[PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Keith Busch]

The PCI interrupt vectors intended to be associated with a queue may
not start at 0; a driver may allocate pre_vectors for special use. This
patch adds an offset parameter so blk-mq may find the intended affinity
mask and updates all drivers using this API accordingly.

Cc: Don Brace <don.brace@microsemi.com>
Cc: <qla2xxx-upstream@qlogic.com>
Cc: <linux-scsi@vger.kernel.org>
Signed-off-by: Keith Busch <keith.busch@intel.com>
---
v1 -> v2:

  Update blk-mq API directly instead of chaining a default parameter to
  a new API, and update all drivers accordingly.

 block/blk-mq-pci.c                    | 6 ++++--
 drivers/nvme/host/pci.c               | 2 +-
 drivers/scsi/qla2xxx/qla_os.c         | 2 +-
 drivers/scsi/smartpqi/smartpqi_init.c | 2 +-
 include/linux/blk-mq-pci.h            | 3 ++-
 5 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/block/blk-mq-pci.c b/block/blk-mq-pci.c
index 76944e3271bf..e233996bb76f 100644
--- a/block/blk-mq-pci.c
+++ b/block/blk-mq-pci.c
@@ -21,6 +21,7 @@
  * blk_mq_pci_map_queues - provide a default queue mapping for PCI device
  * @set:	tagset to provide the mapping for
  * @pdev:	PCI device associated with @set.
+ * @offset:	Offset to use for the pci irq vector
  *
  * This function assumes the PCI device @pdev has at least as many available
  * interrupt vectors as @set has queues.  It will then query the vector
@@ -28,13 +29,14 @@
  * that maps a queue to the CPUs that have irq affinity for the corresponding
  * vector.
  */
-int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev)
+int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
+			    int offset)
 {
 	const struct cpumask *mask;
 	unsigned int queue, cpu;
 
 	for (queue = 0; queue < set->nr_hw_queues; queue++) {
-		mask = pci_irq_get_affinity(pdev, queue);
+		mask = pci_irq_get_affinity(pdev, queue + offset);
 		if (!mask)
 			goto fallback;
 
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index cef5ce851a92..e3b9efca0571 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -414,7 +414,7 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 
-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev));
+	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
 }
 
 /**
diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
index 12ee6e02d146..2c705f3dd265 100644
--- a/drivers/scsi/qla2xxx/qla_os.c
+++ b/drivers/scsi/qla2xxx/qla_os.c
@@ -6805,7 +6805,7 @@ static int qla2xxx_map_queues(struct Scsi_Host *shost)
 	if (USER_CTRL_IRQ(vha->hw))
 		rc = blk_mq_map_queues(&shost->tag_set);
 	else
-		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev);
+		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev, 0);
 	return rc;
 }
 
diff --git a/drivers/scsi/smartpqi/smartpqi_init.c b/drivers/scsi/smartpqi/smartpqi_init.c
index b2880c7709e6..10c94011c8a8 100644
--- a/drivers/scsi/smartpqi/smartpqi_init.c
+++ b/drivers/scsi/smartpqi/smartpqi_init.c
@@ -5348,7 +5348,7 @@ static int pqi_map_queues(struct Scsi_Host *shost)
 {
 	struct pqi_ctrl_info *ctrl_info = shost_to_hba(shost);
 
-	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev);
+	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev, 0);
 }
 
 static int pqi_getpciinfo_ioctl(struct pqi_ctrl_info *ctrl_info,
diff --git a/include/linux/blk-mq-pci.h b/include/linux/blk-mq-pci.h
index 6338551e0fb9..9f4c17f0d2d8 100644
--- a/include/linux/blk-mq-pci.h
+++ b/include/linux/blk-mq-pci.h
@@ -5,6 +5,7 @@
 struct blk_mq_tag_set;
 struct pci_dev;
 
-int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev);
+int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
+			  int offset);
 
 #endif /* _LINUX_BLK_MQ_PCI_H */
-- 
2.14.3

[PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Keith Busch]

The PCI interrupt vectors intended to be associated with a queue may
not start at 0; a driver may allocate pre_vectors for special use. This
patch adds an offset parameter so blk-mq may find the intended affinity
mask and updates all drivers using this API accordingly.

Cc: Don Brace <don.brace at microsemi.com>
Cc: <qla2xxx-upstream at qlogic.com>
Cc: <linux-scsi at vger.kernel.org>
Signed-off-by: Keith Busch <keith.busch at intel.com>
---
v1 -> v2:

  Update blk-mq API directly instead of chaining a default parameter to
  a new API, and update all drivers accordingly.

 block/blk-mq-pci.c                    | 6 ++++--
 drivers/nvme/host/pci.c               | 2 +-
 drivers/scsi/qla2xxx/qla_os.c         | 2 +-
 drivers/scsi/smartpqi/smartpqi_init.c | 2 +-
 include/linux/blk-mq-pci.h            | 3 ++-
 5 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/block/blk-mq-pci.c b/block/blk-mq-pci.c
index 76944e3271bf..e233996bb76f 100644
--- a/block/blk-mq-pci.c
+++ b/block/blk-mq-pci.c
@@ -21,6 +21,7 @@
  * blk_mq_pci_map_queues - provide a default queue mapping for PCI device
  * @set:	tagset to provide the mapping for
  * @pdev:	PCI device associated with @set.
+ * @offset:	Offset to use for the pci irq vector
  *
  * This function assumes the PCI device @pdev has at least as many available
  * interrupt vectors as @set has queues.  It will then query the vector
@@ -28,13 +29,14 @@
  * that maps a queue to the CPUs that have irq affinity for the corresponding
  * vector.
  */
-int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev)
+int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
+			    int offset)
 {
 	const struct cpumask *mask;
 	unsigned int queue, cpu;
 
 	for (queue = 0; queue < set->nr_hw_queues; queue++) {
-		mask = pci_irq_get_affinity(pdev, queue);
+		mask = pci_irq_get_affinity(pdev, queue + offset);
 		if (!mask)
 			goto fallback;
 
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index cef5ce851a92..e3b9efca0571 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -414,7 +414,7 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 
-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev));
+	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
 }
 
 /**
diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
index 12ee6e02d146..2c705f3dd265 100644
--- a/drivers/scsi/qla2xxx/qla_os.c
+++ b/drivers/scsi/qla2xxx/qla_os.c
@@ -6805,7 +6805,7 @@ static int qla2xxx_map_queues(struct Scsi_Host *shost)
 	if (USER_CTRL_IRQ(vha->hw))
 		rc = blk_mq_map_queues(&shost->tag_set);
 	else
-		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev);
+		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev, 0);
 	return rc;
 }
 
diff --git a/drivers/scsi/smartpqi/smartpqi_init.c b/drivers/scsi/smartpqi/smartpqi_init.c
index b2880c7709e6..10c94011c8a8 100644
--- a/drivers/scsi/smartpqi/smartpqi_init.c
+++ b/drivers/scsi/smartpqi/smartpqi_init.c
@@ -5348,7 +5348,7 @@ static int pqi_map_queues(struct Scsi_Host *shost)
 {
 	struct pqi_ctrl_info *ctrl_info = shost_to_hba(shost);
 
-	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev);
+	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev, 0);
 }
 
 static int pqi_getpciinfo_ioctl(struct pqi_ctrl_info *ctrl_info,
diff --git a/include/linux/blk-mq-pci.h b/include/linux/blk-mq-pci.h
index 6338551e0fb9..9f4c17f0d2d8 100644
--- a/include/linux/blk-mq-pci.h
+++ b/include/linux/blk-mq-pci.h
@@ -5,6 +5,7 @@
 struct blk_mq_tag_set;
 struct pci_dev;
 
-int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev);
+int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
+			  int offset);
 
 #endif /* _LINUX_BLK_MQ_PCI_H */
-- 
2.14.3

[PATCH 2/3] nvme-pci: Remove unused queue parameter

[Keith Busch]

All the queue memory is allocated up front. We don't take the node
into consideration when creating queues anymore, so removing the unused
parameter.

Signed-off-by: Keith Busch <keith.busch@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
v1 -> v2:

   Added review.

 drivers/nvme/host/pci.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index e3b9efca0571..cc47fbe32ea5 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1379,8 +1379,7 @@ static int nvme_alloc_sq_cmds(struct nvme_dev *dev, struct nvme_queue *nvmeq,
 	return 0;
 }
 
-static int nvme_alloc_queue(struct nvme_dev *dev, int qid,
-		int depth, int node)
+static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
 {
 	struct nvme_queue *nvmeq = &dev->queues[qid];
 
@@ -1595,8 +1594,7 @@ static int nvme_pci_configure_admin_queue(struct nvme_dev *dev)
 	if (result < 0)
 		return result;
 
-	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH,
-			dev_to_node(dev->dev));
+	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH);
 	if (result)
 		return result;
 
@@ -1629,9 +1627,7 @@ static int nvme_create_io_queues(struct nvme_dev *dev)
 	int ret = 0;
 
 	for (i = dev->ctrl.queue_count; i <= dev->max_qid; i++) {
-		/* vector == qid - 1, match nvme_create_queue */
-		if (nvme_alloc_queue(dev, i, dev->q_depth,
-		     pci_irq_get_node(to_pci_dev(dev->dev), i - 1))) {
+		if (nvme_alloc_queue(dev, i, dev->q_depth)) {
 			ret = -ENOMEM;
 			break;
 		}
-- 
2.14.3

[PATCH 2/3] nvme-pci: Remove unused queue parameter

[Keith Busch]

All the queue memory is allocated up front. We don't take the node
into consideration when creating queues anymore, so removing the unused
parameter.

Signed-off-by: Keith Busch <keith.busch at intel.com>
Reviewed-by: Christoph Hellwig <hch at lst.de>
---
v1 -> v2:

   Added review.

 drivers/nvme/host/pci.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index e3b9efca0571..cc47fbe32ea5 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1379,8 +1379,7 @@ static int nvme_alloc_sq_cmds(struct nvme_dev *dev, struct nvme_queue *nvmeq,
 	return 0;
 }
 
-static int nvme_alloc_queue(struct nvme_dev *dev, int qid,
-		int depth, int node)
+static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
 {
 	struct nvme_queue *nvmeq = &dev->queues[qid];
 
@@ -1595,8 +1594,7 @@ static int nvme_pci_configure_admin_queue(struct nvme_dev *dev)
 	if (result < 0)
 		return result;
 
-	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH,
-			dev_to_node(dev->dev));
+	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH);
 	if (result)
 		return result;
 
@@ -1629,9 +1627,7 @@ static int nvme_create_io_queues(struct nvme_dev *dev)
 	int ret = 0;
 
 	for (i = dev->ctrl.queue_count; i <= dev->max_qid; i++) {
-		/* vector == qid - 1, match nvme_create_queue */
-		if (nvme_alloc_queue(dev, i, dev->q_depth,
-		     pci_irq_get_node(to_pci_dev(dev->dev), i - 1))) {
+		if (nvme_alloc_queue(dev, i, dev->q_depth)) {
 			ret = -ENOMEM;
 			break;
 		}
-- 
2.14.3

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Keith Busch]

The admin and first IO queues shared the first irq vector, which has an
affinity mask including cpu0. If a system allows cpu0 to be offlined,
the admin queue may not be usable if no other CPUs in the affinity mask
are online. This is a problem since unlike IO queues, there is only
one admin queue that always needs to be usable.

To fix, this patch allocates one pre_vector for the admin queue that
is assigned all CPUs, so will always be accessible. The IO queues are
assigned the remaining managed vectors.

In case a controller has only one interrupt vector available, the admin
and IO queues will share the pre_vector with all CPUs assigned.

Cc: Jianchao Wang <jianchao.w.wang@oracle.com>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Keith Busch <keith.busch@intel.com>
---
v1 -> v2:

  Update to use new blk-mq API.

  Removed unnecessary braces, inline functions, and temp variables.

  Amended author (this has evolved significantly from the original).

 drivers/nvme/host/pci.c | 23 +++++++++++++++++------
 1 file changed, 17 insertions(+), 6 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index cc47fbe32ea5..50c8eaf51d92 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -84,6 +84,7 @@ struct nvme_dev {
 	struct dma_pool *prp_small_pool;
 	unsigned online_queues;
 	unsigned max_qid;
+	unsigned int num_vecs;
 	int q_depth;
 	u32 db_stride;
 	void __iomem *bar;
@@ -414,7 +415,8 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 
-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
+	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
+				     dev->num_vecs > 1);
 }
 
 /**
@@ -1455,7 +1457,11 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
 		nvmeq->sq_cmds_io = dev->cmb + offset;
 	}
 
-	nvmeq->cq_vector = qid - 1;
+	/*
+	 * A queue's vector matches the queue identifier unless the controller
+	 * has only one vector available.
+	 */
+	nvmeq->cq_vector = dev->num_vecs == 1 ? 0 : qid;
 	result = adapter_alloc_cq(dev, qid, nvmeq);
 	if (result < 0)
 		goto release_vector;
@@ -1909,6 +1915,10 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	int result, nr_io_queues;
 	unsigned long size;
 
+	struct irq_affinity affd = {
+		.pre_vectors = 1
+	};
+
 	nr_io_queues = num_present_cpus();
 	result = nvme_set_queue_count(&dev->ctrl, &nr_io_queues);
 	if (result < 0)
@@ -1944,11 +1954,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	 * setting up the full range we need.
 	 */
 	pci_free_irq_vectors(pdev);
-	nr_io_queues = pci_alloc_irq_vectors(pdev, 1, nr_io_queues,
-			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY);
-	if (nr_io_queues <= 0)
+	result = pci_alloc_irq_vectors_affinity(pdev, 1, nr_io_queues + 1,
+			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
+	if (result <= 0)
 		return -EIO;
-	dev->max_qid = nr_io_queues;
+	dev->num_vecs = result;
+	dev->max_qid = max(result - 1, 1);
 
 	/*
 	 * Should investigate if there's a performance win from allocating
-- 
2.14.3

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Keith Busch]

The admin and first IO queues shared the first irq vector, which has an
affinity mask including cpu0. If a system allows cpu0 to be offlined,
the admin queue may not be usable if no other CPUs in the affinity mask
are online. This is a problem since unlike IO queues, there is only
one admin queue that always needs to be usable.

To fix, this patch allocates one pre_vector for the admin queue that
is assigned all CPUs, so will always be accessible. The IO queues are
assigned the remaining managed vectors.

In case a controller has only one interrupt vector available, the admin
and IO queues will share the pre_vector with all CPUs assigned.

Cc: Jianchao Wang <jianchao.w.wang at oracle.com>
Cc: Ming Lei <ming.lei at redhat.com>
Signed-off-by: Keith Busch <keith.busch at intel.com>
---
v1 -> v2:

  Update to use new blk-mq API.

  Removed unnecessary braces, inline functions, and temp variables.

  Amended author (this has evolved significantly from the original).

 drivers/nvme/host/pci.c | 23 +++++++++++++++++------
 1 file changed, 17 insertions(+), 6 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index cc47fbe32ea5..50c8eaf51d92 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -84,6 +84,7 @@ struct nvme_dev {
 	struct dma_pool *prp_small_pool;
 	unsigned online_queues;
 	unsigned max_qid;
+	unsigned int num_vecs;
 	int q_depth;
 	u32 db_stride;
 	void __iomem *bar;
@@ -414,7 +415,8 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 
-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
+	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
+				     dev->num_vecs > 1);
 }
 
 /**
@@ -1455,7 +1457,11 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
 		nvmeq->sq_cmds_io = dev->cmb + offset;
 	}
 
-	nvmeq->cq_vector = qid - 1;
+	/*
+	 * A queue's vector matches the queue identifier unless the controller
+	 * has only one vector available.
+	 */
+	nvmeq->cq_vector = dev->num_vecs == 1 ? 0 : qid;
 	result = adapter_alloc_cq(dev, qid, nvmeq);
 	if (result < 0)
 		goto release_vector;
@@ -1909,6 +1915,10 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	int result, nr_io_queues;
 	unsigned long size;
 
+	struct irq_affinity affd = {
+		.pre_vectors = 1
+	};
+
 	nr_io_queues = num_present_cpus();
 	result = nvme_set_queue_count(&dev->ctrl, &nr_io_queues);
 	if (result < 0)
@@ -1944,11 +1954,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	 * setting up the full range we need.
 	 */
 	pci_free_irq_vectors(pdev);
-	nr_io_queues = pci_alloc_irq_vectors(pdev, 1, nr_io_queues,
-			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY);
-	if (nr_io_queues <= 0)
+	result = pci_alloc_irq_vectors_affinity(pdev, 1, nr_io_queues + 1,
+			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
+	if (result <= 0)
 		return -EIO;
-	dev->max_qid = nr_io_queues;
+	dev->num_vecs = result;
+	dev->max_qid = max(result - 1, 1);
 
 	/*
 	 * Should investigate if there's a performance win from allocating
-- 
2.14.3

Re: [PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Ming Lei]

On Tue, Mar 27, 2018 at 09:39:06AM -0600, Keith Busch wrote:
> The PCI interrupt vectors intended to be associated with a queue may
> not start at 0; a driver may allocate pre_vectors for special use. This
> patch adds an offset parameter so blk-mq may find the intended affinity
> mask and updates all drivers using this API accordingly.
> 
> Cc: Don Brace <don.brace@microsemi.com>
> Cc: <qla2xxx-upstream@qlogic.com>
> Cc: <linux-scsi@vger.kernel.org>
> Signed-off-by: Keith Busch <keith.busch@intel.com>
> ---
> v1 -> v2:
> 
>   Update blk-mq API directly instead of chaining a default parameter to
>   a new API, and update all drivers accordingly.
> 
>  block/blk-mq-pci.c                    | 6 ++++--
>  drivers/nvme/host/pci.c               | 2 +-
>  drivers/scsi/qla2xxx/qla_os.c         | 2 +-
>  drivers/scsi/smartpqi/smartpqi_init.c | 2 +-
>  include/linux/blk-mq-pci.h            | 3 ++-
>  5 files changed, 9 insertions(+), 6 deletions(-)
> 
> diff --git a/block/blk-mq-pci.c b/block/blk-mq-pci.c
> index 76944e3271bf..e233996bb76f 100644
> --- a/block/blk-mq-pci.c
> +++ b/block/blk-mq-pci.c
> @@ -21,6 +21,7 @@
>   * blk_mq_pci_map_queues - provide a default queue mapping for PCI device
>   * @set:	tagset to provide the mapping for
>   * @pdev:	PCI device associated with @set.
> + * @offset:	Offset to use for the pci irq vector
>   *
>   * This function assumes the PCI device @pdev has at least as many available
>   * interrupt vectors as @set has queues.  It will then query the vector
> @@ -28,13 +29,14 @@
>   * that maps a queue to the CPUs that have irq affinity for the corresponding
>   * vector.
>   */
> -int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev)
> +int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
> +			    int offset)
>  {
>  	const struct cpumask *mask;
>  	unsigned int queue, cpu;
>  
>  	for (queue = 0; queue < set->nr_hw_queues; queue++) {
> -		mask = pci_irq_get_affinity(pdev, queue);
> +		mask = pci_irq_get_affinity(pdev, queue + offset);
>  		if (!mask)
>  			goto fallback;
>  
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index cef5ce851a92..e3b9efca0571 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -414,7 +414,7 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
>  {
>  	struct nvme_dev *dev = set->driver_data;
>  
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev));
> +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
>  }
>  
>  /**
> diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
> index 12ee6e02d146..2c705f3dd265 100644
> --- a/drivers/scsi/qla2xxx/qla_os.c
> +++ b/drivers/scsi/qla2xxx/qla_os.c
> @@ -6805,7 +6805,7 @@ static int qla2xxx_map_queues(struct Scsi_Host *shost)
>  	if (USER_CTRL_IRQ(vha->hw))
>  		rc = blk_mq_map_queues(&shost->tag_set);
>  	else
> -		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev);
> +		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev, 0);
>  	return rc;
>  }
>  
> diff --git a/drivers/scsi/smartpqi/smartpqi_init.c b/drivers/scsi/smartpqi/smartpqi_init.c
> index b2880c7709e6..10c94011c8a8 100644
> --- a/drivers/scsi/smartpqi/smartpqi_init.c
> +++ b/drivers/scsi/smartpqi/smartpqi_init.c
> @@ -5348,7 +5348,7 @@ static int pqi_map_queues(struct Scsi_Host *shost)
>  {
>  	struct pqi_ctrl_info *ctrl_info = shost_to_hba(shost);
>  
> -	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev);
> +	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev, 0);
>  }
>  
>  static int pqi_getpciinfo_ioctl(struct pqi_ctrl_info *ctrl_info,
> diff --git a/include/linux/blk-mq-pci.h b/include/linux/blk-mq-pci.h
> index 6338551e0fb9..9f4c17f0d2d8 100644
> --- a/include/linux/blk-mq-pci.h
> +++ b/include/linux/blk-mq-pci.h
> @@ -5,6 +5,7 @@
>  struct blk_mq_tag_set;
>  struct pci_dev;
>  
> -int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev);
> +int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
> +			  int offset);
>  
>  #endif /* _LINUX_BLK_MQ_PCI_H */
> -- 
> 2.14.3
> 

Reviewed-by: Ming Lei <ming.lei@redhat.com>

-- 
Ming

[PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Ming Lei]

On Tue, Mar 27, 2018@09:39:06AM -0600, Keith Busch wrote:
> The PCI interrupt vectors intended to be associated with a queue may
> not start at 0; a driver may allocate pre_vectors for special use. This
> patch adds an offset parameter so blk-mq may find the intended affinity
> mask and updates all drivers using this API accordingly.
> 
> Cc: Don Brace <don.brace at microsemi.com>
> Cc: <qla2xxx-upstream at qlogic.com>
> Cc: <linux-scsi at vger.kernel.org>
> Signed-off-by: Keith Busch <keith.busch at intel.com>
> ---
> v1 -> v2:
> 
>   Update blk-mq API directly instead of chaining a default parameter to
>   a new API, and update all drivers accordingly.
> 
>  block/blk-mq-pci.c                    | 6 ++++--
>  drivers/nvme/host/pci.c               | 2 +-
>  drivers/scsi/qla2xxx/qla_os.c         | 2 +-
>  drivers/scsi/smartpqi/smartpqi_init.c | 2 +-
>  include/linux/blk-mq-pci.h            | 3 ++-
>  5 files changed, 9 insertions(+), 6 deletions(-)
> 
> diff --git a/block/blk-mq-pci.c b/block/blk-mq-pci.c
> index 76944e3271bf..e233996bb76f 100644
> --- a/block/blk-mq-pci.c
> +++ b/block/blk-mq-pci.c
> @@ -21,6 +21,7 @@
>   * blk_mq_pci_map_queues - provide a default queue mapping for PCI device
>   * @set:	tagset to provide the mapping for
>   * @pdev:	PCI device associated with @set.
> + * @offset:	Offset to use for the pci irq vector
>   *
>   * This function assumes the PCI device @pdev has at least as many available
>   * interrupt vectors as @set has queues.  It will then query the vector
> @@ -28,13 +29,14 @@
>   * that maps a queue to the CPUs that have irq affinity for the corresponding
>   * vector.
>   */
> -int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev)
> +int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
> +			    int offset)
>  {
>  	const struct cpumask *mask;
>  	unsigned int queue, cpu;
>  
>  	for (queue = 0; queue < set->nr_hw_queues; queue++) {
> -		mask = pci_irq_get_affinity(pdev, queue);
> +		mask = pci_irq_get_affinity(pdev, queue + offset);
>  		if (!mask)
>  			goto fallback;
>  
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index cef5ce851a92..e3b9efca0571 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -414,7 +414,7 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
>  {
>  	struct nvme_dev *dev = set->driver_data;
>  
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev));
> +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
>  }
>  
>  /**
> diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
> index 12ee6e02d146..2c705f3dd265 100644
> --- a/drivers/scsi/qla2xxx/qla_os.c
> +++ b/drivers/scsi/qla2xxx/qla_os.c
> @@ -6805,7 +6805,7 @@ static int qla2xxx_map_queues(struct Scsi_Host *shost)
>  	if (USER_CTRL_IRQ(vha->hw))
>  		rc = blk_mq_map_queues(&shost->tag_set);
>  	else
> -		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev);
> +		rc = blk_mq_pci_map_queues(&shost->tag_set, vha->hw->pdev, 0);
>  	return rc;
>  }
>  
> diff --git a/drivers/scsi/smartpqi/smartpqi_init.c b/drivers/scsi/smartpqi/smartpqi_init.c
> index b2880c7709e6..10c94011c8a8 100644
> --- a/drivers/scsi/smartpqi/smartpqi_init.c
> +++ b/drivers/scsi/smartpqi/smartpqi_init.c
> @@ -5348,7 +5348,7 @@ static int pqi_map_queues(struct Scsi_Host *shost)
>  {
>  	struct pqi_ctrl_info *ctrl_info = shost_to_hba(shost);
>  
> -	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev);
> +	return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev, 0);
>  }
>  
>  static int pqi_getpciinfo_ioctl(struct pqi_ctrl_info *ctrl_info,
> diff --git a/include/linux/blk-mq-pci.h b/include/linux/blk-mq-pci.h
> index 6338551e0fb9..9f4c17f0d2d8 100644
> --- a/include/linux/blk-mq-pci.h
> +++ b/include/linux/blk-mq-pci.h
> @@ -5,6 +5,7 @@
>  struct blk_mq_tag_set;
>  struct pci_dev;
>  
> -int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev);
> +int blk_mq_pci_map_queues(struct blk_mq_tag_set *set, struct pci_dev *pdev,
> +			  int offset);
>  
>  #endif /* _LINUX_BLK_MQ_PCI_H */
> -- 
> 2.14.3
> 

Reviewed-by: Ming Lei <ming.lei at redhat.com>

-- 
Ming

Re: [PATCH 2/3] nvme-pci: Remove unused queue parameter

[Ming Lei]

On Tue, Mar 27, 2018 at 09:39:07AM -0600, Keith Busch wrote:
> All the queue memory is allocated up front. We don't take the node
> into consideration when creating queues anymore, so removing the unused
> parameter.
> 
> Signed-off-by: Keith Busch <keith.busch@intel.com>
> Reviewed-by: Christoph Hellwig <hch@lst.de>
> ---
> v1 -> v2:
> 
>    Added review.
> 
>  drivers/nvme/host/pci.c | 10 +++-------
>  1 file changed, 3 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index e3b9efca0571..cc47fbe32ea5 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -1379,8 +1379,7 @@ static int nvme_alloc_sq_cmds(struct nvme_dev *dev, struct nvme_queue *nvmeq,
>  	return 0;
>  }
>  
> -static int nvme_alloc_queue(struct nvme_dev *dev, int qid,
> -		int depth, int node)
> +static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
>  {
>  	struct nvme_queue *nvmeq = &dev->queues[qid];
>  
> @@ -1595,8 +1594,7 @@ static int nvme_pci_configure_admin_queue(struct nvme_dev *dev)
>  	if (result < 0)
>  		return result;
>  
> -	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH,
> -			dev_to_node(dev->dev));
> +	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH);
>  	if (result)
>  		return result;
>  
> @@ -1629,9 +1627,7 @@ static int nvme_create_io_queues(struct nvme_dev *dev)
>  	int ret = 0;
>  
>  	for (i = dev->ctrl.queue_count; i <= dev->max_qid; i++) {
> -		/* vector == qid - 1, match nvme_create_queue */
> -		if (nvme_alloc_queue(dev, i, dev->q_depth,
> -		     pci_irq_get_node(to_pci_dev(dev->dev), i - 1))) {
> +		if (nvme_alloc_queue(dev, i, dev->q_depth)) {
>  			ret = -ENOMEM;
>  			break;
>  		}
> -- 
> 2.14.3
> 

Reviewed-by: Ming Lei <ming.lei@redhat.com>

-- 
Ming

[PATCH 2/3] nvme-pci: Remove unused queue parameter

[Ming Lei]

On Tue, Mar 27, 2018@09:39:07AM -0600, Keith Busch wrote:
> All the queue memory is allocated up front. We don't take the node
> into consideration when creating queues anymore, so removing the unused
> parameter.
> 
> Signed-off-by: Keith Busch <keith.busch at intel.com>
> Reviewed-by: Christoph Hellwig <hch at lst.de>
> ---
> v1 -> v2:
> 
>    Added review.
> 
>  drivers/nvme/host/pci.c | 10 +++-------
>  1 file changed, 3 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index e3b9efca0571..cc47fbe32ea5 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -1379,8 +1379,7 @@ static int nvme_alloc_sq_cmds(struct nvme_dev *dev, struct nvme_queue *nvmeq,
>  	return 0;
>  }
>  
> -static int nvme_alloc_queue(struct nvme_dev *dev, int qid,
> -		int depth, int node)
> +static int nvme_alloc_queue(struct nvme_dev *dev, int qid, int depth)
>  {
>  	struct nvme_queue *nvmeq = &dev->queues[qid];
>  
> @@ -1595,8 +1594,7 @@ static int nvme_pci_configure_admin_queue(struct nvme_dev *dev)
>  	if (result < 0)
>  		return result;
>  
> -	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH,
> -			dev_to_node(dev->dev));
> +	result = nvme_alloc_queue(dev, 0, NVME_AQ_DEPTH);
>  	if (result)
>  		return result;
>  
> @@ -1629,9 +1627,7 @@ static int nvme_create_io_queues(struct nvme_dev *dev)
>  	int ret = 0;
>  
>  	for (i = dev->ctrl.queue_count; i <= dev->max_qid; i++) {
> -		/* vector == qid - 1, match nvme_create_queue */
> -		if (nvme_alloc_queue(dev, i, dev->q_depth,
> -		     pci_irq_get_node(to_pci_dev(dev->dev), i - 1))) {
> +		if (nvme_alloc_queue(dev, i, dev->q_depth)) {
>  			ret = -ENOMEM;
>  			break;
>  		}
> -- 
> 2.14.3
> 

Reviewed-by: Ming Lei <ming.lei at redhat.com>

-- 
Ming

Re: [PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Ming Lei]

On Tue, Mar 27, 2018 at 09:39:08AM -0600, Keith Busch wrote:
> The admin and first IO queues shared the first irq vector, which has an
> affinity mask including cpu0. If a system allows cpu0 to be offlined,
> the admin queue may not be usable if no other CPUs in the affinity mask
> are online. This is a problem since unlike IO queues, there is only
> one admin queue that always needs to be usable.
> 
> To fix, this patch allocates one pre_vector for the admin queue that
> is assigned all CPUs, so will always be accessible. The IO queues are
> assigned the remaining managed vectors.
> 
> In case a controller has only one interrupt vector available, the admin
> and IO queues will share the pre_vector with all CPUs assigned.
> 
> Cc: Jianchao Wang <jianchao.w.wang@oracle.com>
> Cc: Ming Lei <ming.lei@redhat.com>
> Signed-off-by: Keith Busch <keith.busch@intel.com>
> ---
> v1 -> v2:
> 
>   Update to use new blk-mq API.
> 
>   Removed unnecessary braces, inline functions, and temp variables.
> 
>   Amended author (this has evolved significantly from the original).
> 
>  drivers/nvme/host/pci.c | 23 +++++++++++++++++------
>  1 file changed, 17 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index cc47fbe32ea5..50c8eaf51d92 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -84,6 +84,7 @@ struct nvme_dev {
>  	struct dma_pool *prp_small_pool;
>  	unsigned online_queues;
>  	unsigned max_qid;
> +	unsigned int num_vecs;
>  	int q_depth;
>  	u32 db_stride;
>  	void __iomem *bar;
> @@ -414,7 +415,8 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
>  {
>  	struct nvme_dev *dev = set->driver_data;
>  
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
> +				     dev->num_vecs > 1);
>  }
>  
>  /**
> @@ -1455,7 +1457,11 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
>  		nvmeq->sq_cmds_io = dev->cmb + offset;
>  	}
>  
> -	nvmeq->cq_vector = qid - 1;
> +	/*
> +	 * A queue's vector matches the queue identifier unless the controller
> +	 * has only one vector available.
> +	 */
> +	nvmeq->cq_vector = dev->num_vecs == 1 ? 0 : qid;
>  	result = adapter_alloc_cq(dev, qid, nvmeq);
>  	if (result < 0)
>  		goto release_vector;
> @@ -1909,6 +1915,10 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
>  	int result, nr_io_queues;
>  	unsigned long size;
>  
> +	struct irq_affinity affd = {
> +		.pre_vectors = 1
> +	};
> +
>  	nr_io_queues = num_present_cpus();
>  	result = nvme_set_queue_count(&dev->ctrl, &nr_io_queues);
>  	if (result < 0)
> @@ -1944,11 +1954,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
>  	 * setting up the full range we need.
>  	 */
>  	pci_free_irq_vectors(pdev);
> -	nr_io_queues = pci_alloc_irq_vectors(pdev, 1, nr_io_queues,
> -			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY);
> -	if (nr_io_queues <= 0)
> +	result = pci_alloc_irq_vectors_affinity(pdev, 1, nr_io_queues + 1,
> +			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
> +	if (result <= 0)
>  		return -EIO;
> -	dev->max_qid = nr_io_queues;
> +	dev->num_vecs = result;
> +	dev->max_qid = max(result - 1, 1);
>  
>  	/*
>  	 * Should investigate if there's a performance win from allocating
> -- 
> 2.14.3
> 

Reviewed-by: Ming Lei <ming.lei@redhat.com>

-- 
Ming

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Ming Lei]

On Tue, Mar 27, 2018@09:39:08AM -0600, Keith Busch wrote:
> The admin and first IO queues shared the first irq vector, which has an
> affinity mask including cpu0. If a system allows cpu0 to be offlined,
> the admin queue may not be usable if no other CPUs in the affinity mask
> are online. This is a problem since unlike IO queues, there is only
> one admin queue that always needs to be usable.
> 
> To fix, this patch allocates one pre_vector for the admin queue that
> is assigned all CPUs, so will always be accessible. The IO queues are
> assigned the remaining managed vectors.
> 
> In case a controller has only one interrupt vector available, the admin
> and IO queues will share the pre_vector with all CPUs assigned.
> 
> Cc: Jianchao Wang <jianchao.w.wang at oracle.com>
> Cc: Ming Lei <ming.lei at redhat.com>
> Signed-off-by: Keith Busch <keith.busch at intel.com>
> ---
> v1 -> v2:
> 
>   Update to use new blk-mq API.
> 
>   Removed unnecessary braces, inline functions, and temp variables.
> 
>   Amended author (this has evolved significantly from the original).
> 
>  drivers/nvme/host/pci.c | 23 +++++++++++++++++------
>  1 file changed, 17 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
> index cc47fbe32ea5..50c8eaf51d92 100644
> --- a/drivers/nvme/host/pci.c
> +++ b/drivers/nvme/host/pci.c
> @@ -84,6 +84,7 @@ struct nvme_dev {
>  	struct dma_pool *prp_small_pool;
>  	unsigned online_queues;
>  	unsigned max_qid;
> +	unsigned int num_vecs;
>  	int q_depth;
>  	u32 db_stride;
>  	void __iomem *bar;
> @@ -414,7 +415,8 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
>  {
>  	struct nvme_dev *dev = set->driver_data;
>  
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
> +				     dev->num_vecs > 1);
>  }
>  
>  /**
> @@ -1455,7 +1457,11 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
>  		nvmeq->sq_cmds_io = dev->cmb + offset;
>  	}
>  
> -	nvmeq->cq_vector = qid - 1;
> +	/*
> +	 * A queue's vector matches the queue identifier unless the controller
> +	 * has only one vector available.
> +	 */
> +	nvmeq->cq_vector = dev->num_vecs == 1 ? 0 : qid;
>  	result = adapter_alloc_cq(dev, qid, nvmeq);
>  	if (result < 0)
>  		goto release_vector;
> @@ -1909,6 +1915,10 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
>  	int result, nr_io_queues;
>  	unsigned long size;
>  
> +	struct irq_affinity affd = {
> +		.pre_vectors = 1
> +	};
> +
>  	nr_io_queues = num_present_cpus();
>  	result = nvme_set_queue_count(&dev->ctrl, &nr_io_queues);
>  	if (result < 0)
> @@ -1944,11 +1954,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
>  	 * setting up the full range we need.
>  	 */
>  	pci_free_irq_vectors(pdev);
> -	nr_io_queues = pci_alloc_irq_vectors(pdev, 1, nr_io_queues,
> -			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY);
> -	if (nr_io_queues <= 0)
> +	result = pci_alloc_irq_vectors_affinity(pdev, 1, nr_io_queues + 1,
> +			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
> +	if (result <= 0)
>  		return -EIO;
> -	dev->max_qid = nr_io_queues;
> +	dev->num_vecs = result;
> +	dev->max_qid = max(result - 1, 1);
>  
>  	/*
>  	 * Should investigate if there's a performance win from allocating
> -- 
> 2.14.3
> 

Reviewed-by: Ming Lei <ming.lei at redhat.com>

-- 
Ming

Re: [PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Jens Axboe]

On 3/27/18 9:39 AM, Keith Busch wrote:
> The PCI interrupt vectors intended to be associated with a queue may
> not start at 0; a driver may allocate pre_vectors for special use. This
> patch adds an offset parameter so blk-mq may find the intended affinity
> mask and updates all drivers using this API accordingly.

Looks good to me, I'll queue it up for 4.17.

-- 
Jens Axboe

[PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Jens Axboe]

On 3/27/18 9:39 AM, Keith Busch wrote:
> The PCI interrupt vectors intended to be associated with a queue may
> not start at 0; a driver may allocate pre_vectors for special use. This
> patch adds an offset parameter so blk-mq may find the intended affinity
> mask and updates all drivers using this API accordingly.

Looks good to me, I'll queue it up for 4.17.

-- 
Jens Axboe

Re: [PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Christoph Hellwig]

On Tue, Mar 27, 2018 at 09:39:08AM -0600, Keith Busch wrote:
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
> +				     dev->num_vecs > 1);

Can you turn this into:

-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
			     dev->num_vecs > 1 ? 1 /* admin queue */ : 0);

no functional change, but much easier to understand.

Except for that the whole series looks good:

Reviewed-by: Christoph Hellwig <hch@lst.de>

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Christoph Hellwig]

On Tue, Mar 27, 2018@09:39:08AM -0600, Keith Busch wrote:
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
> +				     dev->num_vecs > 1);

Can you turn this into:

-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
			     dev->num_vecs > 1 ? 1 /* admin queue */ : 0);

no functional change, but much easier to understand.

Except for that the whole series looks good:

Reviewed-by: Christoph Hellwig <hch at lst.de>

RE: [PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Don Brace]

> -----Original Message-----
> From: linux-scsi-owner@vger.kernel.org [mailto:linux-scsi-
> owner@vger.kernel.org] On Behalf Of Keith Busch
> Sent: Tuesday, March 27, 2018 10:39 AM
> To: Linux NVMe <linux-nvme@lists.infradead.org>; Linux Block <linux-
> block@vger.kernel.org>
> Cc: Christoph Hellwig <hch@lst.de>; Sagi Grimberg <sagi@grimberg.me>;
> Jianchao Wang <jianchao.w.wang@oracle.com>; Ming Lei
> <ming.lei@redhat.com>; Jens Axboe <axboe@kernel.dk>; Keith Busch
> <keith.busch@intel.com>; Don Brace <don.brace@microsemi.com>; qla2xxx-
> upstream@qlogic.com; linux-scsi@vger.kernel.org
> Subject: [PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues
>=20
> EXTERNAL EMAIL
>=20
>=20
> The PCI interrupt vectors intended to be associated with a queue may
> not start at 0; a driver may allocate pre_vectors for special use. This
> patch adds an offset parameter so blk-mq may find the intended affinity
> mask and updates all drivers using this API accordingly.
>=20
> Cc: Don Brace <don.brace@microsemi.com>
> Cc: <qla2xxx-upstream@qlogic.com>
> Cc: <linux-scsi@vger.kernel.org>
> Signed-off-by: Keith Busch <keith.busch@intel.com>
> ---
> v1 -> v2:
>=20
>   Update blk-mq API directly instead of chaining a default parameter to
>   a new API, and update all drivers accordingly.
>=20
>  block/blk-mq-pci.c                    | 6 ++++--
>  drivers/nvme/host/pci.c               | 2 +-
>  drivers/scsi/qla2xxx/qla_os.c         | 2 +-
>  drivers/scsi/smartpqi/smartpqi_init.c | 2 +-
>  include/linux/blk-mq-pci.h            | 3 ++-
>  5 files changed, 9 insertions(+), 6 deletions(-)
>=20
> diff --git a/drivers/scsi/smartpqi/smartpqi_init.c
> b/drivers/scsi/smartpqi/smartpqi_init.c
> index b2880c7709e6..10c94011c8a8 100644
> --- a/drivers/scsi/smartpqi/smartpqi_init.c
> +++ b/drivers/scsi/smartpqi/smartpqi_init.c
> @@ -5348,7 +5348,7 @@ static int pqi_map_queues(struct Scsi_Host *shost)
>  {
>         struct pqi_ctrl_info *ctrl_info =3D shost_to_hba(shost);
>=20
> -       return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev)=
;
> +       return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev,=
 0);
>  }
>=20

Acked-by: Don Brace <don.brace@microsemi.com>

[PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues

[Don Brace]

> -----Original Message-----
> From: linux-scsi-owner at vger.kernel.org [mailto:linux-scsi-
> owner at vger.kernel.org] On Behalf Of Keith Busch
> Sent: Tuesday, March 27, 2018 10:39 AM
> To: Linux NVMe <linux-nvme at lists.infradead.org>; Linux Block <linux-
> block at vger.kernel.org>
> Cc: Christoph Hellwig <hch at lst.de>; Sagi Grimberg <sagi at grimberg.me>;
> Jianchao Wang <jianchao.w.wang at oracle.com>; Ming Lei
> <ming.lei at redhat.com>; Jens Axboe <axboe at kernel.dk>; Keith Busch
> <keith.busch at intel.com>; Don Brace <don.brace at microsemi.com>; qla2xxx-
> upstream at qlogic.com; linux-scsi at vger.kernel.org
> Subject: [PATCH 1/3] blk-mq: Allow PCI vector offset for mapping queues
> 
> EXTERNAL EMAIL
> 
> 
> The PCI interrupt vectors intended to be associated with a queue may
> not start at 0; a driver may allocate pre_vectors for special use. This
> patch adds an offset parameter so blk-mq may find the intended affinity
> mask and updates all drivers using this API accordingly.
> 
> Cc: Don Brace <don.brace at microsemi.com>
> Cc: <qla2xxx-upstream at qlogic.com>
> Cc: <linux-scsi at vger.kernel.org>
> Signed-off-by: Keith Busch <keith.busch at intel.com>
> ---
> v1 -> v2:
> 
>   Update blk-mq API directly instead of chaining a default parameter to
>   a new API, and update all drivers accordingly.
> 
>  block/blk-mq-pci.c                    | 6 ++++--
>  drivers/nvme/host/pci.c               | 2 +-
>  drivers/scsi/qla2xxx/qla_os.c         | 2 +-
>  drivers/scsi/smartpqi/smartpqi_init.c | 2 +-
>  include/linux/blk-mq-pci.h            | 3 ++-
>  5 files changed, 9 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/scsi/smartpqi/smartpqi_init.c
> b/drivers/scsi/smartpqi/smartpqi_init.c
> index b2880c7709e6..10c94011c8a8 100644
> --- a/drivers/scsi/smartpqi/smartpqi_init.c
> +++ b/drivers/scsi/smartpqi/smartpqi_init.c
> @@ -5348,7 +5348,7 @@ static int pqi_map_queues(struct Scsi_Host *shost)
>  {
>         struct pqi_ctrl_info *ctrl_info = shost_to_hba(shost);
> 
> -       return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev);
> +       return blk_mq_pci_map_queues(&shost->tag_set, ctrl_info->pci_dev, 0);
>  }
> 

Acked-by: Don Brace <don.brace at microsemi.com>

Re: [PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Keith Busch]

On Wed, Mar 28, 2018 at 09:32:14AM +0200, Christoph Hellwig wrote:
> On Tue, Mar 27, 2018 at 09:39:08AM -0600, Keith Busch wrote:
> > -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> > +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
> > +				     dev->num_vecs > 1);
> 
> Can you turn this into:
> 
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> 			     dev->num_vecs > 1 ? 1 /* admin queue */ : 0);
> 
> no functional change, but much easier to understand.
> 
> Except for that the whole series looks good:

Sounds good, thanks for the reviews, Christoph and Ming.

Updated with your suggestion and applied (Jens picked up the blk-mq part;
nvme-4.17 is rebased to that).

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Keith Busch]

On Wed, Mar 28, 2018@09:32:14AM +0200, Christoph Hellwig wrote:
> On Tue, Mar 27, 2018@09:39:08AM -0600, Keith Busch wrote:
> > -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> > +	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
> > +				     dev->num_vecs > 1);
> 
> Can you turn this into:
> 
> -	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev), 0);
> 			     dev->num_vecs > 1 ? 1 /* admin queue */ : 0);
> 
> no functional change, but much easier to understand.
> 
> Except for that the whole series looks good:

Sounds good, thanks for the reviews, Christoph and Ming.

Updated with your suggestion and applied (Jens picked up the blk-mq part;
nvme-4.17 is rebased to that).

Re: [PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Christoph Hellwig]

> +static inline unsigned int nvme_ioq_vector(struct nvme_dev *dev,
> +		unsigned int qid)

No need for the inline here I think.

> +{
> +	/*
> +	 * A queue's vector matches the queue identifier unless the controller
> +	 * has only one vector available.
> +	 */
> +	return (dev->num_vecs == 1) ? 0 : qid;

and no need for the braces here.

> +	struct irq_affinity affd = {.pre_vectors = 1};

	struct irq_affinity affd = {
		.pre_vectors	= 1
	};

to make it a little more readable.

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Christoph Hellwig]

> +static inline unsigned int nvme_ioq_vector(struct nvme_dev *dev,
> +		unsigned int qid)

No need for the inline here I think.

> +{
> +	/*
> +	 * A queue's vector matches the queue identifier unless the controller
> +	 * has only one vector available.
> +	 */
> +	return (dev->num_vecs == 1) ? 0 : qid;

and no need for the braces here.

> +	struct irq_affinity affd = {.pre_vectors = 1};

	struct irq_affinity affd = {
		.pre_vectors	= 1
	};

to make it a little more readable.

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Keith Busch]

From: Jianchao Wang <jianchao.w.wang@oracle.com>

The admin and first IO queues shared the first irq vector, which has an
affinity mask including cpu0. If a system allows cpu0 to be offlined,
the admin queue may not be usable if no other CPUs in the affinity mask
are online. This is a problem since unlike IO queues, there is only
one admin queue that always needs to be usable.

To fix, this patch allocates one pre_vector for the admin queue that
is assigned all CPUs, so will always be accessible. The IO queues are
assigned the remaining managed vectors.

In case a controller has only one interrupt vector available, the admin
and IO queues will share the pre_vector with all CPUs assigned.

Signed-off-by: Jianchao Wang <jianchao.w.wang@oracle.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
[changelog, code comments, merge, and blk-mq pci vector offset]
Signed-off-by: Keith Busch <keith.busch@intel.com>
---
 drivers/nvme/host/pci.c | 27 +++++++++++++++++++++------
 1 file changed, 21 insertions(+), 6 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 632166f7d8f2..7b31bc01df6c 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -84,6 +84,7 @@ struct nvme_dev {
 	struct dma_pool *prp_small_pool;
 	unsigned online_queues;
 	unsigned max_qid;
+	unsigned int num_vecs;
 	int q_depth;
 	u32 db_stride;
 	void __iomem *bar;
@@ -139,6 +140,16 @@ static inline struct nvme_dev *to_nvme_dev(struct nvme_ctrl *ctrl)
 	return container_of(ctrl, struct nvme_dev, ctrl);
 }
 
+static inline unsigned int nvme_ioq_vector(struct nvme_dev *dev,
+		unsigned int qid)
+{
+	/*
+	 * A queue's vector matches the queue identifier unless the controller
+	 * has only one vector available.
+	 */
+	return (dev->num_vecs == 1) ? 0 : qid;
+}
+
 /*
  * An NVM Express queue.  Each device has at least two (one for admin
  * commands and one for I/O commands).
@@ -414,7 +425,8 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 
-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev));
+	return __blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
+				       dev->num_vecs > 1);
 }
 
 /**
@@ -1455,7 +1467,7 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
 		nvmeq->sq_cmds_io = dev->cmb + offset;
 	}
 
-	nvmeq->cq_vector = qid - 1;
+	nvmeq->cq_vector = nvme_ioq_vector(dev, qid);
 	result = adapter_alloc_cq(dev, qid, nvmeq);
 	if (result < 0)
 		goto release_vector;
@@ -1908,6 +1920,8 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	struct pci_dev *pdev = to_pci_dev(dev->dev);
 	int result, nr_io_queues;
 	unsigned long size;
+	struct irq_affinity affd = {.pre_vectors = 1};
+	int ret;
 
 	nr_io_queues = num_present_cpus();
 	result = nvme_set_queue_count(&dev->ctrl, &nr_io_queues);
@@ -1944,11 +1958,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	 * setting up the full range we need.
 	 */
 	pci_free_irq_vectors(pdev);
-	nr_io_queues = pci_alloc_irq_vectors(pdev, 1, nr_io_queues,
-			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY);
-	if (nr_io_queues <= 0)
+	ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_io_queues + 1),
+			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
+	if (ret <= 0)
 		return -EIO;
-	dev->max_qid = nr_io_queues;
+	dev->num_vecs = ret;
+	dev->max_qid = max(ret - 1, 1);
 
 	/*
 	 * Should investigate if there's a performance win from allocating
-- 
2.14.3

[PATCH 3/3] nvme-pci: Separate IO and admin queue IRQ vectors

[Keith Busch]

From: Jianchao Wang <jianchao.w.wang@oracle.com>

The admin and first IO queues shared the first irq vector, which has an
affinity mask including cpu0. If a system allows cpu0 to be offlined,
the admin queue may not be usable if no other CPUs in the affinity mask
are online. This is a problem since unlike IO queues, there is only
one admin queue that always needs to be usable.

To fix, this patch allocates one pre_vector for the admin queue that
is assigned all CPUs, so will always be accessible. The IO queues are
assigned the remaining managed vectors.

In case a controller has only one interrupt vector available, the admin
and IO queues will share the pre_vector with all CPUs assigned.

Signed-off-by: Jianchao Wang <jianchao.w.wang at oracle.com>
Reviewed-by: Ming Lei <ming.lei at redhat.com>
[changelog, code comments, merge, and blk-mq pci vector offset]
Signed-off-by: Keith Busch <keith.busch at intel.com>
---
 drivers/nvme/host/pci.c | 27 +++++++++++++++++++++------
 1 file changed, 21 insertions(+), 6 deletions(-)

diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index 632166f7d8f2..7b31bc01df6c 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -84,6 +84,7 @@ struct nvme_dev {
 	struct dma_pool *prp_small_pool;
 	unsigned online_queues;
 	unsigned max_qid;
+	unsigned int num_vecs;
 	int q_depth;
 	u32 db_stride;
 	void __iomem *bar;
@@ -139,6 +140,16 @@ static inline struct nvme_dev *to_nvme_dev(struct nvme_ctrl *ctrl)
 	return container_of(ctrl, struct nvme_dev, ctrl);
 }
 
+static inline unsigned int nvme_ioq_vector(struct nvme_dev *dev,
+		unsigned int qid)
+{
+	/*
+	 * A queue's vector matches the queue identifier unless the controller
+	 * has only one vector available.
+	 */
+	return (dev->num_vecs == 1) ? 0 : qid;
+}
+
 /*
  * An NVM Express queue.  Each device has at least two (one for admin
  * commands and one for I/O commands).
@@ -414,7 +425,8 @@ static int nvme_pci_map_queues(struct blk_mq_tag_set *set)
 {
 	struct nvme_dev *dev = set->driver_data;
 
-	return blk_mq_pci_map_queues(set, to_pci_dev(dev->dev));
+	return __blk_mq_pci_map_queues(set, to_pci_dev(dev->dev),
+				       dev->num_vecs > 1);
 }
 
 /**
@@ -1455,7 +1467,7 @@ static int nvme_create_queue(struct nvme_queue *nvmeq, int qid)
 		nvmeq->sq_cmds_io = dev->cmb + offset;
 	}
 
-	nvmeq->cq_vector = qid - 1;
+	nvmeq->cq_vector = nvme_ioq_vector(dev, qid);
 	result = adapter_alloc_cq(dev, qid, nvmeq);
 	if (result < 0)
 		goto release_vector;
@@ -1908,6 +1920,8 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	struct pci_dev *pdev = to_pci_dev(dev->dev);
 	int result, nr_io_queues;
 	unsigned long size;
+	struct irq_affinity affd = {.pre_vectors = 1};
+	int ret;
 
 	nr_io_queues = num_present_cpus();
 	result = nvme_set_queue_count(&dev->ctrl, &nr_io_queues);
@@ -1944,11 +1958,12 @@ static int nvme_setup_io_queues(struct nvme_dev *dev)
 	 * setting up the full range we need.
 	 */
 	pci_free_irq_vectors(pdev);
-	nr_io_queues = pci_alloc_irq_vectors(pdev, 1, nr_io_queues,
-			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY);
-	if (nr_io_queues <= 0)
+	ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_io_queues + 1),
+			PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd);
+	if (ret <= 0)
 		return -EIO;
-	dev->max_qid = nr_io_queues;
+	dev->num_vecs = ret;
+	dev->max_qid = max(ret - 1, 1);
 
 	/*
 	 * Should investigate if there's a performance win from allocating
-- 
2.14.3