linux-rdma.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Leon Romanovsky <leon@kernel.org>
To: Doug Ledford <dledford@redhat.com>, Jason Gunthorpe <jgg@nvidia.com>
Cc: Aharon Landau <aharonl@nvidia.com>,
	"David S. Miller" <davem@davemloft.net>,
	Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>,
	Gal Pressman <galpress@amazon.com>,
	Jakub Kicinski <kuba@kernel.org>,
	linux-kernel@vger.kernel.org, linux-rdma@vger.kernel.org,
	Maor Gottlieb <maorg@nvidia.com>,
	Mark Zhang <markzhang@nvidia.com>,
	Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>,
	Mustafa Ismail <mustafa.ismail@intel.com>,
	Naresh Kumar PBS <nareshkumar.pbs@broadcom.com>,
	Neta Ostrovsky <netao@nvidia.com>,
	netdev@vger.kernel.org, Potnuri Bharat Teja <bharat@chelsio.com>,
	Saeed Mahameed <saeedm@nvidia.com>,
	Selvin Xavier <selvin.xavier@broadcom.com>,
	Shiraz Saleem <shiraz.saleem@intel.com>,
	Yishai Hadas <yishaih@nvidia.com>,
	Zhu Yanjun <zyjzyj2000@gmail.com>
Subject: [PATCH rdma-next v1 05/11] RDMA/counter: Add optional counter support
Date: Wed, 15 Sep 2021 02:07:24 +0300	[thread overview]
Message-ID: <04bd7354c6a375e684712d79915f7eb816efee92.1631660727.git.leonro@nvidia.com> (raw)
In-Reply-To: <cover.1631660727.git.leonro@nvidia.com>

From: Aharon Landau <aharonl@nvidia.com>

An optional counter is a vendor-specific counter that may be dynamically
enabled/disabled.
This enhancement allows us to expose counters which are for example
mutual exclusive and cannot be enabled at the same time, counters that
might degrades performance, optional debug counters, etc.

Optional counters are marked with IB_STAT_FLAG_OPTIONAL flag and not
exported in sysfs.

Signed-off-by: Aharon Landau <aharonl@nvidia.com>
Signed-off-by: Neta Ostrovsky <netao@nvidia.com>
Reviewed-by: Mark Zhang <markzhang@nvidia.com>
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
---
 drivers/infiniband/core/counters.c | 22 ++++++++++++++++++++++
 drivers/infiniband/core/device.c   |  1 +
 drivers/infiniband/core/sysfs.c    | 26 ++++++++++++++++----------
 include/rdma/ib_verbs.h            | 14 +++++++++++++-
 include/rdma/rdma_counter.h        |  2 ++
 5 files changed, 54 insertions(+), 11 deletions(-)

diff --git a/drivers/infiniband/core/counters.c b/drivers/infiniband/core/counters.c
index a9559e33a113..974abc73a033 100644
--- a/drivers/infiniband/core/counters.c
+++ b/drivers/infiniband/core/counters.c
@@ -106,6 +106,28 @@ static int __rdma_counter_bind_qp(struct rdma_counter *counter,
 	return ret;
 }
 
+int rdma_counter_modify(struct ib_device *dev, u32 port, int index, bool enable)
+{
+	struct rdma_hw_stats *stats;
+	int ret;
+
+	if (!dev->ops.modify_hw_stat)
+		return -EOPNOTSUPP;
+
+	stats = ib_get_hw_stats_port(dev, port);
+	if (!stats)
+		return -EINVAL;
+
+	mutex_lock(&stats->lock);
+	ret = dev->ops.modify_hw_stat(dev, port, index, enable);
+	if (!ret)
+		enable ? clear_bit(index, stats->is_disabled) :
+			set_bit(index, stats->is_disabled);
+	mutex_unlock(&stats->lock);
+
+	return ret;
+}
+
 static struct rdma_counter *alloc_and_bind(struct ib_device *dev, u32 port,
 					   struct ib_qp *qp,
 					   enum rdma_nl_counter_mode mode)
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index f4814bb7f082..22a4adda7981 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -2676,6 +2676,7 @@ void ib_set_device_ops(struct ib_device *dev, const struct ib_device_ops *ops)
 	SET_DEVICE_OP(dev_ops, modify_cq);
 	SET_DEVICE_OP(dev_ops, modify_device);
 	SET_DEVICE_OP(dev_ops, modify_flow_action_esp);
+	SET_DEVICE_OP(dev_ops, modify_hw_stat);
 	SET_DEVICE_OP(dev_ops, modify_port);
 	SET_DEVICE_OP(dev_ops, modify_qp);
 	SET_DEVICE_OP(dev_ops, modify_srq);
diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c
index a26bf960f7ef..c5fc86d8ed3e 100644
--- a/drivers/infiniband/core/sysfs.c
+++ b/drivers/infiniband/core/sysfs.c
@@ -938,7 +938,7 @@ int ib_setup_device_attrs(struct ib_device *ibdev)
 {
 	struct hw_stats_device_attribute *attr;
 	struct hw_stats_device_data *data;
-	int i, ret;
+	int i, ret, pos = 0;
 
 	data = alloc_hw_stats_device(ibdev);
 	if (IS_ERR(data)) {
@@ -959,16 +959,19 @@ int ib_setup_device_attrs(struct ib_device *ibdev)
 	data->stats->timestamp = jiffies;
 
 	for (i = 0; i < data->stats->num_counters; i++) {
-		attr = &data->attrs[i];
+		if (data->stats->descs[i].flags & IB_STAT_FLAG_OPTIONAL)
+			continue;
+		attr = &data->attrs[pos];
 		sysfs_attr_init(&attr->attr.attr);
 		attr->attr.attr.name = data->stats->descs[i].name;
 		attr->attr.attr.mode = 0444;
 		attr->attr.show = hw_stat_device_show;
 		attr->show = show_hw_stats;
-		data->group.attrs[i] = &attr->attr.attr;
+		data->group.attrs[pos] = &attr->attr.attr;
+		pos++;
 	}
 
-	attr = &data->attrs[i];
+	attr = &data->attrs[pos];
 	sysfs_attr_init(&attr->attr.attr);
 	attr->attr.attr.name = "lifespan";
 	attr->attr.attr.mode = 0644;
@@ -976,7 +979,7 @@ int ib_setup_device_attrs(struct ib_device *ibdev)
 	attr->show = show_stats_lifespan;
 	attr->attr.store = hw_stat_device_store;
 	attr->store = set_stats_lifespan;
-	data->group.attrs[i] = &attr->attr.attr;
+	data->group.attrs[pos] = &attr->attr.attr;
 	for (i = 0; i != ARRAY_SIZE(ibdev->groups); i++)
 		if (!ibdev->groups[i]) {
 			ibdev->groups[i] = &data->group;
@@ -1032,7 +1035,7 @@ static int setup_hw_port_stats(struct ib_port *port,
 {
 	struct hw_stats_port_attribute *attr;
 	struct hw_stats_port_data *data;
-	int i, ret;
+	int i, ret, pos = 0;
 
 	data = alloc_hw_stats_port(port, group);
 	if (IS_ERR(data))
@@ -1050,16 +1053,19 @@ static int setup_hw_port_stats(struct ib_port *port,
 	data->stats->timestamp = jiffies;
 
 	for (i = 0; i < data->stats->num_counters; i++) {
-		attr = &data->attrs[i];
+		if (data->stats->descs[i].flags & IB_STAT_FLAG_OPTIONAL)
+			continue;
+		attr = &data->attrs[pos];
 		sysfs_attr_init(&attr->attr.attr);
 		attr->attr.attr.name = data->stats->descs[i].name;
 		attr->attr.attr.mode = 0444;
 		attr->attr.show = hw_stat_port_show;
 		attr->show = show_hw_stats;
-		group->attrs[i] = &attr->attr.attr;
+		group->attrs[pos] = &attr->attr.attr;
+		pos++;
 	}
 
-	attr = &data->attrs[i];
+	attr = &data->attrs[pos];
 	sysfs_attr_init(&attr->attr.attr);
 	attr->attr.attr.name = "lifespan";
 	attr->attr.attr.mode = 0644;
@@ -1067,7 +1073,7 @@ static int setup_hw_port_stats(struct ib_port *port,
 	attr->show = show_stats_lifespan;
 	attr->attr.store = hw_stat_port_store;
 	attr->store = set_stats_lifespan;
-	group->attrs[i] = &attr->attr.attr;
+	group->attrs[pos] = &attr->attr.attr;
 
 	port->hw_stats_data = data;
 	return 0;
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index f016bc0cd9de..e825e8e7accf 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -545,13 +545,18 @@ enum ib_port_speed {
 	IB_SPEED_NDR	= 128,
 };
 
+enum ib_stat_flag {
+	IB_STAT_FLAG_OPTIONAL = 1 << 0,
+};
+
 /**
  * struct rdma_stat_desc
  * @name - The name of the counter
- *
+ * @flags - Flags of the counter; For example, IB_STAT_FLAG_OPTIONAL
  */
 struct rdma_stat_desc {
 	const char *name;
+	unsigned int flags;
 };
 
 /**
@@ -2591,6 +2596,13 @@ struct ib_device_ops {
 	int (*get_hw_stats)(struct ib_device *device,
 			    struct rdma_hw_stats *stats, u32 port, int index);
 
+	/**
+	 * modify_hw_stat - Modify the counter configuration
+	 * @enable: true/false when enable/disable a counter
+	 * Return codes - 0 on success or error code otherwise.
+	 */
+	int (*modify_hw_stat)(struct ib_device *device, u32 port,
+			      int counter_index, bool enable);
 	/**
 	 * Allows rdma drivers to add their own restrack attributes.
 	 */
diff --git a/include/rdma/rdma_counter.h b/include/rdma/rdma_counter.h
index 0295b22cd1cd..b21ea39efc6c 100644
--- a/include/rdma/rdma_counter.h
+++ b/include/rdma/rdma_counter.h
@@ -63,4 +63,6 @@ int rdma_counter_get_mode(struct ib_device *dev, u32 port,
 			  enum rdma_nl_counter_mode *mode,
 			  enum rdma_nl_counter_mask *mask);
 
+int rdma_counter_modify(struct ib_device *dev, u32 port, int index,
+			bool is_add);
 #endif /* _RDMA_COUNTER_H_ */
-- 
2.31.1


  parent reply	other threads:[~2021-09-14 23:08 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-14 23:07 [PATCH rdma-next v1 00/11] Optional counter statistics support Leon Romanovsky
2021-09-14 23:07 ` [PATCH mlx5-next v1 01/11] net/mlx5: Add ifc bits to support optional counters Leon Romanovsky
2021-09-14 23:07 ` [PATCH mlx5-next v1 02/11] net/mlx5: Add priorities for counters in RDMA namespaces Leon Romanovsky
2021-09-14 23:07 ` [PATCH rdma-next v1 03/11] RDMA/counter: Add a descriptor in struct rdma_hw_stats Leon Romanovsky
2021-09-14 23:07 ` [PATCH rdma-next v1 04/11] RDMA/counter: Add an is_disabled field " Leon Romanovsky
2021-09-27 16:53   ` Jason Gunthorpe
2021-09-28  3:28     ` Mark Zhang
2021-09-28 11:51       ` Jason Gunthorpe
2021-09-14 23:07 ` Leon Romanovsky [this message]
2021-09-27 17:03   ` [PATCH rdma-next v1 05/11] RDMA/counter: Add optional counter support Jason Gunthorpe
2021-09-28  9:03     ` Mark Zhang
2021-09-28 11:51       ` Jason Gunthorpe
2021-09-29 12:14         ` Leon Romanovsky
2021-09-14 23:07 ` [PATCH rdma-next v1 06/11] RDMA/nldev: Add support to get status of all counters Leon Romanovsky
2021-09-27 17:30   ` Jason Gunthorpe
2021-09-28  9:12     ` Mark Zhang
2021-09-28 11:52       ` Jason Gunthorpe
2021-09-28 12:51         ` Mark Zhang
2021-09-29 12:26         ` Leon Romanovsky
2021-09-29 23:56           ` Jason Gunthorpe
2021-09-14 23:07 ` [PATCH rdma-next v1 07/11] RDMA/nldev: Allow optional-counter status configuration through RDMA netlink Leon Romanovsky
2021-09-27 17:20   ` Jason Gunthorpe
2021-09-29 12:27     ` Leon Romanovsky
2021-09-14 23:07 ` [PATCH rdma-next v1 08/11] RDMA/mlx5: Support optional counters in hw_stats initialization Leon Romanovsky
2021-09-14 23:07 ` [PATCH rdma-next v1 09/11] RDMA/mlx5: Add steering support in optional flow counters Leon Romanovsky
2021-09-14 23:07 ` [PATCH rdma-next v1 10/11] RDMA/mlx5: Add modify_op_stat() support Leon Romanovsky
2021-09-14 23:07 ` [PATCH rdma-next v1 11/11] RDMA/mlx5: Add optional counter support in get_hw_stats callback Leon Romanovsky

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=04bd7354c6a375e684712d79915f7eb816efee92.1631660727.git.leonro@nvidia.com \
    --to=leon@kernel.org \
    --cc=aharonl@nvidia.com \
    --cc=bharat@chelsio.com \
    --cc=davem@davemloft.net \
    --cc=dennis.dalessandro@cornelisnetworks.com \
    --cc=dledford@redhat.com \
    --cc=galpress@amazon.com \
    --cc=jgg@nvidia.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rdma@vger.kernel.org \
    --cc=maorg@nvidia.com \
    --cc=markzhang@nvidia.com \
    --cc=mike.marciniszyn@cornelisnetworks.com \
    --cc=mustafa.ismail@intel.com \
    --cc=nareshkumar.pbs@broadcom.com \
    --cc=netao@nvidia.com \
    --cc=netdev@vger.kernel.org \
    --cc=saeedm@nvidia.com \
    --cc=selvin.xavier@broadcom.com \
    --cc=shiraz.saleem@intel.com \
    --cc=yishaih@nvidia.com \
    --cc=zyjzyj2000@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).