[PATCH 1/4] nvme-pci: split nvme_alloc_admin_tags

[PATCH 1/4] nvme-pci: split nvme_alloc_admin_tags

[Christoph Hellwig]

Split nvme_alloc_admin_tags into a helper to actually allocate the
tag set, and one that just restarts the admin queue.  Add a local
variable for the tag_set to clean up the code a bit.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/nvme/host/pci.c | 60 +++++++++++++++++++++--------------------
 1 file changed, 31 insertions(+), 29 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 375502de4c745..ceccd8331a76e 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1766,37 +1766,35 @@ static void nvme_dev_remove_admin(struct nvme_dev *dev)
 	}
 }
 
-static int nvme_alloc_admin_tags(struct nvme_dev *dev)
+static int nvme_pci_alloc_admin_tag_set(struct nvme_dev *dev)
 {
-	if (!dev->ctrl.admin_q) {
-		dev->admin_tagset.ops = &nvme_mq_admin_ops;
-		dev->admin_tagset.nr_hw_queues = 1;
+	struct blk_mq_tag_set *set = &dev->admin_tagset;
 
-		dev->admin_tagset.queue_depth = NVME_AQ_MQ_TAG_DEPTH;
-		dev->admin_tagset.timeout = NVME_ADMIN_TIMEOUT;
-		dev->admin_tagset.numa_node = dev->ctrl.numa_node;
-		dev->admin_tagset.cmd_size = sizeof(struct nvme_iod);
-		dev->admin_tagset.flags = BLK_MQ_F_NO_SCHED;
-		dev->admin_tagset.driver_data = dev;
+	set->ops = &nvme_mq_admin_ops;
+	set->nr_hw_queues = 1;
 
-		if (blk_mq_alloc_tag_set(&dev->admin_tagset))
-			return -ENOMEM;
-		dev->ctrl.admin_tagset = &dev->admin_tagset;
+	set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
+	set->timeout = NVME_ADMIN_TIMEOUT;
+	set->numa_node = dev->ctrl.numa_node;
+	set->cmd_size = sizeof(struct nvme_iod);
+	set->flags = BLK_MQ_F_NO_SCHED;
+	set->driver_data = dev;
 
-		dev->ctrl.admin_q = blk_mq_init_queue(&dev->admin_tagset);
-		if (IS_ERR(dev->ctrl.admin_q)) {
-			blk_mq_free_tag_set(&dev->admin_tagset);
-			dev->ctrl.admin_q = NULL;
-			return -ENOMEM;
-		}
-		if (!blk_get_queue(dev->ctrl.admin_q)) {
-			nvme_dev_remove_admin(dev);
-			dev->ctrl.admin_q = NULL;
-			return -ENODEV;
-		}
-	} else
-		nvme_start_admin_queue(&dev->ctrl);
+	if (blk_mq_alloc_tag_set(set))
+		return -ENOMEM;
+	dev->ctrl.admin_tagset = set;
 
+	dev->ctrl.admin_q = blk_mq_init_queue(set);
+	if (IS_ERR(dev->ctrl.admin_q)) {
+		blk_mq_free_tag_set(set);
+		dev->ctrl.admin_q = NULL;
+		return -ENOMEM;
+	}
+	if (!blk_get_queue(dev->ctrl.admin_q)) {
+		nvme_dev_remove_admin(dev);
+		dev->ctrl.admin_q = NULL;
+		return -ENODEV;
+	}
 	return 0;
 }
 
@@ -2836,9 +2834,13 @@ static void nvme_reset_work(struct work_struct *work)
 	if (result)
 		goto out_unlock;
 
-	result = nvme_alloc_admin_tags(dev);
-	if (result)
-		goto out_unlock;
+	if (!dev->ctrl.admin_q) {
+		result = nvme_pci_alloc_admin_tag_set(dev);
+		if (result)
+			goto out_unlock;
+	} else {
+		nvme_start_admin_queue(&dev->ctrl);
+	}
 
 	/*
 	 * Limit the max command size to prevent iod->sg allocations going
-- 
2.30.2

[PATCH 2/4] nvme-pci: split nvme_dev_add

[Christoph Hellwig]

Split nvme_dev_add into a helper to actually allocate the tag set, and
one that just update the number of queues.  Add a local variable for
the tag_set to clean up the code a bit.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/nvme/host/pci.c | 72 +++++++++++++++++++++--------------------
 1 file changed, 37 insertions(+), 35 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index ceccd8331a76e..61866951dcce5 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -2533,47 +2533,45 @@ static bool __nvme_disable_io_queues(struct nvme_dev *dev, u8 opcode)
 	return true;
 }
 
-static void nvme_dev_add(struct nvme_dev *dev)
+static void nvme_pci_alloc_tag_set(struct nvme_dev *dev)
 {
+	struct blk_mq_tag_set * set = &dev->tagset;
 	int ret;
 
-	if (!dev->ctrl.tagset) {
-		dev->tagset.ops = &nvme_mq_ops;
-		dev->tagset.nr_hw_queues = dev->online_queues - 1;
-		dev->tagset.nr_maps = 2; /* default + read */
-		if (dev->io_queues[HCTX_TYPE_POLL])
-			dev->tagset.nr_maps++;
-		dev->tagset.timeout = NVME_IO_TIMEOUT;
-		dev->tagset.numa_node = dev->ctrl.numa_node;
-		dev->tagset.queue_depth = min_t(unsigned int, dev->q_depth,
-						BLK_MQ_MAX_DEPTH) - 1;
-		dev->tagset.cmd_size = sizeof(struct nvme_iod);
-		dev->tagset.flags = BLK_MQ_F_SHOULD_MERGE;
-		dev->tagset.driver_data = dev;
-
-		/*
-		 * Some Apple controllers requires tags to be unique
-		 * across admin and IO queue, so reserve the first 32
-		 * tags of the IO queue.
-		 */
-		if (dev->ctrl.quirks & NVME_QUIRK_SHARED_TAGS)
-			dev->tagset.reserved_tags = NVME_AQ_DEPTH;
+	set->ops = &nvme_mq_ops;
+	set->nr_hw_queues = dev->online_queues - 1;
+	set->nr_maps = 2; /* default + read */
+	if (dev->io_queues[HCTX_TYPE_POLL])
+		set->nr_maps++;
+	set->timeout = NVME_IO_TIMEOUT;
+	set->numa_node = dev->ctrl.numa_node;
+	set->queue_depth = min_t(unsigned, dev->q_depth, BLK_MQ_MAX_DEPTH) - 1;
+	set->cmd_size = sizeof(struct nvme_iod);
+	set->flags = BLK_MQ_F_SHOULD_MERGE;
+	set->driver_data = dev;
 
-		ret = blk_mq_alloc_tag_set(&dev->tagset);
-		if (ret) {
-			dev_warn(dev->ctrl.device,
-				"IO queues tagset allocation failed %d\n", ret);
-			return;
-		}
-		dev->ctrl.tagset = &dev->tagset;
-	} else {
-		blk_mq_update_nr_hw_queues(&dev->tagset, dev->online_queues - 1);
+	/*
+	 * Some Apple controllers requires tags to be unique
+	 * across admin and IO queue, so reserve the first 32
+	 * tags of the IO queue.
+	 */
+	if (dev->ctrl.quirks & NVME_QUIRK_SHARED_TAGS)
+		set->reserved_tags = NVME_AQ_DEPTH;
 
-		/* Free previously allocated queues that are no longer usable */
-		nvme_free_queues(dev, dev->online_queues);
+	ret = blk_mq_alloc_tag_set(set);
+	if (ret) {
+		dev_warn(dev->ctrl.device,
+			"IO queues tagset allocation failed %d\n", ret);
+		return;
 	}
+	dev->ctrl.tagset = set;
+}
 
-	nvme_dbbuf_set(dev);
+static void nvme_pci_update_nr_queues(struct nvme_dev *dev)
+{
+	blk_mq_update_nr_hw_queues(&dev->tagset, dev->online_queues - 1);
+	/* free previously allocated queues that are no longer usable */
+	nvme_free_queues(dev, dev->online_queues);
 }
 
 static int nvme_pci_enable(struct nvme_dev *dev)
@@ -2919,7 +2917,11 @@ static void nvme_reset_work(struct work_struct *work)
 	} else {
 		nvme_start_queues(&dev->ctrl);
 		nvme_wait_freeze(&dev->ctrl);
-		nvme_dev_add(dev);
+		if (!dev->ctrl.tagset)
+			nvme_pci_alloc_tag_set(dev);
+		else
+			nvme_pci_update_nr_queues(dev);
+		nvme_dbbuf_set(dev);
 		nvme_unfreeze(&dev->ctrl);
 	}
 
-- 
2.30.2

[PATCH 3/4] nvme-rdma: split nvme_rdma_alloc_tagset

[Christoph Hellwig]

Split nvme_rdma_alloc_tagset into one helper for the admin tag_set and
one for the I/O tag set.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/nvme/host/rdma.c | 92 ++++++++++++++++++++--------------------
 1 file changed, 46 insertions(+), 46 deletions(-)

diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c
index 7d01fb7702842..7f4f80f3b524a 100644
--- a/drivers/nvme/host/rdma.c
+++ b/drivers/nvme/host/rdma.c
@@ -787,50 +787,54 @@ static int nvme_rdma_alloc_io_queues(struct nvme_rdma_ctrl *ctrl)
 	return ret;
 }
 
-static struct blk_mq_tag_set *nvme_rdma_alloc_tagset(struct nvme_ctrl *nctrl,
-		bool admin)
+static int nvme_rdma_alloc_admin_tag_set(struct nvme_ctrl *nctrl)
 {
 	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl);
-	struct blk_mq_tag_set *set;
+	struct blk_mq_tag_set *set = &ctrl->admin_tag_set;
 	int ret;
 
-	if (admin) {
-		set = &ctrl->admin_tag_set;
-		memset(set, 0, sizeof(*set));
-		set->ops = &nvme_rdma_admin_mq_ops;
-		set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
-		set->reserved_tags = NVMF_RESERVED_TAGS;
-		set->numa_node = nctrl->numa_node;
-		set->cmd_size = sizeof(struct nvme_rdma_request) +
-				NVME_RDMA_DATA_SGL_SIZE;
-		set->driver_data = ctrl;
-		set->nr_hw_queues = 1;
-		set->timeout = NVME_ADMIN_TIMEOUT;
-		set->flags = BLK_MQ_F_NO_SCHED;
-	} else {
-		set = &ctrl->tag_set;
-		memset(set, 0, sizeof(*set));
-		set->ops = &nvme_rdma_mq_ops;
-		set->queue_depth = nctrl->sqsize + 1;
-		set->reserved_tags = NVMF_RESERVED_TAGS;
-		set->numa_node = nctrl->numa_node;
-		set->flags = BLK_MQ_F_SHOULD_MERGE;
-		set->cmd_size = sizeof(struct nvme_rdma_request) +
-				NVME_RDMA_DATA_SGL_SIZE;
-		if (nctrl->max_integrity_segments)
-			set->cmd_size += sizeof(struct nvme_rdma_sgl) +
-					 NVME_RDMA_METADATA_SGL_SIZE;
-		set->driver_data = ctrl;
-		set->nr_hw_queues = nctrl->queue_count - 1;
-		set->timeout = NVME_IO_TIMEOUT;
-		set->nr_maps = nctrl->opts->nr_poll_queues ? HCTX_MAX_TYPES : 2;
-	}
-
+	memset(set, 0, sizeof(*set));
+	set->ops = &nvme_rdma_admin_mq_ops;
+	set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
+	set->reserved_tags = NVMF_RESERVED_TAGS;
+	set->numa_node = nctrl->numa_node;
+	set->cmd_size = sizeof(struct nvme_rdma_request) +
+			NVME_RDMA_DATA_SGL_SIZE;
+	set->driver_data = ctrl;
+	set->nr_hw_queues = 1;
+	set->timeout = NVME_ADMIN_TIMEOUT;
+	set->flags = BLK_MQ_F_NO_SCHED;
 	ret = blk_mq_alloc_tag_set(set);
-	if (ret)
-		return ERR_PTR(ret);
+	if (!ret)
+		ctrl->ctrl.admin_tagset = set;
+	return ret;
+}
 
-	return set;
+static int nvme_rdma_alloc_tag_set(struct nvme_ctrl *nctrl)
+{
+	struct nvme_rdma_ctrl *ctrl = to_rdma_ctrl(nctrl);
+	struct blk_mq_tag_set *set = &ctrl->tag_set;
+	int ret;
+
+	memset(set, 0, sizeof(*set));
+	set->ops = &nvme_rdma_mq_ops;
+	set->queue_depth = nctrl->sqsize + 1;
+	set->reserved_tags = NVMF_RESERVED_TAGS;
+	set->numa_node = nctrl->numa_node;
+	set->flags = BLK_MQ_F_SHOULD_MERGE;
+	set->cmd_size = sizeof(struct nvme_rdma_request) +
+			NVME_RDMA_DATA_SGL_SIZE;
+	if (nctrl->max_integrity_segments)
+		set->cmd_size += sizeof(struct nvme_rdma_sgl) +
+				 NVME_RDMA_METADATA_SGL_SIZE;
+	set->driver_data = ctrl;
+	set->nr_hw_queues = nctrl->queue_count - 1;
+	set->timeout = NVME_IO_TIMEOUT;
+	set->nr_maps = nctrl->opts->nr_poll_queues ? HCTX_MAX_TYPES : 2;
+	ret = blk_mq_alloc_tag_set(set);
+	if (!ret)
+		ctrl->ctrl.tagset = set;
+	return ret;
 }
 
 static void nvme_rdma_destroy_admin_queue(struct nvme_rdma_ctrl *ctrl,
@@ -882,11 +886,9 @@ static int nvme_rdma_configure_admin_queue(struct nvme_rdma_ctrl *ctrl,
 		goto out_free_queue;
 
 	if (new) {
-		ctrl->ctrl.admin_tagset = nvme_rdma_alloc_tagset(&ctrl->ctrl, true);
-		if (IS_ERR(ctrl->ctrl.admin_tagset)) {
-			error = PTR_ERR(ctrl->ctrl.admin_tagset);
+		error = nvme_rdma_alloc_admin_tag_set(&ctrl->ctrl);
+		if (error)
 			goto out_free_async_qe;
-		}
 
 		ctrl->ctrl.fabrics_q = blk_mq_init_queue(&ctrl->admin_tag_set);
 		if (IS_ERR(ctrl->ctrl.fabrics_q)) {
@@ -969,11 +971,9 @@ static int nvme_rdma_configure_io_queues(struct nvme_rdma_ctrl *ctrl, bool new)
 		return ret;
 
 	if (new) {
-		ctrl->ctrl.tagset = nvme_rdma_alloc_tagset(&ctrl->ctrl, false);
-		if (IS_ERR(ctrl->ctrl.tagset)) {
-			ret = PTR_ERR(ctrl->ctrl.tagset);
+		ret = nvme_rdma_alloc_tag_set(&ctrl->ctrl);
+		if (ret)
 			goto out_free_io_queues;
-		}
 
 		ret = nvme_ctrl_init_connect_q(&(ctrl->ctrl));
 		if (ret)
-- 
2.30.2

[PATCH 4/4] nvme-tcp: split nvme_tcp_alloc_tagset

[Christoph Hellwig]

Split nvme_tcp_alloc_tagset into one helper for the admin tag_set and
one for the I/O tag set.

Signed-off-by: Christoph Hellwig <hch@lst.de>
---
 drivers/nvme/host/tcp.c | 82 ++++++++++++++++++++---------------------
 1 file changed, 41 insertions(+), 41 deletions(-)

diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
index d6a37c374e974..e9d5b052b2bb6 100644
--- a/drivers/nvme/host/tcp.c
+++ b/drivers/nvme/host/tcp.c
@@ -1688,45 +1688,49 @@ static int nvme_tcp_start_queue(struct nvme_ctrl *nctrl, int idx)
 	return ret;
 }
 
-static struct blk_mq_tag_set *nvme_tcp_alloc_tagset(struct nvme_ctrl *nctrl,
-		bool admin)
+static int nvme_tcp_alloc_admin_tag_set(struct nvme_ctrl *nctrl)
 {
 	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
-	struct blk_mq_tag_set *set;
+	struct blk_mq_tag_set *set = &ctrl->admin_tag_set;
 	int ret;
 
-	if (admin) {
-		set = &ctrl->admin_tag_set;
-		memset(set, 0, sizeof(*set));
-		set->ops = &nvme_tcp_admin_mq_ops;
-		set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
-		set->reserved_tags = NVMF_RESERVED_TAGS;
-		set->numa_node = nctrl->numa_node;
-		set->flags = BLK_MQ_F_BLOCKING;
-		set->cmd_size = sizeof(struct nvme_tcp_request);
-		set->driver_data = ctrl;
-		set->nr_hw_queues = 1;
-		set->timeout = NVME_ADMIN_TIMEOUT;
-	} else {
-		set = &ctrl->tag_set;
-		memset(set, 0, sizeof(*set));
-		set->ops = &nvme_tcp_mq_ops;
-		set->queue_depth = nctrl->sqsize + 1;
-		set->reserved_tags = NVMF_RESERVED_TAGS;
-		set->numa_node = nctrl->numa_node;
-		set->flags = BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_BLOCKING;
-		set->cmd_size = sizeof(struct nvme_tcp_request);
-		set->driver_data = ctrl;
-		set->nr_hw_queues = nctrl->queue_count - 1;
-		set->timeout = NVME_IO_TIMEOUT;
-		set->nr_maps = nctrl->opts->nr_poll_queues ? HCTX_MAX_TYPES : 2;
-	}
-
+	memset(set, 0, sizeof(*set));
+	set->ops = &nvme_tcp_admin_mq_ops;
+	set->queue_depth = NVME_AQ_MQ_TAG_DEPTH;
+	set->reserved_tags = NVMF_RESERVED_TAGS;
+	set->numa_node = nctrl->numa_node;
+	set->flags = BLK_MQ_F_BLOCKING;
+	set->cmd_size = sizeof(struct nvme_tcp_request);
+	set->driver_data = ctrl;
+	set->nr_hw_queues = 1;
+	set->timeout = NVME_ADMIN_TIMEOUT;
 	ret = blk_mq_alloc_tag_set(set);
-	if (ret)
-		return ERR_PTR(ret);
+	if (!ret)
+		nctrl->admin_tagset = set;
+	return ret;
+}
 
-	return set;
+static int nvme_tcp_alloc_tag_set(struct nvme_ctrl *nctrl)
+{
+	struct nvme_tcp_ctrl *ctrl = to_tcp_ctrl(nctrl);
+	struct blk_mq_tag_set *set = &ctrl->tag_set;
+	int ret;
+
+	memset(set, 0, sizeof(*set));
+	set->ops = &nvme_tcp_mq_ops;
+	set->queue_depth = nctrl->sqsize + 1;
+	set->reserved_tags = NVMF_RESERVED_TAGS;
+	set->numa_node = nctrl->numa_node;
+	set->flags = BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_BLOCKING;
+	set->cmd_size = sizeof(struct nvme_tcp_request);
+	set->driver_data = ctrl;
+	set->nr_hw_queues = nctrl->queue_count - 1;
+	set->timeout = NVME_IO_TIMEOUT;
+	set->nr_maps = nctrl->opts->nr_poll_queues ? HCTX_MAX_TYPES : 2;
+	ret = blk_mq_alloc_tag_set(set);
+	if (!ret)
+		nctrl->tagset = set;
+	return ret;
 }
 
 static void nvme_tcp_free_admin_queue(struct nvme_ctrl *ctrl)
@@ -1902,11 +1906,9 @@ static int nvme_tcp_configure_io_queues(struct nvme_ctrl *ctrl, bool new)
 		return ret;
 
 	if (new) {
-		ctrl->tagset = nvme_tcp_alloc_tagset(ctrl, false);
-		if (IS_ERR(ctrl->tagset)) {
-			ret = PTR_ERR(ctrl->tagset);
+		ret = nvme_tcp_alloc_tag_set(ctrl);
+		if (ret)
 			goto out_free_io_queues;
-		}
 
 		ret = nvme_ctrl_init_connect_q(ctrl);
 		if (ret)
@@ -1971,11 +1973,9 @@ static int nvme_tcp_configure_admin_queue(struct nvme_ctrl *ctrl, bool new)
 		return error;
 
 	if (new) {
-		ctrl->admin_tagset = nvme_tcp_alloc_tagset(ctrl, true);
-		if (IS_ERR(ctrl->admin_tagset)) {
-			error = PTR_ERR(ctrl->admin_tagset);
+		error = nvme_tcp_alloc_admin_tag_set(ctrl);
+		if (error)
 			goto out_free_queue;
-		}
 
 		ctrl->fabrics_q = blk_mq_init_queue(ctrl->admin_tagset);
 		if (IS_ERR(ctrl->fabrics_q)) {
-- 
2.30.2

Re: [PATCH 1/4] nvme-pci: split nvme_alloc_admin_tags

[Keith Busch]

Series looks good.

Reviewed-by: Keith Busch <kbusch@kernel.org>

Re: [PATCH 1/4] nvme-pci: split nvme_alloc_admin_tags

[Sagi Grimberg]

Series looks good,

Reviewed-by: Sagi Grimberg <sagi@grimberg.me>