xfs: add support for zoned space reservations

For zoned file systems garbage collection (GC) has to take the iolock
and mmaplock after moving data to a new place to synchronize with
readers.  This means waiting for garbage collection with the iolock can
deadlock.

To avoid this, the worst case required blocks have to be reserved before
taking the iolock, which is done using a new RTAVAILABLE counter that
tracks blocks that are free to write into and don't require garbage
collection.  The new helpers try to take these available blocks, and
if there aren't enough available it wakes and waits for GC.  This is
done using a list of on-stack reservations to ensure fairness.

Co-developed-by: Hans Holmberg <hans.holmberg@wdc.com>
Signed-off-by: Hans Holmberg <hans.holmberg@wdc.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>

Author: Christoph Hellwig

fs/xfs/Makefile

@@ -138,7 +138,8 @@ xfs-$(CONFIG_XFS_QUOTA)		+= xfs_dquot.o \
 
 # xfs_rtbitmap is shared with libxfs
 xfs-$(CONFIG_XFS_RT)		+= xfs_rtalloc.o \
-				   xfs_zone_alloc.o
+				   xfs_zone_alloc.o \
+				   xfs_zone_space_resv.o
 
 xfs-$(CONFIG_XFS_POSIX_ACL)	+= xfs_acl.o
 xfs-$(CONFIG_SYSCTL)		+= xfs_sysctl.o

fs/xfs/libxfs/xfs_bmap.c

@@ -40,6 +40,7 @@
 #include "xfs_symlink_remote.h"
 #include "xfs_inode_util.h"
 #include "xfs_rtgroup.h"
+#include "xfs_zone_alloc.h"
 
 struct kmem_cache		*xfs_bmap_intent_cache;
 
@@ -4788,12 +4789,18 @@ xfs_bmap_del_extent_delay(
 	da_diff = da_old - da_new;
 	fdblocks = da_diff;
 
-	if (bflags & XFS_BMAPI_REMAP)
+	if (bflags & XFS_BMAPI_REMAP) {
 		;
-	else if (isrt)
-		xfs_add_frextents(mp, xfs_blen_to_rtbxlen(mp, del->br_blockcount));
-	else
+	} else if (isrt) {
+		xfs_rtbxlen_t	rtxlen;
+
+		rtxlen = xfs_blen_to_rtbxlen(mp, del->br_blockcount);
+		if (xfs_is_zoned_inode(ip))
+			xfs_zoned_add_available(mp, rtxlen);
+		xfs_add_frextents(mp, rtxlen);
+	} else {
 		fdblocks += del->br_blockcount;
+	}
 
 	xfs_add_fdblocks(mp, fdblocks);
 	xfs_mod_delalloc(ip, -(int64_t)del->br_blockcount, -da_diff);

fs/xfs/libxfs/xfs_types.h

@@ -244,12 +244,22 @@ enum xfs_free_counter {
 	 */
 	XC_FREE_RTEXTENTS,
 
+	/*
+	 * Number of available for use RT extents.
+	 *
+	 * This counter only exists for zoned RT device and indicates the number
+	 * of RT extents that can be directly used by writes.  XC_FREE_RTEXTENTS
+	 * also includes blocks that have been written previously and freed, but
+	 * sit in a rtgroup that still needs a zone reset.
+	 */
+	XC_FREE_RTAVAILABLE,
 	XC_FREE_NR,
 };
 
 #define XFS_FREECOUNTER_STR \
 	{ XC_FREE_BLOCKS,		"blocks" }, \
-	{ XC_FREE_RTEXTENTS,		"rtextents" }
+	{ XC_FREE_RTEXTENTS,		"rtextents" }, \
+	{ XC_FREE_RTAVAILABLE,		"rtavailable" }
 
 /*
  * Type verifier functions

fs/xfs/xfs_mount.c

@@ -465,29 +465,35 @@ xfs_mount_reset_sbqflags(
 static const char *const xfs_free_pool_name[] = {
 	[XC_FREE_BLOCKS]	= "free blocks",
 	[XC_FREE_RTEXTENTS]	= "free rt extents",
+	[XC_FREE_RTAVAILABLE]	= "available rt extents",
 };
 
 uint64_t
 xfs_default_resblks(
 	struct xfs_mount	*mp,
 	enum xfs_free_counter	ctr)
 {
-	uint64_t resblks;
-
-	if (ctr == XC_FREE_RTEXTENTS)
+	switch (ctr) {
+	case XC_FREE_BLOCKS:
+		/*
+		 * Default to 5% or 8192 FSBs of space reserved, whichever is
+		 * smaller.
+		 *
+		 * This is intended to cover concurrent allocation transactions
+		 * when we initially hit ENOSPC.  These each require a 4 block
+		 * reservation. Hence by default we cover roughly 2000
+		 * concurrent allocation reservations.
+		 */
+		return min(div_u64(mp->m_sb.sb_dblocks, 20), 8192ULL);
+	case XC_FREE_RTEXTENTS:
+	case XC_FREE_RTAVAILABLE:
+		if (IS_ENABLED(CONFIG_XFS_RT) && xfs_has_zoned(mp))
+			return xfs_zoned_default_resblks(mp, ctr);
 		return 0;
-
-	/*
-	 * We default to 5% or 8192 fsbs of space reserved, whichever is
-	 * smaller.  This is intended to cover concurrent allocation
-	 * transactions when we initially hit enospc. These each require a 4
-	 * block reservation. Hence by default we cover roughly 2000 concurrent
-	 * allocation reservations.
-	 */
-	resblks = mp->m_sb.sb_dblocks;
-	do_div(resblks, 20);
-	resblks = min_t(uint64_t, resblks, 8192);
-	return resblks;
+	default:
+		ASSERT(0);
+		return 0;
+	}
 }
 
 /* Ensure the summary counts are correct. */

fs/xfs/xfs_trace.h

@@ -363,6 +363,28 @@ DEFINE_EVENT(xfs_zone_alloc_class, name,			\
 	TP_ARGS(oz, rgbno, len))
 DEFINE_ZONE_ALLOC_EVENT(xfs_zone_record_blocks);
 DEFINE_ZONE_ALLOC_EVENT(xfs_zone_alloc_blocks);
+
+TRACE_EVENT(xfs_zones_mount,
+	TP_PROTO(struct xfs_mount *mp),
+	TP_ARGS(mp),
+	TP_STRUCT__entry(
+		__field(dev_t, dev)
+		__field(xfs_rgnumber_t, rgcount)
+		__field(uint32_t, blocks)
+		__field(unsigned int, max_open_zones)
+	),
+	TP_fast_assign(
+		__entry->dev = mp->m_super->s_dev;
+		__entry->rgcount = mp->m_sb.sb_rgcount;
+		__entry->blocks = mp->m_groups[XG_TYPE_RTG].blocks;
+		__entry->max_open_zones = mp->m_max_open_zones;
+	),
+	TP_printk("dev %d:%d zoned %u blocks_per_zone %u, max_open %u",
+		  MAJOR(__entry->dev), MINOR(__entry->dev),
+		__entry->rgcount,
+		__entry->blocks,
+		__entry->max_open_zones)
+);
 #endif /* CONFIG_XFS_RT */
 
 TRACE_EVENT(xfs_inodegc_worker,
@@ -5767,6 +5789,7 @@ TRACE_EVENT(xfs_growfs_check_rtgeom,
 
 TRACE_DEFINE_ENUM(XC_FREE_BLOCKS);
 TRACE_DEFINE_ENUM(XC_FREE_RTEXTENTS);
+TRACE_DEFINE_ENUM(XC_FREE_RTAVAILABLE);
 
 DECLARE_EVENT_CLASS(xfs_freeblocks_resv_class,
 	TP_PROTO(struct xfs_mount *mp, enum xfs_free_counter ctr,

fs/xfs/xfs_zone_alloc.c

@@ -922,6 +922,7 @@ xfs_mount_zones(
 	xfs_info(mp, "%u zones of %u blocks size (%u max open)",
 		 mp->m_sb.sb_rgcount, mp->m_groups[XG_TYPE_RTG].blocks,
 		 mp->m_max_open_zones);
+	trace_xfs_zones_mount(mp);
 
 	if (bdev_is_zoned(bt->bt_bdev)) {
 		error = blkdev_report_zones(bt->bt_bdev,
@@ -939,6 +940,7 @@ xfs_mount_zones(
 		}
 	}
 
+	xfs_set_freecounter(mp, XC_FREE_RTAVAILABLE, iz.available);
 	xfs_set_freecounter(mp, XC_FREE_RTEXTENTS,
 			iz.available + iz.reclaimable);
 	return 0;

fs/xfs/xfs_zone_alloc.h

@@ -5,6 +5,30 @@
 struct iomap_ioend;
 struct xfs_open_zone;
 
+struct xfs_zone_alloc_ctx {
+	struct xfs_open_zone	*open_zone;
+	xfs_filblks_t		reserved_blocks;
+};
+
+/*
+ * Grab any available space, even if it is less than what the caller asked for.
+ */
+#define XFS_ZR_GREEDY		(1U << 0)
+/*
+ * Only grab instantly available space, don't wait or GC.
+ */
+#define XFS_ZR_NOWAIT		(1U << 1)
+/*
+ * Dip into the reserved pool.
+ */
+#define XFS_ZR_RESERVED		(1U << 2)
+
+int xfs_zoned_space_reserve(struct xfs_inode *ip, xfs_filblks_t count_fsb,
+		unsigned int flags, struct xfs_zone_alloc_ctx *ac);
+void xfs_zoned_space_unreserve(struct xfs_inode *ip,
+		struct xfs_zone_alloc_ctx *ac);
+void xfs_zoned_add_available(struct xfs_mount *mp, xfs_filblks_t count_fsb);
+
 void xfs_zone_alloc_and_submit(struct iomap_ioend *ioend,
 		struct xfs_open_zone **oz);
 int xfs_zone_free_blocks(struct xfs_trans *tp, struct xfs_rtgroup *rtg,
@@ -18,6 +42,9 @@ void xfs_zoned_wake_all(struct xfs_mount *mp);
 bool xfs_zone_rgbno_is_valid(struct xfs_rtgroup *rtg, xfs_rgnumber_t rgbno);
 void xfs_mark_rtg_boundary(struct iomap_ioend *ioend);
 
+uint64_t xfs_zoned_default_resblks(struct xfs_mount *mp,
+		enum xfs_free_counter ctr);
+
 #ifdef CONFIG_XFS_RT
 int xfs_mount_zones(struct xfs_mount *mp);
 void xfs_unmount_zones(struct xfs_mount *mp);

fs/xfs/xfs_zone_priv.h

@@ -86,4 +86,6 @@ struct xfs_zone_info {
 
 struct xfs_open_zone *xfs_open_zone(struct xfs_mount *mp, bool is_gc);
 
+void xfs_zoned_resv_wake_all(struct xfs_mount *mp);
+
 #endif /* _XFS_ZONE_PRIV_H */