block: move cache control settings out of queue->flags

Move the cache control settings into the queue_limits so that the flags
can be set atomically with the device queue frozen.

Add new features and flags field for the driver set flags, and internal
(usually sysfs-controlled) flags in the block layer.  Note that we'll
eventually remove enough field from queue_limits to bring it back to the
previous size.

The disable flag is inverted compared to the previous meaning, which
means it now survives a rescan, similar to the max_sectors and
max_discard_sectors user limits.

The FLUSH and FUA flags are now inherited by blk_stack_limits, which
simplified the code in dm a lot, but also causes a slight behavior
change in that dm-switch and dm-unstripe now advertise a write cache
despite setting num_flush_bios to 0.  The I/O path will handle this
gracefully, but as far as I can tell the lack of num_flush_bios
and thus flush support is a pre-existing data integrity bug in those
targets that really needs fixing, after which a non-zero num_flush_bios
should be required in dm for targets that map to underlying devices.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Ulf Hansson <ulf.hansson@linaro.org>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/20240617060532.127975-14-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>

6 files changed, 31 insertions, 44 deletions

diff --git a/block/blk-core.c b/block/blk-core.c
index 82c3ae22d76d..2b45a4df9a1a 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -782,7 +782,7 @@ void submit_bio_noacct(struct bio *bio)
 		if (WARN_ON_ONCE(bio_op(bio) != REQ_OP_WRITE &&
 				 bio_op(bio) != REQ_OP_ZONE_APPEND))
 			goto end_io;
-		if (!test_bit(QUEUE_FLAG_WC, &q->queue_flags)) {
+		if (!bdev_write_cache(bdev)) {
 			bio->bi_opf &= ~(REQ_PREFLUSH | REQ_FUA);
 			if (!bio_sectors(bio)) {
 				status = BLK_STS_OK;
diff --git a/block/blk-flush.c b/block/blk-flush.c
index 2234f8b3fc05..30b9d5033a2b 100644
--- a/block/blk-flush.c
+++ b/block/blk-flush.c
@@ -381,8 +381,8 @@ static void blk_rq_init_flush(struct request *rq)
 bool blk_insert_flush(struct request *rq)
 {
 	struct request_queue *q = rq->q;
-	unsigned long fflags = q->queue_flags;	/* may change, cache */
 	struct blk_flush_queue *fq = blk_get_flush_queue(q, rq->mq_ctx);
+	bool supports_fua = q->limits.features & BLK_FEAT_FUA;
 	unsigned int policy = 0;
 
 	/* FLUSH/FUA request must never be merged */
@@ -394,11 +394,10 @@ bool blk_insert_flush(struct request *rq)
 	/*
 	 * Check which flushes we need to sequence for this operation.
 	 */
-	if (fflags & (1UL << QUEUE_FLAG_WC)) {
+	if (blk_queue_write_cache(q)) {
 		if (rq->cmd_flags & REQ_PREFLUSH)
 			policy |= REQ_FSEQ_PREFLUSH;
-		if (!(fflags & (1UL << QUEUE_FLAG_FUA)) &&
-		    (rq->cmd_flags & REQ_FUA))
+		if ((rq->cmd_flags & REQ_FUA) && !supports_fua)
 			policy |= REQ_FSEQ_POSTFLUSH;
 	}
 
@@ -407,7 +406,7 @@ bool blk_insert_flush(struct request *rq)
 	 * REQ_PREFLUSH and FUA for the driver.
 	 */
 	rq->cmd_flags &= ~REQ_PREFLUSH;
-	if (!(fflags & (1UL << QUEUE_FLAG_FUA)))
+	if (!supports_fua)
 		rq->cmd_flags &= ~REQ_FUA;
 
 	/*
diff --git a/block/blk-mq-debugfs.c b/block/blk-mq-debugfs.c
index 770c0c2b72fa..e8b9db7c30c4 100644
--- a/block/blk-mq-debugfs.c
+++ b/block/blk-mq-debugfs.c
@@ -93,8 +93,6 @@ static const char *const blk_queue_flag_name[] = {
 	QUEUE_FLAG_NAME(INIT_DONE),
 	QUEUE_FLAG_NAME(STABLE_WRITES),
 	QUEUE_FLAG_NAME(POLL),
-	QUEUE_FLAG_NAME(WC),
-	QUEUE_FLAG_NAME(FUA),
 	QUEUE_FLAG_NAME(DAX),
 	QUEUE_FLAG_NAME(STATS),
 	QUEUE_FLAG_NAME(REGISTERED),
diff --git a/block/blk-settings.c b/block/blk-settings.c
index f11c8676eb4c..536ee202fcdc 100644
--- a/block/blk-settings.c
+++ b/block/blk-settings.c
@@ -261,6 +261,9 @@ static int blk_validate_limits(struct queue_limits *lim)
 		lim->misaligned = 0;
 	}
 
+	if (!(lim->features & BLK_FEAT_WRITE_CACHE))
+		lim->features &= ~BLK_FEAT_FUA;
+
 	err = blk_validate_integrity_limits(lim);
 	if (err)
 		return err;
@@ -454,6 +457,8 @@ int blk_stack_limits(struct queue_limits *t, struct queue_limits *b,
 {
 	unsigned int top, bottom, alignment, ret = 0;
 
+	t->features |= (b->features & BLK_FEAT_INHERIT_MASK);
+
 	t->max_sectors = min_not_zero(t->max_sectors, b->max_sectors);
 	t->max_user_sectors = min_not_zero(t->max_user_sectors,
 			b->max_user_sectors);
@@ -711,30 +716,6 @@ void blk_set_queue_depth(struct request_queue *q, unsigned int depth)
 }
 EXPORT_SYMBOL(blk_set_queue_depth);
 
-/**
- * blk_queue_write_cache - configure queue's write cache
- * @q:		the request queue for the device
- * @wc:		write back cache on or off
- * @fua:	device supports FUA writes, if true
- *
- * Tell the block layer about the write cache of @q.
- */
-void blk_queue_write_cache(struct request_queue *q, bool wc, bool fua)
-{
-	if (wc) {
-		blk_queue_flag_set(QUEUE_FLAG_HW_WC, q);
-		blk_queue_flag_set(QUEUE_FLAG_WC, q);
-	} else {
-		blk_queue_flag_clear(QUEUE_FLAG_HW_WC, q);
-		blk_queue_flag_clear(QUEUE_FLAG_WC, q);
-	}
-	if (fua)
-		blk_queue_flag_set(QUEUE_FLAG_FUA, q);
-	else
-		blk_queue_flag_clear(QUEUE_FLAG_FUA, q);
-}
-EXPORT_SYMBOL_GPL(blk_queue_write_cache);
-
 int bdev_alignment_offset(struct block_device *bdev)
 {
 	struct request_queue *q = bdev_get_queue(bdev);
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index 5c787965b7d0..4f524c1d5e08 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -423,32 +423,41 @@ static ssize_t queue_io_timeout_store(struct request_queue *q, const char *page,
 
 static ssize_t queue_wc_show(struct request_queue *q, char *page)
 {
-	if (test_bit(QUEUE_FLAG_WC, &q->queue_flags))
-		return sprintf(page, "write back\n");
-
-	return sprintf(page, "write through\n");
+	if (q->limits.features & BLK_FLAGS_WRITE_CACHE_DISABLED)
+		return sprintf(page, "write through\n");
+	return sprintf(page, "write back\n");
 }
 
 static ssize_t queue_wc_store(struct request_queue *q, const char *page,
 			      size_t count)
 {
+	struct queue_limits lim;
+	bool disable;
+	int err;
+
 	if (!strncmp(page, "write back", 10)) {
-		if (!test_bit(QUEUE_FLAG_HW_WC, &q->queue_flags))
-			return -EINVAL;
-		blk_queue_flag_set(QUEUE_FLAG_WC, q);
+		disable = false;
 	} else if (!strncmp(page, "write through", 13) ||
-		 !strncmp(page, "none", 4)) {
-		blk_queue_flag_clear(QUEUE_FLAG_WC, q);
+		   !strncmp(page, "none", 4)) {
+		disable = true;
 	} else {
 		return -EINVAL;
 	}
 
+	lim = queue_limits_start_update(q);
+	if (disable)
+		lim.flags |= BLK_FLAGS_WRITE_CACHE_DISABLED;
+	else
+		lim.flags &= ~BLK_FLAGS_WRITE_CACHE_DISABLED;
+	err = queue_limits_commit_update(q, &lim);
+	if (err)
+		return err;
 	return count;
 }
 
 static ssize_t queue_fua_show(struct request_queue *q, char *page)
 {
-	return sprintf(page, "%u\n", test_bit(QUEUE_FLAG_FUA, &q->queue_flags));
+	return sprintf(page, "%u\n", !!(q->limits.features & BLK_FEAT_FUA));
 }
 
 static ssize_t queue_dax_show(struct request_queue *q, char *page)
diff --git a/block/blk-wbt.c b/block/blk-wbt.c
index 64472134dd26..1a5e4b049ecd 100644
--- a/block/blk-wbt.c
+++ b/block/blk-wbt.c
@@ -206,8 +206,8 @@ static void wbt_rqw_done(struct rq_wb *rwb, struct rq_wait *rqw,
 	 */
 	if (wb_acct & WBT_DISCARD)
 		limit = rwb->wb_background;
-	else if (test_bit(QUEUE_FLAG_WC, &rwb->rqos.disk->queue->queue_flags) &&
-	         !wb_recent_wait(rwb))
+	else if (blk_queue_write_cache(rwb->rqos.disk->queue) &&
+		 !wb_recent_wait(rwb))
 		limit = 0;
 	else
 		limit = rwb->wb_normal;