Merge tag 'realtime-groups-6.13_2024-11-05' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into staging-merge

xfs: shard the realtime section [v5.5 06/10]

Right now, the realtime section uses a single pair of metadata inodes to
store the free space information.  This presents a scalability problem
since every thread trying to allocate or free rt extents have to lock
these files.  Solve this problem by sharding the realtime section into
separate realtime allocation groups.

While we're at it, define a superblock to be stamped into the start of
the rt section.  This enables utilities such as blkid to identify block
devices containing realtime sections, and avoids the situation where
anything written into block 0 of the realtime extent can be
misinterpreted as file data.

The best advantage for rtgroups will become evident later when we get to
adding rmap and reflink to the realtime volume, since the geometry
constraints are the same for rt groups and AGs.  Hence we can reuse all
that code directly.

This is a very large patchset, but it catches us up with 20 years of
technical debt that have accumulated.

With a bit of luck, this should all go splendidly.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>

1 files changed, 115 insertions, 35 deletions

diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
index caa02caac2cd..7b16cdd72e9d 100644
--- a/fs/xfs/xfs_trace.h
+++ b/fs/xfs/xfs_trace.h
@@ -97,6 +97,7 @@ struct xfs_extent_free_item;
 struct xfs_rmap_intent;
 struct xfs_refcount_intent;
 struct xfs_metadir_update;
+struct xfs_rtgroup;
 
 #define XFS_ATTR_FILTER_FLAGS \
 	{ XFS_ATTR_ROOT,	"ROOT" }, \
@@ -1748,6 +1749,80 @@ TRACE_EVENT(xfs_extent_busy_trim,
 		  __entry->tlen)
 );
 
+#ifdef CONFIG_XFS_RT
+TRACE_EVENT(xfs_rtalloc_extent_busy,
+	TP_PROTO(struct xfs_rtgroup *rtg, xfs_rtxnum_t start,
+		 xfs_rtxlen_t minlen, xfs_rtxlen_t maxlen,
+		 xfs_rtxlen_t len, xfs_rtxlen_t prod, xfs_rtxnum_t rtx,
+		 unsigned busy_gen),
+	TP_ARGS(rtg, start, minlen, maxlen, len, prod, rtx, busy_gen),
+	TP_STRUCT__entry(
+		__field(dev_t, dev)
+		__field(xfs_rgnumber_t, rgno)
+		__field(xfs_rtxnum_t, start)
+		__field(xfs_rtxlen_t, minlen)
+		__field(xfs_rtxlen_t, maxlen)
+		__field(xfs_rtxlen_t, mod)
+		__field(xfs_rtxlen_t, prod)
+		__field(xfs_rtxlen_t, len)
+		__field(xfs_rtxnum_t, rtx)
+		__field(unsigned, busy_gen)
+	),
+	TP_fast_assign(
+		__entry->dev = rtg_mount(rtg)->m_super->s_dev;
+		__entry->rgno = rtg_rgno(rtg);
+		__entry->start = start;
+		__entry->minlen = minlen;
+		__entry->maxlen = maxlen;
+		__entry->prod = prod;
+		__entry->len = len;
+		__entry->rtx = rtx;
+		__entry->busy_gen = busy_gen;
+	),
+	TP_printk("dev %d:%d rgno 0x%x startrtx 0x%llx minlen %u maxlen %u "
+		  "prod %u len %u rtx 0%llx busy_gen 0x%x",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->rgno,
+		  __entry->start,
+		  __entry->minlen,
+		  __entry->maxlen,
+		  __entry->prod,
+		  __entry->len,
+		  __entry->rtx,
+		  __entry->busy_gen)
+)
+
+TRACE_EVENT(xfs_rtalloc_extent_busy_trim,
+	TP_PROTO(struct xfs_rtgroup *rtg, xfs_rtxnum_t old_rtx,
+		 xfs_rtxlen_t old_len, xfs_rtxnum_t new_rtx,
+		 xfs_rtxlen_t new_len),
+	TP_ARGS(rtg, old_rtx, old_len, new_rtx, new_len),
+	TP_STRUCT__entry(
+		__field(dev_t, dev)
+		__field(xfs_rgnumber_t, rgno)
+		__field(xfs_rtxnum_t, old_rtx)
+		__field(xfs_rtxnum_t, new_rtx)
+		__field(xfs_rtxlen_t, old_len)
+		__field(xfs_rtxlen_t, new_len)
+	),
+	TP_fast_assign(
+		__entry->dev = rtg_mount(rtg)->m_super->s_dev;
+		__entry->rgno = rtg_rgno(rtg);
+		__entry->old_rtx = old_rtx;
+		__entry->old_len = old_len;
+		__entry->new_rtx = new_rtx;
+		__entry->new_len = new_len;
+	),
+	TP_printk("dev %d:%d rgno 0x%x rtx 0x%llx rtxcount 0x%x -> rtx 0x%llx rtxcount 0x%x",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __entry->rgno,
+		  __entry->old_rtx,
+		  __entry->old_len,
+		  __entry->new_rtx,
+		  __entry->new_len)
+);
+#endif /* CONFIG_XFS_RT */
+
 DECLARE_EVENT_CLASS(xfs_agf_class,
 	TP_PROTO(struct xfs_mount *mp, struct xfs_agf *agf, int flags,
 		 unsigned long caller_ip),
@@ -2479,23 +2554,26 @@ DEFINE_LOG_RECOVER_ICREATE_ITEM(xfs_log_recover_icreate_cancel);
 DEFINE_LOG_RECOVER_ICREATE_ITEM(xfs_log_recover_icreate_recover);
 
 DECLARE_EVENT_CLASS(xfs_discard_class,
-	TP_PROTO(const struct xfs_perag *pag, xfs_agblock_t agbno,
+	TP_PROTO(const struct xfs_group *xg, xfs_agblock_t agbno,
 		 xfs_extlen_t len),
-	TP_ARGS(pag, agbno, len),
+	TP_ARGS(xg, agbno, len),
 	TP_STRUCT__entry(
 		__field(dev_t, dev)
+		__field(enum xfs_group_type, type)
 		__field(xfs_agnumber_t, agno)
 		__field(xfs_agblock_t, agbno)
 		__field(xfs_extlen_t, len)
 	),
 	TP_fast_assign(
-		__entry->dev = pag_mount(pag)->m_super->s_dev;
-		__entry->agno = pag_agno(pag);
+		__entry->dev = xg->xg_mount->m_super->s_dev;
+		__entry->type = xg->xg_type;
+		__entry->agno = xg->xg_gno;
 		__entry->agbno = agbno;
 		__entry->len = len;
 	),
-	TP_printk("dev %d:%d agno 0x%x agbno 0x%x fsbcount 0x%x",
+	TP_printk("dev %d:%d %sno 0x%x gbno 0x%x fsbcount 0x%x",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __print_symbolic(__entry->type, XG_TYPE_STRINGS),
 		  __entry->agno,
 		  __entry->agbno,
 		  __entry->len)
@@ -2503,9 +2581,9 @@ DECLARE_EVENT_CLASS(xfs_discard_class,
 
 #define DEFINE_DISCARD_EVENT(name) \
 DEFINE_EVENT(xfs_discard_class, name, \
-	TP_PROTO(const struct xfs_perag *pag, xfs_agblock_t agbno, \
-		xfs_extlen_t len), \
-	TP_ARGS(pag, agbno, len))
+	TP_PROTO(const struct xfs_group *xg, xfs_agblock_t agbno, \
+		 xfs_extlen_t len), \
+	TP_ARGS(xg, agbno, len))
 DEFINE_DISCARD_EVENT(xfs_discard_extent);
 DEFINE_DISCARD_EVENT(xfs_discard_toosmall);
 DEFINE_DISCARD_EVENT(xfs_discard_exclude);
@@ -2765,6 +2843,7 @@ DECLARE_EVENT_CLASS(xfs_free_extent_deferred_class,
 	TP_ARGS(mp, free),
 	TP_STRUCT__entry(
 		__field(dev_t, dev)
+		__field(enum xfs_group_type, type)
 		__field(xfs_agnumber_t, agno)
 		__field(xfs_agblock_t, agbno)
 		__field(xfs_extlen_t, len)
@@ -2772,13 +2851,16 @@ DECLARE_EVENT_CLASS(xfs_free_extent_deferred_class,
 	),
 	TP_fast_assign(
 		__entry->dev = mp->m_super->s_dev;
-		__entry->agno = XFS_FSB_TO_AGNO(mp, free->xefi_startblock);
-		__entry->agbno = XFS_FSB_TO_AGBNO(mp, free->xefi_startblock);
+		__entry->type = free->xefi_group->xg_type;
+		__entry->agno = free->xefi_group->xg_gno;
+		__entry->agbno = xfs_fsb_to_gbno(mp, free->xefi_startblock,
+						free->xefi_group->xg_type);
 		__entry->len = free->xefi_blockcount;
 		__entry->flags = free->xefi_flags;
 	),
-	TP_printk("dev %d:%d agno 0x%x agbno 0x%x fsbcount 0x%x flags 0x%x",
+	TP_printk("dev %d:%d %sno 0x%x gbno 0x%x fsbcount 0x%x flags 0x%x",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		  __print_symbolic(__entry->type, XG_TYPE_STRINGS),
 		  __entry->agno,
 		  __entry->agbno,
 		  __entry->len,
@@ -2788,7 +2870,6 @@ DECLARE_EVENT_CLASS(xfs_free_extent_deferred_class,
 DEFINE_EVENT(xfs_free_extent_deferred_class, name, \
 	TP_PROTO(struct xfs_mount *mp, struct xfs_extent_free_item *free), \
 	TP_ARGS(mp, free))
-DEFINE_FREE_EXTENT_DEFERRED_EVENT(xfs_agfl_free_defer);
 DEFINE_FREE_EXTENT_DEFERRED_EVENT(xfs_agfl_free_deferred);
 DEFINE_FREE_EXTENT_DEFERRED_EVENT(xfs_extent_free_defer);
 DEFINE_FREE_EXTENT_DEFERRED_EVENT(xfs_extent_free_deferred);
@@ -3085,11 +3166,10 @@ DECLARE_EVENT_CLASS(xfs_bmap_deferred_class,
 	TP_ARGS(bi),
 	TP_STRUCT__entry(
 		__field(dev_t, dev)
-		__field(dev_t, opdev)
+		__field(enum xfs_group_type, type)
 		__field(xfs_agnumber_t, agno)
 		__field(xfs_ino_t, ino)
-		__field(xfs_agblock_t, agbno)
-		__field(xfs_fsblock_t, rtbno)
+		__field(unsigned long long, gbno)
 		__field(int, whichfork)
 		__field(xfs_fileoff_t, l_loff)
 		__field(xfs_filblks_t, l_len)
@@ -3098,20 +3178,25 @@ DECLARE_EVENT_CLASS(xfs_bmap_deferred_class,
 	),
 	TP_fast_assign(
 		struct xfs_inode	*ip = bi->bi_owner;
+		struct xfs_mount	*mp = ip->i_mount;
 
-		__entry->dev = ip->i_mount->m_super->s_dev;
-		if (xfs_ifork_is_realtime(ip, bi->bi_whichfork)) {
-			__entry->agno = 0;
-			__entry->agbno = 0;
-			__entry->rtbno = bi->bi_bmap.br_startblock;
-			__entry->opdev = ip->i_mount->m_rtdev_targp->bt_dev;
+		__entry->dev = mp->m_super->s_dev;
+		__entry->type = bi->bi_group->xg_type;
+		__entry->agno = bi->bi_group->xg_gno;
+		if (bi->bi_group->xg_type == XG_TYPE_RTG &&
+		    !xfs_has_rtgroups(mp)) {
+			/*
+			 * Legacy rt filesystems do not have allocation groups
+			 * ondisk.  We emulate this incore with one gigantic
+			 * rtgroup whose size can exceed a 32-bit block number.
+			 * For this tracepoint, we report group 0 and a 64-bit
+			 * group block number.
+			 */
+			__entry->gbno = bi->bi_bmap.br_startblock;
 		} else {
-			__entry->agno = XFS_FSB_TO_AGNO(ip->i_mount,
-						bi->bi_bmap.br_startblock);
-			__entry->agbno = XFS_FSB_TO_AGBNO(ip->i_mount,
-						bi->bi_bmap.br_startblock);
-			__entry->rtbno = 0;
-			__entry->opdev = __entry->dev;
+			__entry->gbno = xfs_fsb_to_gbno(mp,
+						bi->bi_bmap.br_startblock,
+						bi->bi_group->xg_type);
 		}
 		__entry->ino = ip->i_ino;
 		__entry->whichfork = bi->bi_whichfork;
@@ -3120,14 +3205,13 @@ DECLARE_EVENT_CLASS(xfs_bmap_deferred_class,
 		__entry->l_state = bi->bi_bmap.br_state;
 		__entry->op = bi->bi_type;
 	),
-	TP_printk("dev %d:%d op %s opdev %d:%d ino 0x%llx agno 0x%x agbno 0x%x rtbno 0x%llx %s fileoff 0x%llx fsbcount 0x%llx state %d",
+	TP_printk("dev %d:%d op %s ino 0x%llx %sno 0x%x gbno 0x%llx %s fileoff 0x%llx fsbcount 0x%llx state %d",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
 		  __print_symbolic(__entry->op, XFS_BMAP_INTENT_STRINGS),
-		  MAJOR(__entry->opdev), MINOR(__entry->opdev),
 		  __entry->ino,
+		  __print_symbolic(__entry->type, XG_TYPE_STRINGS),
 		  __entry->agno,
-		  __entry->agbno,
-		  __entry->rtbno,
+		  __entry->gbno,
 		  __print_symbolic(__entry->whichfork, XFS_WHICHFORK_STRINGS),
 		  __entry->l_loff,
 		  __entry->l_len,
@@ -4259,10 +4343,6 @@ DEFINE_FS_CORRUPT_EVENT(xfs_fs_mark_sick);
 DEFINE_FS_CORRUPT_EVENT(xfs_fs_mark_corrupt);
 DEFINE_FS_CORRUPT_EVENT(xfs_fs_mark_healthy);
 DEFINE_FS_CORRUPT_EVENT(xfs_fs_unfixed_corruption);
-DEFINE_FS_CORRUPT_EVENT(xfs_rt_mark_sick);
-DEFINE_FS_CORRUPT_EVENT(xfs_rt_mark_corrupt);
-DEFINE_FS_CORRUPT_EVENT(xfs_rt_mark_healthy);
-DEFINE_FS_CORRUPT_EVENT(xfs_rt_unfixed_corruption);
 
 DECLARE_EVENT_CLASS(xfs_group_corrupt_class,
 	TP_PROTO(const struct xfs_group *xg, unsigned int flags),