From: Parav Pandit <parav@xxxxxxxxxxxx>
In order to support sysfs entries in multiple net namespaces for a rdma
device, introduce a ib_core_device whose scope is limited to hold core
device and per port sysfs related entires.
This is preparation patch so that multiple ib_core_devices in each net
namespace can be created in subsequent patch who all can share
ib_device.
(a) Move sysfs specific fields to ib_core_device.
(b) Make sysfs and device life cycle related routines to work on
ib_core_device.
(c) Introduce and use rdma_init_coredev() helper to initialize
coredev fields.
(d) Do not access driver data which is supposed to be used by the
provider driver.
Signed-off-by: Parav Pandit <parav@xxxxxxxxxxxx>
Signed-off-by: Leon Romanovsky <leonro@xxxxxxxxxxxx>
---
drivers/infiniband/core/core_priv.h | 2 ++
drivers/infiniband/core/device.c | 28 ++++++++++++++++------
drivers/infiniband/core/sysfs.c | 36 ++++++++++++++---------------
include/rdma/ib_verbs.h | 24 +++++++++++++++----
4 files changed, 60 insertions(+), 30 deletions(-)
diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h
index cc7535c5e192..0da0a26c02b9 100644
--- a/drivers/infiniband/core/core_priv.h
+++ b/drivers/infiniband/core/core_priv.h
@@ -54,6 +54,8 @@ struct pkey_index_qp_list {
struct list_head qp_list;
};
+extern const struct attribute_group ib_dev_attr_group;
+
int ib_device_register_sysfs(struct ib_device *device,
int (*port_callback)(struct ib_device *,
u8, struct kobject *));
diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
index 0027b0d79b09..16e1e854a3b4 100644
--- a/drivers/infiniband/core/device.c
+++ b/drivers/infiniband/core/device.c
@@ -269,6 +269,24 @@ static struct class ib_class = {
.dev_uevent = ib_device_uevent,
};
+static void rdma_init_coredev(struct ib_core_device *coredev,
+ struct ib_device *dev)
+{
+ /* This BUILD_BUG_ON is intended to catch layout change
+ * of union of ib_core_device and device.
+ * dev must be the first element as ib_core and providers
+ * driver uses it. Adding anything in ib_core_device before
+ * device will break this assumption.
+ */
+ BUILD_BUG_ON(offsetof(struct ib_core_device, dev) != 0);
+
+ coredev->dev.class = &ib_class;
+ coredev->dev.groups = dev->groups;
+ device_initialize(&coredev->dev);
+ coredev->owner = dev;
+ INIT_LIST_HEAD(&coredev->port_list);
+}
+
/**
* ib_alloc_device - allocate an IB device struct
* @size:size of structure to allocate
@@ -290,20 +308,16 @@ struct ib_device *ib_alloc_device(size_t size)
if (!device)
return NULL;
- rdma_restrack_init(&device->res);
-
- device->dev.class = &ib_class;
- device_initialize(&device->dev);
-
- dev_set_drvdata(&device->dev, device);
+ device->groups[0] = &ib_dev_attr_group;
+ rdma_init_coredev(&device->coredev, device);
INIT_LIST_HEAD(&device->event_handler_list);
spin_lock_init(&device->event_handler_lock);
rwlock_init(&device->client_data_lock);
INIT_LIST_HEAD(&device->client_data_list);
- INIT_LIST_HEAD(&device->port_list);
refcount_set(&device->refcount, 1);
init_completion(&device->unreg_completion);
+ rdma_restrack_init(&device->res);
return device;
}
diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c
index da4fafb28867..486219cc5ac4 100644
--- a/drivers/infiniband/core/sysfs.c
+++ b/drivers/infiniband/core/sysfs.c
@@ -1015,10 +1015,11 @@ static void setup_hw_stats(struct ib_device *device, struct ib_port *port,
return;
}
-static int add_port(struct ib_device *device, int port_num,
+static int add_port(struct ib_core_device *coredev, int port_num,
int (*port_callback)(struct ib_device *,
u8, struct kobject *))
{
+ struct ib_device *device = rdma_device_to_ibdev(&coredev->dev);
struct ib_port *p;
struct ib_port_attr attr;
int i;
@@ -1036,7 +1037,7 @@ static int add_port(struct ib_device *device, int port_num,
p->port_num = port_num;
ret = kobject_init_and_add(&p->kobj, &port_type,
- device->ports_kobj,
+ coredev->ports_kobj,
"%d", port_num);
if (ret) {
kfree(p);
@@ -1127,7 +1128,7 @@ static int add_port(struct ib_device *device, int port_num,
if (device->alloc_hw_stats && port_num)
setup_hw_stats(device, p, port_num);
- list_add_tail(&p->kobj.entry, &device->port_list);
+ list_add_tail(&p->kobj.entry, &coredev->port_list);
kobject_uevent(&p->kobj, KOBJ_ADD);
return 0;
@@ -1277,15 +1278,15 @@ static struct attribute *ib_dev_attrs[] = {
NULL,
};
-static const struct attribute_group dev_attr_group = {
+const struct attribute_group ib_dev_attr_group = {
.attrs = ib_dev_attrs,
};
-static void ib_free_port_attrs(struct ib_device *device)
+static void ib_free_port_attrs(struct ib_core_device *coredev)
{
struct kobject *p, *t;
- list_for_each_entry_safe(p, t, &device->port_list, entry) {
+ list_for_each_entry_safe(p, t, &coredev->port_list, entry) {
struct ib_port *port = container_of(p, struct ib_port, kobj);
list_del(&p->entry);
@@ -1305,27 +1306,29 @@ static void ib_free_port_attrs(struct ib_device *device)
kobject_put(p);
}
- kobject_put(device->ports_kobj);
+ kobject_put(coredev->ports_kobj);
}
-static int ib_setup_port_attrs(struct ib_device *device,
+static int ib_setup_port_attrs(struct ib_core_device *coredev,
int (*port_callback)(struct ib_device *,
u8, struct kobject *))
{
+ struct ib_device *device = rdma_device_to_ibdev(&coredev->dev);
int ret;
int i;
- device->ports_kobj = kobject_create_and_add("ports", &device->dev.kobj);
- if (!device->ports_kobj)
+ coredev->ports_kobj = kobject_create_and_add("ports",
+ &coredev->dev.kobj);
+ if (!coredev->ports_kobj)
return -ENOMEM;
if (rdma_cap_ib_switch(device)) {
- ret = add_port(device, 0, port_callback);
+ ret = add_port(coredev, 0, port_callback);
if (ret)
goto err_put;
} else {
for (i = 1; i <= device->phys_port_cnt; ++i) {
- ret = add_port(device, i, port_callback);
+ ret = add_port(coredev, i, port_callback);
if (ret)
goto err_put;
}
@@ -1334,7 +1337,7 @@ static int ib_setup_port_attrs(struct ib_device *device,
return 0;
err_put:
- ib_free_port_attrs(device);
+ ib_free_port_attrs(coredev);
return ret;
}
@@ -1344,14 +1347,11 @@ int ib_device_register_sysfs(struct ib_device *device,
{
int ret;
- device->groups[0] = &dev_attr_group;
- device->dev.groups = device->groups;
-
ret = device_add(&device->dev);
if (ret)
return ret;
- ret = ib_setup_port_attrs(device, port_callback);
+ ret = ib_setup_port_attrs(&device->coredev, port_callback);
if (ret) {
device_del(&device->dev);
return ret;
@@ -1371,6 +1371,6 @@ void ib_device_unregister_sysfs(struct ib_device *device)
free_hsag(&device->dev.kobj, device->hw_stats_ag);
kfree(device->hw_stats);
- ib_free_port_attrs(device);
+ ib_free_port_attrs(&device->coredev);
device_unregister(&device->dev);
}
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 446143ed49ae..f9bedf28e292 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -2257,6 +2257,16 @@ struct ib_counters_read_attr {
struct uverbs_attr_bundle;
+struct ib_core_device {
+ /* device must be the first element in structure until,
+ * union of ib_core_device and device exists in ib_device.
+ */
+ struct device dev;
+ struct kobject *ports_kobj;
+ struct list_head port_list;
+ struct ib_device *owner; /* reach back to owner ib_device */
+};
+
struct ib_device {
/* Do not access @dma_device directly from ULP nor from HW drivers. */
struct device *dma_device;
@@ -2553,16 +2563,17 @@ struct ib_device {
struct rdma_netdev_alloc_params *params);
struct module *owner;
- struct device dev;
+ union {
+ struct device dev;
+ struct ib_core_device coredev;
+ };
+
/* First group for device attributes,
* Second group for driver provided attributes (optional).
* It is NULL terminated array.
*/
const struct attribute_group *groups[3];
- struct kobject *ports_kobj;
- struct list_head port_list;
-
enum {
IB_DEV_UNINITIALIZED,
IB_DEV_REGISTERED,
@@ -4262,7 +4273,10 @@ rdma_set_device_sysfs_group(struct ib_device *dev,
*/
static inline struct ib_device *rdma_device_to_ibdev(struct device *device)
{
- return dev_get_drvdata(device);
+ struct ib_core_device *coredev =
+ container_of(device, struct ib_core_device, dev);
+
+ return coredev->owner;
}
/**
--
2.19.1
