@@ -62,6 +62,72 @@ xfs_prealloc_blocks(
}
/*
+ * In order to avoid ENOSPC-related deadlock caused by out-of-order locking of
+ * AGF buffer (PV 947395), we place constraints on the relationship among actual
+ * allocations for data blocks, freelist blocks, and potential file data bmap
+ * btree blocks. However, these restrictions may result in no actual space
+ * allocated for a delayed extent, for example, a data block in a certain AG is
+ * allocated but there is no additional block for the additional bmap btree
+ * block due to a split of the bmap btree of the file. The result of this may
+ * lead to an infinite loop when the file gets flushed to disk and all delayed
+ * extents need to be actually allocated. To get around this, we explicitly set
+ * aside a few blocks which will not be reserved in delayed allocation.
+ *
+ * The minimum number of needed freelist blocks is 4 fsbs _per AG_ when we are
+ * not using rmap btrees a potential split of file's bmap btree requires 1 fsb,
+ * so we set the number of set-aside blocks to 4 + 4*agcount when not using rmap
+ * btrees.
+ *
+ * When rmap btrees are active, we have to consider that using the last block in
+ * the AG can cause a full height rmap btree split and we need enough blocks on
+ * the AGFL to be able to handle this. That means we have, in addition to the
+ * above consideration, another (2 * mp->m_ag_levels) - 1 blocks required to be
+ * available to the free list.
+ */
+unsigned int
+xfs_alloc_set_aside(
+ struct xfs_mount *mp)
+{
+ unsigned int blocks;
+
+ blocks = 4 + (mp->m_sb.sb_agcount * XFS_ALLOC_AGFL_RESERVE);
+ if (!xfs_sb_version_hasrmapbt(&mp->m_sb))
+ return blocks;
+ return blocks + (mp->m_sb.sb_agcount * (2 * mp->m_ag_maxlevels) - 1);
+}
+
+/*
+ * When deciding how much space to allocate out of an AG, we limit the
+ * allocation maximum size to the size the AG. However, we cannot use all the
+ * blocks in the AG - some are permanently used by metadata. These
+ * blocks are generally:
+ * - the AG superblock, AGF, AGI and AGFL
+ * - the AGF (bno and cnt) and AGI btree root blocks, and optionally
+ * the AGI free inode and rmap btree root blocks.
+ * - blocks on the AGFL according to xfs_alloc_set_aside() limits
+ *
+ * The AG headers are sector sized, so the amount of space they take up is
+ * dependent on filesystem geometry. The others are all single blocks.
+ */
+unsigned int
+xfs_alloc_ag_max_usable(struct xfs_mount *mp)
+{
+ unsigned int blocks;
+
+ blocks = XFS_BB_TO_FSB(mp, XFS_FSS_TO_BB(mp, 4)); /* ag headers */
+ blocks += XFS_ALLOC_AGFL_RESERVE;
+ blocks += 3; /* AGF, AGI btree root blocks */
+ if (xfs_sb_version_hasfinobt(&mp->m_sb))
+ blocks++; /* finobt root block */
+ if (xfs_sb_version_hasrmapbt(&mp->m_sb)) {
+ /* rmap root block + full tree split on full AG */
+ blocks += 1 + (2 * mp->m_ag_maxlevels) - 1;
+ }
+
+ return mp->m_sb.sb_agblocks - blocks;
+}
+
+/*
* Lookup the record equal to [bno, len] in the btree given by cur.
*/
STATIC int /* error */
@@ -1911,6 +1977,9 @@ xfs_alloc_min_freelist(
/* space needed by-size freespace btree */
min_free += min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_CNTi] + 1,
mp->m_ag_maxlevels);
+ /* space needed reverse mapping used space btree */
+ min_free += min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_RMAPi] + 1,
+ mp->m_ag_maxlevels);
return min_free;
}
@@ -56,42 +56,6 @@ typedef unsigned int xfs_alloctype_t;
#define XFS_ALLOC_FLAG_FREEING 0x00000002 /* indicate caller is freeing extents*/
/*
- * In order to avoid ENOSPC-related deadlock caused by
- * out-of-order locking of AGF buffer (PV 947395), we place
- * constraints on the relationship among actual allocations for
- * data blocks, freelist blocks, and potential file data bmap
- * btree blocks. However, these restrictions may result in no
- * actual space allocated for a delayed extent, for example, a data
- * block in a certain AG is allocated but there is no additional
- * block for the additional bmap btree block due to a split of the
- * bmap btree of the file. The result of this may lead to an
- * infinite loop in xfssyncd when the file gets flushed to disk and
- * all delayed extents need to be actually allocated. To get around
- * this, we explicitly set aside a few blocks which will not be
- * reserved in delayed allocation. Considering the minimum number of
- * needed freelist blocks is 4 fsbs _per AG_, a potential split of file's bmap
- * btree requires 1 fsb, so we set the number of set-aside blocks
- * to 4 + 4*agcount.
- */
-#define XFS_ALLOC_SET_ASIDE(mp) (4 + ((mp)->m_sb.sb_agcount * 4))
-
-/*
- * When deciding how much space to allocate out of an AG, we limit the
- * allocation maximum size to the size the AG. However, we cannot use all the
- * blocks in the AG - some are permanently used by metadata. These
- * blocks are generally:
- * - the AG superblock, AGF, AGI and AGFL
- * - the AGF (bno and cnt) and AGI btree root blocks
- * - 4 blocks on the AGFL according to XFS_ALLOC_SET_ASIDE() limits
- *
- * The AG headers are sector sized, so the amount of space they take up is
- * dependent on filesystem geometry. The others are all single blocks.
- */
-#define XFS_ALLOC_AG_MAX_USABLE(mp) \
- ((mp)->m_sb.sb_agblocks - XFS_BB_TO_FSB(mp, XFS_FSS_TO_BB(mp, 4)) - 7)
-
-
-/*
* Argument structure for xfs_alloc routines.
* This is turned into a structure to avoid having 20 arguments passed
* down several levels of the stack.
@@ -131,6 +95,11 @@ typedef struct xfs_alloc_arg {
#define XFS_ALLOC_USERDATA 1 /* allocation is for user data*/
#define XFS_ALLOC_INITIAL_USER_DATA 2 /* special case start of file */
+/* freespace limit calculations */
+#define XFS_ALLOC_AGFL_RESERVE 4
+unsigned int xfs_alloc_set_aside(struct xfs_mount *mp);
+unsigned int xfs_alloc_ag_max_usable(struct xfs_mount *mp);
+
xfs_extlen_t xfs_alloc_longest_free_extent(struct xfs_mount *mp,
struct xfs_perag *pag, xfs_extlen_t need);
unsigned int xfs_alloc_min_freelist(struct xfs_mount *mp,
@@ -3716,7 +3716,7 @@ xfs_bmap_btalloc(
args.fsbno = ap->blkno;
/* Trim the allocation back to the maximum an AG can fit. */
- args.maxlen = MIN(ap->length, XFS_ALLOC_AG_MAX_USABLE(mp));
+ args.maxlen = MIN(ap->length, mp->m_ag_max_usable);
args.firstblock = *ap->firstblock;
blen = 0;
if (nullfb) {
@@ -732,6 +732,8 @@ xfs_sb_mount_common(
mp->m_ialloc_min_blks = sbp->sb_spino_align;
else
mp->m_ialloc_min_blks = mp->m_ialloc_blks;
+ mp->m_alloc_set_aside = xfs_alloc_set_aside(mp);
+ mp->m_ag_max_usable = xfs_alloc_ag_max_usable(mp);
}
/*
@@ -179,7 +179,7 @@ xfs_ioc_trim(
* matter as trimming blocks is an advisory interface.
*/
if (range.start >= XFS_FSB_TO_B(mp, mp->m_sb.sb_dblocks) ||
- range.minlen > XFS_FSB_TO_B(mp, XFS_ALLOC_AG_MAX_USABLE(mp)) ||
+ range.minlen > XFS_FSB_TO_B(mp, mp->m_ag_max_usable) ||
range.len < mp->m_sb.sb_blocksize)
return -EINVAL;
@@ -723,7 +723,7 @@ xfs_fs_counts(
cnt->allocino = percpu_counter_read_positive(&mp->m_icount);
cnt->freeino = percpu_counter_read_positive(&mp->m_ifree);
cnt->freedata = percpu_counter_read_positive(&mp->m_fdblocks) -
- XFS_ALLOC_SET_ASIDE(mp);
+ mp->m_alloc_set_aside;
spin_lock(&mp->m_sb_lock);
cnt->freertx = mp->m_sb.sb_frextents;
@@ -796,7 +796,7 @@ retry:
__int64_t free;
free = percpu_counter_sum(&mp->m_fdblocks) -
- XFS_ALLOC_SET_ASIDE(mp);
+ mp->m_alloc_set_aside;
if (!free)
goto out; /* ENOSPC and fdblks_delta = 0 */
@@ -1196,7 +1196,7 @@ xfs_mod_fdblocks(
batch = XFS_FDBLOCKS_BATCH;
__percpu_counter_add(&mp->m_fdblocks, delta, batch);
- if (__percpu_counter_compare(&mp->m_fdblocks, XFS_ALLOC_SET_ASIDE(mp),
+ if (__percpu_counter_compare(&mp->m_fdblocks, mp->m_alloc_set_aside,
XFS_FDBLOCKS_BATCH) >= 0) {
/* we had space! */
return 0;
@@ -96,6 +96,8 @@ typedef struct xfs_mount {
uint m_bm_maxlevels[2]; /* XFS_BM_MAXLEVELS */
uint m_in_maxlevels; /* max inobt btree levels. */
xfs_extlen_t m_ag_prealloc_blocks; /* reserved ag blocks */
+ uint m_alloc_set_aside; /* space we can't use */
+ uint m_ag_max_usable; /* max space per AG */
struct radix_tree_root m_perag_tree; /* per-ag accounting info */
spinlock_t m_perag_lock; /* lock for m_perag_tree */
struct mutex m_growlock; /* growfs mutex */
@@ -1072,7 +1072,7 @@ xfs_fs_statfs(
statp->f_blocks = sbp->sb_dblocks - lsize;
spin_unlock(&mp->m_sb_lock);
- statp->f_bfree = fdblocks - XFS_ALLOC_SET_ASIDE(mp);
+ statp->f_bfree = fdblocks - mp->m_alloc_set_aside;
statp->f_bavail = statp->f_bfree;
fakeinos = statp->f_bfree << sbp->sb_inopblog;