Merge tag 'xfs-fixes-6.16-rc7' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux

Pull xfs fixes from Carlos Maiolino:
 "This contains mostly code clean up, refactoring and comments
  modification.

  The most important patch in this series is the last one that removes
  an unnecessary data structure allocation of xfs busy extents which
  might lead to a memory leak on the zoned allocator code"

* tag 'xfs-fixes-6.16-rc7' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
  xfs: don't allocate the xfs_extent_busy structure for zoned RTGs
  xfs: remove the bt_bdev_file buftarg field
  xfs: rename the bt_bdev_* buftarg fields
  xfs: refactor xfs_calc_atomic_write_unit_max
  xfs: add a xfs_group_type_buftarg helper
  xfs: remove the call to sync_blockdev in xfs_configure_buftarg
  xfs: clean up the initial read logic in xfs_readsb
  xfs: replace strncpy with memcpy in xattr listing

Author: torvalds

fs/xfs/libxfs/xfs_group.c

@@ -163,7 +163,8 @@ xfs_group_free(
 
 	xfs_defer_drain_free(&xg->xg_intents_drain);
 #ifdef __KERNEL__
-	kfree(xg->xg_busy_extents);
+	if (xfs_group_has_extent_busy(xg->xg_mount, xg->xg_type))
+		kfree(xg->xg_busy_extents);
 #endif
 
 	if (uninit)
@@ -189,9 +190,11 @@ xfs_group_insert(
 	xg->xg_type = type;
 
 #ifdef __KERNEL__
-	xg->xg_busy_extents = xfs_extent_busy_alloc();
-	if (!xg->xg_busy_extents)
-		return -ENOMEM;
+	if (xfs_group_has_extent_busy(mp, type)) {
+		xg->xg_busy_extents = xfs_extent_busy_alloc();
+		if (!xg->xg_busy_extents)
+			return -ENOMEM;
+	}
 	spin_lock_init(&xg->xg_state_lock);
 	xfs_hooks_init(&xg->xg_rmap_update_hooks);
 #endif
@@ -210,7 +213,8 @@ xfs_group_insert(
 out_drain:
 	xfs_defer_drain_free(&xg->xg_intents_drain);
 #ifdef __KERNEL__
-	kfree(xg->xg_busy_extents);
+	if (xfs_group_has_extent_busy(xg->xg_mount, xg->xg_type))
+		kfree(xg->xg_busy_extents);
 #endif
 	return error;
 }

fs/xfs/xfs_buf.c

@@ -1683,7 +1683,7 @@ xfs_free_buftarg(
 	fs_put_dax(btp->bt_daxdev, btp->bt_mount);
 	/* the main block device is closed by kill_block_super */
 	if (btp->bt_bdev != btp->bt_mount->m_super->s_bdev)
-		bdev_fput(btp->bt_bdev_file);
+		bdev_fput(btp->bt_file);
 	kfree(btp);
 }
 
@@ -1712,8 +1712,8 @@ xfs_configure_buftarg_atomic_writes(
 		max_bytes = 0;
 	}
 
-	btp->bt_bdev_awu_min = min_bytes;
-	btp->bt_bdev_awu_max = max_bytes;
+	btp->bt_awu_min = min_bytes;
+	btp->bt_awu_max = max_bytes;
 }
 
 /* Configure a buffer target that abstracts a block device. */
@@ -1738,14 +1738,9 @@ xfs_configure_buftarg(
 		return -EINVAL;
 	}
 
-	/*
-	 * Flush the block device pagecache so our bios see anything dirtied
-	 * before mount.
-	 */
 	if (bdev_can_atomic_write(btp->bt_bdev))
 		xfs_configure_buftarg_atomic_writes(btp);
-
-	return sync_blockdev(btp->bt_bdev);
+	return 0;
 }
 
 int
@@ -1803,7 +1798,7 @@ xfs_alloc_buftarg(
 	btp = kzalloc(sizeof(*btp), GFP_KERNEL | __GFP_NOFAIL);
 
 	btp->bt_mount = mp;
-	btp->bt_bdev_file = bdev_file;
+	btp->bt_file = bdev_file;
 	btp->bt_bdev = file_bdev(bdev_file);
 	btp->bt_dev = btp->bt_bdev->bd_dev;
 	btp->bt_daxdev = fs_dax_get_by_bdev(btp->bt_bdev, &btp->bt_dax_part_off,

fs/xfs/xfs_buf.h

@@ -94,7 +94,6 @@ void xfs_buf_cache_destroy(struct xfs_buf_cache *bch);
  */
 struct xfs_buftarg {
 	dev_t			bt_dev;
-	struct file		*bt_bdev_file;
 	struct block_device	*bt_bdev;
 	struct dax_device	*bt_daxdev;
 	struct file		*bt_file;
@@ -112,9 +111,9 @@ struct xfs_buftarg {
 	struct percpu_counter	bt_readahead_count;
 	struct ratelimit_state	bt_ioerror_rl;
 
-	/* Atomic write unit values, bytes */
-	unsigned int		bt_bdev_awu_min;
-	unsigned int		bt_bdev_awu_max;
+	/* Hardware atomic write unit values, bytes */
+	unsigned int		bt_awu_min;
+	unsigned int		bt_awu_max;
 
 	/* built-in cache, if we're not using the perag one */
 	struct xfs_buf_cache	bt_cache[];
@@ -375,7 +374,6 @@ extern void xfs_buftarg_wait(struct xfs_buftarg *);
 extern void xfs_buftarg_drain(struct xfs_buftarg *);
 int xfs_configure_buftarg(struct xfs_buftarg *btp, unsigned int sectorsize);
 
-#define xfs_getsize_buftarg(buftarg)	block_size((buftarg)->bt_bdev)
 #define xfs_readonly_buftarg(buftarg)	bdev_read_only((buftarg)->bt_bdev)
 
 int xfs_buf_reverify(struct xfs_buf *bp, const struct xfs_buf_ops *ops);

fs/xfs/xfs_discard.c

@@ -103,24 +103,6 @@ xfs_discard_endio(
 	bio_put(bio);
 }
 
-static inline struct block_device *
-xfs_group_bdev(
-	const struct xfs_group	*xg)
-{
-	struct xfs_mount	*mp = xg->xg_mount;
-
-	switch (xg->xg_type) {
-	case XG_TYPE_AG:
-		return mp->m_ddev_targp->bt_bdev;
-	case XG_TYPE_RTG:
-		return mp->m_rtdev_targp->bt_bdev;
-	default:
-		ASSERT(0);
-		break;
-	}
-	return NULL;
-}
-
 /*
  * Walk the discard list and issue discards on all the busy extents in the
  * list. We plug and chain the bios so that we only need a single completion
@@ -138,11 +120,14 @@ xfs_discard_extents(
 
 	blk_start_plug(&plug);
 	list_for_each_entry(busyp, &extents->extent_list, list) {
-		trace_xfs_discard_extent(busyp->group, busyp->bno,
-				busyp->length);
+		struct xfs_group	*xg = busyp->group;
+		struct xfs_buftarg	*btp =
+			xfs_group_type_buftarg(xg->xg_mount, xg->xg_type);
+
+		trace_xfs_discard_extent(xg, busyp->bno, busyp->length);
 
-		error = __blkdev_issue_discard(xfs_group_bdev(busyp->group),
-				xfs_gbno_to_daddr(busyp->group, busyp->bno),
+		error = __blkdev_issue_discard(btp->bt_bdev,
+				xfs_gbno_to_daddr(xg, busyp->bno),
 				XFS_FSB_TO_BB(mp, busyp->length),
 				GFP_KERNEL, &bio);
 		if (error && error != -EOPNOTSUPP) {

fs/xfs/xfs_extent_busy.h

@@ -68,4 +68,12 @@ static inline void xfs_extent_busy_sort(struct list_head *list)
 	list_sort(NULL, list, xfs_extent_busy_ag_cmp);
 }
 
+/*
+ * Zoned RTGs don't need to track busy extents, as the actual block freeing only
+ * happens by a zone reset, which forces out all transactions that touched the
+ * to be reset zone first.
+ */
+#define xfs_group_has_extent_busy(mp, type) \
+	((type) == XG_TYPE_AG || !xfs_has_zoned((mp)))
+
 #endif /* __XFS_EXTENT_BUSY_H__ */

fs/xfs/xfs_file.c

@@ -752,7 +752,7 @@ xfs_file_dio_write_atomic(
 	 * HW offload should be faster, so try that first if it is already
 	 * known that the write length is not too large.
 	 */
-	if (ocount > xfs_inode_buftarg(ip)->bt_bdev_awu_max)
+	if (ocount > xfs_inode_buftarg(ip)->bt_awu_max)
 		dops = &xfs_atomic_write_cow_iomap_ops;
 	else
 		dops = &xfs_direct_write_iomap_ops;

fs/xfs/xfs_inode.h

@@ -358,7 +358,7 @@ static inline bool xfs_inode_has_bigrtalloc(const struct xfs_inode *ip)
 
 static inline bool xfs_inode_can_hw_atomic_write(const struct xfs_inode *ip)
 {
-	return xfs_inode_buftarg(ip)->bt_bdev_awu_max > 0;
+	return xfs_inode_buftarg(ip)->bt_awu_max > 0;
 }
 
 /*

fs/xfs/xfs_iomap.c

@@ -827,7 +827,7 @@ xfs_bmap_hw_atomic_write_possible(
 	/*
 	 * The ->iomap_begin caller should ensure this, but check anyway.
 	 */
-	return len <= xfs_inode_buftarg(ip)->bt_bdev_awu_max;
+	return len <= xfs_inode_buftarg(ip)->bt_awu_max;
 }
 
 static int

fs/xfs/xfs_iops.c

@@ -665,7 +665,7 @@ xfs_get_atomic_write_max_opt(
 	 * less than our out of place write limit, but we don't want to exceed
 	 * the awu_max.
 	 */
-	return min(awu_max, xfs_inode_buftarg(ip)->bt_bdev_awu_max);
+	return min(awu_max, xfs_inode_buftarg(ip)->bt_awu_max);
 }
 
 static void

fs/xfs/xfs_mount.c

@@ -171,19 +171,16 @@ xfs_readsb(
 	ASSERT(mp->m_ddev_targp != NULL);
 
 	/*
-	 * For the initial read, we must guess at the sector
-	 * size based on the block device.  It's enough to
-	 * get the sb_sectsize out of the superblock and
-	 * then reread with the proper length.
-	 * We don't verify it yet, because it may not be complete.
+	 * In the first pass, use the device sector size to just read enough
+	 * of the superblock to extract the XFS sector size.
+	 *
+	 * The device sector size must be smaller than or equal to the XFS
+	 * sector size and thus we can always read the superblock.  Once we know
+	 * the XFS sector size, re-read it and run the buffer verifier.
 	 */
-	sector_size = xfs_getsize_buftarg(mp->m_ddev_targp);
+	sector_size = mp->m_ddev_targp->bt_logical_sectorsize;
 	buf_ops = NULL;
 
-	/*
-	 * Allocate a (locked) buffer to hold the superblock. This will be kept
-	 * around at all times to optimize access to the superblock.
-	 */
 reread:
 	error = xfs_buf_read_uncached(mp->m_ddev_targp, XFS_SB_DADDR,
 				      BTOBB(sector_size), &bp, buf_ops);
@@ -247,6 +244,10 @@ xfs_readsb(
 	/* no need to be quiet anymore, so reset the buf ops */
 	bp->b_ops = &xfs_sb_buf_ops;
 
+	/*
+	 * Keep a pointer of the sb buffer around instead of caching it in the
+	 * buffer cache because we access it frequently.
+	 */
 	mp->m_sb_bp = bp;
 	xfs_buf_unlock(bp);
 	return 0;
@@ -678,68 +679,46 @@ static inline unsigned int max_pow_of_two_factor(const unsigned int nr)
 }
 
 /*
- * If the data device advertises atomic write support, limit the size of data
- * device atomic writes to the greatest power-of-two factor of the AG size so
- * that every atomic write unit aligns with the start of every AG.  This is
- * required so that the per-AG allocations for an atomic write will always be
+ * If the underlying device advertises atomic write support, limit the size of
+ * atomic writes to the greatest power-of-two factor of the group size so
+ * that every atomic write unit aligns with the start of every group.  This is
+ * required so that the allocations for an atomic write will always be
  * aligned compatibly with the alignment requirements of the storage.
  *
- * If the data device doesn't advertise atomic writes, then there are no
- * alignment restrictions and the largest out-of-place write we can do
- * ourselves is the number of blocks that user files can allocate from any AG.
- */
-static inline xfs_extlen_t xfs_calc_perag_awu_max(struct xfs_mount *mp)
-{
-	if (mp->m_ddev_targp->bt_bdev_awu_min > 0)
-		return max_pow_of_two_factor(mp->m_sb.sb_agblocks);
-	return rounddown_pow_of_two(mp->m_ag_max_usable);
-}
-
-/*
- * Reflink on the realtime device requires rtgroups, and atomic writes require
- * reflink.
- *
- * If the realtime device advertises atomic write support, limit the size of
- * data device atomic writes to the greatest power-of-two factor of the rtgroup
- * size so that every atomic write unit aligns with the start of every rtgroup.
- * This is required so that the per-rtgroup allocations for an atomic write
- * will always be aligned compatibly with the alignment requirements of the
- * storage.
- *
- * If the rt device doesn't advertise atomic writes, then there are no
- * alignment restrictions and the largest out-of-place write we can do
- * ourselves is the number of blocks that user files can allocate from any
- * rtgroup.
+ * If the device doesn't advertise atomic writes, then there are no alignment
+ * restrictions and the largest out-of-place write we can do ourselves is the
+ * number of blocks that user files can allocate from any group.
  */
-static inline xfs_extlen_t xfs_calc_rtgroup_awu_max(struct xfs_mount *mp)
+static xfs_extlen_t
+xfs_calc_group_awu_max(
+	struct xfs_mount	*mp,
+	enum xfs_group_type	type)
 {
-	struct xfs_groups	*rgs = &mp->m_groups[XG_TYPE_RTG];
+	struct xfs_groups	*g = &mp->m_groups[type];
+	struct xfs_buftarg	*btp = xfs_group_type_buftarg(mp, type);
 
-	if (rgs->blocks == 0)
+	if (g->blocks == 0)
 		return 0;
-	if (mp->m_rtdev_targp && mp->m_rtdev_targp->bt_bdev_awu_min > 0)
-		return max_pow_of_two_factor(rgs->blocks);
-	return rounddown_pow_of_two(rgs->blocks);
+	if (btp && btp->bt_awu_min > 0)
+		return max_pow_of_two_factor(g->blocks);
+	return rounddown_pow_of_two(g->blocks);
 }
 
 /* Compute the maximum atomic write unit size for each section. */
 static inline void
 xfs_calc_atomic_write_unit_max(
-	struct xfs_mount	*mp)
+	struct xfs_mount	*mp,
+	enum xfs_group_type	type)
 {
-	struct xfs_groups	*ags = &mp->m_groups[XG_TYPE_AG];
-	struct xfs_groups	*rgs = &mp->m_groups[XG_TYPE_RTG];
+	struct xfs_groups	*g = &mp->m_groups[type];
 
 	const xfs_extlen_t	max_write = xfs_calc_atomic_write_max(mp);
 	const xfs_extlen_t	max_ioend = xfs_reflink_max_atomic_cow(mp);
-	const xfs_extlen_t	max_agsize = xfs_calc_perag_awu_max(mp);
-	const xfs_extlen_t	max_rgsize = xfs_calc_rtgroup_awu_max(mp);
-
-	ags->awu_max = min3(max_write, max_ioend, max_agsize);
-	rgs->awu_max = min3(max_write, max_ioend, max_rgsize);
+	const xfs_extlen_t	max_gsize = xfs_calc_group_awu_max(mp, type);
 
-	trace_xfs_calc_atomic_write_unit_max(mp, max_write, max_ioend,
-			max_agsize, max_rgsize);
+	g->awu_max = min3(max_write, max_ioend, max_gsize);
+	trace_xfs_calc_atomic_write_unit_max(mp, type, max_write, max_ioend,
+			max_gsize, g->awu_max);
 }
 
 /*
@@ -757,7 +736,8 @@ xfs_set_max_atomic_write_opt(
 		max(mp->m_groups[XG_TYPE_AG].blocks,
 		    mp->m_groups[XG_TYPE_RTG].blocks);
 	const xfs_extlen_t	max_group_write =
-		max(xfs_calc_perag_awu_max(mp), xfs_calc_rtgroup_awu_max(mp));
+		max(xfs_calc_group_awu_max(mp, XG_TYPE_AG),
+		    xfs_calc_group_awu_max(mp, XG_TYPE_RTG));
 	int			error;
 
 	if (new_max_bytes == 0)
@@ -813,7 +793,8 @@ xfs_set_max_atomic_write_opt(
 		return error;
 	}
 
-	xfs_calc_atomic_write_unit_max(mp);
+	xfs_calc_atomic_write_unit_max(mp, XG_TYPE_AG);
+	xfs_calc_atomic_write_unit_max(mp, XG_TYPE_RTG);
 	mp->m_awu_max_bytes = new_max_bytes;
 	return 0;
 }