+
for (i = 0; i < XFS_ATTR_LEAF_MAPSIZE; map++, i++) {
- if (be16_to_cpu(map->base) == tmp) {
- be16_add_cpu(&map->base, sizeof(xfs_attr_leaf_entry_t));
- be16_add_cpu(&map->size,
- -((int)sizeof(xfs_attr_leaf_entry_t)));
+ if (ichdr->freemap[i].base == tmp) {
+ ichdr->freemap[i].base += sizeof(xfs_attr_leaf_entry_t);
+ ichdr->freemap[i].size -= sizeof(xfs_attr_leaf_entry_t);
}
}
- be16_add_cpu(&hdr->usedbytes, xfs_attr_leaf_entsize(leaf, args->index));
- xfs_trans_log_buf(args->trans, bp,
- XFS_DA_LOGRANGE(leaf, hdr, sizeof(*hdr)));
- return(0);
+ ichdr->usedbytes += xfs_attr_leaf_entsize(leaf, args->index);
+ return 0;
}
/*
* Garbage collect a leaf attribute list block by copying it to a new buffer.
*/
STATIC void
-xfs_attr_leaf_compact(
+xfs_attr3_leaf_compact(
struct xfs_da_args *args,
+ struct xfs_attr3_icleaf_hdr *ichdr_d,
struct xfs_buf *bp)
{
xfs_attr_leafblock_t *leaf_s, *leaf_d;
- xfs_attr_leaf_hdr_t *hdr_s, *hdr_d;
+ struct xfs_attr3_icleaf_hdr ichdr_s;
struct xfs_trans *trans = args->trans;
struct xfs_mount *mp = trans->t_mountp;
char *tmpbuffer;
@@ -1286,34 +1456,69 @@ xfs_attr_leaf_compact(
*/
leaf_s = (xfs_attr_leafblock_t *)tmpbuffer;
leaf_d = bp->b_addr;
- hdr_s = &leaf_s->hdr;
- hdr_d = &leaf_d->hdr;
- hdr_d->info = hdr_s->info; /* struct copy */
- hdr_d->firstused = cpu_to_be16(XFS_LBSIZE(mp));
- /* handle truncation gracefully */
- if (!hdr_d->firstused) {
- hdr_d->firstused = cpu_to_be16(
- XFS_LBSIZE(mp) - XFS_ATTR_LEAF_NAME_ALIGN);
- }
- hdr_d->usedbytes = 0;
- hdr_d->count = 0;
- hdr_d->holes = 0;
- hdr_d->freemap[0].base = cpu_to_be16(sizeof(xfs_attr_leaf_hdr_t));
- hdr_d->freemap[0].size = cpu_to_be16(be16_to_cpu(hdr_d->firstused) -
- sizeof(xfs_attr_leaf_hdr_t));
+ ichdr_s = *ichdr_d; /* struct copy */
+ ichdr_d->firstused = XFS_LBSIZE(mp);
+ ichdr_d->usedbytes = 0;
+ ichdr_d->count = 0;
+ ichdr_d->holes = 0;
+ ichdr_d->freemap[0].base = xfs_attr3_leaf_hdr_size(leaf_s);
+ ichdr_d->freemap[0].size = ichdr_d->firstused - ichdr_d->freemap[0].base;
/*
* Copy all entry's in the same (sorted) order,
* but allocate name/value pairs packed and in sequence.
*/
- xfs_attr_leaf_moveents(leaf_s, 0, leaf_d, 0,
- be16_to_cpu(hdr_s->count), mp);
+ xfs_attr3_leaf_moveents(leaf_s, &ichdr_s, 0, leaf_d, ichdr_d, 0,
+ ichdr_s.count, mp);
+ /*
+ * this logs the entire buffer, but the caller must write the header
+ * back to the buffer when it is finished modifying it.
+ */
xfs_trans_log_buf(trans, bp, 0, XFS_LBSIZE(mp) - 1);
kmem_free(tmpbuffer);
}
/*
+ * Compare two leaf blocks "order".
+ * Return 0 unless leaf2 should go before leaf1.
+ */
+static int
+xfs_attr3_leaf_order(
+ struct xfs_buf *leaf1_bp,
+ struct xfs_attr3_icleaf_hdr *leaf1hdr,
+ struct xfs_buf *leaf2_bp,
+ struct xfs_attr3_icleaf_hdr *leaf2hdr)
+{
+ struct xfs_attr_leaf_entry *entries1;
+ struct xfs_attr_leaf_entry *entries2;
+
+ entries1 = xfs_attr3_leaf_entryp(leaf1_bp->b_addr);
+ entries2 = xfs_attr3_leaf_entryp(leaf2_bp->b_addr);
+ if (leaf1hdr->count > 0 && leaf2hdr->count > 0 &&
+ ((be32_to_cpu(entries2[0].hashval) <
+ be32_to_cpu(entries1[0].hashval)) ||
+ (be32_to_cpu(entries2[leaf2hdr->count - 1].hashval) <
+ be32_to_cpu(entries1[leaf1hdr->count - 1].hashval)))) {
+ return 1;
+ }
+ return 0;
+}
+
+int
+xfs_attr_leaf_order(
+ struct xfs_buf *leaf1_bp,
+ struct xfs_buf *leaf2_bp)
+{
+ struct xfs_attr3_icleaf_hdr ichdr1;
+ struct xfs_attr3_icleaf_hdr ichdr2;
+
+ xfs_attr3_leaf_hdr_from_disk(&ichdr1, leaf1_bp->b_addr);
+ xfs_attr3_leaf_hdr_from_disk(&ichdr2, leaf2_bp->b_addr);
+ return xfs_attr3_leaf_order(leaf1_bp, &ichdr1, leaf2_bp, &ichdr2);
+}
+
+/*
* Redistribute the attribute list entries between two leaf nodes,
* taking into account the size of the new entry.
*
@@ -1326,14 +1531,23 @@ xfs_attr_leaf_compact(
* the "new" and "old" values can end up in different blocks.
*/
STATIC void
-xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
- xfs_da_state_blk_t *blk2)
+xfs_attr3_leaf_rebalance(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *blk1,
+ struct xfs_da_state_blk *blk2)
{
- xfs_da_args_t *args;
- xfs_da_state_blk_t *tmp_blk;
- xfs_attr_leafblock_t *leaf1, *leaf2;
- xfs_attr_leaf_hdr_t *hdr1, *hdr2;
- int count, totallen, max, space, swap;
+ struct xfs_da_args *args;
+ struct xfs_attr_leafblock *leaf1;
+ struct xfs_attr_leafblock *leaf2;
+ struct xfs_attr3_icleaf_hdr ichdr1;
+ struct xfs_attr3_icleaf_hdr ichdr2;
+ struct xfs_attr_leaf_entry *entries1;
+ struct xfs_attr_leaf_entry *entries2;
+ int count;
+ int totallen;
+ int max;
+ int space;
+ int swap;
/*
* Set up environment.
@@ -1342,9 +1556,9 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
ASSERT(blk2->magic == XFS_ATTR_LEAF_MAGIC);
leaf1 = blk1->bp->b_addr;
leaf2 = blk2->bp->b_addr;
- ASSERT(leaf1->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- ASSERT(leaf2->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- ASSERT(leaf2->hdr.count == 0);
+ xfs_attr3_leaf_hdr_from_disk(&ichdr1, leaf1);
+ xfs_attr3_leaf_hdr_from_disk(&ichdr2, leaf2);
+ ASSERT(ichdr2.count == 0);
args = state->args;
trace_xfs_attr_leaf_rebalance(args);
@@ -1356,16 +1570,23 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* second block, this code should never set "swap".
*/
swap = 0;
- if (xfs_attr_leaf_order(blk1->bp, blk2->bp)) {
+ if (xfs_attr3_leaf_order(blk1->bp, &ichdr1, blk2->bp, &ichdr2)) {
+ struct xfs_da_state_blk *tmp_blk;
+ struct xfs_attr3_icleaf_hdr tmp_ichdr;
+
tmp_blk = blk1;
blk1 = blk2;
blk2 = tmp_blk;
+
+ /* struct copies to swap them rather than reconverting */
+ tmp_ichdr = ichdr1;
+ ichdr1 = ichdr2;
+ ichdr2 = tmp_ichdr;
+
leaf1 = blk1->bp->b_addr;
leaf2 = blk2->bp->b_addr;
swap = 1;
}
- hdr1 = &leaf1->hdr;
- hdr2 = &leaf2->hdr;
/*
* Examine entries until we reduce the absolute difference in
@@ -1375,41 +1596,39 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* "inleaf" is true if the new entry should be inserted into blk1.
* If "swap" is also true, then reverse the sense of "inleaf".
*/
- state->inleaf = xfs_attr_leaf_figure_balance(state, blk1, blk2,
- &count, &totallen);
+ state->inleaf = xfs_attr3_leaf_figure_balance(state, blk1, &ichdr1,
+ blk2, &ichdr2,
+ &count, &totallen);
if (swap)
state->inleaf = !state->inleaf;
/*
* Move any entries required from leaf to leaf:
*/
- if (count < be16_to_cpu(hdr1->count)) {
+ if (count < ichdr1.count) {
/*
* Figure the total bytes to be added to the destination leaf.
*/
/* number entries being moved */
- count = be16_to_cpu(hdr1->count) - count;
- space = be16_to_cpu(hdr1->usedbytes) - totallen;
+ count = ichdr1.count - count;
+ space = ichdr1.usedbytes - totallen;
space += count * sizeof(xfs_attr_leaf_entry_t);
/*
* leaf2 is the destination, compact it if it looks tight.
*/
- max = be16_to_cpu(hdr2->firstused)
- - sizeof(xfs_attr_leaf_hdr_t);
- max -= be16_to_cpu(hdr2->count) * sizeof(xfs_attr_leaf_entry_t);
+ max = ichdr2.firstused - xfs_attr3_leaf_hdr_size(leaf1);
+ max -= ichdr2.count * sizeof(xfs_attr_leaf_entry_t);
if (space > max)
- xfs_attr_leaf_compact(args, blk2->bp);
+ xfs_attr3_leaf_compact(args, &ichdr2, blk2->bp);
/*
* Move high entries from leaf1 to low end of leaf2.
*/
- xfs_attr_leaf_moveents(leaf1, be16_to_cpu(hdr1->count) - count,
- leaf2, 0, count, state->mp);
+ xfs_attr3_leaf_moveents(leaf1, &ichdr1, ichdr1.count - count,
+ leaf2, &ichdr2, 0, count, state->mp);
- xfs_trans_log_buf(args->trans, blk1->bp, 0, state->blocksize-1);
- xfs_trans_log_buf(args->trans, blk2->bp, 0, state->blocksize-1);
- } else if (count > be16_to_cpu(hdr1->count)) {
+ } else if (count > ichdr1.count) {
/*
* I assert that since all callers pass in an empty
* second buffer, this code should never execute.
@@ -1420,36 +1639,37 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* Figure the total bytes to be added to the destination leaf.
*/
/* number entries being moved */
- count -= be16_to_cpu(hdr1->count);
- space = totallen - be16_to_cpu(hdr1->usedbytes);
+ count -= ichdr1.count;
+ space = totallen - ichdr1.usedbytes;
space += count * sizeof(xfs_attr_leaf_entry_t);
/*
* leaf1 is the destination, compact it if it looks tight.
*/
- max = be16_to_cpu(hdr1->firstused)
- - sizeof(xfs_attr_leaf_hdr_t);
- max -= be16_to_cpu(hdr1->count) * sizeof(xfs_attr_leaf_entry_t);
+ max = ichdr1.firstused - xfs_attr3_leaf_hdr_size(leaf1);
+ max -= ichdr1.count * sizeof(xfs_attr_leaf_entry_t);
if (space > max)
- xfs_attr_leaf_compact(args, blk1->bp);
+ xfs_attr3_leaf_compact(args, &ichdr1, blk1->bp);
/*
* Move low entries from leaf2 to high end of leaf1.
*/
- xfs_attr_leaf_moveents(leaf2, 0, leaf1,
- be16_to_cpu(hdr1->count), count, state->mp);
-
- xfs_trans_log_buf(args->trans, blk1->bp, 0, state->blocksize-1);
- xfs_trans_log_buf(args->trans, blk2->bp, 0, state->blocksize-1);
+ xfs_attr3_leaf_moveents(leaf2, &ichdr2, 0, leaf1, &ichdr1,
+ ichdr1.count, count, state->mp);
}
+ xfs_attr3_leaf_hdr_to_disk(leaf1, &ichdr1);
+ xfs_attr3_leaf_hdr_to_disk(leaf2, &ichdr2);
+ xfs_trans_log_buf(args->trans, blk1->bp, 0, state->blocksize-1);
+ xfs_trans_log_buf(args->trans, blk2->bp, 0, state->blocksize-1);
+
/*
* Copy out last hashval in each block for B-tree code.
*/
- blk1->hashval = be32_to_cpu(
- leaf1->entries[be16_to_cpu(leaf1->hdr.count)-1].hashval);
- blk2->hashval = be32_to_cpu(
- leaf2->entries[be16_to_cpu(leaf2->hdr.count)-1].hashval);
+ entries1 = xfs_attr3_leaf_entryp(leaf1);
+ entries2 = xfs_attr3_leaf_entryp(leaf2);
+ blk1->hashval = be32_to_cpu(entries1[ichdr1.count - 1].hashval);
+ blk2->hashval = be32_to_cpu(entries2[ichdr2.count - 1].hashval);
/*
* Adjust the expected index for insertion.
@@ -1463,12 +1683,12 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* inserting. The index/blkno fields refer to the "old" entry,
* while the index2/blkno2 fields refer to the "new" entry.
*/
- if (blk1->index > be16_to_cpu(leaf1->hdr.count)) {
+ if (blk1->index > ichdr1.count) {
ASSERT(state->inleaf == 0);
- blk2->index = blk1->index - be16_to_cpu(leaf1->hdr.count);
+ blk2->index = blk1->index - ichdr1.count;
args->index = args->index2 = blk2->index;
args->blkno = args->blkno2 = blk2->blkno;
- } else if (blk1->index == be16_to_cpu(leaf1->hdr.count)) {
+ } else if (blk1->index == ichdr1.count) {
if (state->inleaf) {
args->index = blk1->index;
args->blkno = blk1->blkno;
@@ -1480,8 +1700,7 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* is already stored in blkno2/index2, so don't
* overwrite it overwise we corrupt the tree.
*/
- blk2->index = blk1->index
- - be16_to_cpu(leaf1->hdr.count);
+ blk2->index = blk1->index - ichdr1.count;
args->index = blk2->index;
args->blkno = blk2->blkno;
if (!state->extravalid) {
@@ -1509,42 +1728,40 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1,
* GROT: Do a double-split for this case?
*/
STATIC int
-xfs_attr_leaf_figure_balance(xfs_da_state_t *state,
- xfs_da_state_blk_t *blk1,
- xfs_da_state_blk_t *blk2,
- int *countarg, int *usedbytesarg)
+xfs_attr3_leaf_figure_balance(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *blk1,
+ struct xfs_attr3_icleaf_hdr *ichdr1,
+ struct xfs_da_state_blk *blk2,
+ struct xfs_attr3_icleaf_hdr *ichdr2,
+ int *countarg,
+ int *usedbytesarg)
{
- xfs_attr_leafblock_t *leaf1, *leaf2;
- xfs_attr_leaf_hdr_t *hdr1, *hdr2;
- xfs_attr_leaf_entry_t *entry;
- int count, max, index, totallen, half;
- int lastdelta, foundit, tmp;
-
- /*
- * Set up environment.
- */
- leaf1 = blk1->bp->b_addr;
- leaf2 = blk2->bp->b_addr;
- hdr1 = &leaf1->hdr;
- hdr2 = &leaf2->hdr;
- foundit = 0;
- totallen = 0;
+ struct xfs_attr_leafblock *leaf1 = blk1->bp->b_addr;
+ struct xfs_attr_leafblock *leaf2 = blk2->bp->b_addr;
+ struct xfs_attr_leaf_entry *entry;
+ int count;
+ int max;
+ int index;
+ int totallen = 0;
+ int half;
+ int lastdelta;
+ int foundit = 0;
+ int tmp;
/*
* Examine entries until we reduce the absolute difference in
* byte usage between the two blocks to a minimum.
*/
- max = be16_to_cpu(hdr1->count) + be16_to_cpu(hdr2->count);
- half = (max+1) * sizeof(*entry);
- half += be16_to_cpu(hdr1->usedbytes) +
- be16_to_cpu(hdr2->usedbytes) +
- xfs_attr_leaf_newentsize(
- state->args->namelen,
- state->args->valuelen,
- state->blocksize, NULL);
+ max = ichdr1->count + ichdr2->count;
+ half = (max + 1) * sizeof(*entry);
+ half += ichdr1->usedbytes + ichdr2->usedbytes +
+ xfs_attr_leaf_newentsize(state->args->namelen,
+ state->args->valuelen,
+ state->blocksize, NULL);
half /= 2;
lastdelta = state->blocksize;
- entry = &leaf1->entries[0];
+ entry = xfs_attr3_leaf_entryp(leaf1);
for (count = index = 0; count < max; entry++, index++, count++) {
#define XFS_ATTR_ABS(A) (((A) < 0) ? -(A) : (A))
@@ -1567,9 +1784,9 @@ xfs_attr_leaf_figure_balance(xfs_da_state_t *state,
/*
* Wrap around into the second block if necessary.
*/
- if (count == be16_to_cpu(hdr1->count)) {
+ if (count == ichdr1->count) {
leaf1 = leaf2;
- entry = &leaf1->entries[0];
+ entry = xfs_attr3_leaf_entryp(leaf1);
index = 0;
}
@@ -1600,7 +1817,7 @@ xfs_attr_leaf_figure_balance(xfs_da_state_t *state,
*countarg = count;
*usedbytesarg = totallen;
- return(foundit);
+ return foundit;
}
/*========================================================================
@@ -1619,14 +1836,20 @@ xfs_attr_leaf_figure_balance(xfs_da_state_t *state,
* GROT: allow for INCOMPLETE entries in calculation.
*/
int
-xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
+xfs_attr3_leaf_toosmall(
+ struct xfs_da_state *state,
+ int *action)
{
- xfs_attr_leafblock_t *leaf;
- xfs_da_state_blk_t *blk;
- xfs_da_blkinfo_t *info;
- int count, bytes, forward, error, retval, i;
- xfs_dablk_t blkno;
- struct xfs_buf *bp;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_da_state_blk *blk;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_buf *bp;
+ xfs_dablk_t blkno;
+ int bytes;
+ int forward;
+ int error;
+ int retval;
+ int i;
trace_xfs_attr_leaf_toosmall(state->args);
@@ -1636,13 +1859,11 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
* to coalesce with a sibling.
*/
blk = &state->path.blk[ state->path.active-1 ];
- info = blk->bp->b_addr;
- ASSERT(info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- leaf = (xfs_attr_leafblock_t *)info;
- count = be16_to_cpu(leaf->hdr.count);
- bytes = sizeof(xfs_attr_leaf_hdr_t) +
- count * sizeof(xfs_attr_leaf_entry_t) +
- be16_to_cpu(leaf->hdr.usedbytes);
+ leaf = blk->bp->b_addr;
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ bytes = xfs_attr3_leaf_hdr_size(leaf) +
+ ichdr.count * sizeof(xfs_attr_leaf_entry_t) +
+ ichdr.usedbytes;
if (bytes > (state->blocksize >> 1)) {
*action = 0; /* blk over 50%, don't try to join */
return(0);
@@ -1654,12 +1875,12 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
* coalesce it with a sibling block. We choose (arbitrarily)
* to merge with the forward block unless it is NULL.
*/
- if (count == 0) {
+ if (ichdr.count == 0) {
/*
* Make altpath point to the block we want to keep and
* path point to the block we want to drop (this one).
*/
- forward = (info->forw != 0);
+ forward = (ichdr.forw != 0);
memcpy(&state->altpath, &state->path, sizeof(state->path));
error = xfs_da3_path_shift(state, &state->altpath, forward,
0, &retval);
@@ -1670,7 +1891,7 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
} else {
*action = 2;
}
- return(0);
+ return 0;
}
/*
@@ -1681,28 +1902,28 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
* to shrink an attribute list over time.
*/
/* start with smaller blk num */
- forward = (be32_to_cpu(info->forw) < be32_to_cpu(info->back));
+ forward = ichdr.forw < ichdr.back;
for (i = 0; i < 2; forward = !forward, i++) {
+ struct xfs_attr3_icleaf_hdr ichdr2;
if (forward)
- blkno = be32_to_cpu(info->forw);
+ blkno = ichdr.forw;
else
- blkno = be32_to_cpu(info->back);
+ blkno = ichdr.back;
if (blkno == 0)
continue;
- error = xfs_attr_leaf_read(state->args->trans, state->args->dp,
+ error = xfs_attr3_leaf_read(state->args->trans, state->args->dp,
blkno, -1, &bp);
if (error)
return(error);
- leaf = (xfs_attr_leafblock_t *)info;
- count = be16_to_cpu(leaf->hdr.count);
- bytes = state->blocksize - (state->blocksize>>2);
- bytes -= be16_to_cpu(leaf->hdr.usedbytes);
- leaf = bp->b_addr;
- count += be16_to_cpu(leaf->hdr.count);
- bytes -= be16_to_cpu(leaf->hdr.usedbytes);
- bytes -= count * sizeof(xfs_attr_leaf_entry_t);
- bytes -= sizeof(xfs_attr_leaf_hdr_t);
+ xfs_attr3_leaf_hdr_from_disk(&ichdr2, bp->b_addr);
+
+ bytes = state->blocksize - (state->blocksize >> 2) -
+ ichdr.usedbytes - ichdr2.usedbytes -
+ ((ichdr.count + ichdr2.count) *
+ sizeof(xfs_attr_leaf_entry_t)) -
+ xfs_attr3_leaf_hdr_size(leaf);
+
xfs_trans_brelse(state->args->trans, bp);
if (bytes >= 0)
break; /* fits with at least 25% to spare */
@@ -1741,32 +1962,35 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action)
* If two leaves are 37% full, when combined they will leave 25% free.
*/
int
-xfs_attr_leaf_remove(
- struct xfs_buf *bp,
- xfs_da_args_t *args)
+xfs_attr3_leaf_remove(
+ struct xfs_buf *bp,
+ struct xfs_da_args *args)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_hdr_t *hdr;
- xfs_attr_leaf_map_t *map;
- xfs_attr_leaf_entry_t *entry;
- int before, after, smallest, entsize;
- int tablesize, tmp, i;
- xfs_mount_t *mp;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_mount *mp = args->trans->t_mountp;
+ int before;
+ int after;
+ int smallest;
+ int entsize;
+ int tablesize;
+ int tmp;
+ int i;
trace_xfs_attr_leaf_remove(args);
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- hdr = &leaf->hdr;
- mp = args->trans->t_mountp;
- ASSERT((be16_to_cpu(hdr->count) > 0)
- && (be16_to_cpu(hdr->count) < (XFS_LBSIZE(mp)/8)));
- ASSERT((args->index >= 0)
- && (args->index < be16_to_cpu(hdr->count)));
- ASSERT(be16_to_cpu(hdr->firstused) >=
- ((be16_to_cpu(hdr->count) * sizeof(*entry)) + sizeof(*hdr)));
- entry = &leaf->entries[args->index];
- ASSERT(be16_to_cpu(entry->nameidx) >= be16_to_cpu(hdr->firstused));
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+
+ ASSERT(ichdr.count > 0 && ichdr.count < XFS_LBSIZE(mp) / 8);
+ ASSERT(args->index >= 0 && args->index < ichdr.count);
+ ASSERT(ichdr.firstused >= ichdr.count * sizeof(*entry) +
+ xfs_attr3_leaf_hdr_size(leaf));
+
+ entry = &xfs_attr3_leaf_entryp(leaf)[args->index];
+
+ ASSERT(be16_to_cpu(entry->nameidx) >= ichdr.firstused);
ASSERT(be16_to_cpu(entry->nameidx) < XFS_LBSIZE(mp));
/*
@@ -1775,30 +1999,28 @@ xfs_attr_leaf_remove(
* find smallest free region in case we need to replace it,
* adjust any map that borders the entry table,
*/
- tablesize = be16_to_cpu(hdr->count) * sizeof(xfs_attr_leaf_entry_t)
- + sizeof(xfs_attr_leaf_hdr_t);
- map = &hdr->freemap[0];
- tmp = be16_to_cpu(map->size);
+ tablesize = ichdr.count * sizeof(xfs_attr_leaf_entry_t)
+ + xfs_attr3_leaf_hdr_size(leaf);
+ tmp = ichdr.freemap[0].size;
before = after = -1;
smallest = XFS_ATTR_LEAF_MAPSIZE - 1;
entsize = xfs_attr_leaf_entsize(leaf, args->index);
- for (i = 0; i < XFS_ATTR_LEAF_MAPSIZE; map++, i++) {
- ASSERT(be16_to_cpu(map->base) < XFS_LBSIZE(mp));
- ASSERT(be16_to_cpu(map->size) < XFS_LBSIZE(mp));
- if (be16_to_cpu(map->base) == tablesize) {
- be16_add_cpu(&map->base,
- -((int)sizeof(xfs_attr_leaf_entry_t)));
- be16_add_cpu(&map->size, sizeof(xfs_attr_leaf_entry_t));
+ for (i = 0; i < XFS_ATTR_LEAF_MAPSIZE; i++) {
+ ASSERT(ichdr.freemap[i].base < XFS_LBSIZE(mp));
+ ASSERT(ichdr.freemap[i].size < XFS_LBSIZE(mp));
+ if (ichdr.freemap[i].base == tablesize) {
+ ichdr.freemap[i].base -= sizeof(xfs_attr_leaf_entry_t);
+ ichdr.freemap[i].size += sizeof(xfs_attr_leaf_entry_t);
}
- if ((be16_to_cpu(map->base) + be16_to_cpu(map->size))
- == be16_to_cpu(entry->nameidx)) {
+ if (ichdr.freemap[i].base + ichdr.freemap[i].size ==
+ be16_to_cpu(entry->nameidx)) {
before = i;
- } else if (be16_to_cpu(map->base)
- == (be16_to_cpu(entry->nameidx) + entsize)) {
+ } else if (ichdr.freemap[i].base ==
+ (be16_to_cpu(entry->nameidx) + entsize)) {
after = i;
- } else if (be16_to_cpu(map->size) < tmp) {
- tmp = be16_to_cpu(map->size);
+ } else if (ichdr.freemap[i].size < tmp) {
+ tmp = ichdr.freemap[i].size;
smallest = i;
}
}
@@ -1809,36 +2031,30 @@ xfs_attr_leaf_remove(
*/
if ((before >= 0) || (after >= 0)) {
if ((before >= 0) && (after >= 0)) {
- map = &hdr->freemap[before];
- be16_add_cpu(&map->size, entsize);
- be16_add_cpu(&map->size,
- be16_to_cpu(hdr->freemap[after].size));
- hdr->freemap[after].base = 0;
- hdr->freemap[after].size = 0;
+ ichdr.freemap[before].size += entsize;
+ ichdr.freemap[before].size += ichdr.freemap[after].size;
+ ichdr.freemap[after].base = 0;
+ ichdr.freemap[after].size = 0;
} else if (before >= 0) {
- map = &hdr->freemap[before];
- be16_add_cpu(&map->size, entsize);
+ ichdr.freemap[before].size += entsize;
} else {
- map = &hdr->freemap[after];
- /* both on-disk, don't endian flip twice */
- map->base = entry->nameidx;
- be16_add_cpu(&map->size, entsize);
+ ichdr.freemap[after].base = be16_to_cpu(entry->nameidx);
+ ichdr.freemap[after].size += entsize;
}
} else {
/*
* Replace smallest region (if it is smaller than free'd entry)
*/
- map = &hdr->freemap[smallest];
- if (be16_to_cpu(map->size) < entsize) {
- map->base = cpu_to_be16(be16_to_cpu(entry->nameidx));
- map->size = cpu_to_be16(entsize);
+ if (ichdr.freemap[smallest].size < entsize) {
+ ichdr.freemap[smallest].base = be16_to_cpu(entry->nameidx);
+ ichdr.freemap[smallest].size = entsize;
}
}
/*
* Did we remove the first entry?
*/
- if (be16_to_cpu(entry->nameidx) == be16_to_cpu(hdr->firstused))
+ if (be16_to_cpu(entry->nameidx) == ichdr.firstused)
smallest = 1;
else
smallest = 0;
@@ -1846,20 +2062,20 @@ xfs_attr_leaf_remove(
/*
* Compress the remaining entries and zero out the removed stuff.
*/
- memset(xfs_attr_leaf_name(leaf, args->index), 0, entsize);
- be16_add_cpu(&hdr->usedbytes, -entsize);
+ memset(xfs_attr3_leaf_name(leaf, args->index), 0, entsize);
+ ichdr.usedbytes -= entsize;
xfs_trans_log_buf(args->trans, bp,
- XFS_DA_LOGRANGE(leaf, xfs_attr_leaf_name(leaf, args->index),
+ XFS_DA_LOGRANGE(leaf, xfs_attr3_leaf_name(leaf, args->index),
entsize));
- tmp = (be16_to_cpu(hdr->count) - args->index)
- * sizeof(xfs_attr_leaf_entry_t);
- memmove((char *)entry, (char *)(entry+1), tmp);
- be16_add_cpu(&hdr->count, -1);
+ tmp = (ichdr.count - args->index) * sizeof(xfs_attr_leaf_entry_t);
+ memmove(entry, entry + 1, tmp);
+ ichdr.count--;
xfs_trans_log_buf(args->trans, bp,
- XFS_DA_LOGRANGE(leaf, entry, tmp + sizeof(*entry)));
- entry = &leaf->entries[be16_to_cpu(hdr->count)];
- memset((char *)entry, 0, sizeof(xfs_attr_leaf_entry_t));
+ XFS_DA_LOGRANGE(leaf, entry, tmp + sizeof(xfs_attr_leaf_entry_t)));
+
+ entry = &xfs_attr3_leaf_entryp(leaf)[ichdr.count];
+ memset(entry, 0, sizeof(xfs_attr_leaf_entry_t));
/*
* If we removed the first entry, re-find the first used byte
@@ -1869,130 +2085,130 @@ xfs_attr_leaf_remove(
*/
if (smallest) {
tmp = XFS_LBSIZE(mp);
- entry = &leaf->entries[0];
- for (i = be16_to_cpu(hdr->count)-1; i >= 0; entry++, i--) {
- ASSERT(be16_to_cpu(entry->nameidx) >=
- be16_to_cpu(hdr->firstused));
+ entry = xfs_attr3_leaf_entryp(leaf);
+ for (i = ichdr.count - 1; i >= 0; entry++, i--) {
+ ASSERT(be16_to_cpu(entry->nameidx) >= ichdr.firstused);
ASSERT(be16_to_cpu(entry->nameidx) < XFS_LBSIZE(mp));
if (be16_to_cpu(entry->nameidx) < tmp)
tmp = be16_to_cpu(entry->nameidx);
}
- hdr->firstused = cpu_to_be16(tmp);
- if (!hdr->firstused) {
- hdr->firstused = cpu_to_be16(
- tmp - XFS_ATTR_LEAF_NAME_ALIGN);
- }
+ ichdr.firstused = tmp;
+ if (!ichdr.firstused)
+ ichdr.firstused = tmp - XFS_ATTR_LEAF_NAME_ALIGN;
} else {
- hdr->holes = 1; /* mark as needing compaction */
+ ichdr.holes = 1; /* mark as needing compaction */
}
+ xfs_attr3_leaf_hdr_to_disk(leaf, &ichdr);
xfs_trans_log_buf(args->trans, bp,
- XFS_DA_LOGRANGE(leaf, hdr, sizeof(*hdr)));
+ XFS_DA_LOGRANGE(leaf, &leaf->hdr,
+ xfs_attr3_leaf_hdr_size(leaf)));
/*
* Check if leaf is less than 50% full, caller may want to
* "join" the leaf with a sibling if so.
*/
- tmp = sizeof(xfs_attr_leaf_hdr_t);
- tmp += be16_to_cpu(leaf->hdr.count) * sizeof(xfs_attr_leaf_entry_t);
- tmp += be16_to_cpu(leaf->hdr.usedbytes);
- return(tmp < mp->m_attr_magicpct); /* leaf is < 37% full */
+ tmp = ichdr.usedbytes + xfs_attr3_leaf_hdr_size(leaf) +
+ ichdr.count * sizeof(xfs_attr_leaf_entry_t);
+
+ return tmp < mp->m_attr_magicpct; /* leaf is < 37% full */
}
/*
* Move all the attribute list entries from drop_leaf into save_leaf.
*/
void
-xfs_attr_leaf_unbalance(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
- xfs_da_state_blk_t *save_blk)
+xfs_attr3_leaf_unbalance(
+ struct xfs_da_state *state,
+ struct xfs_da_state_blk *drop_blk,
+ struct xfs_da_state_blk *save_blk)
{
- xfs_attr_leafblock_t *drop_leaf, *save_leaf, *tmp_leaf;
- xfs_attr_leaf_hdr_t *drop_hdr, *save_hdr, *tmp_hdr;
- xfs_mount_t *mp;
- char *tmpbuffer;
+ struct xfs_attr_leafblock *drop_leaf = drop_blk->bp->b_addr;
+ struct xfs_attr_leafblock *save_leaf = save_blk->bp->b_addr;
+ struct xfs_attr3_icleaf_hdr drophdr;
+ struct xfs_attr3_icleaf_hdr savehdr;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_mount *mp = state->mp;
trace_xfs_attr_leaf_unbalance(state->args);
- /*
- * Set up environment.
- */
- mp = state->mp;
- ASSERT(drop_blk->magic == XFS_ATTR_LEAF_MAGIC);
- ASSERT(save_blk->magic == XFS_ATTR_LEAF_MAGIC);
drop_leaf = drop_blk->bp->b_addr;
save_leaf = save_blk->bp->b_addr;
- ASSERT(drop_leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- ASSERT(save_leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- drop_hdr = &drop_leaf->hdr;
- save_hdr = &save_leaf->hdr;
+ xfs_attr3_leaf_hdr_from_disk(&drophdr, drop_leaf);
+ xfs_attr3_leaf_hdr_from_disk(&savehdr, save_leaf);
+ entry = xfs_attr3_leaf_entryp(drop_leaf);
/*
* Save last hashval from dying block for later Btree fixup.
*/
- drop_blk->hashval = be32_to_cpu(
- drop_leaf->entries[be16_to_cpu(drop_leaf->hdr.count)-1].hashval);
+ drop_blk->hashval = be32_to_cpu(entry[drophdr.count - 1].hashval);
/*
* Check if we need a temp buffer, or can we do it in place.
* Note that we don't check "leaf" for holes because we will
* always be dropping it, toosmall() decided that for us already.
*/
- if (save_hdr->holes == 0) {
+ if (savehdr.holes == 0) {
/*
* dest leaf has no holes, so we add there. May need
* to make some room in the entry array.
*/
- if (xfs_attr_leaf_order(save_blk->bp, drop_blk->bp)) {
- xfs_attr_leaf_moveents(drop_leaf, 0, save_leaf, 0,
- be16_to_cpu(drop_hdr->count), mp);
+ if (xfs_attr3_leaf_order(save_blk->bp, &savehdr,
+ drop_blk->bp, &drophdr)) {
+ xfs_attr3_leaf_moveents(drop_leaf, &drophdr, 0,
+ save_leaf, &savehdr, 0,
+ drophdr.count, mp);
} else {
- xfs_attr_leaf_moveents(drop_leaf, 0, save_leaf,
- be16_to_cpu(save_hdr->count),
- be16_to_cpu(drop_hdr->count), mp);
+ xfs_attr3_leaf_moveents(drop_leaf, &drophdr, 0,
+ save_leaf, &savehdr,
+ savehdr.count, drophdr.count, mp);
}
} else {
/*
* Destination has holes, so we make a temporary copy
* of the leaf and add them both to that.
*/
- tmpbuffer = kmem_alloc(state->blocksize, KM_SLEEP);
- ASSERT(tmpbuffer != NULL);
- memset(tmpbuffer, 0, state->blocksize);
- tmp_leaf = (xfs_attr_leafblock_t *)tmpbuffer;
- tmp_hdr = &tmp_leaf->hdr;
- tmp_hdr->info = save_hdr->info; /* struct copy */
- tmp_hdr->count = 0;
- tmp_hdr->firstused = cpu_to_be16(state->blocksize);
- if (!tmp_hdr->firstused) {
- tmp_hdr->firstused = cpu_to_be16(
- state->blocksize - XFS_ATTR_LEAF_NAME_ALIGN);
- }
- tmp_hdr->usedbytes = 0;
- if (xfs_attr_leaf_order(save_blk->bp, drop_blk->bp)) {
- xfs_attr_leaf_moveents(drop_leaf, 0, tmp_leaf, 0,
- be16_to_cpu(drop_hdr->count), mp);
- xfs_attr_leaf_moveents(save_leaf, 0, tmp_leaf,
- be16_to_cpu(tmp_leaf->hdr.count),
- be16_to_cpu(save_hdr->count), mp);
+ struct xfs_attr_leafblock *tmp_leaf;
+ struct xfs_attr3_icleaf_hdr tmphdr;
+
+ tmp_leaf = kmem_alloc(state->blocksize, KM_SLEEP);
+ memset(tmp_leaf, 0, state->blocksize);
+ memset(&tmphdr, 0, sizeof(tmphdr));
+
+ tmphdr.magic = savehdr.magic;
+ tmphdr.forw = savehdr.forw;
+ tmphdr.back = savehdr.back;
+ tmphdr.firstused = state->blocksize;
+ if (xfs_attr3_leaf_order(save_blk->bp, &savehdr,
+ drop_blk->bp, &drophdr)) {
+ xfs_attr3_leaf_moveents(drop_leaf, &drophdr, 0,
+ tmp_leaf, &tmphdr, 0,
+ drophdr.count, mp);
+ xfs_attr3_leaf_moveents(save_leaf, &savehdr, 0,
+ tmp_leaf, &tmphdr, tmphdr.count,
+ savehdr.count, mp);
} else {
- xfs_attr_leaf_moveents(save_leaf, 0, tmp_leaf, 0,
- be16_to_cpu(save_hdr->count), mp);
- xfs_attr_leaf_moveents(drop_leaf, 0, tmp_leaf,
- be16_to_cpu(tmp_leaf->hdr.count),
- be16_to_cpu(drop_hdr->count), mp);
+ xfs_attr3_leaf_moveents(save_leaf, &savehdr, 0,
+ tmp_leaf, &tmphdr, 0,
+ savehdr.count, mp);
+ xfs_attr3_leaf_moveents(drop_leaf, &drophdr, 0,
+ tmp_leaf, &tmphdr, tmphdr.count,
+ drophdr.count, mp);
}
- memcpy((char *)save_leaf, (char *)tmp_leaf, state->blocksize);
- kmem_free(tmpbuffer);
+ memcpy(save_leaf, tmp_leaf, state->blocksize);
+ savehdr = tmphdr; /* struct copy */
+ kmem_free(tmp_leaf);
}
+ xfs_attr3_leaf_hdr_to_disk(save_leaf, &savehdr);
xfs_trans_log_buf(state->args->trans, save_blk->bp, 0,
state->blocksize - 1);
/*
* Copy out last hashval in each block for B-tree code.
*/
- save_blk->hashval = be32_to_cpu(
- save_leaf->entries[be16_to_cpu(save_leaf->hdr.count)-1].hashval);
+ entry = xfs_attr3_leaf_entryp(save_leaf);
+ save_blk->hashval = be32_to_cpu(entry[savehdr.count - 1].hashval);
}
/*========================================================================
@@ -2013,31 +2229,33 @@ xfs_attr_leaf_unbalance(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk,
* Don't change the args->value unless we find the attribute.
*/
int
-xfs_attr_leaf_lookup_int(
- struct xfs_buf *bp,
- xfs_da_args_t *args)
+xfs_attr3_leaf_lookup_int(
+ struct xfs_buf *bp,
+ struct xfs_da_args *args)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_entry_t *entry;
- xfs_attr_leaf_name_local_t *name_loc;
- xfs_attr_leaf_name_remote_t *name_rmt;
- int probe, span;
- xfs_dahash_t hashval;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_attr_leaf_entry *entries;
+ struct xfs_attr_leaf_name_local *name_loc;
+ struct xfs_attr_leaf_name_remote *name_rmt;
+ xfs_dahash_t hashval;
+ int probe;
+ int span;
trace_xfs_attr_leaf_lookup(args);
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- ASSERT(be16_to_cpu(leaf->hdr.count)
- < (XFS_LBSIZE(args->dp->i_mount)/8));
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ entries = xfs_attr3_leaf_entryp(leaf);
+ ASSERT(ichdr.count < XFS_LBSIZE(args->dp->i_mount) / 8);
/*
* Binary search. (note: small blocks will skip this loop)
*/
hashval = args->hashval;
- probe = span = be16_to_cpu(leaf->hdr.count) / 2;
- for (entry = &leaf->entries[probe]; span > 4;
- entry = &leaf->entries[probe]) {
+ probe = span = ichdr.count / 2;
+ for (entry = &entries[probe]; span > 4; entry = &entries[probe]) {
span /= 2;
if (be32_to_cpu(entry->hashval) < hashval)
probe += span;
@@ -2046,35 +2264,31 @@ xfs_attr_leaf_lookup_int(
else
break;
}
- ASSERT((probe >= 0) &&
- (!leaf->hdr.count
- || (probe < be16_to_cpu(leaf->hdr.count))));
- ASSERT((span <= 4) || (be32_to_cpu(entry->hashval) == hashval));
+ ASSERT(probe >= 0 && (!ichdr.count || probe < ichdr.count));
+ ASSERT(span <= 4 || be32_to_cpu(entry->hashval) == hashval);
/*
* Since we may have duplicate hashval's, find the first matching
* hashval in the leaf.
*/
- while ((probe > 0) && (be32_to_cpu(entry->hashval) >= hashval)) {
+ while (probe > 0 && be32_to_cpu(entry->hashval) >= hashval) {
entry--;
probe--;
}
- while ((probe < be16_to_cpu(leaf->hdr.count)) &&
- (be32_to_cpu(entry->hashval) < hashval)) {
+ while (probe < ichdr.count &&
+ be32_to_cpu(entry->hashval) < hashval) {
entry++;
probe++;
}
- if ((probe == be16_to_cpu(leaf->hdr.count)) ||
- (be32_to_cpu(entry->hashval) != hashval)) {
+ if (probe == ichdr.count || be32_to_cpu(entry->hashval) != hashval) {
args->index = probe;
- return(XFS_ERROR(ENOATTR));
+ return XFS_ERROR(ENOATTR);
}
/*
* Duplicate keys may be present, so search all of them for a match.
*/
- for ( ; (probe < be16_to_cpu(leaf->hdr.count)) &&
- (be32_to_cpu(entry->hashval) == hashval);
+ for (; probe < ichdr.count && (be32_to_cpu(entry->hashval) == hashval);
entry++, probe++) {
/*
* GROT: Add code to remove incomplete entries.
@@ -2088,21 +2302,22 @@ xfs_attr_leaf_lookup_int(
continue;
}
if (entry->flags & XFS_ATTR_LOCAL) {
- name_loc = xfs_attr_leaf_name_local(leaf, probe);
+ name_loc = xfs_attr3_leaf_name_local(leaf, probe);
if (name_loc->namelen != args->namelen)
continue;
- if (memcmp(args->name, (char *)name_loc->nameval, args->namelen) != 0)
+ if (memcmp(args->name, name_loc->nameval,
+ args->namelen) != 0)
continue;
if (!xfs_attr_namesp_match(args->flags, entry->flags))
continue;
args->index = probe;
- return(XFS_ERROR(EEXIST));
+ return XFS_ERROR(EEXIST);
} else {
- name_rmt = xfs_attr_leaf_name_remote(leaf, probe);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, probe);
if (name_rmt->namelen != args->namelen)
continue;
- if (memcmp(args->name, (char *)name_rmt->name,
- args->namelen) != 0)
+ if (memcmp(args->name, name_rmt->name,
+ args->namelen) != 0)
continue;
if (!xfs_attr_namesp_match(args->flags, entry->flags))
continue;
@@ -2110,11 +2325,11 @@ xfs_attr_leaf_lookup_int(
args->rmtblkno = be32_to_cpu(name_rmt->valueblk);
args->rmtblkcnt = XFS_B_TO_FSB(args->dp->i_mount,
be32_to_cpu(name_rmt->valuelen));
- return(XFS_ERROR(EEXIST));
+ return XFS_ERROR(EEXIST);
}
}
args->index = probe;
- return(XFS_ERROR(ENOATTR));
+ return XFS_ERROR(ENOATTR);
}
/*
@@ -2122,40 +2337,40 @@ xfs_attr_leaf_lookup_int(
* list structure.
*/
int
-xfs_attr_leaf_getvalue(
- struct xfs_buf *bp,
- xfs_da_args_t *args)
+xfs_attr3_leaf_getvalue(
+ struct xfs_buf *bp,
+ struct xfs_da_args *args)
{
- int valuelen;
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_entry_t *entry;
- xfs_attr_leaf_name_local_t *name_loc;
- xfs_attr_leaf_name_remote_t *name_rmt;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_attr_leaf_name_local *name_loc;
+ struct xfs_attr_leaf_name_remote *name_rmt;
+ int valuelen;
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- ASSERT(be16_to_cpu(leaf->hdr.count)
- < (XFS_LBSIZE(args->dp->i_mount)/8));
- ASSERT(args->index < be16_to_cpu(leaf->hdr.count));
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ ASSERT(ichdr.count < XFS_LBSIZE(args->dp->i_mount) / 8);
+ ASSERT(args->index < ichdr.count);
- entry = &leaf->entries[args->index];
+ entry = &xfs_attr3_leaf_entryp(leaf)[args->index];
if (entry->flags & XFS_ATTR_LOCAL) {
- name_loc = xfs_attr_leaf_name_local(leaf, args->index);
+ name_loc = xfs_attr3_leaf_name_local(leaf, args->index);
ASSERT(name_loc->namelen == args->namelen);
ASSERT(memcmp(args->name, name_loc->nameval, args->namelen) == 0);
valuelen = be16_to_cpu(name_loc->valuelen);
if (args->flags & ATTR_KERNOVAL) {
args->valuelen = valuelen;
- return(0);
+ return 0;
}
if (args->valuelen < valuelen) {
args->valuelen = valuelen;
- return(XFS_ERROR(ERANGE));
+ return XFS_ERROR(ERANGE);
}
args->valuelen = valuelen;
memcpy(args->value, &name_loc->nameval[args->namelen], valuelen);
} else {
- name_rmt = xfs_attr_leaf_name_remote(leaf, args->index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, args->index);
ASSERT(name_rmt->namelen == args->namelen);
ASSERT(memcmp(args->name, name_rmt->name, args->namelen) == 0);
valuelen = be32_to_cpu(name_rmt->valuelen);
@@ -2163,15 +2378,15 @@ xfs_attr_leaf_getvalue(
args->rmtblkcnt = XFS_B_TO_FSB(args->dp->i_mount, valuelen);
if (args->flags & ATTR_KERNOVAL) {
args->valuelen = valuelen;
- return(0);
+ return 0;
}
if (args->valuelen < valuelen) {
args->valuelen = valuelen;
- return(XFS_ERROR(ERANGE));
+ return XFS_ERROR(ERANGE);
}
args->valuelen = valuelen;
}
- return(0);
+ return 0;
}
/*========================================================================
@@ -2184,13 +2399,21 @@ xfs_attr_leaf_getvalue(
*/
/*ARGSUSED*/
STATIC void
-xfs_attr_leaf_moveents(xfs_attr_leafblock_t *leaf_s, int start_s,
- xfs_attr_leafblock_t *leaf_d, int start_d,
- int count, xfs_mount_t *mp)
+xfs_attr3_leaf_moveents(
+ struct xfs_attr_leafblock *leaf_s,
+ struct xfs_attr3_icleaf_hdr *ichdr_s,
+ int start_s,
+ struct xfs_attr_leafblock *leaf_d,
+ struct xfs_attr3_icleaf_hdr *ichdr_d,
+ int start_d,
+ int count,
+ struct xfs_mount *mp)
{
- xfs_attr_leaf_hdr_t *hdr_s, *hdr_d;
- xfs_attr_leaf_entry_t *entry_s, *entry_d;
- int desti, tmp, i;
+ struct xfs_attr_leaf_entry *entry_s;
+ struct xfs_attr_leaf_entry *entry_d;
+ int desti;
+ int tmp;
+ int i;
/*
* Check for nothing to do.
@@ -2201,45 +2424,41 @@ xfs_attr_leaf_moveents(xfs_attr_leafblock_t *leaf_s, int start_s,
/*
* Set up environment.
*/
- ASSERT(leaf_s->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- ASSERT(leaf_d->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- hdr_s = &leaf_s->hdr;
- hdr_d = &leaf_d->hdr;
- ASSERT((be16_to_cpu(hdr_s->count) > 0) &&
- (be16_to_cpu(hdr_s->count) < (XFS_LBSIZE(mp)/8)));
- ASSERT(be16_to_cpu(hdr_s->firstused) >=
- ((be16_to_cpu(hdr_s->count)
- * sizeof(*entry_s))+sizeof(*hdr_s)));
- ASSERT(be16_to_cpu(hdr_d->count) < (XFS_LBSIZE(mp)/8));
- ASSERT(be16_to_cpu(hdr_d->firstused) >=
- ((be16_to_cpu(hdr_d->count)
- * sizeof(*entry_d))+sizeof(*hdr_d)));
-
- ASSERT(start_s < be16_to_cpu(hdr_s->count));
- ASSERT(start_d <= be16_to_cpu(hdr_d->count));
- ASSERT(count <= be16_to_cpu(hdr_s->count));
+ ASSERT(ichdr_s->magic == XFS_ATTR_LEAF_MAGIC ||
+ ichdr_s->magic == XFS_ATTR3_LEAF_MAGIC);
+ ASSERT(ichdr_s->magic == ichdr_d->magic);
+ ASSERT(ichdr_s->count > 0 && ichdr_s->count < XFS_LBSIZE(mp) / 8);
+ ASSERT(ichdr_s->firstused >= (ichdr_s->count * sizeof(*entry_s))
+ + xfs_attr3_leaf_hdr_size(leaf_s));
+ ASSERT(ichdr_d->count < XFS_LBSIZE(mp) / 8);
+ ASSERT(ichdr_d->firstused >= (ichdr_d->count * sizeof(*entry_d))
+ + xfs_attr3_leaf_hdr_size(leaf_d));
+
+ ASSERT(start_s < ichdr_s->count);
+ ASSERT(start_d <= ichdr_d->count);
+ ASSERT(count <= ichdr_s->count);
+
/*
* Move the entries in the destination leaf up to make a hole?
*/
- if (start_d < be16_to_cpu(hdr_d->count)) {
- tmp = be16_to_cpu(hdr_d->count) - start_d;
+ if (start_d < ichdr_d->count) {
+ tmp = ichdr_d->count - start_d;
tmp *= sizeof(xfs_attr_leaf_entry_t);
- entry_s = &leaf_d->entries[start_d];
- entry_d = &leaf_d->entries[start_d + count];
- memmove((char *)entry_d, (char *)entry_s, tmp);
+ entry_s = &xfs_attr3_leaf_entryp(leaf_d)[start_d];
+ entry_d = &xfs_attr3_leaf_entryp(leaf_d)[start_d + count];
+ memmove(entry_d, entry_s, tmp);
}
/*
* Copy all entry's in the same (sorted) order,
* but allocate attribute info packed and in sequence.
*/
- entry_s = &leaf_s->entries[start_s];
- entry_d = &leaf_d->entries[start_d];
+ entry_s = &xfs_attr3_leaf_entryp(leaf_s)[start_s];
+ entry_d = &xfs_attr3_leaf_entryp(leaf_d)[start_d];
desti = start_d;
for (i = 0; i < count; entry_s++, entry_d++, desti++, i++) {
- ASSERT(be16_to_cpu(entry_s->nameidx)
- >= be16_to_cpu(hdr_s->firstused));
+ ASSERT(be16_to_cpu(entry_s->nameidx) >= ichdr_s->firstused);
tmp = xfs_attr_leaf_entsize(leaf_s, start_s + i);
#ifdef GROT
/*
@@ -2248,36 +2467,34 @@ xfs_attr_leaf_moveents(xfs_attr_leafblock_t *leaf_s, int start_s,
* off for 6.2, should be revisited later.
*/
if (entry_s->flags & XFS_ATTR_INCOMPLETE) { /* skip partials? */
- memset(xfs_attr_leaf_name(leaf_s, start_s + i), 0, tmp);
- be16_add_cpu(&hdr_s->usedbytes, -tmp);
- be16_add_cpu(&hdr_s->count, -1);
+ memset(xfs_attr3_leaf_name(leaf_s, start_s + i), 0, tmp);
+ ichdr_s->usedbytes -= tmp;
+ ichdr_s->count -= 1;
entry_d--; /* to compensate for ++ in loop hdr */
desti--;
if ((start_s + i) < offset)
result++; /* insertion index adjustment */
} else {
#endif /* GROT */
- be16_add_cpu(&hdr_d->firstused, -tmp);
+ ichdr_d->firstused -= tmp;
/* both on-disk, don't endian flip twice */
entry_d->hashval = entry_s->hashval;
- /* both on-disk, don't endian flip twice */
- entry_d->nameidx = hdr_d->firstused;
+ entry_d->nameidx = cpu_to_be16(ichdr_d->firstused);
entry_d->flags = entry_s->flags;
ASSERT(be16_to_cpu(entry_d->nameidx) + tmp
<= XFS_LBSIZE(mp));
- memmove(xfs_attr_leaf_name(leaf_d, desti),
- xfs_attr_leaf_name(leaf_s, start_s + i), tmp);
+ memmove(xfs_attr3_leaf_name(leaf_d, desti),
+ xfs_attr3_leaf_name(leaf_s, start_s + i), tmp);
ASSERT(be16_to_cpu(entry_s->nameidx) + tmp
<= XFS_LBSIZE(mp));
- memset(xfs_attr_leaf_name(leaf_s, start_s + i), 0, tmp);
- be16_add_cpu(&hdr_s->usedbytes, -tmp);
- be16_add_cpu(&hdr_d->usedbytes, tmp);
- be16_add_cpu(&hdr_s->count, -1);
- be16_add_cpu(&hdr_d->count, 1);
- tmp = be16_to_cpu(hdr_d->count)
- * sizeof(xfs_attr_leaf_entry_t)
- + sizeof(xfs_attr_leaf_hdr_t);
- ASSERT(be16_to_cpu(hdr_d->firstused) >= tmp);
+ memset(xfs_attr3_leaf_name(leaf_s, start_s + i), 0, tmp);
+ ichdr_s->usedbytes -= tmp;
+ ichdr_d->usedbytes += tmp;
+ ichdr_s->count -= 1;
+ ichdr_d->count += 1;
+ tmp = ichdr_d->count * sizeof(xfs_attr_leaf_entry_t)
+ + xfs_attr3_leaf_hdr_size(leaf_d);
+ ASSERT(ichdr_d->firstused >= tmp);
#ifdef GROT
}
#endif /* GROT */
@@ -2286,71 +2503,40 @@ xfs_attr_leaf_moveents(xfs_attr_leafblock_t *leaf_s, int start_s,
/*
* Zero out the entries we just copied.
*/
- if (start_s == be16_to_cpu(hdr_s->count)) {
+ if (start_s == ichdr_s->count) {
tmp = count * sizeof(xfs_attr_leaf_entry_t);
- entry_s = &leaf_s->entries[start_s];
+ entry_s = &xfs_attr3_leaf_entryp(leaf_s)[start_s];
ASSERT(((char *)entry_s + tmp) <=
((char *)leaf_s + XFS_LBSIZE(mp)));
- memset((char *)entry_s, 0, tmp);
+ memset(entry_s, 0, tmp);
} else {
/*
* Move the remaining entries down to fill the hole,
* then zero the entries at the top.
*/
- tmp = be16_to_cpu(hdr_s->count) - count;
- tmp *= sizeof(xfs_attr_leaf_entry_t);
- entry_s = &leaf_s->entries[start_s + count];
- entry_d = &leaf_s->entries[start_s];
- memmove((char *)entry_d, (char *)entry_s, tmp);
+ tmp = (ichdr_s->count - count) - sizeof(xfs_attr_leaf_entry_t);
+ entry_s = &xfs_attr3_leaf_entryp(leaf_s)[start_s + count];
+ entry_d = &xfs_attr3_leaf_entryp(leaf_s)[start_s];
+ memmove(entry_d, entry_s, tmp);
tmp = count * sizeof(xfs_attr_leaf_entry_t);
- entry_s = &leaf_s->entries[be16_to_cpu(hdr_s->count)];
+ entry_s = &xfs_attr3_leaf_entryp(leaf_s)[ichdr_s->count];
ASSERT(((char *)entry_s + tmp) <=
((char *)leaf_s + XFS_LBSIZE(mp)));
- memset((char *)entry_s, 0, tmp);
+ memset(entry_s, 0, tmp);
}
/*
* Fill in the freemap information
*/
- hdr_d->freemap[0].base = cpu_to_be16(sizeof(xfs_attr_leaf_hdr_t));
- be16_add_cpu(&hdr_d->freemap[0].base, be16_to_cpu(hdr_d->count) *
- sizeof(xfs_attr_leaf_entry_t));
- hdr_d->freemap[0].size = cpu_to_be16(be16_to_cpu(hdr_d->firstused)
- - be16_to_cpu(hdr_d->freemap[0].base));
- hdr_d->freemap[1].base = 0;
- hdr_d->freemap[2].base = 0;
- hdr_d->freemap[1].size = 0;
- hdr_d->freemap[2].size = 0;
- hdr_s->holes = 1; /* leaf may not be compact */
-}
-
-/*
- * Compare two leaf blocks "order".
- * Return 0 unless leaf2 should go before leaf1.
- */
-int
-xfs_attr_leaf_order(
- struct xfs_buf *leaf1_bp,
- struct xfs_buf *leaf2_bp)
-{
- xfs_attr_leafblock_t *leaf1, *leaf2;
-
- leaf1 = leaf1_bp->b_addr;
- leaf2 = leaf2_bp->b_addr;
- ASSERT((leaf1->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)) &&
- (leaf2->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)));
- if ((be16_to_cpu(leaf1->hdr.count) > 0) &&
- (be16_to_cpu(leaf2->hdr.count) > 0) &&
- ((be32_to_cpu(leaf2->entries[0].hashval) <
- be32_to_cpu(leaf1->entries[0].hashval)) ||
- (be32_to_cpu(leaf2->entries[
- be16_to_cpu(leaf2->hdr.count)-1].hashval) <
- be32_to_cpu(leaf1->entries[
- be16_to_cpu(leaf1->hdr.count)-1].hashval)))) {
- return(1);
- }
- return(0);
+ ichdr_d->freemap[0].base = xfs_attr3_leaf_hdr_size(leaf_d);
+ ichdr_d->freemap[0].base += ichdr_d->count * sizeof(xfs_attr_leaf_entry_t);
+ ichdr_d->freemap[0].size = ichdr_d->firstused - ichdr_d->freemap[0].base;
+ ichdr_d->freemap[1].base = 0;
+ ichdr_d->freemap[2].base = 0;
+ ichdr_d->freemap[1].size = 0;
+ ichdr_d->freemap[2].size = 0;
+ ichdr_s->holes = 1; /* leaf may not be compact */
}
/*
@@ -2361,15 +2547,16 @@ xfs_attr_leaf_lasthash(
struct xfs_buf *bp,
int *count)
{
- xfs_attr_leafblock_t *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_attr_leaf_entry *entries;
- leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, bp->b_addr);
+ entries = xfs_attr3_leaf_entryp(bp->b_addr);
if (count)
- *count = be16_to_cpu(leaf->hdr.count);
- if (!leaf->hdr.count)
- return(0);
- return be32_to_cpu(leaf->entries[be16_to_cpu(leaf->hdr.count)-1].hashval);
+ *count = ichdr.count;
+ if (!ichdr.count)
+ return 0;
+ return be32_to_cpu(entries[ichdr.count - 1].hashval);
}
/*
@@ -2379,20 +2566,21 @@ xfs_attr_leaf_lasthash(
STATIC int
xfs_attr_leaf_entsize(xfs_attr_leafblock_t *leaf, int index)
{
+ struct xfs_attr_leaf_entry *entries;
xfs_attr_leaf_name_local_t *name_loc;
xfs_attr_leaf_name_remote_t *name_rmt;
int size;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
- if (leaf->entries[index].flags & XFS_ATTR_LOCAL) {
- name_loc = xfs_attr_leaf_name_local(leaf, index);
+ entries = xfs_attr3_leaf_entryp(leaf);
+ if (entries[index].flags & XFS_ATTR_LOCAL) {
+ name_loc = xfs_attr3_leaf_name_local(leaf, index);
size = xfs_attr_leaf_entsize_local(name_loc->namelen,
be16_to_cpu(name_loc->valuelen));
} else {
- name_rmt = xfs_attr_leaf_name_remote(leaf, index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, index);
size = xfs_attr_leaf_entsize_remote(name_rmt->namelen);
}
- return(size);
+ return size;
}
/*
@@ -2417,35 +2605,40 @@ xfs_attr_leaf_newentsize(int namelen, int valuelen, int blocksize, int *local)
*local = 0;
}
}
- return(size);
+ return size;
}
/*
* Copy out attribute list entries for attr_list(), for leaf attribute lists.
*/
int
-xfs_attr_leaf_list_int(
- struct xfs_buf *bp,
- xfs_attr_list_context_t *context)
+xfs_attr3_leaf_list_int(
+ struct xfs_buf *bp,
+ struct xfs_attr_list_context *context)
{
- attrlist_cursor_kern_t *cursor;
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_entry_t *entry;
- int retval, i;
+ struct attrlist_cursor_kern *cursor;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_attr_leaf_entry *entries;
+ struct xfs_attr_leaf_entry *entry;
+ int retval;
+ int i;
+
+ trace_xfs_attr_list_leaf(context);
- ASSERT(bp != NULL);
leaf = bp->b_addr;
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ entries = xfs_attr3_leaf_entryp(leaf);
+
cursor = context->cursor;
cursor->initted = 1;
- trace_xfs_attr_list_leaf(context);
-
/*
* Re-find our place in the leaf block if this is a new syscall.
*/
if (context->resynch) {
- entry = &leaf->entries[0];
- for (i = 0; i < be16_to_cpu(leaf->hdr.count); entry++, i++) {
+ entry = &entries[0];
+ for (i = 0; i < ichdr.count; entry++, i++) {
if (be32_to_cpu(entry->hashval) == cursor->hashval) {
if (cursor->offset == context->dupcnt) {
context->dupcnt = 0;
@@ -2458,12 +2651,12 @@ xfs_attr_leaf_list_int(
break;
}
}
- if (i == be16_to_cpu(leaf->hdr.count)) {
+ if (i == ichdr.count) {
trace_xfs_attr_list_notfound(context);
- return(0);
+ return 0;
}
} else {
- entry = &leaf->entries[0];
+ entry = &entries[0];
i = 0;
}
context->resynch = 0;
@@ -2472,7 +2665,7 @@ xfs_attr_leaf_list_int(
* We have found our place, start copying out the new attributes.
*/
retval = 0;
- for ( ; (i < be16_to_cpu(leaf->hdr.count)); entry++, i++) {
+ for (; i < ichdr.count; entry++, i++) {
if (be32_to_cpu(entry->hashval) != cursor->hashval) {
cursor->hashval = be32_to_cpu(entry->hashval);
cursor->offset = 0;
@@ -2483,7 +2676,7 @@ xfs_attr_leaf_list_int(
if (entry->flags & XFS_ATTR_LOCAL) {
xfs_attr_leaf_name_local_t *name_loc =
- xfs_attr_leaf_name_local(leaf, i);
+ xfs_attr3_leaf_name_local(leaf, i);
retval = context->put_listent(context,
entry->flags,
@@ -2495,7 +2688,7 @@ xfs_attr_leaf_list_int(
return retval;
} else {
xfs_attr_leaf_name_remote_t *name_rmt =
- xfs_attr_leaf_name_remote(leaf, i);
+ xfs_attr3_leaf_name_remote(leaf, i);
int valuelen = be32_to_cpu(name_rmt->valuelen);
@@ -2535,7 +2728,7 @@ xfs_attr_leaf_list_int(
cursor->offset++;
}
trace_xfs_attr_list_leaf_end(context);
- return(retval);
+ return retval;
}
@@ -2547,14 +2740,16 @@ xfs_attr_leaf_list_int(
* Clear the INCOMPLETE flag on an entry in a leaf block.
*/
int
-xfs_attr_leaf_clearflag(xfs_da_args_t *args)
+xfs_attr3_leaf_clearflag(
+ struct xfs_da_args *args)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_entry_t *entry;
- xfs_attr_leaf_name_remote_t *name_rmt;
- struct xfs_buf *bp;
- int error;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_attr_leaf_name_remote *name_rmt;
+ struct xfs_buf *bp;
+ int error;
#ifdef DEBUG
+ struct xfs_attr3_icleaf_hdr ichdr;
xfs_attr_leaf_name_local_t *name_loc;
int namelen;
char *name;
@@ -2564,23 +2759,25 @@ xfs_attr_leaf_clearflag(xfs_da_args_t *args)
/*
* Set up the operation.
*/
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
if (error)
return(error);
leaf = bp->b_addr;
- ASSERT(args->index < be16_to_cpu(leaf->hdr.count));
- ASSERT(args->index >= 0);
- entry = &leaf->entries[ args->index ];
+ entry = &xfs_attr3_leaf_entryp(leaf)[args->index];
ASSERT(entry->flags & XFS_ATTR_INCOMPLETE);
#ifdef DEBUG
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ ASSERT(args->index < ichdr.count);
+ ASSERT(args->index >= 0);
+
if (entry->flags & XFS_ATTR_LOCAL) {
- name_loc = xfs_attr_leaf_name_local(leaf, args->index);
+ name_loc = xfs_attr3_leaf_name_local(leaf, args->index);
namelen = name_loc->namelen;
name = (char *)name_loc->nameval;
} else {
- name_rmt = xfs_attr_leaf_name_remote(leaf, args->index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, args->index);
namelen = name_rmt->namelen;
name = (char *)name_rmt->name;
}
@@ -2595,7 +2792,7 @@ xfs_attr_leaf_clearflag(xfs_da_args_t *args)
if (args->rmtblkno) {
ASSERT((entry->flags & XFS_ATTR_LOCAL) == 0);
- name_rmt = xfs_attr_leaf_name_remote(leaf, args->index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, args->index);
name_rmt->valueblk = cpu_to_be32(args->rmtblkno);
name_rmt->valuelen = cpu_to_be32(args->valuelen);
xfs_trans_log_buf(args->trans, bp,
@@ -2612,34 +2809,41 @@ xfs_attr_leaf_clearflag(xfs_da_args_t *args)
* Set the INCOMPLETE flag on an entry in a leaf block.
*/
int
-xfs_attr_leaf_setflag(xfs_da_args_t *args)
+xfs_attr3_leaf_setflag(
+ struct xfs_da_args *args)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_entry_t *entry;
- xfs_attr_leaf_name_remote_t *name_rmt;
- struct xfs_buf *bp;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_attr_leaf_name_remote *name_rmt;
+ struct xfs_buf *bp;
int error;
+#ifdef DEBUG
+ struct xfs_attr3_icleaf_hdr ichdr;
+#endif
trace_xfs_attr_leaf_setflag(args);
/*
* Set up the operation.
*/
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno, -1, &bp);
if (error)
return(error);
leaf = bp->b_addr;
- ASSERT(args->index < be16_to_cpu(leaf->hdr.count));
+#ifdef DEBUG
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
+ ASSERT(args->index < ichdr.count);
ASSERT(args->index >= 0);
- entry = &leaf->entries[ args->index ];
+#endif
+ entry = &xfs_attr3_leaf_entryp(leaf)[args->index];
ASSERT((entry->flags & XFS_ATTR_INCOMPLETE) == 0);
entry->flags |= XFS_ATTR_INCOMPLETE;
xfs_trans_log_buf(args->trans, bp,
XFS_DA_LOGRANGE(leaf, entry, sizeof(*entry)));
if ((entry->flags & XFS_ATTR_LOCAL) == 0) {
- name_rmt = xfs_attr_leaf_name_remote(leaf, args->index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, args->index);
name_rmt->valueblk = 0;
name_rmt->valuelen = 0;
xfs_trans_log_buf(args->trans, bp,
@@ -2660,14 +2864,20 @@ xfs_attr_leaf_setflag(xfs_da_args_t *args)
* Note that they could be in different blocks, or in the same block.
*/
int
-xfs_attr_leaf_flipflags(xfs_da_args_t *args)
+xfs_attr3_leaf_flipflags(
+ struct xfs_da_args *args)
{
- xfs_attr_leafblock_t *leaf1, *leaf2;
- xfs_attr_leaf_entry_t *entry1, *entry2;
- xfs_attr_leaf_name_remote_t *name_rmt;
- struct xfs_buf *bp1, *bp2;
+ struct xfs_attr_leafblock *leaf1;
+ struct xfs_attr_leafblock *leaf2;
+ struct xfs_attr_leaf_entry *entry1;
+ struct xfs_attr_leaf_entry *entry2;
+ struct xfs_attr_leaf_name_remote *name_rmt;
+ struct xfs_buf *bp1;
+ struct xfs_buf *bp2;
int error;
#ifdef DEBUG
+ struct xfs_attr3_icleaf_hdr ichdr1;
+ struct xfs_attr3_icleaf_hdr ichdr2;
xfs_attr_leaf_name_local_t *name_loc;
int namelen1, namelen2;
char *name1, *name2;
@@ -2678,7 +2888,7 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
/*
* Read the block containing the "old" attr
*/
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp1);
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno, -1, &bp1);
if (error)
return error;
@@ -2686,7 +2896,7 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
* Read the block containing the "new" attr, if it is different
*/
if (args->blkno2 != args->blkno) {
- error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno2,
+ error = xfs_attr3_leaf_read(args->trans, args->dp, args->blkno2,
-1, &bp2);
if (error)
return error;
@@ -2695,31 +2905,35 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
}
leaf1 = bp1->b_addr;
- ASSERT(args->index < be16_to_cpu(leaf1->hdr.count));
- ASSERT(args->index >= 0);
- entry1 = &leaf1->entries[ args->index ];
+ entry1 = &xfs_attr3_leaf_entryp(leaf1)[args->index];
leaf2 = bp2->b_addr;
- ASSERT(args->index2 < be16_to_cpu(leaf2->hdr.count));
- ASSERT(args->index2 >= 0);
- entry2 = &leaf2->entries[ args->index2 ];
+ entry2 = &xfs_attr3_leaf_entryp(leaf2)[args->index2];
#ifdef DEBUG
+ xfs_attr3_leaf_hdr_from_disk(&ichdr1, leaf1);
+ ASSERT(args->index < ichdr1.count);
+ ASSERT(args->index >= 0);
+
+ xfs_attr3_leaf_hdr_from_disk(&ichdr2, leaf2);
+ ASSERT(args->index2 < ichdr2.count);
+ ASSERT(args->index2 >= 0);
+
if (entry1->flags & XFS_ATTR_LOCAL) {
- name_loc = xfs_attr_leaf_name_local(leaf1, args->index);
+ name_loc = xfs_attr3_leaf_name_local(leaf1, args->index);
namelen1 = name_loc->namelen;
name1 = (char *)name_loc->nameval;
} else {
- name_rmt = xfs_attr_leaf_name_remote(leaf1, args->index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf1, args->index);
namelen1 = name_rmt->namelen;
name1 = (char *)name_rmt->name;
}
if (entry2->flags & XFS_ATTR_LOCAL) {
- name_loc = xfs_attr_leaf_name_local(leaf2, args->index2);
+ name_loc = xfs_attr3_leaf_name_local(leaf2, args->index2);
namelen2 = name_loc->namelen;
name2 = (char *)name_loc->nameval;
} else {
- name_rmt = xfs_attr_leaf_name_remote(leaf2, args->index2);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf2, args->index2);
namelen2 = name_rmt->namelen;
name2 = (char *)name_rmt->name;
}
@@ -2736,7 +2950,7 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
XFS_DA_LOGRANGE(leaf1, entry1, sizeof(*entry1)));
if (args->rmtblkno) {
ASSERT((entry1->flags & XFS_ATTR_LOCAL) == 0);
- name_rmt = xfs_attr_leaf_name_remote(leaf1, args->index);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf1, args->index);
name_rmt->valueblk = cpu_to_be32(args->rmtblkno);
name_rmt->valuelen = cpu_to_be32(args->valuelen);
xfs_trans_log_buf(args->trans, bp1,
@@ -2747,7 +2961,7 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
xfs_trans_log_buf(args->trans, bp2,
XFS_DA_LOGRANGE(leaf2, entry2, sizeof(*entry2)));
if ((entry2->flags & XFS_ATTR_LOCAL) == 0) {
- name_rmt = xfs_attr_leaf_name_remote(leaf2, args->index2);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf2, args->index2);
name_rmt->valueblk = 0;
name_rmt->valuelen = 0;
xfs_trans_log_buf(args->trans, bp2,
@@ -2759,7 +2973,7 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
*/
error = xfs_trans_roll(&args->trans, args->dp);
- return(error);
+ return error;
}
/*========================================================================
@@ -2771,12 +2985,14 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args)
* We're doing a depth-first traversal in order to invalidate everything.
*/
int
-xfs_attr_root_inactive(xfs_trans_t **trans, xfs_inode_t *dp)
+xfs_attr3_root_inactive(
+ struct xfs_trans **trans,
+ struct xfs_inode *dp)
{
- xfs_da_blkinfo_t *info;
- xfs_daddr_t blkno;
- struct xfs_buf *bp;
- int error;
+ struct xfs_da_blkinfo *info;
+ struct xfs_buf *bp;
+ xfs_daddr_t blkno;
+ int error;
/*
* Read block 0 to see what we have to work with.
@@ -2786,38 +3002,44 @@ xfs_attr_root_inactive(xfs_trans_t **trans, xfs_inode_t *dp)
*/
error = xfs_da3_node_read(*trans, dp, 0, -1, &bp, XFS_ATTR_FORK);
if (error)
- return(error);
- blkno = XFS_BUF_ADDR(bp);
+ return error;
+ blkno = bp->b_bn;
/*
* Invalidate the tree, even if the "tree" is only a single leaf block.
* This is a depth-first traversal!
*/
info = bp->b_addr;
- if (info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC)) {
- error = xfs_attr_node_inactive(trans, dp, bp, 1);
- } else if (info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)) {
- error = xfs_attr_leaf_inactive(trans, dp, bp);
- } else {
+ switch (info->magic) {
+ case cpu_to_be16(XFS_DA_NODE_MAGIC):
+ case cpu_to_be16(XFS_DA3_NODE_MAGIC):
+ error = xfs_attr3_node_inactive(trans, dp, bp, 1);
+ break;
+ case cpu_to_be16(XFS_ATTR_LEAF_MAGIC):
+ case cpu_to_be16(XFS_ATTR3_LEAF_MAGIC):
+ error = xfs_attr3_leaf_inactive(trans, dp, bp);
+ break;
+ default:
error = XFS_ERROR(EIO);
xfs_trans_brelse(*trans, bp);
+ break;
}
if (error)
- return(error);
+ return error;
/*
* Invalidate the incore copy of the root block.
*/
error = xfs_da_get_buf(*trans, dp, 0, blkno, &bp, XFS_ATTR_FORK);
if (error)
- return(error);
+ return error;
xfs_trans_binval(*trans, bp); /* remove from cache */
/*
* Commit the invalidate and start the next transaction.
*/
error = xfs_trans_roll(trans, dp);
- return (error);
+ return error;
}
/*
@@ -2825,7 +3047,7 @@ xfs_attr_root_inactive(xfs_trans_t **trans, xfs_inode_t *dp)
* We're doing a depth-first traversal in order to invalidate everything.
*/
STATIC int
-xfs_attr_node_inactive(
+xfs_attr3_node_inactive(
struct xfs_trans **trans,
struct xfs_inode *dp,
struct xfs_buf *bp,
@@ -2835,25 +3057,25 @@ xfs_attr_node_inactive(
xfs_da_intnode_t *node;
xfs_dablk_t child_fsb;
xfs_daddr_t parent_blkno, child_blkno;
- int error, count, i;
+ int error, i;
struct xfs_buf *child_bp;
struct xfs_da_node_entry *btree;
+ struct xfs_da3_icnode_hdr ichdr;
/*
* Since this code is recursive (gasp!) we must protect ourselves.
*/
if (level > XFS_DA_NODE_MAXDEPTH) {
xfs_trans_brelse(*trans, bp); /* no locks for later trans */
- return(XFS_ERROR(EIO));
+ return XFS_ERROR(EIO);
}
node = bp->b_addr;
- ASSERT(node->hdr.info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC));
- parent_blkno = XFS_BUF_ADDR(bp); /* save for re-read later */
- count = be16_to_cpu(node->hdr.count);
- if (!count) {
+ xfs_da3_node_hdr_from_disk(&ichdr, node);
+ parent_blkno = bp->b_bn;
+ if (!ichdr.count) {
xfs_trans_brelse(*trans, bp);
- return(0);
+ return 0;
}
btree = xfs_da3_node_tree_p(node);
child_fsb = be32_to_cpu(btree[0].before);
@@ -2864,7 +3086,7 @@ xfs_attr_node_inactive(
* over the leaves removing all of them. If this is higher up
* in the tree, recurse downward.
*/
- for (i = 0; i < count; i++) {
+ for (i = 0; i < ichdr.count; i++) {
/*
* Read the subsidiary block to see what we have to work with.
* Don't do this in a transaction. This is a depth-first
@@ -2883,18 +3105,24 @@ xfs_attr_node_inactive(
* Invalidate the subtree, however we have to.
*/
info = child_bp->b_addr;
- if (info->magic == cpu_to_be16(XFS_DA_NODE_MAGIC)) {
- error = xfs_attr_node_inactive(trans, dp,
- child_bp, level+1);
- } else if (info->magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)) {
- error = xfs_attr_leaf_inactive(trans, dp,
- child_bp);
- } else {
+ switch (info->magic) {
+ case cpu_to_be16(XFS_DA_NODE_MAGIC):
+ case cpu_to_be16(XFS_DA3_NODE_MAGIC):
+ error = xfs_attr3_node_inactive(trans, dp,
+ child_bp, level + 1);
+ break;
+ case cpu_to_be16(XFS_ATTR_LEAF_MAGIC):
+ case cpu_to_be16(XFS_ATTR3_LEAF_MAGIC):
+ error = xfs_attr3_leaf_inactive(trans, dp,
+ child_bp);
+ break;
+ default:
error = XFS_ERROR(EIO);
xfs_trans_brelse(*trans, child_bp);
+ break;
}
if (error)
- return(error);
+ return error;
/*
* Remove the subsidiary block from the cache
@@ -2903,7 +3131,7 @@ xfs_attr_node_inactive(
error = xfs_da_get_buf(*trans, dp, 0, child_blkno,
&child_bp, XFS_ATTR_FORK);
if (error)
- return(error);
+ return error;
xfs_trans_binval(*trans, child_bp);
}
@@ -2911,12 +3139,12 @@ xfs_attr_node_inactive(
* If we're not done, re-read the parent to get the next
* child block number.
*/
- if ((i+1) < count) {
+ if (i + 1 < ichdr.count) {
error = xfs_da3_node_read(*trans, dp, 0, parent_blkno,
&bp, XFS_ATTR_FORK);
if (error)
- return(error);
- child_fsb = be32_to_cpu(btree[i+1].before);
+ return error;
+ child_fsb = be32_to_cpu(btree[i + 1].before);
xfs_trans_brelse(*trans, bp);
}
/*
@@ -2924,10 +3152,10 @@ xfs_attr_node_inactive(
*/
error = xfs_trans_roll(trans, dp);
if (error)
- return (error);
+ return error;
}
- return(0);
+ return 0;
}
/*
@@ -2937,29 +3165,35 @@ xfs_attr_node_inactive(
* caught holding something that the logging code wants to flush to disk.
*/
STATIC int
-xfs_attr_leaf_inactive(
- struct xfs_trans **trans,
- struct xfs_inode *dp,
- struct xfs_buf *bp)
+xfs_attr3_leaf_inactive(
+ struct xfs_trans **trans,
+ struct xfs_inode *dp,
+ struct xfs_buf *bp)
{
- xfs_attr_leafblock_t *leaf;
- xfs_attr_leaf_entry_t *entry;
- xfs_attr_leaf_name_remote_t *name_rmt;
- xfs_attr_inactive_list_t *list, *lp;
- int error, count, size, tmp, i;
+ struct xfs_attr_leafblock *leaf;
+ struct xfs_attr3_icleaf_hdr ichdr;
+ struct xfs_attr_leaf_entry *entry;
+ struct xfs_attr_leaf_name_remote *name_rmt;
+ struct xfs_attr_inactive_list *list;
+ struct xfs_attr_inactive_list *lp;
+ int error;
+ int count;
+ int size;
+ int tmp;
+ int i;
leaf = bp->b_addr;
- ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC));
+ xfs_attr3_leaf_hdr_from_disk(&ichdr, leaf);
/*
* Count the number of "remote" value extents.
*/
count = 0;
- entry = &leaf->entries[0];
- for (i = 0; i < be16_to_cpu(leaf->hdr.count); entry++, i++) {
+ entry = xfs_attr3_leaf_entryp(leaf);
+ for (i = 0; i < ichdr.count; entry++, i++) {
if (be16_to_cpu(entry->nameidx) &&
((entry->flags & XFS_ATTR_LOCAL) == 0)) {
- name_rmt = xfs_attr_leaf_name_remote(leaf, i);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, i);
if (name_rmt->valueblk)
count++;
}
@@ -2970,24 +3204,24 @@ xfs_attr_leaf_inactive(
*/
if (count == 0) {
xfs_trans_brelse(*trans, bp);
- return(0);
+ return 0;
}
/*
* Allocate storage for a list of all the "remote" value extents.
*/
size = count * sizeof(xfs_attr_inactive_list_t);
- list = (xfs_attr_inactive_list_t *)kmem_alloc(size, KM_SLEEP);
+ list = kmem_alloc(size, KM_SLEEP);
/*
* Identify each of the "remote" value extents.
*/
lp = list;
- entry = &leaf->entries[0];
- for (i = 0; i < be16_to_cpu(leaf->hdr.count); entry++, i++) {
+ entry = xfs_attr3_leaf_entryp(leaf);
+ for (i = 0; i < ichdr.count; entry++, i++) {
if (be16_to_cpu(entry->nameidx) &&
((entry->flags & XFS_ATTR_LOCAL) == 0)) {
- name_rmt = xfs_attr_leaf_name_remote(leaf, i);
+ name_rmt = xfs_attr3_leaf_name_remote(leaf, i);
if (name_rmt->valueblk) {
lp->valueblk = be32_to_cpu(name_rmt->valueblk);
lp->valuelen = XFS_B_TO_FSB(dp->i_mount,
@@ -3003,15 +3237,15 @@ xfs_attr_leaf_inactive(
*/
error = 0;
for (lp = list, i = 0; i < count; i++, lp++) {
- tmp = xfs_attr_leaf_freextent(trans, dp,
+ tmp = xfs_attr3_leaf_freextent(trans, dp,
lp->valueblk, lp->valuelen);
if (error == 0)
error = tmp; /* save only the 1st errno */
}
- kmem_free((xfs_caddr_t)list);
- return(error);
+ kmem_free(list);
+ return error;
}
/*
@@ -3019,14 +3253,20 @@ xfs_attr_leaf_inactive(
* invalidate any buffers that are incore/in transactions.
*/
STATIC int
-xfs_attr_leaf_freextent(xfs_trans_t **trans, xfs_inode_t *dp,
- xfs_dablk_t blkno, int blkcnt)
+xfs_attr3_leaf_freextent(
+ struct xfs_trans **trans,
+ struct xfs_inode *dp,
+ xfs_dablk_t blkno,
+ int blkcnt)
{
- xfs_bmbt_irec_t map;
- xfs_dablk_t tblkno;
- int tblkcnt, dblkcnt, nmap, error;
- xfs_daddr_t dblkno;
- xfs_buf_t *bp;
+ struct xfs_bmbt_irec map;
+ struct xfs_buf *bp;
+ xfs_dablk_t tblkno;
+ xfs_daddr_t dblkno;
+ int tblkcnt;
+ int dblkcnt;
+ int nmap;
+ int error;
/*
* Roll through the "value", invalidating the attribute value's
diff --git a/fs/xfs/xfs_attr_leaf.h b/fs/xfs/xfs_attr_leaf.h
index 77de139..5a4558d 100644
--- a/fs/xfs/xfs_attr_leaf.h
+++ b/fs/xfs/xfs_attr_leaf.h
@@ -89,7 +89,7 @@ typedef struct xfs_attr_leaf_hdr { /* constant-structure header block */
typedef struct xfs_attr_leaf_entry { /* sorted on key, not name */
__be32 hashval; /* hash value of name */
- __be16 nameidx; /* index into buffer of name/value */
+ __be16 nameidx; /* index into buffer of name/value */
__u8 flags; /* LOCAL/ROOT/SECURE/INCOMPLETE flag */
__u8 pad2; /* unused pad byte */
} xfs_attr_leaf_entry_t;
@@ -115,6 +115,54 @@ typedef struct xfs_attr_leafblock {
} xfs_attr_leafblock_t;
/*
+ * CRC enabled leaf structures. Called "version 3" structures to match the
+ * version number of the directory and dablk structures for this feature, and
+ * attr2 is already taken by the variable inode attribute fork size feature.
+ */
+struct xfs_attr3_leaf_hdr {
+ struct xfs_da3_blkinfo info;
+ __be16 count;
+ __be16 usedbytes;
+ __be16 firstused;
+ __u8 holes;
+ __u8 pad1;
+ struct xfs_attr_leaf_map freemap[XFS_ATTR_LEAF_MAPSIZE];
+};
+
+#define XFS_ATTR3_LEAF_CRC_OFF (offsetof(struct xfs_attr3_leaf_hdr, info.crc))
+
+struct xfs_attr3_leafblock {
+ struct xfs_attr3_leaf_hdr hdr;
+ struct xfs_attr_leaf_entry entries[1];
+
+ /*
+ * The rest of the block contains the following structures after the
+ * leaf entries, growing from the bottom up. The variables are never
+ * referenced, the locations accessed purely from helper functions.
+ *
+ * struct xfs_attr_leaf_name_local
+ * struct xfs_attr_leaf_name_remote
+ */
+};
+
+/*
+ * incore, neutral version of the attribute leaf header
+ */
+struct xfs_attr3_icleaf_hdr {
+ __uint32_t forw;
+ __uint32_t back;
+ __uint16_t magic;
+ __uint16_t count;
+ __uint16_t usedbytes;
+ __uint16_t firstused;
+ __u8 holes;
+ struct {
+ __uint16_t base;
+ __uint16_t size;
+ } freemap[XFS_ATTR_LEAF_MAPSIZE];
+};
+
+/*
* Flags used in the leaf_entry[i].flags field.
* NOTE: the INCOMPLETE bit must not collide with the flags bits specified
* on the system call, they are "or"ed together for various operations.
@@ -147,26 +195,43 @@ typedef struct xfs_attr_leafblock {
*/
#define XFS_ATTR_LEAF_NAME_ALIGN ((uint)sizeof(xfs_dablk_t))
+static inline int
+xfs_attr3_leaf_hdr_size(struct xfs_attr_leafblock *leafp)
+{
+ if (leafp->hdr.info.magic == cpu_to_be16(XFS_ATTR3_LEAF_MAGIC))
+ return sizeof(struct xfs_attr3_leaf_hdr);
+ return sizeof(struct xfs_attr_leaf_hdr);
+}
+
+static inline struct xfs_attr_leaf_entry *
+xfs_attr3_leaf_entryp(xfs_attr_leafblock_t *leafp)
+{
+ if (leafp->hdr.info.magic == cpu_to_be16(XFS_ATTR3_LEAF_MAGIC))
+ return &((struct xfs_attr3_leafblock *)leafp)->entries[0];
+ return &leafp->entries[0];
+}
+
/*
* Cast typed pointers for "local" and "remote" name/value structs.
*/
-static inline xfs_attr_leaf_name_remote_t *
-xfs_attr_leaf_name_remote(xfs_attr_leafblock_t *leafp, int idx)
+static inline char *
+xfs_attr3_leaf_name(xfs_attr_leafblock_t *leafp, int idx)
{
- return (xfs_attr_leaf_name_remote_t *)
- &((char *)leafp)[be16_to_cpu(leafp->entries[idx].nameidx)];
+ struct xfs_attr_leaf_entry *entries = xfs_attr3_leaf_entryp(leafp);
+
+ return &((char *)leafp)[be16_to_cpu(entries[idx].nameidx)];
}
-static inline xfs_attr_leaf_name_local_t *
-xfs_attr_leaf_name_local(xfs_attr_leafblock_t *leafp, int idx)
+static inline xfs_attr_leaf_name_remote_t *
+xfs_attr3_leaf_name_remote(xfs_attr_leafblock_t *leafp, int idx)
{
- return (xfs_attr_leaf_name_local_t *)
- &((char *)leafp)[be16_to_cpu(leafp->entries[idx].nameidx)];
+ return (xfs_attr_leaf_name_remote_t *)xfs_attr3_leaf_name(leafp, idx);
}
-static inline char *xfs_attr_leaf_name(xfs_attr_leafblock_t *leafp, int idx)
+static inline xfs_attr_leaf_name_local_t *
+xfs_attr3_leaf_name_local(xfs_attr_leafblock_t *leafp, int idx)
{
- return &((char *)leafp)[be16_to_cpu(leafp->entries[idx].nameidx)];
+ return (xfs_attr_leaf_name_local_t *)xfs_attr3_leaf_name(leafp, idx);
}
/*
@@ -221,37 +286,37 @@ int xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes);
/*
* Internal routines when attribute fork size == XFS_LBSIZE(mp).
*/
-int xfs_attr_leaf_to_node(struct xfs_da_args *args);
-int xfs_attr_leaf_to_shortform(struct xfs_buf *bp,
+int xfs_attr3_leaf_to_node(struct xfs_da_args *args);
+int xfs_attr3_leaf_to_shortform(struct xfs_buf *bp,
struct xfs_da_args *args, int forkoff);
-int xfs_attr_leaf_clearflag(struct xfs_da_args *args);
-int xfs_attr_leaf_setflag(struct xfs_da_args *args);
-int xfs_attr_leaf_flipflags(xfs_da_args_t *args);
+int xfs_attr3_leaf_clearflag(struct xfs_da_args *args);
+int xfs_attr3_leaf_setflag(struct xfs_da_args *args);
+int xfs_attr3_leaf_flipflags(struct xfs_da_args *args);
/*
* Routines used for growing the Btree.
*/
-int xfs_attr_leaf_split(struct xfs_da_state *state,
+int xfs_attr3_leaf_split(struct xfs_da_state *state,
struct xfs_da_state_blk *oldblk,
struct xfs_da_state_blk *newblk);
-int xfs_attr_leaf_lookup_int(struct xfs_buf *leaf,
+int xfs_attr3_leaf_lookup_int(struct xfs_buf *leaf,
struct xfs_da_args *args);
-int xfs_attr_leaf_getvalue(struct xfs_buf *bp, struct xfs_da_args *args);
-int xfs_attr_leaf_add(struct xfs_buf *leaf_buffer,
+int xfs_attr3_leaf_getvalue(struct xfs_buf *bp, struct xfs_da_args *args);
+int xfs_attr3_leaf_add(struct xfs_buf *leaf_buffer,
struct xfs_da_args *args);
-int xfs_attr_leaf_remove(struct xfs_buf *leaf_buffer,
+int xfs_attr3_leaf_remove(struct xfs_buf *leaf_buffer,
struct xfs_da_args *args);
-int xfs_attr_leaf_list_int(struct xfs_buf *bp,
+int xfs_attr3_leaf_list_int(struct xfs_buf *bp,
struct xfs_attr_list_context *context);
/*
* Routines used for shrinking the Btree.
*/
-int xfs_attr_leaf_toosmall(struct xfs_da_state *state, int *retval);
-void xfs_attr_leaf_unbalance(struct xfs_da_state *state,
+int xfs_attr3_leaf_toosmall(struct xfs_da_state *state, int *retval);
+void xfs_attr3_leaf_unbalance(struct xfs_da_state *state,
struct xfs_da_state_blk *drop_blk,
struct xfs_da_state_blk *save_blk);
-int xfs_attr_root_inactive(struct xfs_trans **trans, struct xfs_inode *dp);
+int xfs_attr3_root_inactive(struct xfs_trans **trans, struct xfs_inode *dp);
/*
* Utility routines.
@@ -261,10 +326,12 @@ int xfs_attr_leaf_order(struct xfs_buf *leaf1_bp,
struct xfs_buf *leaf2_bp);
int xfs_attr_leaf_newentsize(int namelen, int valuelen, int blocksize,
int *local);
-int xfs_attr_leaf_read(struct xfs_trans *tp, struct xfs_inode *dp,
+int xfs_attr3_leaf_read(struct xfs_trans *tp, struct xfs_inode *dp,
xfs_dablk_t bno, xfs_daddr_t mappedbno,
struct xfs_buf **bpp);
+void xfs_attr3_leaf_hdr_from_disk(struct xfs_attr3_icleaf_hdr *to,
+ struct xfs_attr_leafblock *from);
-extern const struct xfs_buf_ops xfs_attr_leaf_buf_ops;
+extern const struct xfs_buf_ops xfs_attr3_leaf_buf_ops;
#endif /* __XFS_ATTR_LEAF_H__ */
diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c
index f0ffc7c..da619bc 100644
--- a/fs/xfs/xfs_da_btree.c
+++ b/fs/xfs/xfs_da_btree.c
@@ -269,7 +269,7 @@ xfs_da3_node_read_verify(
break;
return;
case XFS_ATTR_LEAF_MAGIC:
- bp->b_ops = &xfs_attr_leaf_buf_ops;
+ bp->b_ops = &xfs_attr3_leaf_buf_ops;
bp->b_ops->verify_read(bp);
return;
case XFS_DIR2_LEAFN_MAGIC:
@@ -399,7 +399,7 @@ xfs_da3_split(
*/
switch (oldblk->magic) {
case XFS_ATTR_LEAF_MAGIC:
- error = xfs_attr_leaf_split(state, oldblk, newblk);
+ error = xfs_attr3_leaf_split(state, oldblk, newblk);
if ((error != 0) && (error != ENOSPC)) {
return(error); /* GROT: attr is inconsistent */
}
@@ -414,12 +414,12 @@ xfs_da3_split(
if (state->inleaf) {
state->extraafter = 0; /* before newblk */
trace_xfs_attr_leaf_split_before(state->args);
- error = xfs_attr_leaf_split(state, oldblk,
+ error = xfs_attr3_leaf_split(state, oldblk,
&state->extrablk);
} else {
state->extraafter = 1; /* after newblk */
trace_xfs_attr_leaf_split_after(state->args);
- error = xfs_attr_leaf_split(state, newblk,
+ error = xfs_attr3_leaf_split(state, newblk,
&state->extrablk);
}
if (error)
@@ -959,12 +959,12 @@ xfs_da3_join(
*/
switch (drop_blk->magic) {
case XFS_ATTR_LEAF_MAGIC:
- error = xfs_attr_leaf_toosmall(state, &action);
+ error = xfs_attr3_leaf_toosmall(state, &action);
if (error)
return(error);
if (action == 0)
return(0);
- xfs_attr_leaf_unbalance(state, drop_blk, save_blk);
+ xfs_attr3_leaf_unbalance(state, drop_blk, save_blk);
break;
case XFS_DIR2_LEAFN_MAGIC:
error = xfs_dir2_leafn_toosmall(state, &action);
@@ -1473,7 +1473,9 @@ xfs_da3_node_lookup_int(
curr = blk->bp->b_addr;
blk->magic = be16_to_cpu(curr->magic);
- if (blk->magic == XFS_ATTR_LEAF_MAGIC) {
+ if (blk->magic == XFS_ATTR_LEAF_MAGIC ||
+ blk->magic == XFS_ATTR3_LEAF_MAGIC) {
+ blk->magic = XFS_ATTR_LEAF_MAGIC;
blk->hashval = xfs_attr_leaf_lasthash(blk->bp, NULL);
break;
}
@@ -1550,7 +1552,7 @@ xfs_da3_node_lookup_int(
retval = xfs_dir2_leafn_lookup_int(blk->bp, args,
&blk->index, state);
} else if (blk->magic == XFS_ATTR_LEAF_MAGIC) {
- retval = xfs_attr_leaf_lookup_int(blk->bp, args);
+ retval = xfs_attr3_leaf_lookup_int(blk->bp, args);
blk->index = args->index;
args->blkno = blk->blkno;
} else {
diff --git a/fs/xfs/xfs_da_btree.h b/fs/xfs/xfs_da_btree.h
index f20741e..fd805d8 100644
--- a/fs/xfs/xfs_da_btree.h
+++ b/fs/xfs/xfs_da_btree.h
@@ -54,6 +54,7 @@ typedef struct xfs_da_blkinfo {
* magic numbers without modification for both v2 and v3 nodes.
*/
#define XFS_DA3_NODE_MAGIC 0x3ebe /* magic number: non-leaf blocks */
+#define XFS_ATTR3_LEAF_MAGIC 0x3bee /* magic number: attribute leaf blks */
#define XFS_DIR3_LEAF1_MAGIC 0x3df1 /* magic number: v2 dirlf single blks */
#define XFS_DIR3_LEAFN_MAGIC 0x3dff /* magic number: v2 dirlf multi blks */
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH 16/16] xfs: add CRC checks to the superblock
2013-02-25 1:31 [PATCH 00/16] xfs: metadata CRCs, second batch Dave Chinner
` (14 preceding siblings ...)
2013-02-25 1:31 ` [PATCH 15/16] xfs: add CRCs to attr leaf blocks Dave Chinner
@ 2013-02-25 1:31 ` Dave Chinner
15 siblings, 0 replies; 23+ messages in thread
From: Dave Chinner @ 2013-02-25 1:31 UTC (permalink / raw)
To: xfs
From: Dave Chinner <dchinner@redhat.com>
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>
---
fs/xfs/xfs_buf_item.h | 4 +-
fs/xfs/xfs_log_recover.c | 8 ++++
fs/xfs/xfs_mount.c | 97 ++++++++++++++++++++++++++++++++++++--------
fs/xfs/xfs_mount.h | 1 +
fs/xfs/xfs_sb.h | 100 ++++++++++++++++++++++++++++++++--------------
5 files changed, 164 insertions(+), 46 deletions(-)
diff --git a/fs/xfs/xfs_buf_item.h b/fs/xfs/xfs_buf_item.h
index 09cab4e..11c4055 100644
--- a/fs/xfs/xfs_buf_item.h
+++ b/fs/xfs/xfs_buf_item.h
@@ -50,6 +50,7 @@ extern kmem_zone_t *xfs_buf_item_zone;
#define XFS_BLF_AGI_BUF (1<<8)
#define XFS_BLF_DINO_BUF (1<<9)
#define XFS_BLF_SYMLINK_BUF (1<<10)
+#define XFS_BLF_SB_BUF (1<<11)
#define XFS_BLF_TYPE_MASK \
(XFS_BLF_UDQUOT_BUF | \
@@ -60,7 +61,8 @@ extern kmem_zone_t *xfs_buf_item_zone;
XFS_BLF_AGFL_BUF | \
XFS_BLF_AGI_BUF | \
XFS_BLF_DINO_BUF | \
- XFS_BLF_SYMLINK_BUF)
+ XFS_BLF_SYMLINK_BUF | \
+ XFS_BLF_SB_BUF)
#define XFS_BLF_CHUNK 128
#define XFS_BLF_SHIFT 7
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 3216ea5..bc30c55 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -2017,6 +2017,14 @@ xlog_recover_do_reg_buffer(
}
bp->b_ops = &xfs_symlink_buf_ops;
break;
+ case XFS_BLF_SB_BUF:
+ if (*(__be32 *)bp->b_addr != cpu_to_be32(XFS_SB_MAGIC)) {
+ xfs_warn(mp, "Bad SB block magic!");
+ ASSERT(0);
+ break;
+ }
+ bp->b_ops = &xfs_sb_buf_ops;
+ break;
default:
break;
}
diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index 3806088..97443c9 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 @@ 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 }
};
@@ -557,6 +565,12 @@ 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_crc = be32_to_cpu(from->sb_crc);
+ to->sb_pquotino = be64_to_cpu(from->sb_pquotino);
+ to->sb_lsn = be64_to_cpu(from->sb_lsn);
}
/*
@@ -612,13 +626,12 @@ 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));
@@ -626,16 +639,46 @@ 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);
+ }
}
/*
@@ -648,11 +691,10 @@ 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;
@@ -663,9 +705,27 @@ 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 = {
@@ -687,7 +747,8 @@ int
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 +787,10 @@ reread:
/*
* 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 +799,18 @@ reread:
* 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 +1697,7 @@ 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_BLF_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 bc90706..157a7d3 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -392,6 +392,7 @@ extern void xfs_set_low_space_thresholds(struct xfs_mount *);
#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 @@ 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)
@@ -500,14 +542,14 @@ 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 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;
}
/*
--
1.7.10
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
^ permalink raw reply related [flat|nested] 23+ messages in thread