From 0d8627cc936de8ea04f3cc1e6921c63fb72cc199 Mon Sep 17 00:00:00 2001
From: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Date: Wed, 22 Oct 2025 13:41:06 +0200
Subject: blktrace: add definitions for blk_user_trace_setup2

Add definitions for a version 2 of the blk_user_trace_setup ioctl. This
new ioctl will enable a different struct layout of the binary data passed
to user-space when using a new version of the blktrace utility requesting
the new struct layout.

Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/blktrace_api.h | 16 ++++++++++++++++
 include/uapi/linux/fs.h           |  1 +
 2 files changed, 17 insertions(+)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/blktrace_api.h b/include/uapi/linux/blktrace_api.h
index 1bfb635e309b..a6958708d477 100644
--- a/include/uapi/linux/blktrace_api.h
+++ b/include/uapi/linux/blktrace_api.h
@@ -129,6 +129,7 @@ enum {
 };
 
 #define BLKTRACE_BDEV_SIZE	32
+#define BLKTRACE_BDEV_SIZE2	64
 
 /*
  * User setup structure passed with BLKTRACESETUP
@@ -143,4 +144,19 @@ struct blk_user_trace_setup {
 	__u32 pid;
 };
 
+/*
+ * User setup structure passed with BLKTRACESETUP2
+ */
+struct blk_user_trace_setup2 {
+	char name[BLKTRACE_BDEV_SIZE2];		/* output */
+	__u64 act_mask;				/* input */
+	__u32 buf_size;				/* input */
+	__u32 buf_nr;				/* input */
+	__u64 start_lba;
+	__u64 end_lba;
+	__u32 pid;
+	__u32 flags;		/* currently unused */
+	__u64 reserved[11];
+};
+
 #endif /* _UAPIBLKTRACE_H */
diff --git a/include/uapi/linux/fs.h b/include/uapi/linux/fs.h
index beb4c2d1e41c..957ce3343a4f 100644
--- a/include/uapi/linux/fs.h
+++ b/include/uapi/linux/fs.h
@@ -300,6 +300,7 @@ struct file_attr {
 #define BLKGETDISKSEQ _IOR(0x12,128,__u64)
 /* 130-136 are used by zoned block device ioctls (uapi/linux/blkzoned.h) */
 /* 137-141 are used by blk-crypto ioctls (uapi/linux/blk-crypto.h) */
+#define BLKTRACESETUP2 _IOWR(0x12, 142, struct blk_user_trace_setup2)
 
 #define BMAP_IOCTL 1		/* obsolete - kept for compatibility */
 #define FIBMAP	   _IO(0x00,1)	/* bmap access */
-- 
cgit v1.2.3


From c44347d606260f36a81f6d8415a5af33cb3015fa Mon Sep 17 00:00:00 2001
From: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Date: Wed, 22 Oct 2025 13:41:08 +0200
Subject: blktrace: add definitions for struct blk_io_trace2

Add definitions for the extended version of the blktrace protocol using a
wider action type to be able to record new actions in the kernel.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/blktrace_api.h | 16 ++++++++++++++++
 kernel/trace/blktrace.c           |  1 +
 2 files changed, 17 insertions(+)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/blktrace_api.h b/include/uapi/linux/blktrace_api.h
index a6958708d477..9f9834d76e00 100644
--- a/include/uapi/linux/blktrace_api.h
+++ b/include/uapi/linux/blktrace_api.h
@@ -94,6 +94,7 @@ enum blktrace_notify {
 
 #define BLK_IO_TRACE_MAGIC	0x65617400
 #define BLK_IO_TRACE_VERSION	0x07
+#define BLK_IO_TRACE2_VERSION	0x08
 
 /*
  * The trace itself
@@ -113,6 +114,21 @@ struct blk_io_trace {
 	/* cgroup id will be stored here if exists */
 };
 
+struct blk_io_trace2 {
+	__u32 magic;		/* MAGIC << 8 | BLK_IO_TRACE2_VERSION */
+	__u32 sequence;		/* event number */
+	__u64 time;		/* in nanoseconds */
+	__u64 sector;		/* disk offset */
+	__u32 bytes;		/* transfer length */
+	__u32 pid;		/* who did it */
+	__u64 action;		/* what happened */
+	__u32 device;		/* device number */
+	__u32 cpu;		/* on what cpu did it happen */
+	__u16 error;		/* completion error */
+	__u16 pdu_len;		/* length of data after this trace */
+	__u8 pad[12];
+	/* cgroup id will be stored here if it exists */
+};
 /*
  * The remap event
  */
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index d1532df84cc8..185f19c9f772 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -1616,6 +1616,7 @@ static int __init init_blk_tracer(void)
 
 	BUILD_BUG_ON(__alignof__(struct blk_user_trace_setup2) %
 		     __alignof__(long));
+	BUILD_BUG_ON(__alignof__(struct blk_io_trace2) % __alignof__(long));
 
 	return 0;
 }
-- 
cgit v1.2.3


From f9ee38bbf70fb20584625849a253c8652176fa66 Mon Sep 17 00:00:00 2001
From: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Date: Wed, 22 Oct 2025 13:41:12 +0200
Subject: blktrace: add block trace commands for zone operations

Add block trace commands for zone operations. These commands can only be
handled with version 2 of the blktrace protocol. For version 1, warn if a
command that does not fit into the 16 bits reserved for the command in
this version is passed in.

Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/blktrace_api.h | 13 +++++++++++--
 kernel/trace/blktrace.c           | 29 +++++++++++++++++++++++++----
 2 files changed, 36 insertions(+), 6 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/blktrace_api.h b/include/uapi/linux/blktrace_api.h
index 9f9834d76e00..190a3c5ab0a0 100644
--- a/include/uapi/linux/blktrace_api.h
+++ b/include/uapi/linux/blktrace_api.h
@@ -26,11 +26,20 @@ enum blktrace_cat {
 	BLK_TC_DRV_DATA	= 1 << 14,	/* binary per-driver data */
 	BLK_TC_FUA	= 1 << 15,	/* fua requests */
 
-	BLK_TC_END	= 1 << 15,	/* we've run out of bits! */
+	BLK_TC_END_V1	= 1 << 15,	/* we've run out of bits! */
+
+	BLK_TC_ZONE_APPEND	= 1ull << 16,  	/* zone append */
+	BLK_TC_ZONE_RESET	= 1ull << 17,	/* zone reset */
+	BLK_TC_ZONE_RESET_ALL	= 1ull << 18,	/* zone reset all */
+	BLK_TC_ZONE_FINISH	= 1ull << 19,	/* zone finish */
+	BLK_TC_ZONE_OPEN	= 1ull << 20,	/* zone open */
+	BLK_TC_ZONE_CLOSE	= 1ull << 21,	/* zone close */
+
+	BLK_TC_END_V2		= 1ull << 21,
 };
 
 #define BLK_TC_SHIFT		(16)
-#define BLK_TC_ACT(act)		((act) << BLK_TC_SHIFT)
+#define BLK_TC_ACT(act)		((u64)(act) << BLK_TC_SHIFT)
 
 /*
  * Basic trace actions
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index 49f73cb3cb33..fb5935885abc 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -163,8 +163,8 @@ static void relay_blktrace_event(struct blk_trace *bt, unsigned long sequence,
 					     bytes, what, error, cgid, cgid_len,
 					     pdu_data, pdu_len);
 	return relay_blktrace_event1(bt, sequence, pid, cpu, sector, bytes,
-				     lower_32_bits(what), error, cgid, cgid_len,
-				     pdu_data, pdu_len);
+				     what, error, cgid, cgid_len, pdu_data,
+				     pdu_len);
 }
 
 /*
@@ -342,10 +342,32 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
 	case REQ_OP_FLUSH:
 		what |= BLK_TC_ACT(BLK_TC_FLUSH);
 		break;
+	case REQ_OP_ZONE_APPEND:
+		what |= BLK_TC_ACT(BLK_TC_ZONE_APPEND);
+		break;
+	case REQ_OP_ZONE_RESET:
+		what |= BLK_TC_ACT(BLK_TC_ZONE_RESET);
+		break;
+	case REQ_OP_ZONE_RESET_ALL:
+		what |= BLK_TC_ACT(BLK_TC_ZONE_RESET_ALL);
+		break;
+	case REQ_OP_ZONE_FINISH:
+		what |= BLK_TC_ACT(BLK_TC_ZONE_FINISH);
+		break;
+	case REQ_OP_ZONE_OPEN:
+		what |= BLK_TC_ACT(BLK_TC_ZONE_OPEN);
+		break;
+	case REQ_OP_ZONE_CLOSE:
+		what |= BLK_TC_ACT(BLK_TC_ZONE_CLOSE);
+		break;
 	default:
 		break;
 	}
 
+	if (WARN_ON_ONCE(bt->version == 1 &&
+		     (what >> BLK_TC_SHIFT) > BLK_TC_END_V1))
+		return;
+
 	if (cgid)
 		what |= __BLK_TA_CGROUP;
 
@@ -386,8 +408,7 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
 	sequence = per_cpu_ptr(bt->sequence, cpu);
 	(*sequence)++;
 	relay_blktrace_event(bt, *sequence, pid, cpu, sector, bytes,
-			     lower_32_bits(what), error, cgid, cgid_len,
-			     pdu_data, pdu_len);
+			     what, error, cgid, cgid_len, pdu_data, pdu_len);
 	local_irq_restore(flags);
 }
 
-- 
cgit v1.2.3


From 1c164fcc1b08e75f1cad1532718f09cddc0ddebe Mon Sep 17 00:00:00 2001
From: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Date: Wed, 22 Oct 2025 13:41:13 +0200
Subject: blktrace: expose ZONE APPEND completions to blktrace

Expose ZONE APPEND completions as a block trace completion action to
blktrace.

As tracing of zoned block commands needs the upper 32bit of the widened
64bit action, only add traces to blktrace if user-space has requested
version 2 of the blktrace protocol.

Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/blktrace_api.h |  3 +++
 kernel/trace/blktrace.c           | 21 +++++++++++++++++++++
 2 files changed, 24 insertions(+)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/blktrace_api.h b/include/uapi/linux/blktrace_api.h
index 190a3c5ab0a0..289872e51fc5 100644
--- a/include/uapi/linux/blktrace_api.h
+++ b/include/uapi/linux/blktrace_api.h
@@ -97,6 +97,9 @@ enum blktrace_notify {
 #define BLK_TA_ABORT		(__BLK_TA_ABORT | BLK_TC_ACT(BLK_TC_QUEUE))
 #define BLK_TA_DRV_DATA	(__BLK_TA_DRV_DATA | BLK_TC_ACT(BLK_TC_DRV_DATA))
 
+#define BLK_TA_ZONE_APPEND	(__BLK_TA_COMPLETE |\
+				 BLK_TC_ACT(BLK_TC_ZONE_APPEND))
+
 #define BLK_TN_PROCESS		(__BLK_TN_PROCESS | BLK_TC_ACT(BLK_TC_NOTIFY))
 #define BLK_TN_TIMESTAMP	(__BLK_TN_TIMESTAMP | BLK_TC_ACT(BLK_TC_NOTIFY))
 #define BLK_TN_MESSAGE		(__BLK_TN_MESSAGE | BLK_TC_ACT(BLK_TC_NOTIFY))
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index fb5935885abc..c83577096607 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -978,6 +978,22 @@ static void blk_add_trace_rq_complete(void *ignore, struct request *rq,
 			 blk_trace_request_get_cgid(rq));
 }
 
+static void blk_add_trace_zone_update_request(void *ignore, struct request *rq)
+{
+	struct blk_trace *bt;
+
+	rcu_read_lock();
+	bt = rcu_dereference(rq->q->blk_trace);
+	if (likely(!bt) || bt->version < 2) {
+		rcu_read_unlock();
+		return;
+	}
+	rcu_read_unlock();
+
+	blk_add_trace_rq(rq, 0, blk_rq_bytes(rq), BLK_TA_ZONE_APPEND,
+			 blk_trace_request_get_cgid(rq));
+}
+
 /**
  * blk_add_trace_bio - Add a trace for a bio oriented action
  * @q:		queue the io is for
@@ -1208,6 +1224,9 @@ static void blk_register_tracepoints(void)
 	WARN_ON(ret);
 	ret = register_trace_block_getrq(blk_add_trace_getrq, NULL);
 	WARN_ON(ret);
+	ret = register_trace_blk_zone_append_update_request_bio(
+		blk_add_trace_zone_update_request, NULL);
+	WARN_ON(ret);
 	ret = register_trace_block_plug(blk_add_trace_plug, NULL);
 	WARN_ON(ret);
 	ret = register_trace_block_unplug(blk_add_trace_unplug, NULL);
@@ -1227,6 +1246,8 @@ static void blk_unregister_tracepoints(void)
 	unregister_trace_block_split(blk_add_trace_split, NULL);
 	unregister_trace_block_unplug(blk_add_trace_unplug, NULL);
 	unregister_trace_block_plug(blk_add_trace_plug, NULL);
+	unregister_trace_blk_zone_append_update_request_bio(
+		blk_add_trace_zone_update_request, NULL);
 	unregister_trace_block_getrq(blk_add_trace_getrq, NULL);
 	unregister_trace_block_bio_queue(blk_add_trace_bio_queue, NULL);
 	unregister_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge, NULL);
-- 
cgit v1.2.3


From 3f6722816a73e2017599d965683dbe71833afd7a Mon Sep 17 00:00:00 2001
From: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Date: Wed, 22 Oct 2025 13:41:14 +0200
Subject: blktrace: trace zone write plugging operations

Trace zone write plugging operations on block devices.

As tracing of zoned block commands needs the upper 32bit of the widened
64bit action, only add traces to blktrace if user-space has requested
version 2 of the blktrace protocol.

Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/blktrace_api.h |  5 +++++
 kernel/trace/blktrace.c           | 39 +++++++++++++++++++++++++++++++++++++++
 2 files changed, 44 insertions(+)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/blktrace_api.h b/include/uapi/linux/blktrace_api.h
index 289872e51fc5..30f3d2589365 100644
--- a/include/uapi/linux/blktrace_api.h
+++ b/include/uapi/linux/blktrace_api.h
@@ -62,6 +62,8 @@ enum blktrace_act {
 	__BLK_TA_REMAP,			/* bio was remapped */
 	__BLK_TA_ABORT,			/* request aborted */
 	__BLK_TA_DRV_DATA,		/* driver-specific binary data */
+	__BLK_TA_ZONE_PLUG,		/* zone write plug was plugged */
+	__BLK_TA_ZONE_UNPLUG,		/* zone write plug was unplugged */
 	__BLK_TA_CGROUP = 1 << 8,	/* from a cgroup*/
 };
 
@@ -99,6 +101,9 @@ enum blktrace_notify {
 
 #define BLK_TA_ZONE_APPEND	(__BLK_TA_COMPLETE |\
 				 BLK_TC_ACT(BLK_TC_ZONE_APPEND))
+#define BLK_TA_ZONE_PLUG	(__BLK_TA_ZONE_PLUG | BLK_TC_ACT(BLK_TC_QUEUE))
+#define BLK_TA_ZONE_UNPLUG	(__BLK_TA_ZONE_UNPLUG |\
+				 BLK_TC_ACT(BLK_TC_QUEUE))
 
 #define BLK_TN_PROCESS		(__BLK_TN_PROCESS | BLK_TC_ACT(BLK_TC_NOTIFY))
 #define BLK_TN_TIMESTAMP	(__BLK_TN_TIMESTAMP | BLK_TC_ACT(BLK_TC_NOTIFY))
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index c83577096607..6bfe1b36a1d3 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -1084,6 +1084,37 @@ static void blk_add_trace_unplug(void *ignore, struct request_queue *q,
 	rcu_read_unlock();
 }
 
+static void blk_add_trace_zone_plug(void *ignore, struct request_queue *q,
+				    unsigned int zno, sector_t sector,
+				    unsigned int sectors)
+{
+	struct blk_trace *bt;
+
+	rcu_read_lock();
+	bt = rcu_dereference(q->blk_trace);
+	if (bt && bt->version >= 2)
+		__blk_add_trace(bt, sector, sectors << SECTOR_SHIFT, 0,
+				BLK_TA_ZONE_PLUG, 0, 0, NULL, 0);
+	rcu_read_unlock();
+
+	return;
+}
+
+static void blk_add_trace_zone_unplug(void *ignore, struct request_queue *q,
+				      unsigned int zno, sector_t sector,
+				      unsigned int sectors)
+{
+	struct blk_trace *bt;
+
+	rcu_read_lock();
+	bt = rcu_dereference(q->blk_trace);
+	if (bt && bt->version >= 2)
+		__blk_add_trace(bt, sector, sectors << SECTOR_SHIFT, 0,
+				BLK_TA_ZONE_UNPLUG, 0, 0, NULL, 0);
+	rcu_read_unlock();
+	return;
+}
+
 static void blk_add_trace_split(void *ignore, struct bio *bio, unsigned int pdu)
 {
 	struct request_queue *q = bio->bi_bdev->bd_disk->queue;
@@ -1227,6 +1258,12 @@ static void blk_register_tracepoints(void)
 	ret = register_trace_blk_zone_append_update_request_bio(
 		blk_add_trace_zone_update_request, NULL);
 	WARN_ON(ret);
+	ret = register_trace_disk_zone_wplug_add_bio(blk_add_trace_zone_plug,
+						     NULL);
+	WARN_ON(ret);
+	ret = register_trace_blk_zone_wplug_bio(blk_add_trace_zone_unplug,
+						NULL);
+	WARN_ON(ret);
 	ret = register_trace_block_plug(blk_add_trace_plug, NULL);
 	WARN_ON(ret);
 	ret = register_trace_block_unplug(blk_add_trace_unplug, NULL);
@@ -1246,6 +1283,8 @@ static void blk_unregister_tracepoints(void)
 	unregister_trace_block_split(blk_add_trace_split, NULL);
 	unregister_trace_block_unplug(blk_add_trace_unplug, NULL);
 	unregister_trace_block_plug(blk_add_trace_plug, NULL);
+	unregister_trace_blk_zone_wplug_bio(blk_add_trace_zone_unplug, NULL);
+	unregister_trace_disk_zone_wplug_add_bio(blk_add_trace_zone_plug, NULL);
 	unregister_trace_blk_zone_append_update_request_bio(
 		blk_add_trace_zone_update_request, NULL);
 	unregister_trace_block_getrq(blk_add_trace_getrq, NULL);
-- 
cgit v1.2.3


From bc49af56eea866c34d21bf582f65b02fc8c06ec3 Mon Sep 17 00:00:00 2001
From: Chaitanya Kulkarni <ckulkarnilinux@gmail.com>
Date: Tue, 28 Oct 2025 20:34:23 -0700
Subject: blktrace: add support for REQ_OP_WRITE_ZEROES tracing
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Currently, REQ_OP_WRITE_ZEROES operations are not handled in the
blktrace infrastructure, resulting in incorrect or missing operation
labels in ftrace blktrace output. This manifests as write-zeroes
operations appearing with incorrect labels like "N" instead of a
proper "WZ" designation.

This patch adds complete support for REQ_OP_WRITE_ZEROES across the
blktrace infrastructure:

Add BLK_TC_WRITE_ZEROES trace category in blktrace_api.h and update
BLK_TC_END_V2 marker accordingly
Map REQ_OP_WRITE_ZEROES to BLK_TC_WRITE_ZEROES in __blk_add_trace()
to ensure proper trace event categorization
Update fill_rwbs() to generate "WZ" label for write-zeroes operations
in ftrace output, making them easily identifiable
Add "write-zeroes" string mapping in act_to_str array for debugfs
filter interface
Update blk_fill_rwbs() to handle REQ_OP_WRITE_ZEROES for block layer
event tracing

With this fix, write-zeroes operations are now correctly traced and
displayed.

===========================================================
BEFORE THIS PATCH
===========================================================
blkdiscard -z -o 0 -l 40960 /dev/nvme0n1
   blkdiscard-3809 [030] .....  1212.253701: block_bio_queue: 259,0 NS 0 + 80 [blkdiscard]
   blkdiscard-3809 [030] .....  1212.253703: block_getrq: 259,0 NS 0 + 80 [blkdiscard]
   blkdiscard-3809 [030] .....  1212.253704: block_io_start: 259,0 NS 40960 () 0 + 80 be,0,4 [blkdiscard]
   blkdiscard-3809 [030] .....  1212.253704: block_plug: [blkdiscard]
   blkdiscard-3809 [030] .....  1212.253706: block_unplug: [blkdiscard] 1
   blkdiscard-3809 [030] .....  1212.253706: block_rq_insert: 259,0 NS 40960 () 0 + 80 be,0,4 [blkdiscard]
kworker/30:1H-566  [030] .....  1212.253726: block_rq_issue: 259,0 NS 40960 () 0 + 80 be,0,4 [kworker/30:1H]
       <idle>-0    [030] d.h1.  1212.253957: block_rq_complete: 259,0 NS () 0 + 80 be,0,4 [0]
       <idle>-0    [030] dNh1.  1212.253960: block_io_done: 259,0 NS 0 () 0 + 0 none,0,0 [swapper/30]

Trace Event Breakdown:
 Event             | Device | Op  | Sector | Sectors | Byte Size | Calculation

 block_bio_queue   | 259,0  | NS  | 0      | 80      | -         | 80 × 512 = 40,960
 block_getrq       | 259,0  | NS  | 0      | 80      | -         | 80 × 512 = 40,960
 block_io_start    | 259,0  | NS  | 0      | 80      | 40960     | Direct from trace
 block_rq_insert   | 259,0  | NS  | 0      | 80      | 40960     | Direct from trace
 block_rq_issue    | 259,0  | NS  | 0      | 80      | 40960     | Direct from trace
 block_rq_complete | 259,0  | NS  | 0      | 80      | -         | 80 × 512 = 40,960
 block_io_done     | 259,0  | NS  | 0      | 0       | 0         | Completion (no data)

  Total Bytes Transferred: Sectors: 80 Bytes: 80 × 512 = 40,960 bytes

===========================================================
AFTER THIS PATCH
===========================================================
blkdiscard -z -o 0 -l 40960 /dev/nvme0n1

   blkdiscard-2477 [020] .....   960.989131: block_bio_queue: 259,0 WZS 0 + 80 [blkdiscard]
   blkdiscard-2477 [020] .....   960.989134: block_getrq: 259,0 WZS 0 + 80 [blkdiscard]
   blkdiscard-2477 [020] .....   960.989135: block_io_start: 259,0 WZS 40960 () 0 + 80 be,0,4 [blkdiscard]
   blkdiscard-2477 [020] .....   960.989138: block_plug: [blkdiscard]
   blkdiscard-2477 [020] .....   960.989140: block_unplug: [blkdiscard] 1
   blkdiscard-2477 [020] .....   960.989141: block_rq_insert: 259,0 WZS 40960 () 0 + 80 be,0,4 [blkdiscard]
kworker/20:1H-736  [020] .....   960.989166: block_rq_issue: 259,0 WZS 40960 () 0 + 80 be,0,4 [kworker/20:1H]
       <idle>-0    [020] d.h1.   960.989476: block_rq_complete: 259,0 WZS () 0 + 80 be,0,4 [0]
       <idle>-0    [020] dNh1.   960.989482: block_io_done: 259,0 WZS 0 () 0 + 0 none,0,0 [swapper/20]

Trace Event Breakdown:
 Event             | Device | Op  | Sector | Sectors | Byte Size | Calculation

 block_bio_queue   | 259,0  | WZS | 0      | 80      | -         | 80 × 512 = 40,960
 block_getrq       | 259,0  | WZS | 0      | 80      | -         | 80 × 512 = 40,960
 block_io_start    | 259,0  | WZS | 0      | 80      | 40960     | Direct from trace
 block_rq_insert   | 259,0  | WZS | 0      | 80      | 40960     | Direct from trace
 block_rq_issue    | 259,0  | WZS | 0      | 80      | 40960     | Direct from trace
 block_rq_complete | 259,0  | WZS | 0      | 80      | -         | 80 × 512 = 40,960
 block_io_done     | 259,0  | WZS | 0      | 0       | 0         | Completion (no data)

  Total Bytes Transferred: Sectors: 80 Bytes: 80 × 512 = 40,960 bytes

Tested with ftrace blktrace on NVMe devices using blkdiscard with
the -z (write-zeroes) flag.

Signed-off-by: Chaitanya Kulkarni <ckulkarnilinux@gmail.com>
Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 include/uapi/linux/blktrace_api.h |  4 +++-
 kernel/trace/blktrace.c           | 13 ++++++++++++-
 2 files changed, 15 insertions(+), 2 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/include/uapi/linux/blktrace_api.h b/include/uapi/linux/blktrace_api.h
index 30f3d2589365..7c092d9f3aa4 100644
--- a/include/uapi/linux/blktrace_api.h
+++ b/include/uapi/linux/blktrace_api.h
@@ -35,7 +35,9 @@ enum blktrace_cat {
 	BLK_TC_ZONE_OPEN	= 1ull << 20,	/* zone open */
 	BLK_TC_ZONE_CLOSE	= 1ull << 21,	/* zone close */
 
-	BLK_TC_END_V2		= 1ull << 21,
+	BLK_TC_WRITE_ZEROES	= 1ull << 22,	/* write-zeroes */
+
+	BLK_TC_END_V2		= 1ull << 22,
 };
 
 #define BLK_TC_SHIFT		(16)
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index e4f26ddb7ee2..af8cbc8e1a7c 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -360,6 +360,9 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
 	case REQ_OP_ZONE_CLOSE:
 		what |= BLK_TC_ACT(BLK_TC_ZONE_CLOSE);
 		break;
+	case REQ_OP_WRITE_ZEROES:
+		what |= BLK_TC_ACT(BLK_TC_WRITE_ZEROES);
+		break;
 	default:
 		break;
 	}
@@ -1408,7 +1411,10 @@ static void fill_rwbs(char *rwbs, const struct blk_io_trace2 *t)
 
 	if (tc & BLK_TC_DISCARD)
 		rwbs[i++] = 'D';
-	else if (tc & BLK_TC_WRITE)
+	else if (tc & BLK_TC_WRITE_ZEROES) {
+		rwbs[i++] = 'W';
+		rwbs[i++] = 'Z';
+	} else if (tc & BLK_TC_WRITE)
 		rwbs[i++] = 'W';
 	else if (t->bytes)
 		rwbs[i++] = 'R';
@@ -1951,6 +1957,7 @@ static const struct {
 	{ BLK_TC_DISCARD,	"discard"	},
 	{ BLK_TC_DRV_DATA,	"drv_data"	},
 	{ BLK_TC_FUA,		"fua"		},
+	{ BLK_TC_WRITE_ZEROES,	"write-zeroes"	},
 };
 
 static int blk_trace_str2mask(const char *str)
@@ -2164,6 +2171,10 @@ void blk_fill_rwbs(char *rwbs, blk_opf_t opf)
 		rwbs[i++] = 'Z';
 		rwbs[i++] = 'C';
 		break;
+	case REQ_OP_WRITE_ZEROES:
+		rwbs[i++] = 'W';
+		rwbs[i++] = 'Z';
+		break;
 	default:
 		rwbs[i++] = 'N';
 	}
-- 
cgit v1.2.3


From 0bf0e2e4666822b62d7ad6473dc37fd6b377b5f1 Mon Sep 17 00:00:00 2001
From: Damien Le Moal <dlemoal@kernel.org>
Date: Wed, 5 Nov 2025 06:22:41 +0900
Subject: block: track zone conditions

The function blk_revalidate_zone_cond() already caches the condition of
all zones of a zoned block device in the zones_cond array of a gendisk.
However, the zone conditions are updated only when the device is scanned
or revalidated.

Implement tracking of the runtime changes to zone conditions using
the new cond field in struct blk_zone_wplug. The size of this structure
remains 112 Bytes as the new field replaces the 4 Bytes padding at the
end of the structure.

Beause zones that do not have a zone write plug can be in the empty,
implicit open, explicit open or full condition, the zones_cond array of
a disk is used to track the conditions, of zones that do not have a zone
write plug. The condition of such zone is updated in the disk zones_cond
array when a zone reset, reset all or finish operation is executed, and
also when a zone write plug is removed from the disk hash table when the
zone becomes full.

Since a device may automatically close an implicitly open zone when
writing to an empty or closed zone, if the total number of open zones
has reached the device limit, the BLK_ZONE_COND_IMP_OPEN and
BLK_ZONE_COND_CLOSED zone conditions cannot be precisely tracked. To
overcome this, the zone condition BLK_ZONE_COND_ACTIVE is introduced to
represent a zone that has the condition BLK_ZONE_COND_IMP_OPEN,
BLK_ZONE_COND_EXP_OPEN or BLK_ZONE_COND_CLOSED.  This follows the
definition of an active zone as defined in the NVMe Zoned Namespace
specifications. As such, for a zoned device that has a limit on the
maximum number of open zones, we will never have more zones in the
BLK_ZONE_COND_ACTIVE condition than the device limit. This is compatible
with the SCSI ZBC and ATA ZAC specifications for SMR HDDs as these
devices do not have a limit on the number of active zones.

The function disk_zone_wplug_set_wp_offset() is modified to use the new
helper disk_zone_wplug_update_cond() to update a zone write plug
condition whenever a zone write plug write offset is updated on
submission or merging of write BIOs to a zone.

The functions blk_zone_reset_bio_endio(), blk_zone_reset_all_bio_endio()
and blk_zone_finish_bio_endio() are modified to update the condition of
the zones targeted by reset, reset_all and finish operations, either
using though disk_zone_wplug_set_wp_offset() for zones that have a
zone write plug, or using the disk_zone_set_cond() helper to update the
zones_cond array of the disk for zones that do not have a zone write
plug.

When a zone write plug is removed from the disk hash table (when the
zone becomes empty or full), the condition of struct blk_zone_wplug is
used to update the disk zones_cond array. Conversely, when a zone write
plug is added to the disk hash table, the zones_cond array is used to
initialize the zone write plug condition.

Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-zoned.c             | 120 +++++++++++++++++++++++++++++++++++++++---
 include/uapi/linux/blkzoned.h |  11 ++++
 2 files changed, 125 insertions(+), 6 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/block/blk-zoned.c b/block/blk-zoned.c
index f62862274f9a..c5fa303093a9 100644
--- a/block/blk-zoned.c
+++ b/block/blk-zoned.c
@@ -33,6 +33,7 @@ static const char *const zone_cond_name[] = {
 	ZONE_COND_NAME(READONLY),
 	ZONE_COND_NAME(FULL),
 	ZONE_COND_NAME(OFFLINE),
+	ZONE_COND_NAME(ACTIVE),
 };
 #undef ZONE_COND_NAME
 
@@ -57,6 +58,7 @@ static const char *const zone_cond_name[] = {
  * @zone_no: The number of the zone the plug is managing.
  * @wp_offset: The zone write pointer location relative to the start of the zone
  *             as a number of 512B sectors.
+ * @cond: Condition of the zone
  */
 struct blk_zone_wplug {
 	struct hlist_node	node;
@@ -69,6 +71,7 @@ struct blk_zone_wplug {
 	unsigned int		flags;
 	unsigned int		zone_no;
 	unsigned int		wp_offset;
+	enum blk_zone_cond	cond;
 };
 
 static inline unsigned int disk_zone_wplugs_hash_size(struct gendisk *disk)
@@ -114,6 +117,57 @@ const char *blk_zone_cond_str(enum blk_zone_cond zone_cond)
 }
 EXPORT_SYMBOL_GPL(blk_zone_cond_str);
 
+static void blk_zone_set_cond(u8 *zones_cond, unsigned int zno,
+			      enum blk_zone_cond cond)
+{
+	if (!zones_cond)
+		return;
+
+	switch (cond) {
+	case BLK_ZONE_COND_IMP_OPEN:
+	case BLK_ZONE_COND_EXP_OPEN:
+	case BLK_ZONE_COND_CLOSED:
+		zones_cond[zno] = BLK_ZONE_COND_ACTIVE;
+		return;
+	case BLK_ZONE_COND_NOT_WP:
+	case BLK_ZONE_COND_EMPTY:
+	case BLK_ZONE_COND_FULL:
+	case BLK_ZONE_COND_OFFLINE:
+	case BLK_ZONE_COND_READONLY:
+	default:
+		zones_cond[zno] = cond;
+		return;
+	}
+}
+
+static void disk_zone_set_cond(struct gendisk *disk, sector_t sector,
+			       enum blk_zone_cond cond)
+{
+	u8 *zones_cond;
+
+	rcu_read_lock();
+	zones_cond = rcu_dereference(disk->zones_cond);
+	if (zones_cond) {
+		unsigned int zno = disk_zone_no(disk, sector);
+
+		/*
+		 * The condition of a conventional, readonly and offline zones
+		 * never changes, so do nothing if the target zone is in one of
+		 * these conditions.
+		 */
+		switch (zones_cond[zno]) {
+		case BLK_ZONE_COND_NOT_WP:
+		case BLK_ZONE_COND_READONLY:
+		case BLK_ZONE_COND_OFFLINE:
+			break;
+		default:
+			blk_zone_set_cond(zones_cond, zno, cond);
+			break;
+		}
+	}
+	rcu_read_unlock();
+}
+
 /**
  * bdev_zone_is_seq - check if a sector belongs to a sequential write zone
  * @bdev:       block device to check
@@ -416,6 +470,7 @@ static bool disk_insert_zone_wplug(struct gendisk *disk,
 {
 	struct blk_zone_wplug *zwplg;
 	unsigned long flags;
+	u8 *zones_cond;
 	unsigned int idx =
 		hash_32(zwplug->zone_no, disk->zone_wplugs_hash_bits);
 
@@ -431,6 +486,20 @@ static bool disk_insert_zone_wplug(struct gendisk *disk,
 			return false;
 		}
 	}
+
+	/*
+	 * Set the zone condition: if we do not yet have a zones_cond array
+	 * attached to the disk, then this is a zone write plug insert from the
+	 * first call to blk_revalidate_disk_zones(), in which case the zone is
+	 * necessarilly in the active condition.
+	 */
+	zones_cond = rcu_dereference_check(disk->zones_cond,
+				lockdep_is_held(&disk->zone_wplugs_lock));
+	if (zones_cond)
+		zwplug->cond = zones_cond[zwplug->zone_no];
+	else
+		zwplug->cond = BLK_ZONE_COND_ACTIVE;
+
 	hlist_add_head_rcu(&zwplug->node, &disk->zone_wplugs_hash[idx]);
 	atomic_inc(&disk->nr_zone_wplugs);
 	spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
@@ -530,10 +599,15 @@ static void disk_remove_zone_wplug(struct gendisk *disk,
 
 	/*
 	 * Mark the zone write plug as unhashed and drop the extra reference we
-	 * took when the plug was inserted in the hash table.
+	 * took when the plug was inserted in the hash table. Also update the
+	 * disk zone condition array with the current condition of the zone
+	 * write plug.
 	 */
 	zwplug->flags |= BLK_ZONE_WPLUG_UNHASHED;
 	spin_lock_irqsave(&disk->zone_wplugs_lock, flags);
+	blk_zone_set_cond(rcu_dereference_check(disk->zones_cond,
+				lockdep_is_held(&disk->zone_wplugs_lock)),
+			  zwplug->zone_no, zwplug->cond);
 	hlist_del_init_rcu(&zwplug->node);
 	atomic_dec(&disk->nr_zone_wplugs);
 	spin_unlock_irqrestore(&disk->zone_wplugs_lock, flags);
@@ -635,6 +709,22 @@ static void disk_zone_wplug_abort(struct blk_zone_wplug *zwplug)
 		blk_zone_wplug_bio_io_error(zwplug, bio);
 }
 
+/*
+ * Update a zone write plug condition based on the write pointer offset.
+ */
+static void disk_zone_wplug_update_cond(struct gendisk *disk,
+					struct blk_zone_wplug *zwplug)
+{
+	lockdep_assert_held(&zwplug->lock);
+
+	if (disk_zone_wplug_is_full(disk, zwplug))
+		zwplug->cond = BLK_ZONE_COND_FULL;
+	else if (!zwplug->wp_offset)
+		zwplug->cond = BLK_ZONE_COND_EMPTY;
+	else
+		zwplug->cond = BLK_ZONE_COND_ACTIVE;
+}
+
 /*
  * Set a zone write plug write pointer offset to the specified value.
  * This aborts all plugged BIOs, which is fine as this function is called for
@@ -650,6 +740,8 @@ static void disk_zone_wplug_set_wp_offset(struct gendisk *disk,
 	/* Update the zone write pointer and abort all plugged BIOs. */
 	zwplug->flags &= ~BLK_ZONE_WPLUG_NEED_WP_UPDATE;
 	zwplug->wp_offset = wp_offset;
+	disk_zone_wplug_update_cond(disk, zwplug);
+
 	disk_zone_wplug_abort(zwplug);
 
 	/*
@@ -733,6 +825,7 @@ EXPORT_SYMBOL_GPL(disk_report_zone);
 static void blk_zone_reset_bio_endio(struct bio *bio)
 {
 	struct gendisk *disk = bio->bi_bdev->bd_disk;
+	sector_t sector = bio->bi_iter.bi_sector;
 	struct blk_zone_wplug *zwplug;
 
 	/*
@@ -741,7 +834,7 @@ static void blk_zone_reset_bio_endio(struct bio *bio)
 	 * resetting zones while writes are still in-flight will result in the
 	 * writes failing anyway.
 	 */
-	zwplug = disk_get_zone_wplug(disk, bio->bi_iter.bi_sector);
+	zwplug = disk_get_zone_wplug(disk, sector);
 	if (zwplug) {
 		unsigned long flags;
 
@@ -749,14 +842,18 @@ static void blk_zone_reset_bio_endio(struct bio *bio)
 		disk_zone_wplug_set_wp_offset(disk, zwplug, 0);
 		spin_unlock_irqrestore(&zwplug->lock, flags);
 		disk_put_zone_wplug(zwplug);
+	} else {
+		disk_zone_set_cond(disk, sector, BLK_ZONE_COND_EMPTY);
 	}
 }
 
 static void blk_zone_reset_all_bio_endio(struct bio *bio)
 {
 	struct gendisk *disk = bio->bi_bdev->bd_disk;
+	sector_t capacity = get_capacity(disk);
 	struct blk_zone_wplug *zwplug;
 	unsigned long flags;
+	sector_t sector;
 	unsigned int i;
 
 	/* Update the condition of all zone write plugs. */
@@ -770,12 +867,18 @@ static void blk_zone_reset_all_bio_endio(struct bio *bio)
 		}
 	}
 	rcu_read_unlock();
+
+	/* Update the cached zone conditions. */
+	for (sector = 0; sector < capacity;
+	     sector += bdev_zone_sectors(bio->bi_bdev))
+		disk_zone_set_cond(disk, sector, BLK_ZONE_COND_EMPTY);
 }
 
 static void blk_zone_finish_bio_endio(struct bio *bio)
 {
 	struct block_device *bdev = bio->bi_bdev;
 	struct gendisk *disk = bdev->bd_disk;
+	sector_t sector = bio->bi_iter.bi_sector;
 	struct blk_zone_wplug *zwplug;
 
 	/*
@@ -784,7 +887,7 @@ static void blk_zone_finish_bio_endio(struct bio *bio)
 	 * is fine as resetting zones while writes are still in-flight will
 	 * result in the writes failing anyway.
 	 */
-	zwplug = disk_get_zone_wplug(disk, bio->bi_iter.bi_sector);
+	zwplug = disk_get_zone_wplug(disk, sector);
 	if (zwplug) {
 		unsigned long flags;
 
@@ -793,6 +896,8 @@ static void blk_zone_finish_bio_endio(struct bio *bio)
 					      bdev_zone_sectors(bdev));
 		spin_unlock_irqrestore(&zwplug->lock, flags);
 		disk_put_zone_wplug(zwplug);
+	} else {
+		disk_zone_set_cond(disk, sector, BLK_ZONE_COND_FULL);
 	}
 }
 
@@ -888,6 +993,7 @@ static inline void disk_zone_wplug_add_bio(struct gendisk *disk,
  */
 void blk_zone_write_plug_bio_merged(struct bio *bio)
 {
+	struct gendisk *disk = bio->bi_bdev->bd_disk;
 	struct blk_zone_wplug *zwplug;
 	unsigned long flags;
 
@@ -909,13 +1015,13 @@ void blk_zone_write_plug_bio_merged(struct bio *bio)
 	 * have at least one request and one BIO referencing the zone write
 	 * plug. So this should not fail.
 	 */
-	zwplug = disk_get_zone_wplug(bio->bi_bdev->bd_disk,
-				     bio->bi_iter.bi_sector);
+	zwplug = disk_get_zone_wplug(disk, bio->bi_iter.bi_sector);
 	if (WARN_ON_ONCE(!zwplug))
 		return;
 
 	spin_lock_irqsave(&zwplug->lock, flags);
 	zwplug->wp_offset += bio_sectors(bio);
+	disk_zone_wplug_update_cond(disk, zwplug);
 	spin_unlock_irqrestore(&zwplug->lock, flags);
 }
 
@@ -974,6 +1080,7 @@ void blk_zone_write_plug_init_request(struct request *req)
 		/* Drop the reference taken by disk_zone_wplug_add_bio(). */
 		blk_queue_exit(q);
 		zwplug->wp_offset += bio_sectors(bio);
+		disk_zone_wplug_update_cond(disk, zwplug);
 
 		req_back_sector += bio_sectors(bio);
 	}
@@ -1037,6 +1144,7 @@ static bool blk_zone_wplug_prepare_bio(struct blk_zone_wplug *zwplug,
 
 	/* Advance the zone write pointer offset. */
 	zwplug->wp_offset += bio_sectors(bio);
+	disk_zone_wplug_update_cond(disk, zwplug);
 
 	return true;
 }
@@ -1683,7 +1791,7 @@ static int blk_revalidate_zone_cond(struct blk_zone *zone, unsigned int idx,
 		return -ENODEV;
 	}
 
-	args->zones_cond[idx] = cond;
+	blk_zone_set_cond(args->zones_cond, idx, cond);
 
 	return 0;
 
diff --git a/include/uapi/linux/blkzoned.h b/include/uapi/linux/blkzoned.h
index f85743ef6e7d..5c7662971414 100644
--- a/include/uapi/linux/blkzoned.h
+++ b/include/uapi/linux/blkzoned.h
@@ -48,6 +48,8 @@ enum blk_zone_type {
  *                      FINISH ZONE command.
  * @BLK_ZONE_COND_READONLY: The zone is read-only.
  * @BLK_ZONE_COND_OFFLINE: The zone is offline (sectors cannot be read/written).
+ * @BLK_ZONE_COND_ACTIVE: The zone is either implicitly open, explicitly open,
+ *			  or closed.
  *
  * The Zone Condition state machine in the ZBC/ZAC standards maps the above
  * deinitions as:
@@ -61,6 +63,13 @@ enum blk_zone_type {
  *
  * Conditions 0x5 to 0xC are reserved by the current ZBC/ZAC spec and should
  * be considered invalid.
+ *
+ * The condition BLK_ZONE_COND_ACTIVE is used only with cached zone reports.
+ * It is used to report any of the BLK_ZONE_COND_IMP_OPEN,
+ * BLK_ZONE_COND_EXP_OPEN and BLK_ZONE_COND_CLOSED conditions. Conversely, a
+ * regular zone report will never report a zone condition using
+ * BLK_ZONE_COND_ACTIVE and instead use the conditions BLK_ZONE_COND_IMP_OPEN,
+ * BLK_ZONE_COND_EXP_OPEN or BLK_ZONE_COND_CLOSED as reported by the device.
  */
 enum blk_zone_cond {
 	BLK_ZONE_COND_NOT_WP	= 0x0,
@@ -71,6 +80,8 @@ enum blk_zone_cond {
 	BLK_ZONE_COND_READONLY	= 0xD,
 	BLK_ZONE_COND_FULL	= 0xE,
 	BLK_ZONE_COND_OFFLINE	= 0xF,
+
+	BLK_ZONE_COND_ACTIVE	= 0xFF,
 };
 
 /**
-- 
cgit v1.2.3


From b30ffcdc0c15a88f8866529d3532454e02571221 Mon Sep 17 00:00:00 2001
From: Damien Le Moal <dlemoal@kernel.org>
Date: Wed, 5 Nov 2025 06:22:45 +0900
Subject: block: introduce BLKREPORTZONESV2 ioctl

Introduce the new BLKREPORTZONESV2 ioctl command to allow user
applications access to the fast zone report implemented by
blkdev_report_zones_cached(). This new ioctl is defined as number 142
and is documented in include/uapi/linux/fs.h.

Unlike the existing BLKREPORTZONES ioctl, this new ioctl uses the flags
field of struct blk_zone_report also as an input. If the user sets the
BLK_ZONE_REP_CACHED flag as an input, then blkdev_report_zones_cached()
is used to generate the zone report using cached zone information. If
this flag is not set, then BLKREPORTZONESV2 behaves in the same manner
as BLKREPORTZONES and the zone report is generated by accessing the
zoned device.

Signed-off-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
---
 block/blk-zoned.c             | 25 ++++++++++++++++++++++---
 block/ioctl.c                 |  1 +
 include/uapi/linux/blkzoned.h | 35 ++++++++++++++++++++++++++++++-----
 include/uapi/linux/fs.h       |  2 +-
 4 files changed, 54 insertions(+), 9 deletions(-)

(limited to 'include/uapi/linux')

diff --git a/block/blk-zoned.c b/block/blk-zoned.c
index bbd105b11843..7a7b0704f095 100644
--- a/block/blk-zoned.c
+++ b/block/blk-zoned.c
@@ -357,7 +357,12 @@ static int blkdev_copy_zone_to_user(struct blk_zone *zone, unsigned int idx,
 }
 
 /*
- * BLKREPORTZONE ioctl processing.
+ * Mask of valid input flags for BLKREPORTZONEV2 ioctl.
+ */
+#define BLK_ZONE_REPV2_INPUT_FLAGS	BLK_ZONE_REP_CACHED
+
+/*
+ * BLKREPORTZONE and BLKREPORTZONEV2 ioctl processing.
  * Called from blkdev_ioctl.
  */
 int blkdev_report_zones_ioctl(struct block_device *bdev, unsigned int cmd,
@@ -381,8 +386,22 @@ int blkdev_report_zones_ioctl(struct block_device *bdev, unsigned int cmd,
 		return -EINVAL;
 
 	args.zones = argp + sizeof(struct blk_zone_report);
-	ret = blkdev_report_zones(bdev, rep.sector, rep.nr_zones,
-				  blkdev_copy_zone_to_user, &args);
+
+	switch (cmd) {
+	case BLKREPORTZONE:
+		ret = blkdev_report_zones(bdev, rep.sector, rep.nr_zones,
+					  blkdev_copy_zone_to_user, &args);
+		break;
+	case BLKREPORTZONEV2:
+		if (rep.flags & ~BLK_ZONE_REPV2_INPUT_FLAGS)
+			return -EINVAL;
+		ret = blkdev_report_zones_cached(bdev, rep.sector, rep.nr_zones,
+					 blkdev_copy_zone_to_user, &args);
+		break;
+	default:
+		return -EINVAL;
+	}
+
 	if (ret < 0)
 		return ret;
 
diff --git a/block/ioctl.c b/block/ioctl.c
index 3927ca4707d0..698629e4c619 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -581,6 +581,7 @@ static int blkdev_common_ioctl(struct block_device *bdev, blk_mode_t mode,
 	case BLKGETDISKSEQ:
 		return put_u64(argp, bdev->bd_disk->diskseq);
 	case BLKREPORTZONE:
+	case BLKREPORTZONEV2:
 		return blkdev_report_zones_ioctl(bdev, cmd, arg);
 	case BLKRESETZONE:
 	case BLKOPENZONE:
diff --git a/include/uapi/linux/blkzoned.h b/include/uapi/linux/blkzoned.h
index 5c7662971414..e33f02703350 100644
--- a/include/uapi/linux/blkzoned.h
+++ b/include/uapi/linux/blkzoned.h
@@ -87,10 +87,20 @@ enum blk_zone_cond {
 /**
  * enum blk_zone_report_flags - Feature flags of reported zone descriptors.
  *
- * @BLK_ZONE_REP_CAPACITY: Zone descriptor has capacity field.
+ * @BLK_ZONE_REP_CAPACITY: Output only. Indicates that zone descriptors in a
+ *			   zone report have a valid capacity field.
+ * @BLK_ZONE_REP_CACHED: Input only. Indicates that the zone report should be
+ *			 generated using cached zone information. In this case,
+ *			 the implicit open, explicit open and closed zone
+ *			 conditions are all reported with the
+ *			 BLK_ZONE_COND_ACTIVE condition.
  */
 enum blk_zone_report_flags {
-	BLK_ZONE_REP_CAPACITY	= (1 << 0),
+	/* Output flags */
+	BLK_ZONE_REP_CAPACITY	= (1U << 0),
+
+	/* Input flags */
+	BLK_ZONE_REP_CACHED	= (1U << 31),
 };
 
 /**
@@ -133,6 +143,10 @@ struct blk_zone {
  * @sector: starting sector of report
  * @nr_zones: IN maximum / OUT actual
  * @flags: one or more flags as defined by enum blk_zone_report_flags.
+ * @flags: one or more flags as defined by enum blk_zone_report_flags.
+ *	   With BLKREPORTZONE, this field is ignored as an input and is valid
+ *	   only as an output. Using BLKREPORTZONEV2, this field is used as both
+ *	   input and output.
  * @zones: Space to hold @nr_zones @zones entries on reply.
  *
  * The array of at most @nr_zones must follow this structure in memory.
@@ -159,9 +173,19 @@ struct blk_zone_range {
 /**
  * Zoned block device ioctl's:
  *
- * @BLKREPORTZONE: Get zone information. Takes a zone report as argument.
- *                 The zone report will start from the zone containing the
- *                 sector specified in the report request structure.
+ * @BLKREPORTZONE: Get zone information from a zoned device. Takes a zone report
+ *		   as argument. The zone report will start from the zone
+ *		   containing the sector specified in struct blk_zone_report.
+ *		   The flags field of struct blk_zone_report is used as an
+ *		   output only and ignored as an input.
+ *		   DEPRECATED, use BLKREPORTZONEV2 instead.
+ * @BLKREPORTZONEV2: Same as @BLKREPORTZONE but uses the flags field of
+ *		     struct blk_zone_report as an input, allowing to get a zone
+ *		     report using cached zone information if the flag
+ *		     BLK_ZONE_REP_CACHED is set. In such case, the zone report
+ *		     may include zones with the condition @BLK_ZONE_COND_ACTIVE
+ *		     (c.f. the description of this condition above for more
+ *		     details).
  * @BLKRESETZONE: Reset the write pointer of the zones in the specified
  *                sector range. The sector range must be zone aligned.
  * @BLKGETZONESZ: Get the device zone size in number of 512 B sectors.
@@ -180,5 +204,6 @@ struct blk_zone_range {
 #define BLKOPENZONE	_IOW(0x12, 134, struct blk_zone_range)
 #define BLKCLOSEZONE	_IOW(0x12, 135, struct blk_zone_range)
 #define BLKFINISHZONE	_IOW(0x12, 136, struct blk_zone_range)
+#define BLKREPORTZONEV2	_IOWR(0x12, 142, struct blk_zone_report)
 
 #endif /* _UAPI_BLKZONED_H */
diff --git a/include/uapi/linux/fs.h b/include/uapi/linux/fs.h
index 957ce3343a4f..66ca526cf786 100644
--- a/include/uapi/linux/fs.h
+++ b/include/uapi/linux/fs.h
@@ -298,7 +298,7 @@ struct file_attr {
 #define BLKROTATIONAL _IO(0x12,126)
 #define BLKZEROOUT _IO(0x12,127)
 #define BLKGETDISKSEQ _IOR(0x12,128,__u64)
-/* 130-136 are used by zoned block device ioctls (uapi/linux/blkzoned.h) */
+/* 130-136 and 142 are used by zoned block device ioctls (uapi/linux/blkzoned.h) */
 /* 137-141 are used by blk-crypto ioctls (uapi/linux/blk-crypto.h) */
 #define BLKTRACESETUP2 _IOWR(0x12, 142, struct blk_user_trace_setup2)
 
-- 
cgit v1.2.3


From 62ed1b58224636185fa689db81224b8c8af46473 Mon Sep 17 00:00:00 2001
From: Li Nan <linan122@huawei.com>
Date: Mon, 3 Nov 2025 20:57:57 +0800
Subject: md: allow configuring logical block size

Previously, raid array used the maximum logical block size (LBS)
of all member disks. Adding a larger LBS disk at runtime could
unexpectedly increase RAID's LBS, risking corruption of existing
partitions. This can be reproduced by:

```
  # LBS of sd[de] is 512 bytes, sdf is 4096 bytes.
  mdadm -CRq /dev/md0 -l1 -n3 /dev/sd[de] missing --assume-clean

  # LBS is 512
  cat /sys/block/md0/queue/logical_block_size

  # create partition md0p1
  parted -s /dev/md0 mklabel gpt mkpart primary 1MiB 100%
  lsblk | grep md0p1

  # LBS becomes 4096 after adding sdf
  mdadm --add -q /dev/md0 /dev/sdf
  cat /sys/block/md0/queue/logical_block_size

  # partition lost
  partprobe /dev/md0
  lsblk | grep md0p1
```

Simply restricting larger-LBS disks is inflexible. In some scenarios,
only disks with 512 bytes LBS are available currently, but later, disks
with 4KB LBS may be added to the array.

Making LBS configurable is the best way to solve this scenario.
After this patch, the raid will:
  - store LBS in disk metadata
  - add a read-write sysfs 'mdX/logical_block_size'

Future mdadm should support setting LBS via metadata field during RAID
creation and the new sysfs. Though the kernel allows runtime LBS changes,
users should avoid modifying it after creating partitions or filesystems
to prevent compatibility issues.

Only 1.x metadata supports configurable LBS. 0.90 metadata inits all
fields to default values at auto-detect. Supporting 0.90 would require
more extensive changes and no such use case has been observed.

Note that many RAID paths rely on PAGE_SIZE alignment, including for
metadata I/O. A larger LBS than PAGE_SIZE will result in metadata
read/write failures. So this config should be prevented.

Link: https://lore.kernel.org/linux-raid/20251103125757.1405796-6-linan666@huaweicloud.com
Signed-off-by: Li Nan <linan122@huawei.com>
Reviewed-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Yu Kuai <yukuai@fnnas.com>
---
 Documentation/admin-guide/md.rst | 10 ++++++
 drivers/md/md-linear.c           |  1 +
 drivers/md/md.c                  | 77 ++++++++++++++++++++++++++++++++++++++++
 drivers/md/md.h                  |  1 +
 drivers/md/raid0.c               |  1 +
 drivers/md/raid1.c               |  1 +
 drivers/md/raid10.c              |  1 +
 drivers/md/raid5.c               |  1 +
 include/uapi/linux/raid/md_p.h   |  3 +-
 9 files changed, 95 insertions(+), 1 deletion(-)

(limited to 'include/uapi/linux')

diff --git a/Documentation/admin-guide/md.rst b/Documentation/admin-guide/md.rst
index deed823eab01..dc7eab191caa 100644
--- a/Documentation/admin-guide/md.rst
+++ b/Documentation/admin-guide/md.rst
@@ -238,6 +238,16 @@ All md devices contain:
      the number of devices in a raid4/5/6, or to support external
      metadata formats which mandate such clipping.
 
+  logical_block_size
+     Configure the array's logical block size in bytes. This attribute
+     is only supported for 1.x meta. Write the value before starting
+     array. The final array LBS uses the maximum between this
+     configuration and LBS of all combined devices. Note that
+     LBS cannot exceed PAGE_SIZE before RAID supports folio.
+     WARNING: Arrays created on new kernel cannot be assembled at old
+     kernel due to padding check, Set module parameter 'check_new_feature'
+     to false to bypass, but data loss may occur.
+
   reshape_position
      This is either ``none`` or a sector number within the devices of
      the array where ``reshape`` is up to.  If this is set, the three
diff --git a/drivers/md/md-linear.c b/drivers/md/md-linear.c
index 25a6ddedea65..8d7b82c4a723 100644
--- a/drivers/md/md-linear.c
+++ b/drivers/md/md-linear.c
@@ -72,6 +72,7 @@ static int linear_set_limits(struct mddev *mddev)
 
 	md_init_stacking_limits(&lim);
 	lim.max_hw_sectors = mddev->chunk_sectors;
+	lim.logical_block_size = mddev->logical_block_size;
 	lim.max_write_zeroes_sectors = mddev->chunk_sectors;
 	lim.max_hw_wzeroes_unmap_sectors = mddev->chunk_sectors;
 	lim.io_min = mddev->chunk_sectors << 9;
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 9676e2477df6..7b5c5967568f 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -1999,6 +1999,7 @@ static int super_1_validate(struct mddev *mddev, struct md_rdev *freshest, struc
 		mddev->layout = le32_to_cpu(sb->layout);
 		mddev->raid_disks = le32_to_cpu(sb->raid_disks);
 		mddev->dev_sectors = le64_to_cpu(sb->size);
+		mddev->logical_block_size = le32_to_cpu(sb->logical_block_size);
 		mddev->events = ev1;
 		mddev->bitmap_info.offset = 0;
 		mddev->bitmap_info.space = 0;
@@ -2208,6 +2209,7 @@ static void super_1_sync(struct mddev *mddev, struct md_rdev *rdev)
 	sb->chunksize = cpu_to_le32(mddev->chunk_sectors);
 	sb->level = cpu_to_le32(mddev->level);
 	sb->layout = cpu_to_le32(mddev->layout);
+	sb->logical_block_size = cpu_to_le32(mddev->logical_block_size);
 	if (test_bit(FailFast, &rdev->flags))
 		sb->devflags |= FailFast1;
 	else
@@ -5936,6 +5938,68 @@ static struct md_sysfs_entry md_serialize_policy =
 __ATTR(serialize_policy, S_IRUGO | S_IWUSR, serialize_policy_show,
        serialize_policy_store);
 
+static int mddev_set_logical_block_size(struct mddev *mddev,
+				unsigned int lbs)
+{
+	int err = 0;
+	struct queue_limits lim;
+
+	if (queue_logical_block_size(mddev->gendisk->queue) >= lbs) {
+		pr_err("%s: Cannot set LBS smaller than mddev LBS %u\n",
+		       mdname(mddev), lbs);
+		return -EINVAL;
+	}
+
+	lim = queue_limits_start_update(mddev->gendisk->queue);
+	lim.logical_block_size = lbs;
+	pr_info("%s: logical_block_size is changed, data may be lost\n",
+		mdname(mddev));
+	err = queue_limits_commit_update(mddev->gendisk->queue, &lim);
+	if (err)
+		return err;
+
+	mddev->logical_block_size = lbs;
+	/* New lbs will be written to superblock after array is running */
+	set_bit(MD_SB_CHANGE_DEVS, &mddev->sb_flags);
+	return 0;
+}
+
+static ssize_t
+lbs_show(struct mddev *mddev, char *page)
+{
+	return sprintf(page, "%u\n", mddev->logical_block_size);
+}
+
+static ssize_t
+lbs_store(struct mddev *mddev, const char *buf, size_t len)
+{
+	unsigned int lbs;
+	int err = -EBUSY;
+
+	/* Only 1.x meta supports configurable LBS */
+	if (mddev->major_version == 0)
+		return -EINVAL;
+
+	if (mddev->pers)
+		return -EBUSY;
+
+	err = kstrtouint(buf, 10, &lbs);
+	if (err < 0)
+		return -EINVAL;
+
+	err = mddev_lock(mddev);
+	if (err)
+		goto unlock;
+
+	err = mddev_set_logical_block_size(mddev, lbs);
+
+unlock:
+	mddev_unlock(mddev);
+	return err ?: len;
+}
+
+static struct md_sysfs_entry md_logical_block_size =
+__ATTR(logical_block_size, 0644, lbs_show, lbs_store);
 
 static struct attribute *md_default_attrs[] = {
 	&md_level.attr,
@@ -5958,6 +6022,7 @@ static struct attribute *md_default_attrs[] = {
 	&md_consistency_policy.attr,
 	&md_fail_last_dev.attr,
 	&md_serialize_policy.attr,
+	&md_logical_block_size.attr,
 	NULL,
 };
 
@@ -6088,6 +6153,17 @@ int mddev_stack_rdev_limits(struct mddev *mddev, struct queue_limits *lim,
 			return -EINVAL;
 	}
 
+	/*
+	 * Before RAID adding folio support, the logical_block_size
+	 * should be smaller than the page size.
+	 */
+	if (lim->logical_block_size > PAGE_SIZE) {
+		pr_err("%s: logical_block_size must not larger than PAGE_SIZE\n",
+			mdname(mddev));
+		return -EINVAL;
+	}
+	mddev->logical_block_size = lim->logical_block_size;
+
 	return 0;
 }
 EXPORT_SYMBOL_GPL(mddev_stack_rdev_limits);
@@ -6699,6 +6775,7 @@ static void md_clean(struct mddev *mddev)
 	mddev->chunk_sectors = 0;
 	mddev->ctime = mddev->utime = 0;
 	mddev->layout = 0;
+	mddev->logical_block_size = 0;
 	mddev->max_disks = 0;
 	mddev->events = 0;
 	mddev->can_decrease_events = 0;
diff --git a/drivers/md/md.h b/drivers/md/md.h
index fd6e001c1d38..6985f2829bbd 100644
--- a/drivers/md/md.h
+++ b/drivers/md/md.h
@@ -433,6 +433,7 @@ struct mddev {
 	sector_t			array_sectors; /* exported array size */
 	int				external_size; /* size managed
 							* externally */
+	unsigned int			logical_block_size;
 	__u64				events;
 	/* If the last 'event' was simply a clean->dirty transition, and
 	 * we didn't write it to the spares, then it is safe and simple
diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c
index fbf763401521..47aee1b1d4d1 100644
--- a/drivers/md/raid0.c
+++ b/drivers/md/raid0.c
@@ -380,6 +380,7 @@ static int raid0_set_limits(struct mddev *mddev)
 	lim.max_hw_sectors = mddev->chunk_sectors;
 	lim.max_write_zeroes_sectors = mddev->chunk_sectors;
 	lim.max_hw_wzeroes_unmap_sectors = mddev->chunk_sectors;
+	lim.logical_block_size = mddev->logical_block_size;
 	lim.io_min = mddev->chunk_sectors << 9;
 	lim.io_opt = lim.io_min * mddev->raid_disks;
 	lim.chunk_sectors = mddev->chunk_sectors;
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 592a40233004..57d50465eed1 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -3213,6 +3213,7 @@ static int raid1_set_limits(struct mddev *mddev)
 	md_init_stacking_limits(&lim);
 	lim.max_write_zeroes_sectors = 0;
 	lim.max_hw_wzeroes_unmap_sectors = 0;
+	lim.logical_block_size = mddev->logical_block_size;
 	lim.features |= BLK_FEAT_ATOMIC_WRITES;
 	err = mddev_stack_rdev_limits(mddev, &lim, MDDEV_STACK_INTEGRITY);
 	if (err)
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 14dcd5142eb4..84be4cc7e873 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -4000,6 +4000,7 @@ static int raid10_set_queue_limits(struct mddev *mddev)
 	md_init_stacking_limits(&lim);
 	lim.max_write_zeroes_sectors = 0;
 	lim.max_hw_wzeroes_unmap_sectors = 0;
+	lim.logical_block_size = mddev->logical_block_size;
 	lim.io_min = mddev->chunk_sectors << 9;
 	lim.chunk_sectors = mddev->chunk_sectors;
 	lim.io_opt = lim.io_min * raid10_nr_stripes(conf);
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 24b32a0c95b4..cdbc7eba5c54 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -7745,6 +7745,7 @@ static int raid5_set_limits(struct mddev *mddev)
 	stripe = roundup_pow_of_two(data_disks * (mddev->chunk_sectors << 9));
 
 	md_init_stacking_limits(&lim);
+	lim.logical_block_size = mddev->logical_block_size;
 	lim.io_min = mddev->chunk_sectors << 9;
 	lim.io_opt = lim.io_min * (conf->raid_disks - conf->max_degraded);
 	lim.features |= BLK_FEAT_RAID_PARTIAL_STRIPES_EXPENSIVE;
diff --git a/include/uapi/linux/raid/md_p.h b/include/uapi/linux/raid/md_p.h
index ac74133a4768..310068bb2a1d 100644
--- a/include/uapi/linux/raid/md_p.h
+++ b/include/uapi/linux/raid/md_p.h
@@ -291,7 +291,8 @@ struct mdp_superblock_1 {
 	__le64	resync_offset;	/* data before this offset (from data_offset) known to be in sync */
 	__le32	sb_csum;	/* checksum up to devs[max_dev] */
 	__le32	max_dev;	/* size of devs[] array to consider */
-	__u8	pad3[64-32];	/* set to 0 when writing */
+	__le32  logical_block_size;	/* same as q->limits->logical_block_size */
+	__u8	pad3[64-36];	/* set to 0 when writing */
 
 	/* device state information. Indexed by dev_number.
 	 * 2 bytes per device
-- 
cgit v1.2.3