blk-mq: make the sysfs mq/ layout reflect current mappings

Currently blk-mq registers all the hardware queues in sysfs,
regardless of whether it uses them (e.g. they have CPU mappings)
or not. The unused hardware queues lack the cpux/ directories,
and the other sysfs entries (like active, pending, etc) are all
zeroes.

Change this so that sysfs correctly reflects the current mappings
of the hardware queues.
Signed-off-by: Jens Axboe <axboe@fb.com>

block/blk-mq-sysfs.c

@@ -327,6 +327,42 @@ static struct kobj_type blk_mq_hw_ktype = {
 	.release	= blk_mq_sysfs_release,
 };
 
+void blk_mq_unregister_hctx(struct blk_mq_hw_ctx *hctx)
+{
+	struct blk_mq_ctx *ctx;
+	int i;
+
+	if (!hctx->nr_ctx || !(hctx->flags & BLK_MQ_F_SYSFS_UP))
+		return;
+
+	hctx_for_each_ctx(hctx, ctx, i)
+		kobject_del(&ctx->kobj);
+
+	kobject_del(&hctx->kobj);
+}
+
+int blk_mq_register_hctx(struct blk_mq_hw_ctx *hctx)
+{
+	struct request_queue *q = hctx->queue;
+	struct blk_mq_ctx *ctx;
+	int i, ret;
+
+	if (!hctx->nr_ctx || !(hctx->flags & BLK_MQ_F_SYSFS_UP))
+		return 0;
+
+	ret = kobject_add(&hctx->kobj, &q->mq_kobj, "%u", hctx->queue_num);
+	if (ret)
+		return ret;
+
+	hctx_for_each_ctx(hctx, ctx, i) {
+		ret = kobject_add(&ctx->kobj, &hctx->kobj, "cpu%u", ctx->cpu);
+		if (ret)
+			break;
+	}
+
+	return ret;
+}
+
 void blk_mq_unregister_disk(struct gendisk *disk)
 {
 	struct request_queue *q = disk->queue;
@@ -335,11 +371,11 @@ void blk_mq_unregister_disk(struct gendisk *disk)
 	int i, j;
 
 	queue_for_each_hw_ctx(q, hctx, i) {
-		hctx_for_each_ctx(hctx, ctx, j) {
-			kobject_del(&ctx->kobj);
+		blk_mq_unregister_hctx(hctx);
+
+		hctx_for_each_ctx(hctx, ctx, j)
 			kobject_put(&ctx->kobj);
-		}
-		kobject_del(&hctx->kobj);
+
 		kobject_put(&hctx->kobj);
 	}
 
@@ -350,15 +386,30 @@ void blk_mq_unregister_disk(struct gendisk *disk)
 	kobject_put(&disk_to_dev(disk)->kobj);
 }
 
+static void blk_mq_sysfs_init(struct request_queue *q)
+{
+	struct blk_mq_hw_ctx *hctx;
+	struct blk_mq_ctx *ctx;
+	int i, j;
+
+	kobject_init(&q->mq_kobj, &blk_mq_ktype);
+
+	queue_for_each_hw_ctx(q, hctx, i) {
+		kobject_init(&hctx->kobj, &blk_mq_hw_ktype);
+
+		hctx_for_each_ctx(hctx, ctx, j)
+			kobject_init(&ctx->kobj, &blk_mq_ctx_ktype);
+	}
+}
+
 int blk_mq_register_disk(struct gendisk *disk)
 {
 	struct device *dev = disk_to_dev(disk);
 	struct request_queue *q = disk->queue;
 	struct blk_mq_hw_ctx *hctx;
-	struct blk_mq_ctx *ctx;
-	int ret, i, j;
+	int ret, i;
 
-	kobject_init(&q->mq_kobj, &blk_mq_ktype);
+	blk_mq_sysfs_init(q);
 
 	ret = kobject_add(&q->mq_kobj, kobject_get(&dev->kobj), "%s", "mq");
 	if (ret < 0)
@@ -367,20 +418,10 @@ int blk_mq_register_disk(struct gendisk *disk)
 	kobject_uevent(&q->mq_kobj, KOBJ_ADD);
 
 	queue_for_each_hw_ctx(q, hctx, i) {
-		kobject_init(&hctx->kobj, &blk_mq_hw_ktype);
-		ret = kobject_add(&hctx->kobj, &q->mq_kobj, "%u", i);
+		hctx->flags |= BLK_MQ_F_SYSFS_UP;
+		ret = blk_mq_register_hctx(hctx);
 		if (ret)
 			break;
-
-		if (!hctx->nr_ctx)
-			continue;
-
-		hctx_for_each_ctx(hctx, ctx, j) {
-			kobject_init(&ctx->kobj, &blk_mq_ctx_ktype);
-			ret = kobject_add(&ctx->kobj, &hctx->kobj, "cpu%u", ctx->cpu);
-			if (ret)
-				break;
-		}
 	}
 
 	if (ret) {
@@ -390,3 +431,26 @@ int blk_mq_register_disk(struct gendisk *disk)
 
 	return 0;
 }
+
+void blk_mq_sysfs_unregister(struct request_queue *q)
+{
+	struct blk_mq_hw_ctx *hctx;
+	int i;
+
+	queue_for_each_hw_ctx(q, hctx, i)
+		blk_mq_unregister_hctx(hctx);
+}
+
+int blk_mq_sysfs_register(struct request_queue *q)
+{
+	struct blk_mq_hw_ctx *hctx;
+	int i, ret = 0;
+
+	queue_for_each_hw_ctx(q, hctx, i) {
+		ret = blk_mq_register_hctx(hctx);
+		if (ret)
+			break;
+	}
+
+	return ret;
+}

block/blk-mq.c

@@ -1924,6 +1924,8 @@ static void blk_mq_queue_reinit(struct request_queue *q)
 {
 	blk_mq_freeze_queue(q);
 
+	blk_mq_sysfs_unregister(q);
+
 	blk_mq_update_queue_map(q->mq_map, q->nr_hw_queues);
 
 	/*
@@ -1934,6 +1936,8 @@ static void blk_mq_queue_reinit(struct request_queue *q)
 
 	blk_mq_map_swqueue(q);
 
+	blk_mq_sysfs_register(q);
+
 	blk_mq_unfreeze_queue(q);
 }
 

block/blk-mq.h

@@ -54,6 +54,12 @@ extern unsigned int *blk_mq_make_queue_map(struct blk_mq_tag_set *set);
 extern int blk_mq_update_queue_map(unsigned int *map, unsigned int nr_queues);
 extern int blk_mq_hw_queue_to_node(unsigned int *map, unsigned int);
 
+/*
+ * sysfs helpers
+ */
+extern int blk_mq_sysfs_register(struct request_queue *q);
+extern void blk_mq_sysfs_unregister(struct request_queue *q);
+
 /*
  * Basic implementation of sparser bitmap, allowing the user to spread
  * the bits over more cachelines.

include/linux/blk-mq.h

@@ -130,6 +130,7 @@ enum {
 	BLK_MQ_F_SHOULD_SORT	= 1 << 1,
 	BLK_MQ_F_TAG_SHARED	= 1 << 2,
 	BLK_MQ_F_SG_MERGE	= 1 << 3,
+	BLK_MQ_F_SYSFS_UP	= 1 << 4,
 
 	BLK_MQ_S_STOPPED	= 0,
 	BLK_MQ_S_TAG_ACTIVE	= 1,