[PATCH] [RFC] xfs: increase inode cluster size for v5 filesystems

[Dave Chinner <david@xxxxxxxxxxxxx>]

From: Dave Chinner <dchinner@xxxxxxxxxx>

v5 filesystems use 512 byte inodes as a minimum, so read inodes in
clusters that are effectively half the size of a v4 filesystem with
256 byte inodes. For v5 fielsystems, scale the inode cluster size
with the size of the inode so that we keep a constant 32 inodes per
cluster ratio for all inode IO.

This only works if mkfs.xfs sets the inode alignment appropriately
for larger inode clusters, so this functionality is made conditional
on mkfs doing the right thing. xfs_repair needs to know about
the inode alignment changes, too.

FWIW, results with lookaside cache size of 37 entries with this
patch are (Note: finobt enabled on v5 filesystems, v4 using defaults
including 256 byte inode size):

Wall time:
        create  bulkstat        find+stat       ls -R   unlink
v4      237s    161s            173s            201s    299s
v5      235s    163s            205s             31s    356s
patched 234s    160s            182s             29s    317s

System time:
        create  bulkstat        find+stat       ls -R   unlink
v4      2601s   2490s           1653s           1656s   2960s
v5      2637s   2497s           1681s             20s   3216s
patched 2613s   2451s           1658s             20s   3007s

Lookaside cache hit rate:
        create  bulkstat        find+stat       ls -R   unlink
v4      0.73    0.91            0.71            0.70    0.71
v5      0.76    0.88            0.68            0.10    0.75
patched 0.81    0.93            0.70            0.08    0.84

So, wall time same or down across the board, system time same or
down across the board, and cache hit rates all improve except for
the ls -R case which is a pure cold cache directory read workload
on v5 filesystems...

So, this patch removes most of the performance and CPU usage
differential between v4 and v5 filesystems on traversal related
workloads.

Signed-off-by: Dave Chinner <dchinner@xxxxxxxxxx>
---
 fs/xfs/xfs_mount.c      | 14 ++++++++++++++
 fs/xfs/xfs_mount.h      |  2 +-
 fs/xfs/xfs_trans_resv.c |  3 +--
 3 files changed, 16 insertions(+), 3 deletions(-)

diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c
index 8ac98c7..788d666d 100644
--- a/fs/xfs/xfs_mount.c
+++ b/fs/xfs/xfs_mount.c
@@ -722,8 +722,22 @@ xfs_mountfs(
         * Set the inode cluster size.
         * This may still be overridden by the file system
         * block size if it is larger than the chosen cluster size.
+        *
+        * For v5 filesystems, scale the cluster size with the inode size to
+        * keep a constant ratio of inode per cluster buffer, but only if mkfs
+        * has set the inode alignment value appropriately for larger cluster
+        * sizes.
         */
        mp->m_inode_cluster_size = XFS_INODE_BIG_CLUSTER_SIZE;
+       if (xfs_sb_version_hascrc(&mp->m_sb)) {
+               int     new_size = mp->m_inode_cluster_size;
+
+               new_size *= mp->m_sb.sb_inodesize / XFS_DINODE_MIN_SIZE;
+               if (mp->m_sb.sb_inoalignmt >= XFS_B_TO_FSBT(mp, new_size))
+                       mp->m_inode_cluster_size = new_size;
+               xfs_info(mp, "Using inode cluster size of %d bytes",
+                        mp->m_inode_cluster_size);
+       }
 
        /*
         * Set inode alignment fields
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 2a997dc..a4f7f94 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -111,7 +111,7 @@ typedef struct xfs_mount {
        __uint8_t               m_blkbb_log;    /* blocklog - BBSHIFT */
        __uint8_t               m_agno_log;     /* log #ag's */
        __uint8_t               m_agino_log;    /* #bits for agino in inum */
-       __uint16_t              m_inode_cluster_size;/* min inode buf size */
+       uint                    m_inode_cluster_size;/* min inode buf size */
        uint                    m_blockmask;    /* sb_blocksize-1 */
        uint                    m_blockwsize;   /* sb_blocksize in words */
        uint                    m_blockwmask;   /* blockwsize-1 */
diff --git a/fs/xfs/xfs_trans_resv.c b/fs/xfs/xfs_trans_resv.c
index ae7a185..1494f62 100644
--- a/fs/xfs/xfs_trans_resv.c
+++ b/fs/xfs/xfs_trans_resv.c
@@ -403,8 +403,7 @@ xfs_calc_ifree_reservation(
                xfs_calc_inode_res(mp, 1) +
                xfs_calc_buf_res(2, mp->m_sb.sb_sectsize) +
                xfs_calc_buf_res(2, XFS_FSB_TO_B(mp, 1)) +
-               MAX((__uint16_t)XFS_FSB_TO_B(mp, 1),
-                   XFS_INODE_CLUSTER_SIZE(mp)) +
+               MAX(XFS_FSB_TO_B(mp, 1), XFS_INODE_CLUSTER_SIZE(mp)) +
                xfs_calc_buf_res(1, 0) +
                xfs_calc_buf_res(2 + XFS_IALLOC_BLOCKS(mp) +
                                 mp->m_in_maxlevels, 0) +
-- 
1.8.3.2