[PATCH net-next RFC v5 00/15] Add devlink reload action and limit level options

[PATCH net-next RFC v5 00/15] Add devlink reload action and limit level options

[Moshe Shemesh]

Introduce new options on devlink reload API to enable the user to select
the reload action required and contrains limits on these actions that he
may want to ensure. Complete support for reload actions in mlx5.
The following reload actions are supported:
  driver_reinit: driver entities re-initialization, applying devlink-param
                 and devlink-resource values.
  fw_activate: firmware activate.

The uAPI is backward compatible, if the reload action option is omitted
from the reload command, the driver reinit action will be used.
Note that when required to do firmware activation some drivers may need
to reload the driver. On the other hand some drivers may need to reset
the firmware to reinitialize the driver entities. Therefore, the devlink
reload command returns the actions which were actually performed.

By default reload actions are not limited and driver implementation may
include reset or downtime as needed to perform the actions.
However, if limit_level is selected, the driver should perform only if
it can do it while keeping the limit level constrains.
Reload action limit level added:
  no_reset: No reset allowed, no down time allowed, no link flap and no
            configuration is lost.

Each driver which supports devlink reload command should expose the
reload actions and limit levels supported.

Add reload action stats to hold the history per reload action per limit
level. For example, the number of times fw_activate has been done on
this device since the driver module was added or if the firmware
activation was done with or without reset.

Patch 1-2 add the new API reload action and reload action limit level
          option to devlink reload.
Patch 3 adds reload actions stats.
Patch 4 exposes the reload actions stats on devlink dev get.
Patches 5-10 add support on mlx5 for devlink reload action fw_activate
            and handle the firmware reset events.
Patches 11-12 add devlink enable remote dev reset parameter and use it
             in mlx5.
Patches 13-14 mlx5 add devlink reload action limit level no_reset
              support for fw_activate reload action.
Patch 15 adds documentation file devlink-reload.rst 


Moshe Shemesh (15):
  devlink: Add reload action option to devlink reload command
  devlink: Add reload action limit level
  devlink: Add reload action stats
  devlink: Add reload actions stats to dev get
  net/mlx5: Add functions to set/query MFRL register
  net/mlx5: Set cap for pci sync for fw update event
  net/mlx5: Handle sync reset request event
  net/mlx5: Handle sync reset now event
  net/mlx5: Handle sync reset abort event
  net/mlx5: Add support for devlink reload action fw activate
  devlink: Add enable_remote_dev_reset generic parameter
  net/mlx5: Add devlink param enable_remote_dev_reset support
  net/mlx5: Add support for fw live patch event
  net/mlx5: Add support for devlink reload action limit level no reset
  devlink: Add Documentation/networking/devlink/devlink-reload.rst

 .../networking/devlink/devlink-params.rst     |   6 +
 .../networking/devlink/devlink-reload.rst     |  79 +++
 Documentation/networking/devlink/index.rst    |   1 +
 drivers/net/ethernet/mellanox/mlx4/main.c     |  16 +-
 .../net/ethernet/mellanox/mlx5/core/Makefile  |   2 +-
 .../net/ethernet/mellanox/mlx5/core/devlink.c | 122 ++++-
 .../mellanox/mlx5/core/diag/fw_tracer.c       |  31 ++
 .../mellanox/mlx5/core/diag/fw_tracer.h       |   1 +
 .../ethernet/mellanox/mlx5/core/fw_reset.c    | 454 ++++++++++++++++++
 .../ethernet/mellanox/mlx5/core/fw_reset.h    |  19 +
 .../net/ethernet/mellanox/mlx5/core/health.c  |  35 +-
 .../net/ethernet/mellanox/mlx5/core/main.c    |  13 +
 .../ethernet/mellanox/mlx5/core/mlx5_core.h   |   2 +
 drivers/net/ethernet/mellanox/mlxsw/core.c    |  27 +-
 drivers/net/netdevsim/dev.c                   |  17 +-
 include/linux/mlx5/device.h                   |   1 +
 include/linux/mlx5/driver.h                   |   4 +
 include/net/devlink.h                         |  21 +-
 include/uapi/linux/devlink.h                  |  41 ++
 net/core/devlink.c                            | 339 ++++++++++++-
 20 files changed, 1179 insertions(+), 52 deletions(-)
 create mode 100644 Documentation/networking/devlink/devlink-reload.rst
 create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
 create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h

-- 
2.17.1

[PATCH net-next RFC v5 01/15] devlink: Add reload action option to devlink reload command

[Moshe Shemesh]

Add devlink reload action to allow the user to request a specific reload
action. The action parameter is optional, if not specified then devlink
driver re-init action is used (backward compatible).
Note that when required to do firmware activation some drivers may need
to reload the driver. On the other hand some drivers may need to reset
the firmware to reinitialize the driver entities. Therefore, the devlink
reload command returns the actions which were actually performed.
Reload actions supported are:
driver_reinit: driver entities re-initialization, applying devlink-param
               and devlink-resource values.
fw_activate: firmware activate.

command examples:
$devlink dev reload pci/0000:82:00.0 action driver_reinit
reload_actions_performed:
  driver_reinit

$devlink dev reload pci/0000:82:00.0 action fw_activate
reload_actions_performed:
  driver_reinit fw_activate

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v4 -> v5:
- Always pass actions_performed to unload_up() instead of checking in
  each driver
- Verify returned actions_performed includes the requested action
- Changed  devlink_reload_actions_verify(devlink) to get ops
- Changed  devlink_reload_actions_verify() to return bool and rename to
  devlink_reload_actions_valid()
-  Only generate the reply if request uses new attributes
v3 -> v4:
- Removed fw_activate_no_reset as an action (next patch adds limit
  levels instead).
- Renamed actions_done to actions_performed
v2 -> v3:
- Replace fw_live_patch action by fw_activate_no_reset
- Devlink reload returns the actions done over netlink reply
v1 -> v2:
- Instead of reload levels driver,fw_reset,fw_live_patch have reload
  actions driver_reinit,fw_activate,fw_live_patch
- Remove driver default level, the action driver_reinit is the default
  action for all drivers
---
 drivers/net/ethernet/mellanox/mlx4/main.c     |  13 +-
 .../net/ethernet/mellanox/mlx5/core/devlink.c |  14 +-
 drivers/net/ethernet/mellanox/mlxsw/core.c    |  24 +++-
 drivers/net/netdevsim/dev.c                   |  15 +-
 include/net/devlink.h                         |   7 +-
 include/uapi/linux/devlink.h                  |  19 +++
 net/core/devlink.c                            | 136 ++++++++++++++++--
 7 files changed, 192 insertions(+), 36 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
index 70cf24ba71e4..1a482120cc0a 100644
--- a/drivers/net/ethernet/mellanox/mlx4/main.c
+++ b/drivers/net/ethernet/mellanox/mlx4/main.c
@@ -3946,6 +3946,7 @@ static int mlx4_restart_one_up(struct pci_dev *pdev, bool reload,
 			       struct devlink *devlink);
 
 static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
+				    enum devlink_reload_action action,
 				    struct netlink_ext_ack *extack)
 {
 	struct mlx4_priv *priv = devlink_priv(devlink);
@@ -3962,8 +3963,8 @@ static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
 	return 0;
 }
 
-static int mlx4_devlink_reload_up(struct devlink *devlink,
-				  struct netlink_ext_ack *extack)
+static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+				  struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	struct mlx4_priv *priv = devlink_priv(devlink);
 	struct mlx4_dev *dev = &priv->dev;
@@ -3971,15 +3972,19 @@ static int mlx4_devlink_reload_up(struct devlink *devlink,
 	int err;
 
 	err = mlx4_restart_one_up(persist->pdev, true, devlink);
-	if (err)
+	if (err) {
 		mlx4_err(persist->dev, "mlx4_restart_one_up failed, ret=%d\n",
 			 err);
+		return err;
+	}
+	*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
 
-	return err;
+	return 0;
 }
 
 static const struct devlink_ops mlx4_devlink_ops = {
 	.port_type_set	= mlx4_devlink_port_type_set,
+	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
 	.reload_down	= mlx4_devlink_reload_down,
 	.reload_up	= mlx4_devlink_reload_up,
 };
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index c709e9a385f6..ffc0525cea64 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -89,6 +89,7 @@ mlx5_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
 }
 
 static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
+				    enum devlink_reload_action action,
 				    struct netlink_ext_ack *extack)
 {
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
@@ -97,12 +98,18 @@ static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 	return 0;
 }
 
-static int mlx5_devlink_reload_up(struct devlink *devlink,
-				  struct netlink_ext_ack *extack)
+static int mlx5_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+				  struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
+	int err;
 
-	return mlx5_load_one(dev, false);
+	err = mlx5_load_one(dev, false);
+	if (err)
+		return err;
+	*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
+
+	return 0;
 }
 
 static const struct devlink_ops mlx5_devlink_ops = {
@@ -118,6 +125,7 @@ static const struct devlink_ops mlx5_devlink_ops = {
 #endif
 	.flash_update = mlx5_devlink_flash_update,
 	.info_get = mlx5_devlink_info_get,
+	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
 	.reload_down = mlx5_devlink_reload_down,
 	.reload_up = mlx5_devlink_reload_up,
 };
diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c
index 1bb21fe295b9..19f4486d5faf 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/core.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/core.c
@@ -1409,7 +1409,7 @@ mlxsw_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
 
 static int
 mlxsw_devlink_core_bus_device_reload_down(struct devlink *devlink,
-					  bool netns_change,
+					  bool netns_change, enum devlink_reload_action action,
 					  struct netlink_ext_ack *extack)
 {
 	struct mlxsw_core *mlxsw_core = devlink_priv(devlink);
@@ -1422,15 +1422,23 @@ mlxsw_devlink_core_bus_device_reload_down(struct devlink *devlink,
 }
 
 static int
-mlxsw_devlink_core_bus_device_reload_up(struct devlink *devlink,
-					struct netlink_ext_ack *extack)
+mlxsw_devlink_core_bus_device_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+					struct netlink_ext_ack *extack,
+					unsigned long *actions_performed)
 {
 	struct mlxsw_core *mlxsw_core = devlink_priv(devlink);
+	int err;
 
-	return mlxsw_core_bus_device_register(mlxsw_core->bus_info,
-					      mlxsw_core->bus,
-					      mlxsw_core->bus_priv, true,
-					      devlink, extack);
+	err = mlxsw_core_bus_device_register(mlxsw_core->bus_info,
+					     mlxsw_core->bus,
+					     mlxsw_core->bus_priv, true,
+					     devlink, extack);
+	if (err)
+		return err;
+	*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
+			     BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE);
+
+	return 0;
 }
 
 static int mlxsw_devlink_flash_update(struct devlink *devlink,
@@ -1560,6 +1568,8 @@ mlxsw_devlink_trap_policer_counter_get(struct devlink *devlink,
 }
 
 static const struct devlink_ops mlxsw_devlink_ops = {
+	.supported_reload_actions	= BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
+					  BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE),
 	.reload_down		= mlxsw_devlink_core_bus_device_reload_down,
 	.reload_up		= mlxsw_devlink_core_bus_device_reload_up,
 	.port_type_set			= mlxsw_devlink_port_type_set,
diff --git a/drivers/net/netdevsim/dev.c b/drivers/net/netdevsim/dev.c
index 32f339fedb21..0eb522f6a718 100644
--- a/drivers/net/netdevsim/dev.c
+++ b/drivers/net/netdevsim/dev.c
@@ -697,7 +697,7 @@ static int nsim_dev_reload_create(struct nsim_dev *nsim_dev,
 static void nsim_dev_reload_destroy(struct nsim_dev *nsim_dev);
 
 static int nsim_dev_reload_down(struct devlink *devlink, bool netns_change,
-				struct netlink_ext_ack *extack)
+				enum devlink_reload_action action, struct netlink_ext_ack *extack)
 {
 	struct nsim_dev *nsim_dev = devlink_priv(devlink);
 
@@ -713,10 +713,11 @@ static int nsim_dev_reload_down(struct devlink *devlink, bool netns_change,
 	return 0;
 }
 
-static int nsim_dev_reload_up(struct devlink *devlink,
-			      struct netlink_ext_ack *extack)
+static int nsim_dev_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+			      struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	struct nsim_dev *nsim_dev = devlink_priv(devlink);
+	int err;
 
 	if (nsim_dev->fail_reload) {
 		/* For testing purposes, user set debugfs fail_reload
@@ -726,7 +727,12 @@ static int nsim_dev_reload_up(struct devlink *devlink,
 		return -EINVAL;
 	}
 
-	return nsim_dev_reload_create(nsim_dev, extack);
+	err = nsim_dev_reload_create(nsim_dev, extack);
+	if (err)
+		return err;
+	*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
+
+	return 0;
 }
 
 static int nsim_dev_info_get(struct devlink *devlink,
@@ -875,6 +881,7 @@ nsim_dev_devlink_trap_policer_counter_get(struct devlink *devlink,
 }
 
 static const struct devlink_ops nsim_dev_devlink_ops = {
+	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
 	.reload_down = nsim_dev_reload_down,
 	.reload_up = nsim_dev_reload_up,
 	.info_get = nsim_dev_info_get,
diff --git a/include/net/devlink.h b/include/net/devlink.h
index 48b1c1ef1ebd..37abc3e08e9e 100644
--- a/include/net/devlink.h
+++ b/include/net/devlink.h
@@ -1014,10 +1014,11 @@ enum devlink_trap_group_generic_id {
 	}
 
 struct devlink_ops {
+	unsigned long supported_reload_actions;
 	int (*reload_down)(struct devlink *devlink, bool netns_change,
-			   struct netlink_ext_ack *extack);
-	int (*reload_up)(struct devlink *devlink,
-			 struct netlink_ext_ack *extack);
+			   enum devlink_reload_action action, struct netlink_ext_ack *extack);
+	int (*reload_up)(struct devlink *devlink, enum devlink_reload_action action,
+			 struct netlink_ext_ack *extack, unsigned long *actions_performed);
 	int (*port_type_set)(struct devlink_port *devlink_port,
 			     enum devlink_port_type port_type);
 	int (*port_split)(struct devlink *devlink, unsigned int port_index,
diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
index 631f5bdf1707..fdba7ab58a79 100644
--- a/include/uapi/linux/devlink.h
+++ b/include/uapi/linux/devlink.h
@@ -274,6 +274,21 @@ enum {
 	DEVLINK_ATTR_TRAP_METADATA_TYPE_FA_COOKIE,
 };
 
+/**
+ * enum devlink_reload_action - Reload action.
+ * @DEVLINK_RELOAD_ACTION_DRIVER_REINIT: Driver entities re-instantiation.
+ * @DEVLINK_RELOAD_ACTION_FW_ACTIVATE: FW activate.
+ */
+enum devlink_reload_action {
+	DEVLINK_RELOAD_ACTION_UNSPEC,
+	DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
+	DEVLINK_RELOAD_ACTION_FW_ACTIVATE,
+
+	/* Add new reload actions above */
+	__DEVLINK_RELOAD_ACTION_MAX,
+	DEVLINK_RELOAD_ACTION_MAX = __DEVLINK_RELOAD_ACTION_MAX - 1
+};
+
 enum devlink_attr {
 	/* don't change the order or add anything between, this is ABI! */
 	DEVLINK_ATTR_UNSPEC,
@@ -462,6 +477,10 @@ enum devlink_attr {
 
 	DEVLINK_ATTR_PORT_EXTERNAL,		/* u8 */
 	DEVLINK_ATTR_PORT_CONTROLLER_NUMBER,	/* u32 */
+
+	DEVLINK_ATTR_RELOAD_ACTION,		/* u8 */
+	DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,	/* nested */
+
 	/* add new attributes above here, update the policy in devlink.c */
 
 	__DEVLINK_ATTR_MAX,
diff --git a/net/core/devlink.c b/net/core/devlink.c
index e5b71f3c2d4d..318ef29f81f2 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -462,6 +462,12 @@ static int devlink_nl_put_handle(struct sk_buff *msg, struct devlink *devlink)
 	return 0;
 }
 
+static bool
+devlink_reload_action_is_supported(struct devlink *devlink, enum devlink_reload_action action)
+{
+	return test_bit(action, &devlink->ops->supported_reload_actions);
+}
+
 static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
 			   enum devlink_command cmd, u32 portid,
 			   u32 seq, int flags)
@@ -2948,11 +2954,11 @@ static void devlink_reload_netns_change(struct devlink *devlink,
 				     DEVLINK_CMD_PARAM_NEW);
 }
 
-static bool devlink_reload_supported(const struct devlink *devlink)
+static bool devlink_reload_supported(const struct devlink_ops *ops)
 {
-	return devlink->ops->reload_down && devlink->ops->reload_up;
+	return ops->reload_down && ops->reload_up;
 }
-
+
 static void devlink_reload_failed_set(struct devlink *devlink,
 				      bool reload_failed)
 {
@@ -2969,32 +2975,79 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
 EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
 
 static int devlink_reload(struct devlink *devlink, struct net *dest_net,
-			  struct netlink_ext_ack *extack)
+			  enum devlink_reload_action action, struct netlink_ext_ack *extack,
+			  unsigned long *actions_performed)
 {
 	int err;
 
 	if (!devlink->reload_enabled)
 		return -EOPNOTSUPP;
 
-	err = devlink->ops->reload_down(devlink, !!dest_net, extack);
+	err = devlink->ops->reload_down(devlink, !!dest_net, action, extack);
 	if (err)
 		return err;
 
 	if (dest_net && !net_eq(dest_net, devlink_net(devlink)))
 		devlink_reload_netns_change(devlink, dest_net);
 
-	err = devlink->ops->reload_up(devlink, extack);
+	err = devlink->ops->reload_up(devlink, action, extack, actions_performed);
 	devlink_reload_failed_set(devlink, !!err);
-	return err;
+	if (err)
+		return err;
+
+	WARN_ON(!test_bit(action, actions_performed));
+	return 0;
+}
+
+static int
+devlink_nl_reload_actions_performed_fill(struct sk_buff *msg,
+					 struct devlink *devlink,
+					 unsigned long actions_performed,
+					 enum devlink_command cmd, u32 portid,
+					 u32 seq, int flags)
+{
+	struct nlattr *actions_performed_attr;
+	void *hdr;
+	int i;
+
+	hdr = genlmsg_put(msg, portid, seq, &devlink_nl_family, flags, cmd);
+	if (!hdr)
+		return -EMSGSIZE;
+
+	if (devlink_nl_put_handle(msg, devlink))
+		goto genlmsg_cancel;
+
+	actions_performed_attr = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED);
+	if (!actions_performed_attr)
+		goto genlmsg_cancel;
+
+	for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
+		if (!test_bit(i, &actions_performed))
+			continue;
+		if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION, i))
+			goto actions_performed_nest_cancel;
+	}
+	nla_nest_end(msg, actions_performed_attr);
+	genlmsg_end(msg, hdr);
+	return 0;
+
+actions_performed_nest_cancel:
+	nla_nest_cancel(msg, actions_performed_attr);
+genlmsg_cancel:
+	genlmsg_cancel(msg, hdr);
+	return -EMSGSIZE;
 }
 
 static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
 {
 	struct devlink *devlink = info->user_ptr[0];
+	enum devlink_reload_action action;
+	unsigned long actions_performed;
 	struct net *dest_net = NULL;
+	struct sk_buff *msg;
 	int err;
 
-	if (!devlink_reload_supported(devlink))
+	if (!devlink_reload_supported(devlink->ops))
 		return -EOPNOTSUPP;
 
 	err = devlink_resources_validate(devlink, NULL, info);
@@ -3011,12 +3064,43 @@ static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
 			return PTR_ERR(dest_net);
 	}
 
-	err = devlink_reload(devlink, dest_net, info->extack);
+	if (info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
+		action = nla_get_u8(info->attrs[DEVLINK_ATTR_RELOAD_ACTION]);
+	else
+		action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT;
+
+	if (action == DEVLINK_RELOAD_ACTION_UNSPEC) {
+		NL_SET_ERR_MSG_MOD(info->extack, "Invalid reload action");
+		return -EINVAL;
+	} else if (!devlink_reload_action_is_supported(devlink, action)) {
+		NL_SET_ERR_MSG_MOD(info->extack, "Requested reload action is not supported by the driver");
+		return -EOPNOTSUPP;
+	}
+
+	err = devlink_reload(devlink, dest_net, action, info->extack, &actions_performed);
 
 	if (dest_net)
 		put_net(dest_net);
 
-	return err;
+	if (err)
+		return err;
+	/* For backward compatibility generate reply only if attributes used by user */
+	if (!info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
+		return 0;
+
+	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+	if (!msg)
+		return -ENOMEM;
+
+	err = devlink_nl_reload_actions_performed_fill(msg, devlink, actions_performed,
+						       DEVLINK_CMD_RELOAD, info->snd_portid,
+						       info->snd_seq, 0);
+	if (err) {
+		nlmsg_free(msg);
+		return err;
+	}
+
+	return genlmsg_reply(msg, info);
 }
 
 static int devlink_nl_flash_update_fill(struct sk_buff *msg,
@@ -7069,6 +7153,7 @@ static const struct nla_policy devlink_nl_policy[DEVLINK_ATTR_MAX + 1] = {
 	[DEVLINK_ATTR_TRAP_POLICER_RATE] = { .type = NLA_U64 },
 	[DEVLINK_ATTR_TRAP_POLICER_BURST] = { .type = NLA_U64 },
 	[DEVLINK_ATTR_PORT_FUNCTION] = { .type = NLA_NESTED },
+	[DEVLINK_ATTR_RELOAD_ACTION] = { .type = NLA_U8 },
 };
 
 static const struct genl_ops devlink_nl_ops[] = {
@@ -7402,6 +7487,20 @@ static struct genl_family devlink_nl_family __ro_after_init = {
 	.n_mcgrps	= ARRAY_SIZE(devlink_nl_mcgrps),
 };
 
+static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
+{
+	if (!devlink_reload_supported(ops)) {
+		if (WARN_ON(ops->supported_reload_actions))
+			return false;
+		return true;
+	}
+
+	if (WARN_ON(ops->supported_reload_actions >= BIT(__DEVLINK_RELOAD_ACTION_MAX) ||
+		    ops->supported_reload_actions <= BIT(DEVLINK_RELOAD_ACTION_UNSPEC)))
+		return false;
+	return true;
+}
+
 /**
  *	devlink_alloc - Allocate new devlink instance resources
  *
@@ -7418,10 +7517,14 @@ struct devlink *devlink_alloc(const struct devlink_ops *ops, size_t priv_size)
 	if (WARN_ON(!ops))
 		return NULL;
 
+	if (!devlink_reload_actions_valid(ops))
+		return NULL;
+
 	devlink = kzalloc(sizeof(*devlink) + priv_size, GFP_KERNEL);
 	if (!devlink)
 		return NULL;
 	devlink->ops = ops;
+
 	xa_init_flags(&devlink->snapshot_ids, XA_FLAGS_ALLOC);
 	__devlink_net_set(devlink, &init_net);
 	INIT_LIST_HEAD(&devlink->port_list);
@@ -7466,7 +7569,7 @@ EXPORT_SYMBOL_GPL(devlink_register);
 void devlink_unregister(struct devlink *devlink)
 {
 	mutex_lock(&devlink_mutex);
-	WARN_ON(devlink_reload_supported(devlink) &&
+	WARN_ON(devlink_reload_supported(devlink->ops) &&
 		devlink->reload_enabled);
 	devlink_notify(devlink, DEVLINK_CMD_DEL);
 	list_del(&devlink->list);
@@ -8503,7 +8606,7 @@ __devlink_param_driverinit_value_set(struct devlink *devlink,
 int devlink_param_driverinit_value_get(struct devlink *devlink, u32 param_id,
 				       union devlink_param_value *init_val)
 {
-	if (!devlink_reload_supported(devlink))
+	if (!devlink_reload_supported(devlink->ops))
 		return -EOPNOTSUPP;
 
 	return __devlink_param_driverinit_value_get(&devlink->param_list,
@@ -8550,7 +8653,7 @@ int devlink_port_param_driverinit_value_get(struct devlink_port *devlink_port,
 {
 	struct devlink *devlink = devlink_port->devlink;
 
-	if (!devlink_reload_supported(devlink))
+	if (!devlink_reload_supported(devlink->ops))
 		return -EOPNOTSUPP;
 
 	return __devlink_param_driverinit_value_get(&devlink_port->param_list,
@@ -9676,6 +9779,7 @@ int devlink_compat_switch_id_get(struct net_device *dev,
 
 static void __net_exit devlink_pernet_pre_exit(struct net *net)
 {
+	unsigned long actions_performed;
 	struct devlink *devlink;
 	int err;
 
@@ -9685,9 +9789,11 @@ static void __net_exit devlink_pernet_pre_exit(struct net *net)
 	mutex_lock(&devlink_mutex);
 	list_for_each_entry(devlink, &devlink_list, list) {
 		if (net_eq(devlink_net(devlink), net)) {
-			if (WARN_ON(!devlink_reload_supported(devlink)))
+			if (WARN_ON(!devlink_reload_supported(devlink->ops)))
 				continue;
-			err = devlink_reload(devlink, &init_net, NULL);
+			err = devlink_reload(devlink, &init_net,
+					     DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
+					     NULL, &actions_performed);
 			if (err && err != -EOPNOTSUPP)
 				pr_warn("Failed to reload devlink instance into init_net\n");
 		}
-- 
2.17.1

[PATCH net-next RFC v5 02/15] devlink: Add reload action limit level

[Moshe Shemesh]

Add reload action limit level to demand restrictions on actions.
Reload action limit levels supported:
none (default): No constrains on actions. Driver implementation may
                include reset or downtime as needed to perform the
                actions.
no_reset: No reset allowed, no down time allowed, no link flap and no
          configuration is lost.

Some combinations of action and limit level are invalid. For example,
driver can not reinitialize its entities without any downtime.

The no_reset limit level will have usecase in this patchset to
implement restricted fw_activate on mlx5.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v4 -> v5:
- Remove check DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX
- Added list of invalid action-limit_level combinations and add check to
  supported actions and levels and check user request
v3 -> v4:
- New patch
---
 drivers/net/ethernet/mellanox/mlx4/main.c     |  3 +
 .../net/ethernet/mellanox/mlx5/core/devlink.c |  3 +
 drivers/net/ethernet/mellanox/mlxsw/core.c    |  3 +
 drivers/net/netdevsim/dev.c                   |  6 +-
 include/net/devlink.h                         |  6 +-
 include/uapi/linux/devlink.h                  | 17 +++++
 net/core/devlink.c                            | 76 +++++++++++++++++--
 7 files changed, 107 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
index 1a482120cc0a..f0ef295af477 100644
--- a/drivers/net/ethernet/mellanox/mlx4/main.c
+++ b/drivers/net/ethernet/mellanox/mlx4/main.c
@@ -3947,6 +3947,7 @@ static int mlx4_restart_one_up(struct pci_dev *pdev, bool reload,
 
 static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
 				    enum devlink_reload_action action,
+				    enum devlink_reload_action_limit_level limit_level,
 				    struct netlink_ext_ack *extack)
 {
 	struct mlx4_priv *priv = devlink_priv(devlink);
@@ -3964,6 +3965,7 @@ static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
 }
 
 static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+				  enum devlink_reload_action_limit_level limit_level,
 				  struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	struct mlx4_priv *priv = devlink_priv(devlink);
@@ -3985,6 +3987,7 @@ static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_a
 static const struct devlink_ops mlx4_devlink_ops = {
 	.port_type_set	= mlx4_devlink_port_type_set,
 	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
+	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),
 	.reload_down	= mlx4_devlink_reload_down,
 	.reload_up	= mlx4_devlink_reload_up,
 };
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index ffc0525cea64..38b00b4a3ce8 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -90,6 +90,7 @@ mlx5_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
 
 static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 				    enum devlink_reload_action action,
+				    enum devlink_reload_action_limit_level limit_level,
 				    struct netlink_ext_ack *extack)
 {
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
@@ -99,6 +100,7 @@ static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 }
 
 static int mlx5_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+				  enum devlink_reload_action_limit_level limit_level,
 				  struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
@@ -126,6 +128,7 @@ static const struct devlink_ops mlx5_devlink_ops = {
 	.flash_update = mlx5_devlink_flash_update,
 	.info_get = mlx5_devlink_info_get,
 	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
+	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),
 	.reload_down = mlx5_devlink_reload_down,
 	.reload_up = mlx5_devlink_reload_up,
 };
diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c
index 19f4486d5faf..fd20b77f70cf 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/core.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/core.c
@@ -1410,6 +1410,7 @@ mlxsw_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
 static int
 mlxsw_devlink_core_bus_device_reload_down(struct devlink *devlink,
 					  bool netns_change, enum devlink_reload_action action,
+					  enum devlink_reload_action_limit_level limit_level,
 					  struct netlink_ext_ack *extack)
 {
 	struct mlxsw_core *mlxsw_core = devlink_priv(devlink);
@@ -1423,6 +1424,7 @@ mlxsw_devlink_core_bus_device_reload_down(struct devlink *devlink,
 
 static int
 mlxsw_devlink_core_bus_device_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+					enum devlink_reload_action_limit_level limit_level,
 					struct netlink_ext_ack *extack,
 					unsigned long *actions_performed)
 {
@@ -1570,6 +1572,7 @@ mlxsw_devlink_trap_policer_counter_get(struct devlink *devlink,
 static const struct devlink_ops mlxsw_devlink_ops = {
 	.supported_reload_actions	= BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
 					  BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE),
+	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),
 	.reload_down		= mlxsw_devlink_core_bus_device_reload_down,
 	.reload_up		= mlxsw_devlink_core_bus_device_reload_up,
 	.port_type_set			= mlxsw_devlink_port_type_set,
diff --git a/drivers/net/netdevsim/dev.c b/drivers/net/netdevsim/dev.c
index 0eb522f6a718..f59af3e85e3a 100644
--- a/drivers/net/netdevsim/dev.c
+++ b/drivers/net/netdevsim/dev.c
@@ -697,7 +697,9 @@ static int nsim_dev_reload_create(struct nsim_dev *nsim_dev,
 static void nsim_dev_reload_destroy(struct nsim_dev *nsim_dev);
 
 static int nsim_dev_reload_down(struct devlink *devlink, bool netns_change,
-				enum devlink_reload_action action, struct netlink_ext_ack *extack)
+				enum devlink_reload_action action,
+				enum devlink_reload_action_limit_level limit_level,
+				struct netlink_ext_ack *extack)
 {
 	struct nsim_dev *nsim_dev = devlink_priv(devlink);
 
@@ -714,6 +716,7 @@ static int nsim_dev_reload_down(struct devlink *devlink, bool netns_change,
 }
 
 static int nsim_dev_reload_up(struct devlink *devlink, enum devlink_reload_action action,
+			      enum devlink_reload_action_limit_level limit_level,
 			      struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	struct nsim_dev *nsim_dev = devlink_priv(devlink);
@@ -882,6 +885,7 @@ nsim_dev_devlink_trap_policer_counter_get(struct devlink *devlink,
 
 static const struct devlink_ops nsim_dev_devlink_ops = {
 	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
+	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),
 	.reload_down = nsim_dev_reload_down,
 	.reload_up = nsim_dev_reload_up,
 	.info_get = nsim_dev_info_get,
diff --git a/include/net/devlink.h b/include/net/devlink.h
index 37abc3e08e9e..d8c62d605381 100644
--- a/include/net/devlink.h
+++ b/include/net/devlink.h
@@ -1015,9 +1015,13 @@ enum devlink_trap_group_generic_id {
 
 struct devlink_ops {
 	unsigned long supported_reload_actions;
+	unsigned long supported_reload_action_limit_levels;
 	int (*reload_down)(struct devlink *devlink, bool netns_change,
-			   enum devlink_reload_action action, struct netlink_ext_ack *extack);
+			   enum devlink_reload_action action,
+			   enum devlink_reload_action_limit_level limit_level,
+			   struct netlink_ext_ack *extack);
 	int (*reload_up)(struct devlink *devlink, enum devlink_reload_action action,
+			 enum devlink_reload_action_limit_level limit_level,
 			 struct netlink_ext_ack *extack, unsigned long *actions_performed);
 	int (*port_type_set)(struct devlink_port *devlink_port,
 			     enum devlink_port_type port_type);
diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
index fdba7ab58a79..0c5d942dcbd5 100644
--- a/include/uapi/linux/devlink.h
+++ b/include/uapi/linux/devlink.h
@@ -289,6 +289,22 @@ enum devlink_reload_action {
 	DEVLINK_RELOAD_ACTION_MAX = __DEVLINK_RELOAD_ACTION_MAX - 1
 };
 
+/**
+ * enum devlink_reload_action_limit_level - Reload action limit level.
+ * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE: No constrains on action. Action may include
+ *                                          reset or downtime as needed.
+ * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET: No reset allowed, no down time allowed,
+ *                                              no link flap and no configuration is lost.
+ */
+enum devlink_reload_action_limit_level {
+	DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
+	DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
+
+	/* Add new reload actions limit level above */
+	__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX,
+	DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX = __DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX - 1
+};
+
 enum devlink_attr {
 	/* don't change the order or add anything between, this is ABI! */
 	DEVLINK_ATTR_UNSPEC,
@@ -480,6 +496,7 @@ enum devlink_attr {
 
 	DEVLINK_ATTR_RELOAD_ACTION,		/* u8 */
 	DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,	/* nested */
+	DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL,	/* u8 */
 
 	/* add new attributes above here, update the policy in devlink.c */
 
diff --git a/net/core/devlink.c b/net/core/devlink.c
index 318ef29f81f2..fee6fcc7dead 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -462,12 +462,45 @@ static int devlink_nl_put_handle(struct sk_buff *msg, struct devlink *devlink)
 	return 0;
 }
 
+struct devlink_reload_combination {
+	enum devlink_reload_action action;
+	enum devlink_reload_action_limit_level limit_level;
+};
+
+static const struct devlink_reload_combination devlink_reload_invalid_combinations[] = {
+	{
+		/* can't reinitialize driver with no down time */
+		.action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
+		.limit_level = DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
+	},
+};
+
+static bool
+devlink_reload_combination_is_invalid(enum devlink_reload_action action,
+				      enum devlink_reload_action_limit_level limit_level)
+{
+	int i;
+
+	for (i = 0 ; i <  ARRAY_SIZE(devlink_reload_invalid_combinations) ; i++)
+		if (devlink_reload_invalid_combinations[i].action == action &&
+		    devlink_reload_invalid_combinations[i].limit_level == limit_level)
+			return true;
+	return false;
+}
+
 static bool
 devlink_reload_action_is_supported(struct devlink *devlink, enum devlink_reload_action action)
 {
 	return test_bit(action, &devlink->ops->supported_reload_actions);
 }
 
+static bool
+devlink_reload_action_limit_level_is_supported(struct devlink *devlink,
+					       enum devlink_reload_action_limit_level limit_level)
+{
+	return test_bit(limit_level, &devlink->ops->supported_reload_action_limit_levels);
+}
+
 static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
 			   enum devlink_command cmd, u32 portid,
 			   u32 seq, int flags)
@@ -2975,22 +3008,23 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
 EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
 
 static int devlink_reload(struct devlink *devlink, struct net *dest_net,
-			  enum devlink_reload_action action, struct netlink_ext_ack *extack,
-			  unsigned long *actions_performed)
+			  enum devlink_reload_action action,
+			  enum devlink_reload_action_limit_level limit_level,
+			  struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	int err;
 
 	if (!devlink->reload_enabled)
 		return -EOPNOTSUPP;
 
-	err = devlink->ops->reload_down(devlink, !!dest_net, action, extack);
+	err = devlink->ops->reload_down(devlink, !!dest_net, action, limit_level, extack);
 	if (err)
 		return err;
 
 	if (dest_net && !net_eq(dest_net, devlink_net(devlink)))
 		devlink_reload_netns_change(devlink, dest_net);
 
-	err = devlink->ops->reload_up(devlink, action, extack, actions_performed);
+	err = devlink->ops->reload_up(devlink, action, limit_level, extack, actions_performed);
 	devlink_reload_failed_set(devlink, !!err);
 	if (err)
 		return err;
@@ -3040,6 +3074,7 @@ devlink_nl_reload_actions_performed_fill(struct sk_buff *msg,
 
 static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
 {
+	enum devlink_reload_action_limit_level limit_level;
 	struct devlink *devlink = info->user_ptr[0];
 	enum devlink_reload_action action;
 	unsigned long actions_performed;
@@ -3077,7 +3112,21 @@ static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
 		return -EOPNOTSUPP;
 	}
 
-	err = devlink_reload(devlink, dest_net, action, info->extack, &actions_performed);
+	if (info->attrs[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL])
+		limit_level = nla_get_u8(info->attrs[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL]);
+	else
+		limit_level = DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE;
+
+	if (!devlink_reload_action_limit_level_is_supported(devlink, limit_level)) {
+		NL_SET_ERR_MSG_MOD(info->extack, "Requested limit level is not supported by the driver");
+		return -EOPNOTSUPP;
+	}
+	if (devlink_reload_combination_is_invalid(action, limit_level)) {
+		NL_SET_ERR_MSG_MOD(info->extack, "Requested limit level is invalid for this action");
+		return -EINVAL;
+	}
+	err = devlink_reload(devlink, dest_net, action, limit_level, info->extack,
+			     &actions_performed);
 
 	if (dest_net)
 		put_net(dest_net);
@@ -7154,6 +7203,7 @@ static const struct nla_policy devlink_nl_policy[DEVLINK_ATTR_MAX + 1] = {
 	[DEVLINK_ATTR_TRAP_POLICER_BURST] = { .type = NLA_U64 },
 	[DEVLINK_ATTR_PORT_FUNCTION] = { .type = NLA_NESTED },
 	[DEVLINK_ATTR_RELOAD_ACTION] = { .type = NLA_U8 },
+	[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL] = { .type = NLA_U8 },
 };
 
 static const struct genl_ops devlink_nl_ops[] = {
@@ -7489,6 +7539,9 @@ static struct genl_family devlink_nl_family __ro_after_init = {
 
 static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
 {
+	const struct devlink_reload_combination *comb;
+	int i;
+
 	if (!devlink_reload_supported(ops)) {
 		if (WARN_ON(ops->supported_reload_actions))
 			return false;
@@ -7498,6 +7551,18 @@ static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
 	if (WARN_ON(ops->supported_reload_actions >= BIT(__DEVLINK_RELOAD_ACTION_MAX) ||
 		    ops->supported_reload_actions <= BIT(DEVLINK_RELOAD_ACTION_UNSPEC)))
 		return false;
+
+	if (WARN_ON(!ops->supported_reload_action_limit_levels ||
+		    ops->supported_reload_action_limit_levels >=
+		    BIT(__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX)))
+		return false;
+
+	for (i = 0; i <  ARRAY_SIZE(devlink_reload_invalid_combinations); i++)  {
+		comb = &devlink_reload_invalid_combinations[i];
+		if (ops->supported_reload_actions == BIT(comb->action) &&
+		    ops->supported_reload_action_limit_levels == BIT(comb->limit_level))
+			return false;
+	}
 	return true;
 }
 
@@ -9793,6 +9858,7 @@ static void __net_exit devlink_pernet_pre_exit(struct net *net)
 				continue;
 			err = devlink_reload(devlink, &init_net,
 					     DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
+					     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
 					     NULL, &actions_performed);
 			if (err && err != -EOPNOTSUPP)
 				pr_warn("Failed to reload devlink instance into init_net\n");
-- 
2.17.1

[PATCH net-next RFC v5 03/15] devlink: Add reload action stats

[Moshe Shemesh]

Add reload action stats to hold the history per reload action type and
limit level.

For example, the number of times fw_activate has been performed on this
device since the driver module was added or if the firmware activation
was performed with or without reset.

The function devlink_remote_reload_actions_performed() is exported to
enable also drivers update on reload actions performed, for example in
case firmware activation with reset finished successfully but was
initiated by remote host.

Add devlink notification on stats update.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v4 -> v5:
- Add separate reload action stats for updating on remote actions
- Protect  from updating remote actions stats during reload_down()/up()
v3 -> v4:
- Renamed reload_actions_cnts to reload_action_stats
- Add devlink notifications on stats update
- Renamed devlink_reload_actions_implicit_actions_performed() and add
  function comment in code
v2 -> v3:
- New patch
---
 include/net/devlink.h |  8 ++++++
 net/core/devlink.c    | 62 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 70 insertions(+)

diff --git a/include/net/devlink.h b/include/net/devlink.h
index d8c62d605381..f09f55a47d09 100644
--- a/include/net/devlink.h
+++ b/include/net/devlink.h
@@ -20,6 +20,9 @@
 #include <uapi/linux/devlink.h>
 #include <linux/xarray.h>
 
+#define DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE \
+	(__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX * __DEVLINK_RELOAD_ACTION_MAX)
+
 struct devlink_ops;
 
 struct devlink {
@@ -38,6 +41,8 @@ struct devlink {
 	struct list_head trap_policer_list;
 	const struct devlink_ops *ops;
 	struct xarray snapshot_ids;
+	u32 reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
+	u32 remote_reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
 	struct device *dev;
 	possible_net_t _net;
 	struct mutex lock; /* Serializes access to devlink instance specific objects such as
@@ -1400,6 +1405,9 @@ void
 devlink_health_reporter_recovery_done(struct devlink_health_reporter *reporter);
 
 bool devlink_is_reload_failed(const struct devlink *devlink);
+void devlink_remote_reload_actions_performed(struct devlink *devlink,
+					     enum devlink_reload_action_limit_level limit_level,
+					     unsigned long actions_performed);
 
 void devlink_flash_update_begin_notify(struct devlink *devlink);
 void devlink_flash_update_end_notify(struct devlink *devlink);
diff --git a/net/core/devlink.c b/net/core/devlink.c
index fee6fcc7dead..1509c2ffb98b 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -3007,16 +3007,74 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
 }
 EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
 
+static void
+__devlink_reload_action_stats_update(struct devlink *devlink,
+				     u32 *reload_action_stats,
+				     enum devlink_reload_action_limit_level limit_level,
+				     unsigned long actions_performed)
+{
+	int stat_idx;
+	int action;
+
+	if (!actions_performed)
+		return;
+
+	if (WARN_ON(limit_level > DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX))
+		return;
+	for (action = 0; action <= DEVLINK_RELOAD_ACTION_MAX; action++) {
+		if (!test_bit(action, &actions_performed))
+			continue;
+		stat_idx = limit_level * __DEVLINK_RELOAD_ACTION_MAX + action;
+		reload_action_stats[stat_idx]++;
+	}
+	devlink_notify(devlink, DEVLINK_CMD_NEW);
+}
+
+static void
+devlink_reload_action_stats_update(struct devlink *devlink,
+				   enum devlink_reload_action_limit_level limit_level,
+				   unsigned long actions_performed)
+{
+	__devlink_reload_action_stats_update(devlink, devlink->reload_action_stats,
+					     limit_level, actions_performed);
+}
+
+/**
+ *	devlink_remote_reload_actions_performed - Update devlink on reload actions
+ *	  performed which are not a direct result of devlink reload call.
+ *
+ *	This should be called by a driver after performing reload actions in case it was not
+ *	a result of devlink reload call. For example fw_activate was performed as a result
+ *	of devlink reload triggered fw_activate on another host.
+ *	The motivation for this function is to keep data on reload actions performed on this
+ *	function whether it was done due to direct devlink reload call or not.
+ *
+ *	@devlink: devlink
+ *	@limit_level: reload action limit level
+ *	@actions_performed: bitmask of actions performed
+ */
+void devlink_remote_reload_actions_performed(struct devlink *devlink,
+					     enum devlink_reload_action_limit_level limit_level,
+					     unsigned long actions_performed)
+{
+	__devlink_reload_action_stats_update(devlink, devlink->remote_reload_action_stats,
+					     limit_level, actions_performed);
+}
+EXPORT_SYMBOL_GPL(devlink_remote_reload_actions_performed);
+
 static int devlink_reload(struct devlink *devlink, struct net *dest_net,
 			  enum devlink_reload_action action,
 			  enum devlink_reload_action_limit_level limit_level,
 			  struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
+	u32 remote_reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
 	int err;
 
 	if (!devlink->reload_enabled)
 		return -EOPNOTSUPP;
 
+	memcpy(remote_reload_action_stats, devlink->remote_reload_action_stats,
+	       sizeof(remote_reload_action_stats));
 	err = devlink->ops->reload_down(devlink, !!dest_net, action, limit_level, extack);
 	if (err)
 		return err;
@@ -3030,6 +3088,10 @@ static int devlink_reload(struct devlink *devlink, struct net *dest_net,
 		return err;
 
 	WARN_ON(!test_bit(action, actions_performed));
+	/* protect from driver updating the remote action within devlink reload */
+	WARN_ON(memcmp(remote_reload_action_stats, devlink->remote_reload_action_stats,
+		       sizeof(remote_reload_action_stats)));
+	devlink_reload_action_stats_update(devlink, limit_level, *actions_performed);
 	return 0;
 }
 
-- 
2.17.1

[PATCH net-next RFC v5 04/15] devlink: Add reload actions stats to dev get

[Moshe Shemesh]

Expose devlink reload actions stats to the user through devlink dev
get command.

Examples:
$ devlink dev show
pci/0000:82:00.0:
  stats:
      reload_action_stats:
        driver_reinit 2
        fw_activate 1
        fw_activate_no_reset 0
        remote_driver_reinit 0
        remote_fw_activate 0
        remote_fw_activate_no_reset 0
pci/0000:82:00.1:
  stats:
      reload_action_stats:
        driver_reinit 0
        fw_activate 0
        fw_activate_no_reset 0
        remote_driver_reinit 1
        remote_fw_activate 1
        remote_fw_activate_no_reset 0

$ devlink dev show -jp
{
    "dev": {
        "pci/0000:82:00.0": {
            "stats": {
                "reload_action_stats": [ {
                        "driver_reinit": 2
                    },{
                        "fw_activate": 1
                    },{
                        "fw_activate_no_reset": 0
                    },{
                        "remote_driver_reinit": 0
                    },{
                        "remote_fw_activate": 0
                    },{
                        "remote_fw_activate_no_reset": 0
                    } ]
            }
        },
        "pci/0000:82:00.1": {
            "stats": {
                "reload_action_stats": [ {
                        "driver_reinit": 0
                    },{
                        "fw_activate": 0
                    },{
                        "fw_activate_no_reset": 0
                    },{
                        "remote_driver_reinit": 1
                    },{
                        "remote_fw_activate": 1
                    },{
                        "remote_fw_activate_no_reset": 0
                    } ]
            }
        }
    }
}

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v4 -> v5:
- Add remote actions stats
- If devlink reload is not supported, show only remote_stats
v3 -> v4:
- Renamed DEVLINK_ATTR_RELOAD_ACTION_CNT to
  DEVLINK_ATTR_RELOAD_ACTION_STAT
- Add stats per action per limit level
v2 -> v3:
- Add reload actions counters instead of supported reload actions
  (reload actions counters are only for supported action so no need for
   both)
v1 -> v2:
- Removed DEVLINK_ATTR_RELOAD_DEFAULT_LEVEL
- Removed DEVLINK_ATTR_RELOAD_LEVELS_INFO
- Have actions instead of levels
---
 include/uapi/linux/devlink.h |  5 +++
 net/core/devlink.c           | 70 ++++++++++++++++++++++++++++++++++++
 2 files changed, 75 insertions(+)

diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
index 0c5d942dcbd5..648d53be691e 100644
--- a/include/uapi/linux/devlink.h
+++ b/include/uapi/linux/devlink.h
@@ -497,7 +497,12 @@ enum devlink_attr {
 	DEVLINK_ATTR_RELOAD_ACTION,		/* u8 */
 	DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,	/* nested */
 	DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL,	/* u8 */
+	DEVLINK_ATTR_RELOAD_ACTION_STATS,	/* nested */
+	DEVLINK_ATTR_RELOAD_ACTION_STAT,	/* nested */
+	DEVLINK_ATTR_RELOAD_ACTION_STAT_REMOTE,	/* flag */
+	DEVLINK_ATTR_RELOAD_ACTION_STAT_VALUE,	/* u32 */
 
+	DEVLINK_ATTR_DEV_STATS,			/* nested */
 	/* add new attributes above here, update the policy in devlink.c */
 
 	__DEVLINK_ATTR_MAX,
diff --git a/net/core/devlink.c b/net/core/devlink.c
index 1509c2ffb98b..71aeda259e6a 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -501,10 +501,39 @@ devlink_reload_action_limit_level_is_supported(struct devlink *devlink,
 	return test_bit(limit_level, &devlink->ops->supported_reload_action_limit_levels);
 }
 
+static int devlink_reload_action_stat_put(struct sk_buff *msg, enum devlink_reload_action action,
+					  enum devlink_reload_action_limit_level limit_level,
+					  bool is_remote, u32 value)
+{
+	struct nlattr *reload_action_stat;
+
+	reload_action_stat = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT);
+	if (!reload_action_stat)
+		return -EMSGSIZE;
+
+	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION, action))
+		goto nla_put_failure;
+	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL, limit_level))
+		goto nla_put_failure;
+	if (is_remote && nla_put_flag(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT_REMOTE))
+		goto nla_put_failure;
+	if (nla_put_u32(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT_VALUE, value))
+		goto nla_put_failure;
+	nla_nest_end(msg, reload_action_stat);
+	return 0;
+
+nla_put_failure:
+	nla_nest_cancel(msg, reload_action_stat);
+	return -EMSGSIZE;
+}
+
 static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
 			   enum devlink_command cmd, u32 portid,
 			   u32 seq, int flags)
 {
+	struct nlattr *dev_stats, *reload_action_stats;
+	int i, j, stat_idx;
+	u32 value;
 	void *hdr;
 
 	hdr = genlmsg_put(msg, portid, seq, &devlink_nl_family, flags, cmd);
@@ -516,9 +545,50 @@ static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
 	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_FAILED, devlink->reload_failed))
 		goto nla_put_failure;
 
+	dev_stats = nla_nest_start(msg, DEVLINK_ATTR_DEV_STATS);
+	if (!dev_stats)
+		goto nla_put_failure;
+	reload_action_stats = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTION_STATS);
+	if (!reload_action_stats)
+		goto dev_stats_nest_cancel;
+
+	for (j = 0; j <= DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX; j++) {
+		if (!devlink_reload_action_limit_level_is_supported(devlink, j))
+			continue;
+		for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
+			if (!devlink_reload_action_is_supported(devlink, i) ||
+			    devlink_reload_combination_is_invalid(i, j))
+				continue;
+
+			stat_idx = j * __DEVLINK_RELOAD_ACTION_MAX + i;
+			value = devlink->reload_action_stats[stat_idx];
+			if (devlink_reload_action_stat_put(msg, i, j, false, value))
+				goto reload_action_stats_nest_cancel;
+		}
+	}
+
+	for (j = 0; j <= DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX; j++) {
+		for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
+			if (i == DEVLINK_RELOAD_ACTION_UNSPEC ||
+			    devlink_reload_combination_is_invalid(i, j))
+				continue;
+
+			stat_idx = j * __DEVLINK_RELOAD_ACTION_MAX + i;
+			value = devlink->remote_reload_action_stats[stat_idx];
+			if (devlink_reload_action_stat_put(msg, i, j, true, value))
+				goto reload_action_stats_nest_cancel;
+		}
+	}
+
+	nla_nest_end(msg, reload_action_stats);
+	nla_nest_end(msg, dev_stats);
 	genlmsg_end(msg, hdr);
 	return 0;
 
+dev_stats_nest_cancel:
+	nla_nest_cancel(msg, dev_stats);
+reload_action_stats_nest_cancel:
+	nla_nest_cancel(msg, reload_action_stats);
 nla_put_failure:
 	genlmsg_cancel(msg, hdr);
 	return -EMSGSIZE;
-- 
2.17.1

[PATCH net-next RFC v5 05/15] net/mlx5: Add functions to set/query MFRL register

[Moshe Shemesh]

Add functions to query and set the MFRL reset options supported by
firmware.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/Makefile  |  2 +-
 .../ethernet/mellanox/mlx5/core/fw_reset.c    | 46 +++++++++++++++++++
 .../ethernet/mellanox/mlx5/core/fw_reset.h    | 13 ++++++
 3 files changed, 60 insertions(+), 1 deletion(-)
 create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
 create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
index 0b3eaa102751..fa0f01151f2d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile
@@ -16,7 +16,7 @@ mlx5_core-y :=	main.o cmd.o debugfs.o fw.o eq.o uar.o pagealloc.o \
 		transobj.o vport.o sriov.o fs_cmd.o fs_core.o pci_irq.o \
 		fs_counters.o rl.o lag.o dev.o events.o wq.o lib/gid.o \
 		lib/devcom.o lib/pci_vsc.o lib/dm.o diag/fs_tracepoint.o \
-		diag/fw_tracer.o diag/crdump.o devlink.o diag/rsc_dump.o
+		diag/fw_tracer.o diag/crdump.o devlink.o diag/rsc_dump.o fw_reset.o
 
 #
 # Netdev basic
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
new file mode 100644
index 000000000000..76d2cece29ac
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -0,0 +1,46 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+/* Copyright (c) 2020, Mellanox Technologies inc.  All rights reserved. */
+
+#include "fw_reset.h"
+
+static int mlx5_reg_mfrl_set(struct mlx5_core_dev *dev, u8 reset_level,
+			     u8 reset_type_sel, u8 sync_resp, bool sync_start)
+{
+	u32 out[MLX5_ST_SZ_DW(mfrl_reg)] = {};
+	u32 in[MLX5_ST_SZ_DW(mfrl_reg)] = {};
+
+	MLX5_SET(mfrl_reg, in, reset_level, reset_level);
+	MLX5_SET(mfrl_reg, in, rst_type_sel, reset_type_sel);
+	MLX5_SET(mfrl_reg, in, pci_sync_for_fw_update_resp, sync_resp);
+	MLX5_SET(mfrl_reg, in, pci_sync_for_fw_update_start, sync_start);
+
+	return mlx5_core_access_reg(dev, in, sizeof(in), out, sizeof(out), MLX5_REG_MFRL, 0, 1);
+}
+
+int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_type)
+{
+	u32 out[MLX5_ST_SZ_DW(mfrl_reg)] = {};
+	u32 in[MLX5_ST_SZ_DW(mfrl_reg)] = {};
+	int err;
+
+	err = mlx5_core_access_reg(dev, in, sizeof(in), out, sizeof(out), MLX5_REG_MFRL, 0, 0);
+	if (err)
+		return err;
+
+	if (reset_level)
+		*reset_level = MLX5_GET(mfrl_reg, out, reset_level);
+	if (reset_type)
+		*reset_type = MLX5_GET(mfrl_reg, out, reset_type);
+
+	return 0;
+}
+
+int mlx5_fw_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel)
+{
+	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, reset_type_sel, 0, true);
+}
+
+int mlx5_fw_set_live_patch(struct mlx5_core_dev *dev)
+{
+	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL0, 0, 0, false);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
new file mode 100644
index 000000000000..1bbd95182ca6
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */
+/* Copyright (c) 2020, Mellanox Technologies inc.  All rights reserved. */
+
+#ifndef __MLX5_FW_RESET_H
+#define __MLX5_FW_RESET_H
+
+#include "mlx5_core.h"
+
+int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_type);
+int mlx5_fw_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel);
+int mlx5_fw_set_live_patch(struct mlx5_core_dev *dev);
+
+#endif
-- 
2.17.1

[PATCH net-next RFC v5 06/15] net/mlx5: Set cap for pci sync for fw update event

[Moshe Shemesh]

Set capability to notify the firmware that this host driver is capable
of handling pci sync for firmware update events.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/main.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index ce43e3feccd9..871d28b09f8a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -548,6 +548,9 @@ static int handle_hca_cap(struct mlx5_core_dev *dev, void *set_ctx)
 	if (MLX5_CAP_GEN_MAX(dev, dct))
 		MLX5_SET(cmd_hca_cap, set_hca_cap, dct, 1);
 
+	if (MLX5_CAP_GEN_MAX(dev, pci_sync_for_fw_update_event))
+		MLX5_SET(cmd_hca_cap, set_hca_cap, pci_sync_for_fw_update_event, 1);
+
 	if (MLX5_CAP_GEN_MAX(dev, num_vhca_ports))
 		MLX5_SET(cmd_hca_cap,
 			 set_hca_cap,
-- 
2.17.1

[PATCH net-next RFC v5 07/15] net/mlx5: Handle sync reset request event

[Moshe Shemesh]

Once the driver gets sync_reset_request from firmware it prepares for the
coming reset and sends acknowledge.
After getting this event the driver expects device reset, either it will
trigger PCI reset on sync_reset_now event or such PCI reset will be
triggered by another PF of the same device. So it moves to reset
requested mode and if it gets PCI reset triggered by the other PF it
detect the reset and reloads.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v1 -> v2:
- Moved handling of sync reset recovery from health to fw_reset
---
 .../ethernet/mellanox/mlx5/core/fw_reset.c    | 167 ++++++++++++++++++
 .../ethernet/mellanox/mlx5/core/fw_reset.h    |   3 +
 .../net/ethernet/mellanox/mlx5/core/health.c  |  35 ++--
 .../net/ethernet/mellanox/mlx5/core/main.c    |  10 ++
 .../ethernet/mellanox/mlx5/core/mlx5_core.h   |   2 +
 include/linux/mlx5/driver.h                   |   4 +
 6 files changed, 206 insertions(+), 15 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
index 76d2cece29ac..0f224454b4a2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -3,6 +3,20 @@
 
 #include "fw_reset.h"
 
+enum {
+	MLX5_FW_RESET_FLAGS_RESET_REQUESTED,
+};
+
+struct mlx5_fw_reset {
+	struct mlx5_core_dev *dev;
+	struct mlx5_nb nb;
+	struct workqueue_struct *wq;
+	struct work_struct reset_request_work;
+	struct work_struct reset_reload_work;
+	unsigned long reset_flags;
+	struct timer_list timer;
+};
+
 static int mlx5_reg_mfrl_set(struct mlx5_core_dev *dev, u8 reset_level,
 			     u8 reset_type_sel, u8 sync_resp, bool sync_start)
 {
@@ -44,3 +58,156 @@ int mlx5_fw_set_live_patch(struct mlx5_core_dev *dev)
 {
 	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL0, 0, 0, false);
 }
+
+static int mlx5_fw_set_reset_sync_ack(struct mlx5_core_dev *dev)
+{
+	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, 0, 1, false);
+}
+
+static void mlx5_sync_reset_reload_work(struct work_struct *work)
+{
+	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
+						      reset_reload_work);
+	struct mlx5_core_dev *dev = fw_reset->dev;
+
+	mlx5_enter_error_state(dev, true);
+	mlx5_unload_one(dev, false);
+	if (mlx5_health_wait_pci_up(dev)) {
+		mlx5_core_err(dev, "reset reload flow aborted, PCI reads still not working\n");
+		return;
+	}
+	mlx5_load_one(dev, false);
+}
+
+static void mlx5_stop_sync_reset_poll(struct mlx5_core_dev *dev)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	del_timer(&fw_reset->timer);
+}
+
+static void mlx5_sync_reset_clear_reset_requested(struct mlx5_core_dev *dev, bool poll_health)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	mlx5_stop_sync_reset_poll(dev);
+	clear_bit(MLX5_FW_RESET_FLAGS_RESET_REQUESTED, &fw_reset->reset_flags);
+	if (poll_health)
+		mlx5_start_health_poll(dev);
+}
+
+#define MLX5_RESET_POLL_INTERVAL	(HZ / 10)
+static void poll_sync_reset(struct timer_list *t)
+{
+	struct mlx5_fw_reset *fw_reset = from_timer(fw_reset, t, timer);
+	struct mlx5_core_dev *dev = fw_reset->dev;
+	u32 fatal_error;
+
+	if (!test_bit(MLX5_FW_RESET_FLAGS_RESET_REQUESTED, &fw_reset->reset_flags))
+		return;
+
+	fatal_error = mlx5_health_check_fatal_sensors(dev);
+
+	if (fatal_error) {
+		mlx5_core_warn(dev, "Got Device Reset\n");
+		mlx5_sync_reset_clear_reset_requested(dev, false);
+		queue_work(fw_reset->wq, &fw_reset->reset_reload_work);
+		return;
+	}
+
+	mod_timer(&fw_reset->timer, round_jiffies(jiffies + MLX5_RESET_POLL_INTERVAL));
+}
+
+static void mlx5_start_sync_reset_poll(struct mlx5_core_dev *dev)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	timer_setup(&fw_reset->timer, poll_sync_reset, 0);
+	fw_reset->timer.expires = round_jiffies(jiffies + MLX5_RESET_POLL_INTERVAL);
+	add_timer(&fw_reset->timer);
+}
+
+static void mlx5_sync_reset_set_reset_requested(struct mlx5_core_dev *dev)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	mlx5_stop_health_poll(dev, true);
+	set_bit(MLX5_FW_RESET_FLAGS_RESET_REQUESTED, &fw_reset->reset_flags);
+	mlx5_start_sync_reset_poll(dev);
+}
+
+static void mlx5_sync_reset_request_event(struct work_struct *work)
+{
+	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
+						      reset_request_work);
+	struct mlx5_core_dev *dev = fw_reset->dev;
+
+	mlx5_sync_reset_set_reset_requested(dev);
+	if (mlx5_fw_set_reset_sync_ack(dev))
+		mlx5_core_warn(dev, "PCI Sync FW Update Reset Ack Failed.\n");
+	else
+		mlx5_core_warn(dev, "PCI Sync FW Update Reset Ack. Device reset is expected.\n");
+}
+
+static void mlx5_sync_reset_events_handle(struct mlx5_fw_reset *fw_reset, struct mlx5_eqe *eqe)
+{
+	struct mlx5_eqe_sync_fw_update *sync_fw_update_eqe;
+	u8 sync_event_rst_type;
+
+	sync_fw_update_eqe = &eqe->data.sync_fw_update;
+	sync_event_rst_type = sync_fw_update_eqe->sync_rst_state & SYNC_RST_STATE_MASK;
+	switch (sync_event_rst_type) {
+	case MLX5_SYNC_RST_STATE_RESET_REQUEST:
+		queue_work(fw_reset->wq, &fw_reset->reset_request_work);
+		break;
+	}
+}
+
+static int fw_reset_event_notifier(struct notifier_block *nb, unsigned long action, void *data)
+{
+	struct mlx5_fw_reset *fw_reset = mlx5_nb_cof(nb, struct mlx5_fw_reset, nb);
+	struct mlx5_eqe *eqe = data;
+
+	switch (eqe->sub_type) {
+	case MLX5_GENERAL_SUBTYPE_PCI_SYNC_FOR_FW_UPDATE_EVENT:
+		mlx5_sync_reset_events_handle(fw_reset, eqe);
+		break;
+	default:
+		return NOTIFY_DONE;
+	}
+
+	return NOTIFY_OK;
+}
+
+int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev)
+{
+	struct mlx5_fw_reset *fw_reset = kzalloc(sizeof(*fw_reset), GFP_KERNEL);
+
+	if (!fw_reset)
+		return -ENOMEM;
+	fw_reset->wq = create_singlethread_workqueue("mlx5_fw_reset_events");
+	if (!fw_reset->wq) {
+		kfree(fw_reset);
+		return -ENOMEM;
+	}
+
+	fw_reset->dev = dev;
+	dev->priv.fw_reset = fw_reset;
+
+	INIT_WORK(&fw_reset->reset_request_work, mlx5_sync_reset_request_event);
+	INIT_WORK(&fw_reset->reset_reload_work, mlx5_sync_reset_reload_work);
+
+	MLX5_NB_INIT(&fw_reset->nb, fw_reset_event_notifier, GENERAL_EVENT);
+	mlx5_eq_notifier_register(dev, &fw_reset->nb);
+
+	return 0;
+}
+
+void mlx5_fw_reset_events_cleanup(struct mlx5_core_dev *dev)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	mlx5_eq_notifier_unregister(dev, &fw_reset->nb);
+	destroy_workqueue(fw_reset->wq);
+	kvfree(dev->priv.fw_reset);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
index 1bbd95182ca6..278f538ea92a 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
@@ -10,4 +10,7 @@ int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_ty
 int mlx5_fw_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel);
 int mlx5_fw_set_live_patch(struct mlx5_core_dev *dev);
 
+int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev);
+void mlx5_fw_reset_events_cleanup(struct mlx5_core_dev *dev);
+
 #endif
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/health.c b/drivers/net/ethernet/mellanox/mlx5/core/health.c
index b31f769d2df9..54523bed16cd 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/health.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/health.c
@@ -110,7 +110,7 @@ static bool sensor_fw_synd_rfr(struct mlx5_core_dev *dev)
 	return rfr && synd;
 }
 
-static u32 check_fatal_sensors(struct mlx5_core_dev *dev)
+u32 mlx5_health_check_fatal_sensors(struct mlx5_core_dev *dev)
 {
 	if (sensor_pci_not_working(dev))
 		return MLX5_SENSOR_PCI_COMM_ERR;
@@ -173,7 +173,7 @@ static bool reset_fw_if_needed(struct mlx5_core_dev *dev)
 	 * Check again to avoid a redundant 2nd reset. If the fatal erros was
 	 * PCI related a reset won't help.
 	 */
-	fatal_error = check_fatal_sensors(dev);
+	fatal_error = mlx5_health_check_fatal_sensors(dev);
 	if (fatal_error == MLX5_SENSOR_PCI_COMM_ERR ||
 	    fatal_error == MLX5_SENSOR_NIC_DISABLED ||
 	    fatal_error == MLX5_SENSOR_NIC_SW_RESET) {
@@ -195,7 +195,7 @@ void mlx5_enter_error_state(struct mlx5_core_dev *dev, bool force)
 	bool err_detected = false;
 
 	/* Mark the device as fatal in order to abort FW commands */
-	if ((check_fatal_sensors(dev) || force) &&
+	if ((mlx5_health_check_fatal_sensors(dev) || force) &&
 	    dev->state == MLX5_DEVICE_STATE_UP) {
 		dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
 		err_detected = true;
@@ -208,7 +208,7 @@ void mlx5_enter_error_state(struct mlx5_core_dev *dev, bool force)
 		goto unlock;
 	}
 
-	if (check_fatal_sensors(dev) || force) { /* protected state setting */
+	if (mlx5_health_check_fatal_sensors(dev) || force) { /* protected state setting */
 		dev->state = MLX5_DEVICE_STATE_INTERNAL_ERROR;
 		mlx5_cmd_flush(dev);
 	}
@@ -231,7 +231,7 @@ void mlx5_error_sw_reset(struct mlx5_core_dev *dev)
 
 	mlx5_core_err(dev, "start\n");
 
-	if (check_fatal_sensors(dev) == MLX5_SENSOR_FW_SYND_RFR) {
+	if (mlx5_health_check_fatal_sensors(dev) == MLX5_SENSOR_FW_SYND_RFR) {
 		/* Get cr-dump and reset FW semaphore */
 		lock = lock_sem_sw_reset(dev, true);
 
@@ -308,26 +308,31 @@ static void mlx5_handle_bad_state(struct mlx5_core_dev *dev)
 
 /* How much time to wait until health resetting the driver (in msecs) */
 #define MLX5_RECOVERY_WAIT_MSECS 60000
-static int mlx5_health_try_recover(struct mlx5_core_dev *dev)
+int mlx5_health_wait_pci_up(struct mlx5_core_dev *dev)
 {
 	unsigned long end;
 
-	mlx5_core_warn(dev, "handling bad device here\n");
-	mlx5_handle_bad_state(dev);
 	end = jiffies + msecs_to_jiffies(MLX5_RECOVERY_WAIT_MSECS);
 	while (sensor_pci_not_working(dev)) {
-		if (time_after(jiffies, end)) {
-			mlx5_core_err(dev,
-				      "health recovery flow aborted, PCI reads still not working\n");
-			return -EIO;
-		}
+		if (time_after(jiffies, end))
+			return -ETIMEDOUT;
 		msleep(100);
 	}
+	return 0;
+}
 
+static int mlx5_health_try_recover(struct mlx5_core_dev *dev)
+{
+	mlx5_core_warn(dev, "handling bad device here\n");
+	mlx5_handle_bad_state(dev);
+	if (mlx5_health_wait_pci_up(dev)) {
+		mlx5_core_err(dev, "health recovery flow aborted, PCI reads still not working\n");
+		return -EIO;
+	}
 	mlx5_core_err(dev, "starting health recovery flow\n");
 	mlx5_recover_device(dev);
 	if (!test_bit(MLX5_INTERFACE_STATE_UP, &dev->intf_state) ||
-	    check_fatal_sensors(dev)) {
+	    mlx5_health_check_fatal_sensors(dev)) {
 		mlx5_core_err(dev, "health recovery failed\n");
 		return -EIO;
 	}
@@ -696,7 +701,7 @@ static void poll_health(struct timer_list *t)
 	if (dev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR)
 		goto out;
 
-	fatal_error = check_fatal_sensors(dev);
+	fatal_error = mlx5_health_check_fatal_sensors(dev);
 
 	if (fatal_error && !health->fatal_error) {
 		mlx5_core_err(dev, "Fatal error %u detected\n", fatal_error);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index 871d28b09f8a..e833db424f11 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -57,6 +57,7 @@
 #include "lib/mpfs.h"
 #include "eswitch.h"
 #include "devlink.h"
+#include "fw_reset.h"
 #include "lib/mlx5.h"
 #include "fpga/core.h"
 #include "fpga/ipsec.h"
@@ -835,6 +836,12 @@ static int mlx5_init_once(struct mlx5_core_dev *dev)
 		goto err_eq_cleanup;
 	}
 
+	err = mlx5_fw_reset_events_init(dev);
+	if (err) {
+		mlx5_core_err(dev, "failed to initialize fw reset events\n");
+		goto err_events_cleanup;
+	}
+
 	mlx5_cq_debugfs_init(dev);
 
 	mlx5_init_reserved_gids(dev);
@@ -896,6 +903,8 @@ static int mlx5_init_once(struct mlx5_core_dev *dev)
 	mlx5_geneve_destroy(dev->geneve);
 	mlx5_vxlan_destroy(dev->vxlan);
 	mlx5_cq_debugfs_cleanup(dev);
+	mlx5_fw_reset_events_cleanup(dev);
+err_events_cleanup:
 	mlx5_events_cleanup(dev);
 err_eq_cleanup:
 	mlx5_eq_table_cleanup(dev);
@@ -923,6 +932,7 @@ static void mlx5_cleanup_once(struct mlx5_core_dev *dev)
 	mlx5_cleanup_clock(dev);
 	mlx5_cleanup_reserved_gids(dev);
 	mlx5_cq_debugfs_cleanup(dev);
+	mlx5_fw_reset_events_cleanup(dev);
 	mlx5_events_cleanup(dev);
 	mlx5_eq_table_cleanup(dev);
 	mlx5_irq_table_cleanup(dev);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
index fc1649dac11b..d07a32165792 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h
@@ -123,6 +123,8 @@ int mlx5_cmd_force_teardown_hca(struct mlx5_core_dev *dev);
 int mlx5_cmd_fast_teardown_hca(struct mlx5_core_dev *dev);
 void mlx5_enter_error_state(struct mlx5_core_dev *dev, bool force);
 void mlx5_error_sw_reset(struct mlx5_core_dev *dev);
+u32 mlx5_health_check_fatal_sensors(struct mlx5_core_dev *dev);
+int mlx5_health_wait_pci_up(struct mlx5_core_dev *dev);
 void mlx5_disable_device(struct mlx5_core_dev *dev);
 void mlx5_recover_device(struct mlx5_core_dev *dev);
 int mlx5_sriov_init(struct mlx5_core_dev *dev);
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 8dc3da6e6480..80e31a7684e0 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -501,6 +501,7 @@ struct mlx5_mpfs;
 struct mlx5_eswitch;
 struct mlx5_lag;
 struct mlx5_devcom;
+struct mlx5_fw_reset;
 struct mlx5_eq_table;
 struct mlx5_irq_table;
 
@@ -578,6 +579,7 @@ struct mlx5_priv {
 	struct mlx5_core_sriov	sriov;
 	struct mlx5_lag		*lag;
 	struct mlx5_devcom	*devcom;
+	struct mlx5_fw_reset	*fw_reset;
 	struct mlx5_core_roce	roce;
 	struct mlx5_fc_stats		fc_stats;
 	struct mlx5_rl_table            rl_table;
@@ -943,6 +945,8 @@ void mlx5_start_health_poll(struct mlx5_core_dev *dev);
 void mlx5_stop_health_poll(struct mlx5_core_dev *dev, bool disable_health);
 void mlx5_drain_health_wq(struct mlx5_core_dev *dev);
 void mlx5_trigger_health_work(struct mlx5_core_dev *dev);
+void mlx5_health_set_reset_requested_mode(struct mlx5_core_dev *dev);
+void mlx5_health_clear_reset_requested_mode(struct mlx5_core_dev *dev);
 int mlx5_buf_alloc(struct mlx5_core_dev *dev,
 		   int size, struct mlx5_frag_buf *buf);
 void mlx5_buf_free(struct mlx5_core_dev *dev, struct mlx5_frag_buf *buf);
-- 
2.17.1

[PATCH net-next RFC v5 08/15] net/mlx5: Handle sync reset now event

[Moshe Shemesh]

On sync_reset_now event the driver does reload and PCI link toggle to
activate firmware upgrade reset. When the firmware sends this event it
syncs the event on all PFs, so all PFs will do PCI link toggle at once.
To do PCI link toggle, the driver ensures that no other device ID under
the same bridge by checking that all the PF functions under the same PCI
bridge have same device ID. If no other device it uses PCI bridge link
control to turn link down and up.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
 .../ethernet/mellanox/mlx5/core/fw_reset.c    | 121 ++++++++++++++++++
 1 file changed, 121 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
index 0f224454b4a2..f9e293de9de3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -13,6 +13,7 @@ struct mlx5_fw_reset {
 	struct workqueue_struct *wq;
 	struct work_struct reset_request_work;
 	struct work_struct reset_reload_work;
+	struct work_struct reset_now_work;
 	unsigned long reset_flags;
 	struct timer_list timer;
 };
@@ -149,6 +150,122 @@ static void mlx5_sync_reset_request_event(struct work_struct *work)
 		mlx5_core_warn(dev, "PCI Sync FW Update Reset Ack. Device reset is expected.\n");
 }
 
+#define MLX5_PCI_LINK_UP_TIMEOUT 2000
+
+static int mlx5_pci_link_toggle(struct mlx5_core_dev *dev)
+{
+	struct pci_bus *bridge_bus = dev->pdev->bus;
+	struct pci_dev *bridge = bridge_bus->self;
+	u16 reg16, dev_id, sdev_id;
+	unsigned long timeout;
+	struct pci_dev *sdev;
+	int cap, err;
+	u32 reg32;
+
+	/* Check that all functions under the pci bridge are PFs of
+	 * this device otherwise fail this function.
+	 */
+	err = pci_read_config_word(dev->pdev, PCI_DEVICE_ID, &dev_id);
+	if (err)
+		return err;
+	list_for_each_entry(sdev, &bridge_bus->devices, bus_list) {
+		err = pci_read_config_word(sdev, PCI_DEVICE_ID, &sdev_id);
+		if (err)
+			return err;
+		if (sdev_id != dev_id)
+			return -EPERM;
+	}
+
+	cap = pci_find_capability(bridge, PCI_CAP_ID_EXP);
+	if (!cap)
+		return -EOPNOTSUPP;
+
+	list_for_each_entry(sdev, &bridge_bus->devices, bus_list) {
+		pci_save_state(sdev);
+		pci_cfg_access_lock(sdev);
+	}
+	/* PCI link toggle */
+	err = pci_read_config_word(bridge, cap + PCI_EXP_LNKCTL, &reg16);
+	if (err)
+		return err;
+	reg16 |= PCI_EXP_LNKCTL_LD;
+	err = pci_write_config_word(bridge, cap + PCI_EXP_LNKCTL, reg16);
+	if (err)
+		return err;
+	msleep(500);
+	reg16 &= ~PCI_EXP_LNKCTL_LD;
+	err = pci_write_config_word(bridge, cap + PCI_EXP_LNKCTL, reg16);
+	if (err)
+		return err;
+
+	/* Check link */
+	err = pci_read_config_dword(bridge, cap + PCI_EXP_LNKCAP, &reg32);
+	if (err)
+		return err;
+	if (!(reg32 & PCI_EXP_LNKCAP_DLLLARC)) {
+		mlx5_core_warn(dev, "No PCI link reporting capability (0x%08x)\n", reg32);
+		msleep(1000);
+		goto restore;
+	}
+
+	timeout = jiffies + msecs_to_jiffies(MLX5_PCI_LINK_UP_TIMEOUT);
+	do {
+		err = pci_read_config_word(bridge, cap + PCI_EXP_LNKSTA, &reg16);
+		if (err)
+			return err;
+		if (reg16 & PCI_EXP_LNKSTA_DLLLA)
+			break;
+		msleep(20);
+	} while (!time_after(jiffies, timeout));
+
+	if (reg16 & PCI_EXP_LNKSTA_DLLLA) {
+		mlx5_core_info(dev, "PCI Link up\n");
+	} else {
+		mlx5_core_err(dev, "PCI link not ready (0x%04x) after %d ms\n",
+			      reg16, MLX5_PCI_LINK_UP_TIMEOUT);
+		err = -ETIMEDOUT;
+	}
+
+restore:
+	list_for_each_entry(sdev, &bridge_bus->devices, bus_list) {
+		pci_cfg_access_unlock(sdev);
+		pci_restore_state(sdev);
+	}
+
+	return err;
+}
+
+static void mlx5_sync_reset_now_event(struct work_struct *work)
+{
+	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
+						      reset_now_work);
+	struct mlx5_core_dev *dev = fw_reset->dev;
+	int err;
+
+	mlx5_sync_reset_clear_reset_requested(dev, false);
+
+	mlx5_core_warn(dev, "Sync Reset now. Device is going to reset.\n");
+
+	err = mlx5_cmd_fast_teardown_hca(dev);
+	if (err) {
+		mlx5_core_warn(dev, "Fast teardown failed, no reset done, err %d\n", err);
+		goto done;
+	}
+
+	err = mlx5_pci_link_toggle(dev);
+	if (err) {
+		mlx5_core_warn(dev, "mlx5_pci_link_toggle failed, no reset done, err %d\n", err);
+		goto done;
+	}
+
+	mlx5_enter_error_state(dev, true);
+	mlx5_unload_one(dev, false);
+done:
+	if (err)
+		mlx5_start_health_poll(dev);
+	mlx5_load_one(dev, false);
+}
+
 static void mlx5_sync_reset_events_handle(struct mlx5_fw_reset *fw_reset, struct mlx5_eqe *eqe)
 {
 	struct mlx5_eqe_sync_fw_update *sync_fw_update_eqe;
@@ -160,6 +277,9 @@ static void mlx5_sync_reset_events_handle(struct mlx5_fw_reset *fw_reset, struct
 	case MLX5_SYNC_RST_STATE_RESET_REQUEST:
 		queue_work(fw_reset->wq, &fw_reset->reset_request_work);
 		break;
+	case MLX5_SYNC_RST_STATE_RESET_NOW:
+		queue_work(fw_reset->wq, &fw_reset->reset_now_work);
+		break;
 	}
 }
 
@@ -196,6 +316,7 @@ int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev)
 
 	INIT_WORK(&fw_reset->reset_request_work, mlx5_sync_reset_request_event);
 	INIT_WORK(&fw_reset->reset_reload_work, mlx5_sync_reset_reload_work);
+	INIT_WORK(&fw_reset->reset_now_work, mlx5_sync_reset_now_event);
 
 	MLX5_NB_INIT(&fw_reset->nb, fw_reset_event_notifier, GENERAL_EVENT);
 	mlx5_eq_notifier_register(dev, &fw_reset->nb);
-- 
2.17.1

[PATCH net-next RFC v5 09/15] net/mlx5: Handle sync reset abort event

[Moshe Shemesh]

If firmware sends sync_reset_abort to driver the driver should clear the
reset requested mode as reset is not expected any more.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
 .../net/ethernet/mellanox/mlx5/core/fw_reset.c    | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
index f9e293de9de3..61237f4836cc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -14,6 +14,7 @@ struct mlx5_fw_reset {
 	struct work_struct reset_request_work;
 	struct work_struct reset_reload_work;
 	struct work_struct reset_now_work;
+	struct work_struct reset_abort_work;
 	unsigned long reset_flags;
 	struct timer_list timer;
 };
@@ -266,6 +267,16 @@ static void mlx5_sync_reset_now_event(struct work_struct *work)
 	mlx5_load_one(dev, false);
 }
 
+static void mlx5_sync_reset_abort_event(struct work_struct *work)
+{
+	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
+						      reset_abort_work);
+	struct mlx5_core_dev *dev = fw_reset->dev;
+
+	mlx5_sync_reset_clear_reset_requested(dev, true);
+	mlx5_core_warn(dev, "PCI Sync FW Update Reset Aborted.\n");
+}
+
 static void mlx5_sync_reset_events_handle(struct mlx5_fw_reset *fw_reset, struct mlx5_eqe *eqe)
 {
 	struct mlx5_eqe_sync_fw_update *sync_fw_update_eqe;
@@ -280,6 +291,9 @@ static void mlx5_sync_reset_events_handle(struct mlx5_fw_reset *fw_reset, struct
 	case MLX5_SYNC_RST_STATE_RESET_NOW:
 		queue_work(fw_reset->wq, &fw_reset->reset_now_work);
 		break;
+	case MLX5_SYNC_RST_STATE_RESET_ABORT:
+		queue_work(fw_reset->wq, &fw_reset->reset_abort_work);
+		break;
 	}
 }
 
@@ -317,6 +331,7 @@ int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev)
 	INIT_WORK(&fw_reset->reset_request_work, mlx5_sync_reset_request_event);
 	INIT_WORK(&fw_reset->reset_reload_work, mlx5_sync_reset_reload_work);
 	INIT_WORK(&fw_reset->reset_now_work, mlx5_sync_reset_now_event);
+	INIT_WORK(&fw_reset->reset_abort_work, mlx5_sync_reset_abort_event);
 
 	MLX5_NB_INIT(&fw_reset->nb, fw_reset_event_notifier, GENERAL_EVENT);
 	mlx5_eq_notifier_register(dev, &fw_reset->nb);
-- 
2.17.1

[PATCH net-next RFC v5 10/15] net/mlx5: Add support for devlink reload action fw activate

[Moshe Shemesh]

Add support for devlink reload action fw_activate. To activate firmware
image the mlx5 driver resets the firmware and reloads it from flash. If
a new image was stored on flash it will be loaded. Once this reload
command is executed the driver initiates fw sync reset flow, where the
firmware synchronizes all PFs on coming reset and driver reload.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v4 -> v5:
- If fw activation was called by another PF update the remote actions
  stats
v3 -> v4:
- Renamed actions_done to actions_performed
v2 -> v3:
- Return the reload actions done
- Update reload action counters if reset initiated by remote host
v1 -> v2:
- Have fw_activate action instead of fw_reset level
---
 .../net/ethernet/mellanox/mlx5/core/devlink.c | 61 ++++++++++++++++---
 .../ethernet/mellanox/mlx5/core/fw_reset.c    | 60 ++++++++++++++++--
 .../ethernet/mellanox/mlx5/core/fw_reset.h    |  1 +
 3 files changed, 108 insertions(+), 14 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index 38b00b4a3ce8..9b95434873ed 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -4,6 +4,7 @@
 #include <devlink.h>
 
 #include "mlx5_core.h"
+#include "fw_reset.h"
 #include "fs_core.h"
 #include "eswitch.h"
 
@@ -88,6 +89,32 @@ mlx5_devlink_info_get(struct devlink *devlink, struct devlink_info_req *req,
 	return 0;
 }
 
+static int mlx5_devlink_reload_fw_activate(struct devlink *devlink, struct netlink_ext_ack *extack)
+{
+	struct mlx5_core_dev *dev = devlink_priv(devlink);
+	u8 reset_level, reset_type, net_port_alive;
+	int err;
+
+	err = mlx5_reg_mfrl_query(dev, &reset_level, &reset_type);
+	if (err)
+		return err;
+	if (!(reset_level & MLX5_MFRL_REG_RESET_LEVEL3)) {
+		NL_SET_ERR_MSG_MOD(extack, "FW activate requires reboot");
+		return -EINVAL;
+	}
+
+	net_port_alive = !!(reset_type & MLX5_MFRL_REG_RESET_TYPE_NET_PORT_ALIVE);
+	err = mlx5_fw_set_reset_sync(dev, net_port_alive);
+	if (err)
+		goto out;
+
+	err = mlx5_fw_wait_fw_reset_done(dev);
+out:
+	if (err)
+		NL_SET_ERR_MSG_MOD(extack, "FW activate command failed");
+	return err;
+}
+
 static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 				    enum devlink_reload_action action,
 				    enum devlink_reload_action_limit_level limit_level,
@@ -95,8 +122,17 @@ static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 {
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
 
-	mlx5_unload_one(dev, false);
-	return 0;
+	switch (action) {
+	case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
+		mlx5_unload_one(dev, false);
+		return 0;
+	case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
+		return mlx5_devlink_reload_fw_activate(devlink, extack);
+	default:
+		/* Unsupported action should not get to this function */
+		WARN_ON(1);
+		return -EOPNOTSUPP;
+	}
 }
 
 static int mlx5_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
@@ -104,12 +140,20 @@ static int mlx5_devlink_reload_up(struct devlink *devlink, enum devlink_reload_a
 				  struct netlink_ext_ack *extack, unsigned long *actions_performed)
 {
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
-	int err;
 
-	err = mlx5_load_one(dev, false);
-	if (err)
-		return err;
-	*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
+	*actions_performed = BIT(action);
+
+	switch (action) {
+	case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
+	case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
+		/* On fw_activate action, also driver is reloaded and reinit performed */
+		*actions_performed |= BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
+		return mlx5_load_one(dev, false);
+	default:
+		/* Unsupported action should not get to this function */
+		WARN_ON(1);
+		return -EOPNOTSUPP;
+	}
 
 	return 0;
 }
@@ -127,7 +171,8 @@ static const struct devlink_ops mlx5_devlink_ops = {
 #endif
 	.flash_update = mlx5_devlink_flash_update,
 	.info_get = mlx5_devlink_info_get,
-	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
+	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
+				    BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE),
 	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),
 	.reload_down = mlx5_devlink_reload_down,
 	.reload_up = mlx5_devlink_reload_up,
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
index 61237f4836cc..5519af94feb6 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -5,6 +5,7 @@
 
 enum {
 	MLX5_FW_RESET_FLAGS_RESET_REQUESTED,
+	MLX5_FW_RESET_FLAGS_PENDING_COMP
 };
 
 struct mlx5_fw_reset {
@@ -17,6 +18,8 @@ struct mlx5_fw_reset {
 	struct work_struct reset_abort_work;
 	unsigned long reset_flags;
 	struct timer_list timer;
+	struct completion done;
+	int ret;
 };
 
 static int mlx5_reg_mfrl_set(struct mlx5_core_dev *dev, u8 reset_level,
@@ -53,7 +56,14 @@ int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_ty
 
 int mlx5_fw_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel)
 {
-	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, reset_type_sel, 0, true);
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+	int err;
+
+	set_bit(MLX5_FW_RESET_FLAGS_PENDING_COMP, &fw_reset->reset_flags);
+	err =  mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, reset_type_sel, 0, true);
+	if (err)
+		clear_bit(MLX5_FW_RESET_FLAGS_PENDING_COMP, &fw_reset->reset_flags);
+	return err;
 }
 
 int mlx5_fw_set_live_patch(struct mlx5_core_dev *dev)
@@ -66,19 +76,36 @@ static int mlx5_fw_set_reset_sync_ack(struct mlx5_core_dev *dev)
 	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, 0, 1, false);
 }
 
+static void mlx5_fw_reset_complete_reload(struct mlx5_core_dev *dev)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	/* if this is the driver that initiated the fw reset, devlink completed the reload */
+	if (test_bit(MLX5_FW_RESET_FLAGS_PENDING_COMP, &fw_reset->reset_flags)) {
+		complete(&fw_reset->done);
+	} else {
+		mlx5_load_one(dev, false);
+		devlink_remote_reload_actions_performed(priv_to_devlink(dev),
+							DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
+							BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
+							BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE));
+	}
+}
+
 static void mlx5_sync_reset_reload_work(struct work_struct *work)
 {
 	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
 						      reset_reload_work);
 	struct mlx5_core_dev *dev = fw_reset->dev;
+	int err;
 
 	mlx5_enter_error_state(dev, true);
 	mlx5_unload_one(dev, false);
-	if (mlx5_health_wait_pci_up(dev)) {
+	err = mlx5_health_wait_pci_up(dev);
+	if (err)
 		mlx5_core_err(dev, "reset reload flow aborted, PCI reads still not working\n");
-		return;
-	}
-	mlx5_load_one(dev, false);
+	fw_reset->ret = err;
+	mlx5_fw_reset_complete_reload(dev);
 }
 
 static void mlx5_stop_sync_reset_poll(struct mlx5_core_dev *dev)
@@ -264,7 +291,8 @@ static void mlx5_sync_reset_now_event(struct work_struct *work)
 done:
 	if (err)
 		mlx5_start_health_poll(dev);
-	mlx5_load_one(dev, false);
+	fw_reset->ret = err;
+	mlx5_fw_reset_complete_reload(dev);
 }
 
 static void mlx5_sync_reset_abort_event(struct work_struct *work)
@@ -313,6 +341,25 @@ static int fw_reset_event_notifier(struct notifier_block *nb, unsigned long acti
 	return NOTIFY_OK;
 }
 
+#define MLX5_FW_RESET_TIMEOUT_MSEC 5000
+int mlx5_fw_wait_fw_reset_done(struct mlx5_core_dev *dev)
+{
+	unsigned long timeout = msecs_to_jiffies(MLX5_FW_RESET_TIMEOUT_MSEC);
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+	int err;
+
+	if (!wait_for_completion_timeout(&fw_reset->done, timeout)) {
+		mlx5_core_warn(dev, "FW sync reset timeout after %d seconds\n",
+			       MLX5_FW_RESET_TIMEOUT_MSEC / 1000);
+		err = -ETIMEDOUT;
+		goto out;
+	}
+	err = fw_reset->ret;
+out:
+	clear_bit(MLX5_FW_RESET_FLAGS_PENDING_COMP, &fw_reset->reset_flags);
+	return err;
+}
+
 int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev)
 {
 	struct mlx5_fw_reset *fw_reset = kzalloc(sizeof(*fw_reset), GFP_KERNEL);
@@ -336,6 +383,7 @@ int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev)
 	MLX5_NB_INIT(&fw_reset->nb, fw_reset_event_notifier, GENERAL_EVENT);
 	mlx5_eq_notifier_register(dev, &fw_reset->nb);
 
+	init_completion(&fw_reset->done);
 	return 0;
 }
 
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
index 278f538ea92a..d7ee951a2258 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
@@ -10,6 +10,7 @@ int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_ty
 int mlx5_fw_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel);
 int mlx5_fw_set_live_patch(struct mlx5_core_dev *dev);
 
+int mlx5_fw_wait_fw_reset_done(struct mlx5_core_dev *dev);
 int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev);
 void mlx5_fw_reset_events_cleanup(struct mlx5_core_dev *dev);
 
-- 
2.17.1

[PATCH net-next RFC v5 11/15] devlink: Add enable_remote_dev_reset generic parameter

[Moshe Shemesh]

The enable_remote_dev_reset devlink param flags that the host admin
allows device resets that can be initiated by other hosts. This
parameter is useful for setups where a device is shared by different
hosts, such as multi-host setup. Once the user set this parameter to
false, the driver should NACK any attempt to reset the device while the
driver is loaded.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
---
 Documentation/networking/devlink/devlink-params.rst | 6 ++++++
 include/net/devlink.h                               | 4 ++++
 net/core/devlink.c                                  | 5 +++++
 3 files changed, 15 insertions(+)

diff --git a/Documentation/networking/devlink/devlink-params.rst b/Documentation/networking/devlink/devlink-params.rst
index d075fd090b3d..54c9f107c4b0 100644
--- a/Documentation/networking/devlink/devlink-params.rst
+++ b/Documentation/networking/devlink/devlink-params.rst
@@ -108,3 +108,9 @@ own name.
    * - ``region_snapshot_enable``
      - Boolean
      - Enable capture of ``devlink-region`` snapshots.
+   * - ``enable_remote_dev_reset``
+     - Boolean
+     - Enable device reset by remote host. When cleared, the device driver
+       will NACK any attempt of other host to reset the device. This parameter
+       is useful for setups where a device is shared by different hosts, such
+       as multi-host setup.
diff --git a/include/net/devlink.h b/include/net/devlink.h
index f09f55a47d09..8ab5ac5c365f 100644
--- a/include/net/devlink.h
+++ b/include/net/devlink.h
@@ -445,6 +445,7 @@ enum devlink_param_generic_id {
 	DEVLINK_PARAM_GENERIC_ID_FW_LOAD_POLICY,
 	DEVLINK_PARAM_GENERIC_ID_RESET_DEV_ON_DRV_PROBE,
 	DEVLINK_PARAM_GENERIC_ID_ENABLE_ROCE,
+	DEVLINK_PARAM_GENERIC_ID_ENABLE_REMOTE_DEV_RESET,
 
 	/* add new param generic ids above here*/
 	__DEVLINK_PARAM_GENERIC_ID_MAX,
@@ -482,6 +483,9 @@ enum devlink_param_generic_id {
 #define DEVLINK_PARAM_GENERIC_ENABLE_ROCE_NAME "enable_roce"
 #define DEVLINK_PARAM_GENERIC_ENABLE_ROCE_TYPE DEVLINK_PARAM_TYPE_BOOL
 
+#define DEVLINK_PARAM_GENERIC_ENABLE_REMOTE_DEV_RESET_NAME "enable_remote_dev_reset"
+#define DEVLINK_PARAM_GENERIC_ENABLE_REMOTE_DEV_RESET_TYPE DEVLINK_PARAM_TYPE_BOOL
+
 #define DEVLINK_PARAM_GENERIC(_id, _cmodes, _get, _set, _validate)	\
 {									\
 	.id = DEVLINK_PARAM_GENERIC_ID_##_id,				\
diff --git a/net/core/devlink.c b/net/core/devlink.c
index 71aeda259e6a..565fa786d01f 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -3458,6 +3458,11 @@ static const struct devlink_param devlink_param_generic[] = {
 		.name = DEVLINK_PARAM_GENERIC_ENABLE_ROCE_NAME,
 		.type = DEVLINK_PARAM_GENERIC_ENABLE_ROCE_TYPE,
 	},
+	{
+		.id = DEVLINK_PARAM_GENERIC_ID_ENABLE_REMOTE_DEV_RESET,
+		.name = DEVLINK_PARAM_GENERIC_ENABLE_REMOTE_DEV_RESET_NAME,
+		.type = DEVLINK_PARAM_GENERIC_ENABLE_REMOTE_DEV_RESET_TYPE,
+	},
 };
 
 static int devlink_param_generic_verify(const struct devlink_param *param)
-- 
2.17.1

[PATCH net-next RFC v5 12/15] net/mlx5: Add devlink param enable_remote_dev_reset support

[Moshe Shemesh]

The enable_remote_dev_reset devlink param flags that the host admin
allows resets by other hosts. In case it is cleared mlx5 host PF driver
will send NACK on pci sync for firmware update reset request and the
command will fail.
By default enable_remote_dev_reset parameter is true, so pci sync for
firmware update reset is enabled.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v1 -> v2:
- Have MLX5_FW_RESET_FLAGS_NACK_RESET_REQUEST instead of
  MLX5_HEALTH_RESET_FLAGS_NACK_RESET_REQUEST
---
 .../net/ethernet/mellanox/mlx5/core/devlink.c | 21 +++++++++++++
 .../ethernet/mellanox/mlx5/core/fw_reset.c    | 30 +++++++++++++++++++
 .../ethernet/mellanox/mlx5/core/fw_reset.h    |  2 ++
 3 files changed, 53 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index 9b95434873ed..d2e7b52619cc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -284,6 +284,24 @@ static int mlx5_devlink_large_group_num_validate(struct devlink *devlink, u32 id
 }
 #endif
 
+static int mlx5_devlink_enable_remote_dev_reset_set(struct devlink *devlink, u32 id,
+						    struct devlink_param_gset_ctx *ctx)
+{
+	struct mlx5_core_dev *dev = devlink_priv(devlink);
+
+	mlx5_fw_enable_remote_dev_reset_set(dev, ctx->val.vbool);
+	return 0;
+}
+
+static int mlx5_devlink_enable_remote_dev_reset_get(struct devlink *devlink, u32 id,
+						    struct devlink_param_gset_ctx *ctx)
+{
+	struct mlx5_core_dev *dev = devlink_priv(devlink);
+
+	ctx->val.vbool = mlx5_fw_enable_remote_dev_reset_get(dev);
+	return 0;
+}
+
 static const struct devlink_param mlx5_devlink_params[] = {
 	DEVLINK_PARAM_DRIVER(MLX5_DEVLINK_PARAM_ID_FLOW_STEERING_MODE,
 			     "flow_steering_mode", DEVLINK_PARAM_TYPE_STRING,
@@ -299,6 +317,9 @@ static const struct devlink_param mlx5_devlink_params[] = {
 			     NULL, NULL,
 			     mlx5_devlink_large_group_num_validate),
 #endif
+	DEVLINK_PARAM_GENERIC(ENABLE_REMOTE_DEV_RESET, BIT(DEVLINK_PARAM_CMODE_RUNTIME),
+			      mlx5_devlink_enable_remote_dev_reset_get,
+			      mlx5_devlink_enable_remote_dev_reset_set, NULL),
 };
 
 static void mlx5_devlink_set_params_init_values(struct devlink *devlink)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
index 5519af94feb6..95a02e0e3a41 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -5,6 +5,7 @@
 
 enum {
 	MLX5_FW_RESET_FLAGS_RESET_REQUESTED,
+	MLX5_FW_RESET_FLAGS_NACK_RESET_REQUEST,
 	MLX5_FW_RESET_FLAGS_PENDING_COMP
 };
 
@@ -22,6 +23,23 @@ struct mlx5_fw_reset {
 	int ret;
 };
 
+void mlx5_fw_enable_remote_dev_reset_set(struct mlx5_core_dev *dev, bool enable)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	if (enable)
+		clear_bit(MLX5_FW_RESET_FLAGS_NACK_RESET_REQUEST, &fw_reset->reset_flags);
+	else
+		set_bit(MLX5_FW_RESET_FLAGS_NACK_RESET_REQUEST, &fw_reset->reset_flags);
+}
+
+bool mlx5_fw_enable_remote_dev_reset_get(struct mlx5_core_dev *dev)
+{
+	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
+
+	return !test_bit(MLX5_FW_RESET_FLAGS_NACK_RESET_REQUEST, &fw_reset->reset_flags);
+}
+
 static int mlx5_reg_mfrl_set(struct mlx5_core_dev *dev, u8 reset_level,
 			     u8 reset_type_sel, u8 sync_resp, bool sync_start)
 {
@@ -76,6 +94,11 @@ static int mlx5_fw_set_reset_sync_ack(struct mlx5_core_dev *dev)
 	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, 0, 1, false);
 }
 
+static int mlx5_fw_set_reset_sync_nack(struct mlx5_core_dev *dev)
+{
+	return mlx5_reg_mfrl_set(dev, MLX5_MFRL_REG_RESET_LEVEL3, 0, 2, false);
+}
+
 static void mlx5_fw_reset_complete_reload(struct mlx5_core_dev *dev)
 {
 	struct mlx5_fw_reset *fw_reset = dev->priv.fw_reset;
@@ -170,7 +193,14 @@ static void mlx5_sync_reset_request_event(struct work_struct *work)
 	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
 						      reset_request_work);
 	struct mlx5_core_dev *dev = fw_reset->dev;
+	int err;
 
+	if (test_bit(MLX5_FW_RESET_FLAGS_NACK_RESET_REQUEST, &fw_reset->reset_flags)) {
+		err = mlx5_fw_set_reset_sync_nack(dev);
+		mlx5_core_warn(dev, "PCI Sync FW Update Reset Nack %s",
+			       err ? "Failed" : "Sent");
+		return;
+	}
 	mlx5_sync_reset_set_reset_requested(dev);
 	if (mlx5_fw_set_reset_sync_ack(dev))
 		mlx5_core_warn(dev, "PCI Sync FW Update Reset Ack Failed.\n");
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
index d7ee951a2258..fd558dfe93fc 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
@@ -6,6 +6,8 @@
 
 #include "mlx5_core.h"
 
+void mlx5_fw_enable_remote_dev_reset_set(struct mlx5_core_dev *dev, bool enable);
+bool mlx5_fw_enable_remote_dev_reset_get(struct mlx5_core_dev *dev);
 int mlx5_reg_mfrl_query(struct mlx5_core_dev *dev, u8 *reset_level, u8 *reset_type);
 int mlx5_fw_set_reset_sync(struct mlx5_core_dev *dev, u8 reset_type_sel);
 int mlx5_fw_set_live_patch(struct mlx5_core_dev *dev);
-- 
2.17.1

[PATCH net-next RFC v5 13/15] net/mlx5: Add support for fw live patch event

[Moshe Shemesh]

Firmware live patch event notifies the driver that the firmware was just
updated using live patch. In such case the driver should not reload or
re-initiate entities, part to updating the firmware version and
re-initiate the firmware tracer which can be updated by live patch with
new strings database to help debugging an issue.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
 .../mellanox/mlx5/core/diag/fw_tracer.c       | 31 +++++++++++++++++++
 .../mellanox/mlx5/core/diag/fw_tracer.h       |  1 +
 .../ethernet/mellanox/mlx5/core/fw_reset.c    | 27 ++++++++++++++++
 include/linux/mlx5/device.h                   |  1 +
 4 files changed, 60 insertions(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.c b/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.c
index ad3594c4afcb..08dae045d185 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.c
@@ -1064,6 +1064,37 @@ void mlx5_fw_tracer_destroy(struct mlx5_fw_tracer *tracer)
 	kvfree(tracer);
 }
 
+int mlx5_fw_tracer_recreate_strings_db(struct mlx5_fw_tracer *tracer)
+{
+	struct mlx5_core_dev *dev;
+	int err;
+
+	if (IS_ERR_OR_NULL(tracer))
+		return -EINVAL;
+
+	cancel_work_sync(&tracer->read_fw_strings_work);
+	mlx5_fw_tracer_clean_ready_list(tracer);
+	mlx5_fw_tracer_clean_print_hash(tracer);
+	mlx5_fw_tracer_clean_saved_traces_array(tracer);
+	mlx5_fw_tracer_free_strings_db(tracer);
+
+	dev = tracer->dev;
+	err = mlx5_query_mtrc_caps(tracer);
+	if (err) {
+		mlx5_core_dbg(dev, "FWTracer: Failed to query capabilities %d\n", err);
+		return err;
+	}
+
+	err = mlx5_fw_tracer_allocate_strings_db(tracer);
+	if (err) {
+		mlx5_core_warn(dev, "FWTracer: Allocate strings DB failed %d\n", err);
+		return err;
+	}
+	mlx5_fw_tracer_init_saved_traces_array(tracer);
+
+	return 0;
+}
+
 static int fw_tracer_event(struct notifier_block *nb, unsigned long action, void *data)
 {
 	struct mlx5_fw_tracer *tracer = mlx5_nb_cof(nb, struct mlx5_fw_tracer, nb);
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.h b/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.h
index 40601fba80ba..1a755098aeeb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.h
+++ b/drivers/net/ethernet/mellanox/mlx5/core/diag/fw_tracer.h
@@ -191,5 +191,6 @@ void mlx5_fw_tracer_destroy(struct mlx5_fw_tracer *tracer);
 int mlx5_fw_tracer_trigger_core_dump_general(struct mlx5_core_dev *dev);
 int mlx5_fw_tracer_get_saved_traces_objects(struct mlx5_fw_tracer *tracer,
 					    struct devlink_fmsg *fmsg);
+int mlx5_fw_tracer_recreate_strings_db(struct mlx5_fw_tracer *tracer);
 
 #endif
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
index 95a02e0e3a41..cc838e5c0172 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
@@ -2,6 +2,7 @@
 /* Copyright (c) 2020, Mellanox Technologies inc.  All rights reserved. */
 
 #include "fw_reset.h"
+#include "diag/fw_tracer.h"
 
 enum {
 	MLX5_FW_RESET_FLAGS_RESET_REQUESTED,
@@ -13,6 +14,7 @@ struct mlx5_fw_reset {
 	struct mlx5_core_dev *dev;
 	struct mlx5_nb nb;
 	struct workqueue_struct *wq;
+	struct work_struct fw_live_patch_work;
 	struct work_struct reset_request_work;
 	struct work_struct reset_reload_work;
 	struct work_struct reset_now_work;
@@ -188,6 +190,27 @@ static void mlx5_sync_reset_set_reset_requested(struct mlx5_core_dev *dev)
 	mlx5_start_sync_reset_poll(dev);
 }
 
+static void mlx5_fw_live_patch_event(struct work_struct *work)
+{
+	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
+						      fw_live_patch_work);
+	struct mlx5_core_dev *dev = fw_reset->dev;
+	struct mlx5_fw_tracer *tracer;
+
+	mlx5_core_info(dev, "Live patch updated firmware version: %d.%d.%d\n", fw_rev_maj(dev),
+		       fw_rev_min(dev), fw_rev_sub(dev));
+
+	tracer = dev->tracer;
+	if (IS_ERR_OR_NULL(tracer))
+		return;
+
+	mlx5_fw_tracer_cleanup(tracer);
+	if (mlx5_fw_tracer_recreate_strings_db(tracer))
+		mlx5_core_err(dev, "Failed to recreate FW tracer strings DB\n");
+	if (mlx5_fw_tracer_init(tracer))
+		mlx5_core_err(dev, "Failed to re-initialize FW tracer\n");
+}
+
 static void mlx5_sync_reset_request_event(struct work_struct *work)
 {
 	struct mlx5_fw_reset *fw_reset = container_of(work, struct mlx5_fw_reset,
@@ -361,6 +384,9 @@ static int fw_reset_event_notifier(struct notifier_block *nb, unsigned long acti
 	struct mlx5_eqe *eqe = data;
 
 	switch (eqe->sub_type) {
+	case MLX5_GENERAL_SUBTYPE_FW_LIVE_PATCH_EVENT:
+			queue_work(fw_reset->wq, &fw_reset->fw_live_patch_work);
+		break;
 	case MLX5_GENERAL_SUBTYPE_PCI_SYNC_FOR_FW_UPDATE_EVENT:
 		mlx5_sync_reset_events_handle(fw_reset, eqe);
 		break;
@@ -405,6 +431,7 @@ int mlx5_fw_reset_events_init(struct mlx5_core_dev *dev)
 	fw_reset->dev = dev;
 	dev->priv.fw_reset = fw_reset;
 
+	INIT_WORK(&fw_reset->fw_live_patch_work, mlx5_fw_live_patch_event);
 	INIT_WORK(&fw_reset->reset_request_work, mlx5_sync_reset_request_event);
 	INIT_WORK(&fw_reset->reset_reload_work, mlx5_sync_reset_reload_work);
 	INIT_WORK(&fw_reset->reset_now_work, mlx5_sync_reset_now_event);
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 81ca5989009b..cf824366a7d1 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -366,6 +366,7 @@ enum {
 enum {
 	MLX5_GENERAL_SUBTYPE_DELAY_DROP_TIMEOUT = 0x1,
 	MLX5_GENERAL_SUBTYPE_PCI_POWER_CHANGE_EVENT = 0x5,
+	MLX5_GENERAL_SUBTYPE_FW_LIVE_PATCH_EVENT = 0x7,
 	MLX5_GENERAL_SUBTYPE_PCI_SYNC_FOR_FW_UPDATE_EVENT = 0x8,
 };
 
-- 
2.17.1

[PATCH net-next RFC v5 14/15] net/mlx5: Add support for devlink reload action limit level no reset

[Moshe Shemesh]

Add support for devlink reload action fw_activate with limit level
no_reset which does firmware live patching, updating the firmware image
without reset, no downtime and no configuration lose. The driver checks
if the firmware is capable of handling the pending firmware changes as a
live patch. If it is then it triggers firmware live patching flow.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v3 -> v4:
- Have action fw_activate with limit level no_reset instead of action
  fw_activate_no_reset
v2 -> v3:
- Replace fw_live_patch action by fw_activate_no_reset
v1 -> v2:
- Have fw_live_patch action instead of level
---
 .../net/ethernet/mellanox/mlx5/core/devlink.c | 37 ++++++++++++++++++-
 1 file changed, 36 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
index d2e7b52619cc..02fe57fb507f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/devlink.c
@@ -115,6 +115,29 @@ static int mlx5_devlink_reload_fw_activate(struct devlink *devlink, struct netli
 	return err;
 }
 
+static int mlx5_devlink_trigger_fw_live_patch(struct devlink *devlink,
+					      struct netlink_ext_ack *extack)
+{
+	struct mlx5_core_dev *dev = devlink_priv(devlink);
+	u8 reset_level;
+	int err;
+
+	err = mlx5_reg_mfrl_query(dev, &reset_level, NULL);
+	if (err)
+		return err;
+	if (!(reset_level & MLX5_MFRL_REG_RESET_LEVEL0)) {
+		NL_SET_ERR_MSG_MOD(extack,
+				   "FW upgrade to the stored FW can't be done by FW live patching");
+		return -EINVAL;
+	}
+
+	err = mlx5_fw_set_live_patch(dev);
+	if (err)
+		return err;
+
+	return 0;
+}
+
 static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 				    enum devlink_reload_action action,
 				    enum devlink_reload_action_limit_level limit_level,
@@ -122,11 +145,19 @@ static int mlx5_devlink_reload_down(struct devlink *devlink, bool netns_change,
 {
 	struct mlx5_core_dev *dev = devlink_priv(devlink);
 
+	if (limit_level == DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET &&
+	    action != DEVLINK_RELOAD_ACTION_FW_ACTIVATE) {
+		NL_SET_ERR_MSG_MOD(extack, "Requested limit level is not supported");
+		return -EOPNOTSUPP;
+	}
+
 	switch (action) {
 	case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
 		mlx5_unload_one(dev, false);
 		return 0;
 	case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
+		if (limit_level == DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET)
+			return mlx5_devlink_trigger_fw_live_patch(devlink, extack);
 		return mlx5_devlink_reload_fw_activate(devlink, extack);
 	default:
 		/* Unsupported action should not get to this function */
@@ -145,7 +176,10 @@ static int mlx5_devlink_reload_up(struct devlink *devlink, enum devlink_reload_a
 
 	switch (action) {
 	case DEVLINK_RELOAD_ACTION_DRIVER_REINIT:
+		return mlx5_load_one(dev, false);
 	case DEVLINK_RELOAD_ACTION_FW_ACTIVATE:
+		if (limit_level == DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET)
+			break;
 		/* On fw_activate action, also driver is reloaded and reinit performed */
 		*actions_performed |= BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
 		return mlx5_load_one(dev, false);
@@ -173,7 +207,8 @@ static const struct devlink_ops mlx5_devlink_ops = {
 	.info_get = mlx5_devlink_info_get,
 	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT) |
 				    BIT(DEVLINK_RELOAD_ACTION_FW_ACTIVATE),
-	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),
+	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE) |
+						BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET),
 	.reload_down = mlx5_devlink_reload_down,
 	.reload_up = mlx5_devlink_reload_up,
 };
-- 
2.17.1

[PATCH net-next RFC v5 15/15] devlink: Add Documentation/networking/devlink/devlink-reload.rst

[Moshe Shemesh]

Add devlink reload rst documentation file.
Update index file to include it.

Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
---
v4 -> v5:
- Rephrase namespace chnage section
- Rephrase note on actions performed
v3 -> v4:
- Remove reload action fw_activate_no_reset
- Add reload actions limit levels and document the no_reset limit level
  constrains
v2 -> v3:
- Devlink reload returns the actions done
- Replace fw_live_patch action by fw_activate_no_reset
- Explain fw_activate meaning
v1 -> v2:
- Instead of reload levels driver,fw_reset,fw_live_patch have reload
  actions driver_reinit,fw_activate,fw_live_patch
---
 .../networking/devlink/devlink-reload.rst     | 79 +++++++++++++++++++
 Documentation/networking/devlink/index.rst    |  1 +
 2 files changed, 80 insertions(+)
 create mode 100644 Documentation/networking/devlink/devlink-reload.rst

diff --git a/Documentation/networking/devlink/devlink-reload.rst b/Documentation/networking/devlink/devlink-reload.rst
new file mode 100644
index 000000000000..ffc72c0f635f
--- /dev/null
+++ b/Documentation/networking/devlink/devlink-reload.rst
@@ -0,0 +1,79 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+==============
+Devlink Reload
+==============
+
+``devlink-reload`` provides mechanism to either reinit driver entities,
+applying ``devlink-params`` and ``devlink-resources`` new values or firmware
+activation depends on reload action selected.
+
+Reload actions
+==============
+
+User may select a reload action.
+By default ``driver_reinit`` action is selected.
+
+.. list-table:: Possible reload actions
+   :widths: 5 90
+
+   * - Name
+     - Description
+   * - ``driver-reinit``
+     - Devlink driver entities re-initialization, including applying
+       new values to devlink entities which are used during driver
+       load such as ``devlink-params`` in configuration mode
+       ``driverinit`` or ``devlink-resources``
+   * - ``fw_activate``
+     - Firmware activate. Activates new firmware if such image is stored and
+       pending activation. This action involves firmware reset, if no new image
+       pending this action will reload current firmware image.
+
+Note that even though user asks for a specific action, the driver
+implementation might require to perform another action alongside with
+it. For example, some driver do not support driver reinitialization
+being performed without fw activation. Therefore, the devlink reload
+command return the list of actions which were actrually performed.
+
+Reload action limit levels
+==========================
+
+By default reload actions are not limited and driver implementation may
+include reset or downtime as needed to perform the actions.
+
+However, some drivers support action limit levels, which limits the action
+implementation to specific constrains.
+
+.. list-table:: Possible reload action limit levels
+   :widths: 5 90
+
+   * - Name
+     - Description
+   * - ``no_reset``
+     - No reset allowed, no down time allowed, no link flap and no
+       configuration is lost.
+
+Change namespace
+================
+
+The netns option allow user to be able to move devlink instances into
+namespaces during devlink reload operation.
+By default all devlink instances are created in init_net and stay there.
+
+example usage
+-------------
+
+.. code:: shell
+
+    $ devlink dev reload help
+    $ devlink dev reload DEV [ netns { PID | NAME | ID } ] [ action { driver_reinit | fw_activate } ] [limit_level no_reset]
+
+    # Run reload command for devlink driver entities re-initialization:
+    $ devlink dev reload pci/0000:82:00.0 action driver_reinit
+    reload_actions_performed:
+      driver_reinit
+
+    # Run reload command to activate firmware:
+    # Note that mlx5 driver reloads the driver while activating firmware
+    $ devlink dev reload pci/0000:82:00.0 action fw_activate
+    reload_actions_performed:
+      driver_reinit fw_activate
diff --git a/Documentation/networking/devlink/index.rst b/Documentation/networking/devlink/index.rst
index 7684ae5c4a4a..d82874760ae2 100644
--- a/Documentation/networking/devlink/index.rst
+++ b/Documentation/networking/devlink/index.rst
@@ -20,6 +20,7 @@ general.
    devlink-params
    devlink-region
    devlink-resource
+   devlink-reload
    devlink-trap
 
 Driver-specific documentation
-- 
2.17.1

Re: [PATCH net-next RFC v5 00/15] Add devlink reload action and limit level options

[Moshe Shemesh]

I see no more comments, we reached consensus on API.

I will finalize my work and re-send as a feature to net-next.


Thanks,

Moshe.

On 9/18/2020 7:06 PM, Moshe Shemesh wrote:
> Introduce new options on devlink reload API to enable the user to select
> the reload action required and contrains limits on these actions that he
> may want to ensure. Complete support for reload actions in mlx5.
> The following reload actions are supported:
>    driver_reinit: driver entities re-initialization, applying devlink-param
>                   and devlink-resource values.
>    fw_activate: firmware activate.
>
> The uAPI is backward compatible, if the reload action option is omitted
> from the reload command, the driver reinit action will be used.
> Note that when required to do firmware activation some drivers may need
> to reload the driver. On the other hand some drivers may need to reset
> the firmware to reinitialize the driver entities. Therefore, the devlink
> reload command returns the actions which were actually performed.
>
> By default reload actions are not limited and driver implementation may
> include reset or downtime as needed to perform the actions.
> However, if limit_level is selected, the driver should perform only if
> it can do it while keeping the limit level constrains.
> Reload action limit level added:
>    no_reset: No reset allowed, no down time allowed, no link flap and no
>              configuration is lost.
>
> Each driver which supports devlink reload command should expose the
> reload actions and limit levels supported.
>
> Add reload action stats to hold the history per reload action per limit
> level. For example, the number of times fw_activate has been done on
> this device since the driver module was added or if the firmware
> activation was done with or without reset.
>
> Patch 1-2 add the new API reload action and reload action limit level
>            option to devlink reload.
> Patch 3 adds reload actions stats.
> Patch 4 exposes the reload actions stats on devlink dev get.
> Patches 5-10 add support on mlx5 for devlink reload action fw_activate
>              and handle the firmware reset events.
> Patches 11-12 add devlink enable remote dev reset parameter and use it
>               in mlx5.
> Patches 13-14 mlx5 add devlink reload action limit level no_reset
>                support for fw_activate reload action.
> Patch 15 adds documentation file devlink-reload.rst
>
>
> Moshe Shemesh (15):
>    devlink: Add reload action option to devlink reload command
>    devlink: Add reload action limit level
>    devlink: Add reload action stats
>    devlink: Add reload actions stats to dev get
>    net/mlx5: Add functions to set/query MFRL register
>    net/mlx5: Set cap for pci sync for fw update event
>    net/mlx5: Handle sync reset request event
>    net/mlx5: Handle sync reset now event
>    net/mlx5: Handle sync reset abort event
>    net/mlx5: Add support for devlink reload action fw activate
>    devlink: Add enable_remote_dev_reset generic parameter
>    net/mlx5: Add devlink param enable_remote_dev_reset support
>    net/mlx5: Add support for fw live patch event
>    net/mlx5: Add support for devlink reload action limit level no reset
>    devlink: Add Documentation/networking/devlink/devlink-reload.rst
>
>   .../networking/devlink/devlink-params.rst     |   6 +
>   .../networking/devlink/devlink-reload.rst     |  79 +++
>   Documentation/networking/devlink/index.rst    |   1 +
>   drivers/net/ethernet/mellanox/mlx4/main.c     |  16 +-
>   .../net/ethernet/mellanox/mlx5/core/Makefile  |   2 +-
>   .../net/ethernet/mellanox/mlx5/core/devlink.c | 122 ++++-
>   .../mellanox/mlx5/core/diag/fw_tracer.c       |  31 ++
>   .../mellanox/mlx5/core/diag/fw_tracer.h       |   1 +
>   .../ethernet/mellanox/mlx5/core/fw_reset.c    | 454 ++++++++++++++++++
>   .../ethernet/mellanox/mlx5/core/fw_reset.h    |  19 +
>   .../net/ethernet/mellanox/mlx5/core/health.c  |  35 +-
>   .../net/ethernet/mellanox/mlx5/core/main.c    |  13 +
>   .../ethernet/mellanox/mlx5/core/mlx5_core.h   |   2 +
>   drivers/net/ethernet/mellanox/mlxsw/core.c    |  27 +-
>   drivers/net/netdevsim/dev.c                   |  17 +-
>   include/linux/mlx5/device.h                   |   1 +
>   include/linux/mlx5/driver.h                   |   4 +
>   include/net/devlink.h                         |  21 +-
>   include/uapi/linux/devlink.h                  |  41 ++
>   net/core/devlink.c                            | 339 ++++++++++++-
>   20 files changed, 1179 insertions(+), 52 deletions(-)
>   create mode 100644 Documentation/networking/devlink/devlink-reload.rst
>   create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.c
>   create mode 100644 drivers/net/ethernet/mellanox/mlx5/core/fw_reset.h
>

Re: [PATCH net-next RFC v5 01/15] devlink: Add reload action option to devlink reload command

[Jakub Kicinski]

On Fri, 18 Sep 2020 19:06:37 +0300 Moshe Shemesh wrote:
> Add devlink reload action to allow the user to request a specific reload
> action. The action parameter is optional, if not specified then devlink
> driver re-init action is used (backward compatible).
> Note that when required to do firmware activation some drivers may need
> to reload the driver. On the other hand some drivers may need to reset
> the firmware to reinitialize the driver entities. Therefore, the devlink
> reload command returns the actions which were actually performed.
> Reload actions supported are:
> driver_reinit: driver entities re-initialization, applying devlink-param
>                and devlink-resource values.
> fw_activate: firmware activate.
> 
> command examples:
> $devlink dev reload pci/0000:82:00.0 action driver_reinit
> reload_actions_performed:
>   driver_reinit
> 
> $devlink dev reload pci/0000:82:00.0 action fw_activate
> reload_actions_performed:
>   driver_reinit fw_activate
> 
> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>

> @@ -3971,15 +3972,19 @@ static int mlx4_devlink_reload_up(struct devlink *devlink,
>  	int err;
>  
>  	err = mlx4_restart_one_up(persist->pdev, true, devlink);
> -	if (err)
> +	if (err) {
>  		mlx4_err(persist->dev, "mlx4_restart_one_up failed, ret=%d\n",
>  			 err);
> +		return err;
> +	}
> +	*actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);

FWIW I think drivers should be able to assign this even if they return
an error. On error there is no certainty what actions were actually
performed (e.g. when timeout happened but the device did the reset a
little later) so this argument should not be interpreted in presence of
errors, anyway.

Also consider providing a second enum for the BIT(xyz)s.

> -static bool devlink_reload_supported(const struct devlink *devlink)
> +static bool devlink_reload_supported(const struct devlink_ops *ops)
>  {
> -	return devlink->ops->reload_down && devlink->ops->reload_up;
> +	return ops->reload_down && ops->reload_up;
>  }

Please make the change to devlink_reload_supported() a separate patch.

> -
> +

What is this white space funk? 🤔

>  static void devlink_reload_failed_set(struct devlink *devlink,
>  				      bool reload_failed)
>  {
> @@ -2969,32 +2975,79 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
>  EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
>  
>  static int devlink_reload(struct devlink *devlink, struct net *dest_net,
> -			  struct netlink_ext_ack *extack)
> +			  enum devlink_reload_action action, struct netlink_ext_ack *extack,
> +			  unsigned long *actions_performed)
>  {
>  	int err;
>  
>  	if (!devlink->reload_enabled)
>  		return -EOPNOTSUPP;
>  
> -	err = devlink->ops->reload_down(devlink, !!dest_net, extack);
> +	err = devlink->ops->reload_down(devlink, !!dest_net, action, extack);
>  	if (err)
>  		return err;
>  
>  	if (dest_net && !net_eq(dest_net, devlink_net(devlink)))
>  		devlink_reload_netns_change(devlink, dest_net);
>  
> -	err = devlink->ops->reload_up(devlink, extack);
> +	err = devlink->ops->reload_up(devlink, action, extack, actions_performed);
>  	devlink_reload_failed_set(devlink, !!err);
> -	return err;
> +	if (err)
> +		return err;
> +
> +	WARN_ON(!test_bit(action, actions_performed));
> +	return 0;
> +}
> +
> +static int
> +devlink_nl_reload_actions_performed_fill(struct sk_buff *msg,
> +					 struct devlink *devlink,
> +					 unsigned long actions_performed,
> +					 enum devlink_command cmd, u32 portid,
> +					 u32 seq, int flags)
> +{
> +	struct nlattr *actions_performed_attr;
> +	void *hdr;
> +	int i;
> +
> +	hdr = genlmsg_put(msg, portid, seq, &devlink_nl_family, flags, cmd);
> +	if (!hdr)
> +		return -EMSGSIZE;
> +
> +	if (devlink_nl_put_handle(msg, devlink))
> +		goto genlmsg_cancel;
> +
> +	actions_performed_attr = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED);
> +	if (!actions_performed_attr)
> +		goto genlmsg_cancel;
> +
> +	for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
> +		if (!test_bit(i, &actions_performed))
> +			continue;
> +		if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION, i))
> +			goto actions_performed_nest_cancel;

Why not just return a mask? You need a special attribute for the nest,
anyway..

User space would probably actually prefer to have a single attr than an
iteration over a nest...

> +	}
> +	nla_nest_end(msg, actions_performed_attr);
> +	genlmsg_end(msg, hdr);
> +	return 0;
> +
> +actions_performed_nest_cancel:
> +	nla_nest_cancel(msg, actions_performed_attr);
> +genlmsg_cancel:
> +	genlmsg_cancel(msg, hdr);
> +	return -EMSGSIZE;
>  }
>  
>  static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>  {
>  	struct devlink *devlink = info->user_ptr[0];
> +	enum devlink_reload_action action;
> +	unsigned long actions_performed;
>  	struct net *dest_net = NULL;
> +	struct sk_buff *msg;
>  	int err;
>  
> -	if (!devlink_reload_supported(devlink))
> +	if (!devlink_reload_supported(devlink->ops))
>  		return -EOPNOTSUPP;
>  
>  	err = devlink_resources_validate(devlink, NULL, info);
> @@ -3011,12 +3064,43 @@ static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>  			return PTR_ERR(dest_net);
>  	}
>  
> -	err = devlink_reload(devlink, dest_net, info->extack);
> +	if (info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
> +		action = nla_get_u8(info->attrs[DEVLINK_ATTR_RELOAD_ACTION]);
> +	else
> +		action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT;
> +
> +	if (action == DEVLINK_RELOAD_ACTION_UNSPEC) {
> +		NL_SET_ERR_MSG_MOD(info->extack, "Invalid reload action");
> +		return -EINVAL;
> +	} else if (!devlink_reload_action_is_supported(devlink, action)) {
> +		NL_SET_ERR_MSG_MOD(info->extack, "Requested reload action is not supported by the driver");
> +		return -EOPNOTSUPP;
> +	}
> +
> +	err = devlink_reload(devlink, dest_net, action, info->extack, &actions_performed);

Perhaps we can pass the requested action to the driver via
actions_performed already, and then all the drivers which 
only do what they're asked to don't have to touch it?

>  	if (dest_net)
>  		put_net(dest_net);
>  
> -	return err;
> +	if (err)
> +		return err;
> +	/* For backward compatibility generate reply only if attributes used by user */
> +	if (!info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
> +		return 0;
> +
> +	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
> +	if (!msg)
> +		return -ENOMEM;
> +
> +	err = devlink_nl_reload_actions_performed_fill(msg, devlink, actions_performed,
> +						       DEVLINK_CMD_RELOAD, info->snd_portid,
> +						       info->snd_seq, 0);
> +	if (err) {
> +		nlmsg_free(msg);
> +		return err;
> +	}
> +
> +	return genlmsg_reply(msg, info);

Are you using devlink_nl_reload_actions_performed_fill() somewhere else?
I'd move the nlmsg_new() / genlmsg_reply() into the helper.

>  }
>  
>  static int devlink_nl_flash_update_fill(struct sk_buff *msg,
> @@ -7069,6 +7153,7 @@ static const struct nla_policy devlink_nl_policy[DEVLINK_ATTR_MAX + 1] = {
>  	[DEVLINK_ATTR_TRAP_POLICER_RATE] = { .type = NLA_U64 },
>  	[DEVLINK_ATTR_TRAP_POLICER_BURST] = { .type = NLA_U64 },
>  	[DEVLINK_ATTR_PORT_FUNCTION] = { .type = NLA_NESTED },
> +	[DEVLINK_ATTR_RELOAD_ACTION] = { .type = NLA_U8 },

Why not just range validation here?

>  };
>  
>  static const struct genl_ops devlink_nl_ops[] = {
> @@ -7402,6 +7487,20 @@ static struct genl_family devlink_nl_family __ro_after_init = {
>  	.n_mcgrps	= ARRAY_SIZE(devlink_nl_mcgrps),
>  };
>  
> +static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
> +{
> +	if (!devlink_reload_supported(ops)) {
> +		if (WARN_ON(ops->supported_reload_actions))
> +			return false;
> +		return true;
> +	}
> +
> +	if (WARN_ON(ops->supported_reload_actions >= BIT(__DEVLINK_RELOAD_ACTION_MAX) ||
> +		    ops->supported_reload_actions <= BIT(DEVLINK_RELOAD_ACTION_UNSPEC)))

This won't protect you from ACTION_UNSPEC being set..

WARN_ON(ops->supported_reload_actions & ~GENMASK(...))

> +		return false;
> +	return true;
> +}

Re: [PATCH net-next RFC v5 02/15] devlink: Add reload action limit level

[Jakub Kicinski]

On Fri, 18 Sep 2020 19:06:38 +0300 Moshe Shemesh wrote:
> Add reload action limit level to demand restrictions on actions.
> Reload action limit levels supported:
> none (default): No constrains on actions. Driver implementation may
>                 include reset or downtime as needed to perform the
>                 actions.
> no_reset: No reset allowed, no down time allowed, no link flap and no
>           configuration is lost.
> 
> Some combinations of action and limit level are invalid. For example,
> driver can not reinitialize its entities without any downtime.
> 
> The no_reset limit level will have usecase in this patchset to
> implement restricted fw_activate on mlx5.
> 
> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
> ---
> v4 -> v5:
> - Remove check DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX
> - Added list of invalid action-limit_level combinations and add check to
>   supported actions and levels and check user request
> v3 -> v4:
> - New patch
> ---
>  drivers/net/ethernet/mellanox/mlx4/main.c     |  3 +
>  .../net/ethernet/mellanox/mlx5/core/devlink.c |  3 +
>  drivers/net/ethernet/mellanox/mlxsw/core.c    |  3 +
>  drivers/net/netdevsim/dev.c                   |  6 +-
>  include/net/devlink.h                         |  6 +-
>  include/uapi/linux/devlink.h                  | 17 +++++
>  net/core/devlink.c                            | 76 +++++++++++++++++--
>  7 files changed, 107 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
> index 1a482120cc0a..f0ef295af477 100644
> --- a/drivers/net/ethernet/mellanox/mlx4/main.c
> +++ b/drivers/net/ethernet/mellanox/mlx4/main.c
> @@ -3947,6 +3947,7 @@ static int mlx4_restart_one_up(struct pci_dev *pdev, bool reload,
>  
>  static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
>  				    enum devlink_reload_action action,
> +				    enum devlink_reload_action_limit_level limit_level,
>  				    struct netlink_ext_ack *extack)
>  {
>  	struct mlx4_priv *priv = devlink_priv(devlink);
> @@ -3964,6 +3965,7 @@ static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
>  }
>  
>  static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
> +				  enum devlink_reload_action_limit_level limit_level,
>  				  struct netlink_ext_ack *extack, unsigned long *actions_performed)
>  {
>  	struct mlx4_priv *priv = devlink_priv(devlink);
> @@ -3985,6 +3987,7 @@ static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_a
>  static const struct devlink_ops mlx4_devlink_ops = {
>  	.port_type_set	= mlx4_devlink_port_type_set,
>  	.supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
> +	.supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),

Please cut down the name lenghts, this is just lazy.

'supported_reload_limits' or 'cap_reload_limits' is perfectly
sufficient.

'reload_limits' would be even better. Cause what else would it be if
not a capability.

Besides I don't think drivers should have to fill negative attributes
(that they don't support something). Everyone is always going to
support NONE, since it's "unspecified" / "pick your favorite", right?

>  	.reload_down	= mlx4_devlink_reload_down,
>  	.reload_up	= mlx4_devlink_reload_up,
>  };

> diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
> index fdba7ab58a79..0c5d942dcbd5 100644
> --- a/include/uapi/linux/devlink.h
> +++ b/include/uapi/linux/devlink.h
> @@ -289,6 +289,22 @@ enum devlink_reload_action {
>  	DEVLINK_RELOAD_ACTION_MAX = __DEVLINK_RELOAD_ACTION_MAX - 1
>  };
>  
> +/**
> + * enum devlink_reload_action_limit_level - Reload action limit level.
> + * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE: No constrains on action. Action may include
> + *                                          reset or downtime as needed.
> + * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET: No reset allowed, no down time allowed,
> + *                                              no link flap and no configuration is lost.
> + */
> +enum devlink_reload_action_limit_level {

You reserved UNSPEC for actions but not for limit level?

> +	DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
> +	DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
> +
> +	/* Add new reload actions limit level above */
> +	__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX,
> +	DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX = __DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX - 1
> +};
> +
>  enum devlink_attr {
>  	/* don't change the order or add anything between, this is ABI! */
>  	DEVLINK_ATTR_UNSPEC,
> @@ -480,6 +496,7 @@ enum devlink_attr {
>  
>  	DEVLINK_ATTR_RELOAD_ACTION,		/* u8 */
>  	DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,	/* nested */
> +	DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL,	/* u8 */
>  
>  	/* add new attributes above here, update the policy in devlink.c */
>  
> diff --git a/net/core/devlink.c b/net/core/devlink.c
> index 318ef29f81f2..fee6fcc7dead 100644
> --- a/net/core/devlink.c
> +++ b/net/core/devlink.c
> @@ -462,12 +462,45 @@ static int devlink_nl_put_handle(struct sk_buff *msg, struct devlink *devlink)
>  	return 0;
>  }
>  
> +struct devlink_reload_combination {
> +	enum devlink_reload_action action;
> +	enum devlink_reload_action_limit_level limit_level;
> +};
> +
> +static const struct devlink_reload_combination devlink_reload_invalid_combinations[] = {
> +	{
> +		/* can't reinitialize driver with no down time */
> +		.action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
> +		.limit_level = DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
> +	},
> +};
> +
> +static bool
> +devlink_reload_combination_is_invalid(enum devlink_reload_action action,
> +				      enum devlink_reload_action_limit_level limit_level)
> +{
> +	int i;
> +
> +	for (i = 0 ; i <  ARRAY_SIZE(devlink_reload_invalid_combinations) ; i++)

Whitespace. Did you checkpatch?

> +		if (devlink_reload_invalid_combinations[i].action == action &&
> +		    devlink_reload_invalid_combinations[i].limit_level == limit_level)
> +			return true;
> +	return false;
> +}
> +
>  static bool
>  devlink_reload_action_is_supported(struct devlink *devlink, enum devlink_reload_action action)
>  {
>  	return test_bit(action, &devlink->ops->supported_reload_actions);
>  }
>  
> +static bool
> +devlink_reload_action_limit_level_is_supported(struct devlink *devlink,
> +					       enum devlink_reload_action_limit_level limit_level)
> +{
> +	return test_bit(limit_level, &devlink->ops->supported_reload_action_limit_levels);
> +}

This single-use helper just grows LoC and muddies the code IMHO.

>  static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
>  			   enum devlink_command cmd, u32 portid,
>  			   u32 seq, int flags)
> @@ -2975,22 +3008,23 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
>  EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
>  
>  static int devlink_reload(struct devlink *devlink, struct net *dest_net,
> -			  enum devlink_reload_action action, struct netlink_ext_ack *extack,
> -			  unsigned long *actions_performed)
> +			  enum devlink_reload_action action,
> +			  enum devlink_reload_action_limit_level limit_level,
> +			  struct netlink_ext_ack *extack, unsigned long *actions_performed)
>  {
>  	int err;
>  
>  	if (!devlink->reload_enabled)
>  		return -EOPNOTSUPP;
>  
> -	err = devlink->ops->reload_down(devlink, !!dest_net, action, extack);
> +	err = devlink->ops->reload_down(devlink, !!dest_net, action, limit_level, extack);
>  	if (err)
>  		return err;
>  
>  	if (dest_net && !net_eq(dest_net, devlink_net(devlink)))
>  		devlink_reload_netns_change(devlink, dest_net);
>  
> -	err = devlink->ops->reload_up(devlink, action, extack, actions_performed);
> +	err = devlink->ops->reload_up(devlink, action, limit_level, extack, actions_performed);
>  	devlink_reload_failed_set(devlink, !!err);
>  	if (err)
>  		return err;
> @@ -3040,6 +3074,7 @@ devlink_nl_reload_actions_performed_fill(struct sk_buff *msg,
>  
>  static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>  {
> +	enum devlink_reload_action_limit_level limit_level;
>  	struct devlink *devlink = info->user_ptr[0];
>  	enum devlink_reload_action action;
>  	unsigned long actions_performed;
> @@ -3077,7 +3112,21 @@ static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>  		return -EOPNOTSUPP;
>  	}
>  
> -	err = devlink_reload(devlink, dest_net, action, info->extack, &actions_performed);
> +	if (info->attrs[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL])
> +		limit_level = nla_get_u8(info->attrs[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL]);
> +	else
> +		limit_level = DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE;
> +
> +	if (!devlink_reload_action_limit_level_is_supported(devlink, limit_level)) {
> +		NL_SET_ERR_MSG_MOD(info->extack, "Requested limit level is not supported by the driver");
> +		return -EOPNOTSUPP;
> +	}
> +	if (devlink_reload_combination_is_invalid(action, limit_level)) {
> +		NL_SET_ERR_MSG_MOD(info->extack, "Requested limit level is invalid for this action");
> +		return -EINVAL;
> +	}
> +	err = devlink_reload(devlink, dest_net, action, limit_level, info->extack,
> +			     &actions_performed);
>  
>  	if (dest_net)
>  		put_net(dest_net);
> @@ -7154,6 +7203,7 @@ static const struct nla_policy devlink_nl_policy[DEVLINK_ATTR_MAX + 1] = {
>  	[DEVLINK_ATTR_TRAP_POLICER_BURST] = { .type = NLA_U64 },
>  	[DEVLINK_ATTR_PORT_FUNCTION] = { .type = NLA_NESTED },
>  	[DEVLINK_ATTR_RELOAD_ACTION] = { .type = NLA_U8 },
> +	[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL] = { .type = NLA_U8 },

range

>  };
>  
>  static const struct genl_ops devlink_nl_ops[] = {
> @@ -7489,6 +7539,9 @@ static struct genl_family devlink_nl_family __ro_after_init = {
>  
>  static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
>  {
> +	const struct devlink_reload_combination *comb;
> +	int i;
> +
>  	if (!devlink_reload_supported(ops)) {
>  		if (WARN_ON(ops->supported_reload_actions))
>  			return false;
> @@ -7498,6 +7551,18 @@ static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
>  	if (WARN_ON(ops->supported_reload_actions >= BIT(__DEVLINK_RELOAD_ACTION_MAX) ||
>  		    ops->supported_reload_actions <= BIT(DEVLINK_RELOAD_ACTION_UNSPEC)))
>  		return false;
> +
> +	if (WARN_ON(!ops->supported_reload_action_limit_levels ||
> +		    ops->supported_reload_action_limit_levels >=
> +		    BIT(__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX)))
> +		return false;
> +
> +	for (i = 0; i <  ARRAY_SIZE(devlink_reload_invalid_combinations); i++)  {

White space, again.

> +		comb = &devlink_reload_invalid_combinations[i];
> +		if (ops->supported_reload_actions == BIT(comb->action) &&
> +		    ops->supported_reload_action_limit_levels == BIT(comb->limit_level))
> +			return false;
> +	}
>  	return true;
>  }
>  
> @@ -9793,6 +9858,7 @@ static void __net_exit devlink_pernet_pre_exit(struct net *net)
>  				continue;
>  			err = devlink_reload(devlink, &init_net,
>  					     DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
> +					     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
>  					     NULL, &actions_performed);
>  			if (err && err != -EOPNOTSUPP)
>  				pr_warn("Failed to reload devlink instance into init_net\n");

Re: [PATCH net-next RFC v5 03/15] devlink: Add reload action stats

[Jakub Kicinski]

On Fri, 18 Sep 2020 19:06:39 +0300 Moshe Shemesh wrote:
> Add reload action stats to hold the history per reload action type and
> limit level.
> 
> For example, the number of times fw_activate has been performed on this
> device since the driver module was added or if the firmware activation
> was performed with or without reset.
> 
> The function devlink_remote_reload_actions_performed() is exported to
> enable also drivers update on reload actions performed, for example in
> case firmware activation with reset finished successfully but was
> initiated by remote host.
> 
> Add devlink notification on stats update.
> 
> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>

I'd split this and the next patch by stat type (remote vs local).
Rather than split them by collect / report.

> diff --git a/include/net/devlink.h b/include/net/devlink.h
> index d8c62d605381..f09f55a47d09 100644
> --- a/include/net/devlink.h
> +++ b/include/net/devlink.h
> @@ -20,6 +20,9 @@
>  #include <uapi/linux/devlink.h>
>  #include <linux/xarray.h>
>  
> +#define DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE \
> +	(__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX * __DEVLINK_RELOAD_ACTION_MAX)
> +
>  struct devlink_ops;
>  
>  struct devlink {
> @@ -38,6 +41,8 @@ struct devlink {
>  	struct list_head trap_policer_list;
>  	const struct devlink_ops *ops;
>  	struct xarray snapshot_ids;
> +	u32 reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
> +	u32 remote_reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
>  	struct device *dev;
>  	possible_net_t _net;
>  	struct mutex lock; /* Serializes access to devlink instance specific objects such as
> @@ -1400,6 +1405,9 @@ void
>  devlink_health_reporter_recovery_done(struct devlink_health_reporter *reporter);
>  
>  bool devlink_is_reload_failed(const struct devlink *devlink);
> +void devlink_remote_reload_actions_performed(struct devlink *devlink,
> +					     enum devlink_reload_action_limit_level limit_level,
> +					     unsigned long actions_performed);
>  
>  void devlink_flash_update_begin_notify(struct devlink *devlink);
>  void devlink_flash_update_end_notify(struct devlink *devlink);
> diff --git a/net/core/devlink.c b/net/core/devlink.c
> index fee6fcc7dead..1509c2ffb98b 100644
> --- a/net/core/devlink.c
> +++ b/net/core/devlink.c
> @@ -3007,16 +3007,74 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
>  }
>  EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
>  
> +static void
> +__devlink_reload_action_stats_update(struct devlink *devlink,
> +				     u32 *reload_action_stats,
> +				     enum devlink_reload_action_limit_level limit_level,
> +				     unsigned long actions_performed)
> +{
> +	int stat_idx;
> +	int action;
> +
> +	if (!actions_performed)
> +		return;
> +
> +	if (WARN_ON(limit_level > DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX))
> +		return;
> +	for (action = 0; action <= DEVLINK_RELOAD_ACTION_MAX; action++) {
> +		if (!test_bit(action, &actions_performed))
> +			continue;
> +		stat_idx = limit_level * __DEVLINK_RELOAD_ACTION_MAX + action;
> +		reload_action_stats[stat_idx]++;
> +	}
> +	devlink_notify(devlink, DEVLINK_CMD_NEW);
> +}
> +
> +static void
> +devlink_reload_action_stats_update(struct devlink *devlink,
> +				   enum devlink_reload_action_limit_level limit_level,
> +				   unsigned long actions_performed)
> +{
> +	__devlink_reload_action_stats_update(devlink, devlink->reload_action_stats,
> +					     limit_level, actions_performed);
> +}
> +
> +/**
> + *	devlink_remote_reload_actions_performed - Update devlink on reload actions
> + *	  performed which are not a direct result of devlink reload call.
> + *
> + *	This should be called by a driver after performing reload actions in case it was not
> + *	a result of devlink reload call. For example fw_activate was performed as a result
> + *	of devlink reload triggered fw_activate on another host.
> + *	The motivation for this function is to keep data on reload actions performed on this
> + *	function whether it was done due to direct devlink reload call or not.
> + *
> + *	@devlink: devlink
> + *	@limit_level: reload action limit level
> + *	@actions_performed: bitmask of actions performed
> + */
> +void devlink_remote_reload_actions_performed(struct devlink *devlink,
> +					     enum devlink_reload_action_limit_level limit_level,
> +					     unsigned long actions_performed)
> +{
> +	__devlink_reload_action_stats_update(devlink, devlink->remote_reload_action_stats,
> +					     limit_level, actions_performed);
> +}
> +EXPORT_SYMBOL_GPL(devlink_remote_reload_actions_performed);
> +
>  static int devlink_reload(struct devlink *devlink, struct net *dest_net,
>  			  enum devlink_reload_action action,
>  			  enum devlink_reload_action_limit_level limit_level,
>  			  struct netlink_ext_ack *extack, unsigned long *actions_performed)
>  {
> +	u32 remote_reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
>  	int err;
>  
>  	if (!devlink->reload_enabled)
>  		return -EOPNOTSUPP;
>  
> +	memcpy(remote_reload_action_stats, devlink->remote_reload_action_stats,
> +	       sizeof(remote_reload_action_stats));
>  	err = devlink->ops->reload_down(devlink, !!dest_net, action, limit_level, extack);
>  	if (err)
>  		return err;
> @@ -3030,6 +3088,10 @@ static int devlink_reload(struct devlink *devlink, struct net *dest_net,
>  		return err;
>  
>  	WARN_ON(!test_bit(action, actions_performed));
> +	/* protect from driver updating the remote action within devlink reload */

s/protect from/catch/

> +	WARN_ON(memcmp(remote_reload_action_stats, devlink->remote_reload_action_stats,
> +		       sizeof(remote_reload_action_stats)));
> +	devlink_reload_action_stats_update(devlink, limit_level, *actions_performed);
>  	return 0;
>  }
>  

Re: [PATCH net-next RFC v5 04/15] devlink: Add reload actions stats to dev get

[Jakub Kicinski]

On Fri, 18 Sep 2020 19:06:40 +0300 Moshe Shemesh wrote:
> Expose devlink reload actions stats to the user through devlink dev
> get command.
> 
> Examples:
> $ devlink dev show
> pci/0000:82:00.0:
>   stats:
>       reload_action_stats:
>         driver_reinit 2
>         fw_activate 1
>         fw_activate_no_reset 0
>         remote_driver_reinit 0
>         remote_fw_activate 0
>         remote_fw_activate_no_reset 0
> pci/0000:82:00.1:
>   stats:
>       reload_action_stats:
>         driver_reinit 0
>         fw_activate 0
>         fw_activate_no_reset 0
>         remote_driver_reinit 1
>         remote_fw_activate 1
>         remote_fw_activate_no_reset 0
> 
> $ devlink dev show -jp
> {
>     "dev": {
>         "pci/0000:82:00.0": {
>             "stats": {
>                 "reload_action_stats": [ {
>                         "driver_reinit": 2
>                     },{
>                         "fw_activate": 1
>                     },{
>                         "fw_activate_no_reset": 0
>                     },{
>                         "remote_driver_reinit": 0
>                     },{
>                         "remote_fw_activate": 0
>                     },{
>                         "remote_fw_activate_no_reset": 0
>                     } ]
>             }
>         },
>         "pci/0000:82:00.1": {
>             "stats": {
>                 "reload_action_stats": [ {
>                         "driver_reinit": 0
>                     },{
>                         "fw_activate": 0
>                     },{
>                         "fw_activate_no_reset": 0
>                     },{
>                         "remote_driver_reinit": 1
>                     },{
>                         "remote_fw_activate": 1
>                     },{
>                         "remote_fw_activate_no_reset": 0
>                     } ]
>             }
>         }
>     }
> }
> 
> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
> ---
> v4 -> v5:
> - Add remote actions stats
> - If devlink reload is not supported, show only remote_stats
> v3 -> v4:
> - Renamed DEVLINK_ATTR_RELOAD_ACTION_CNT to
>   DEVLINK_ATTR_RELOAD_ACTION_STAT
> - Add stats per action per limit level
> v2 -> v3:
> - Add reload actions counters instead of supported reload actions
>   (reload actions counters are only for supported action so no need for
>    both)
> v1 -> v2:
> - Removed DEVLINK_ATTR_RELOAD_DEFAULT_LEVEL
> - Removed DEVLINK_ATTR_RELOAD_LEVELS_INFO
> - Have actions instead of levels
> ---
>  include/uapi/linux/devlink.h |  5 +++
>  net/core/devlink.c           | 70 ++++++++++++++++++++++++++++++++++++
>  2 files changed, 75 insertions(+)
> 
> diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
> index 0c5d942dcbd5..648d53be691e 100644
> --- a/include/uapi/linux/devlink.h
> +++ b/include/uapi/linux/devlink.h
> @@ -497,7 +497,12 @@ enum devlink_attr {
>  	DEVLINK_ATTR_RELOAD_ACTION,		/* u8 */
>  	DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,	/* nested */
>  	DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL,	/* u8 */
> +	DEVLINK_ATTR_RELOAD_ACTION_STATS,	/* nested */
> +	DEVLINK_ATTR_RELOAD_ACTION_STAT,	/* nested */
> +	DEVLINK_ATTR_RELOAD_ACTION_STAT_REMOTE,	/* flag */
> +	DEVLINK_ATTR_RELOAD_ACTION_STAT_VALUE,	/* u32 */
>  
> +	DEVLINK_ATTR_DEV_STATS,			/* nested */
>  	/* add new attributes above here, update the policy in devlink.c */

I'd propose this nesting:

	[DEV_STATS]
		[RELOAD_STATS]
			[DEV_STATS_ENTRY]
				[ACTION]
				[LIMIT]
				[VALUE]
			[DEV_STATS_ENTRY]
				[...]
		[REMOTE_RELOAD_STATS]
			[DEV_STATS_ENTRY]
				[ACTION]
				[LIMIT]
				[VALUE]
			[DEV_STATS_ENTRY]
				[...]

Then you can fill in the inside of the [REMOTE_]RELOAD_STATS nests with
a helper, and similarly user space can separate the two in JSON more
cleanly than string concat.

>  	__DEVLINK_ATTR_MAX,
> diff --git a/net/core/devlink.c b/net/core/devlink.c
> index 1509c2ffb98b..71aeda259e6a 100644
> --- a/net/core/devlink.c
> +++ b/net/core/devlink.c
> @@ -501,10 +501,39 @@ devlink_reload_action_limit_level_is_supported(struct devlink *devlink,
>  	return test_bit(limit_level, &devlink->ops->supported_reload_action_limit_levels);
>  }
>  
> +static int devlink_reload_action_stat_put(struct sk_buff *msg, enum devlink_reload_action action,
> +					  enum devlink_reload_action_limit_level limit_level,
> +					  bool is_remote, u32 value)
> +{
> +	struct nlattr *reload_action_stat;
> +
> +	reload_action_stat = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT);
> +	if (!reload_action_stat)
> +		return -EMSGSIZE;
> +
> +	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION, action))
> +		goto nla_put_failure;
> +	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL, limit_level))
> +		goto nla_put_failure;
> +	if (is_remote && nla_put_flag(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT_REMOTE))
> +		goto nla_put_failure;
> +	if (nla_put_u32(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT_VALUE, value))
> +		goto nla_put_failure;
> +	nla_nest_end(msg, reload_action_stat);
> +	return 0;
> +
> +nla_put_failure:
> +	nla_nest_cancel(msg, reload_action_stat);
> +	return -EMSGSIZE;
> +}
> +
>  static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
>  			   enum devlink_command cmd, u32 portid,
>  			   u32 seq, int flags)
>  {
> +	struct nlattr *dev_stats, *reload_action_stats;
> +	int i, j, stat_idx;
> +	u32 value;
>  	void *hdr;
>  
>  	hdr = genlmsg_put(msg, portid, seq, &devlink_nl_family, flags, cmd);
> @@ -516,9 +545,50 @@ static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
>  	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_FAILED, devlink->reload_failed))
>  		goto nla_put_failure;
>  
> +	dev_stats = nla_nest_start(msg, DEVLINK_ATTR_DEV_STATS);
> +	if (!dev_stats)
> +		goto nla_put_failure;
> +	reload_action_stats = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTION_STATS);
> +	if (!reload_action_stats)
> +		goto dev_stats_nest_cancel;
> +
> +	for (j = 0; j <= DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX; j++) {
> +		if (!devlink_reload_action_limit_level_is_supported(devlink, j))
> +			continue;
> +		for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
> +			if (!devlink_reload_action_is_supported(devlink, i) ||
> +			    devlink_reload_combination_is_invalid(i, j))
> +				continue;
> +
> +			stat_idx = j * __DEVLINK_RELOAD_ACTION_MAX + i;
> +			value = devlink->reload_action_stats[stat_idx];
> +			if (devlink_reload_action_stat_put(msg, i, j, false, value))
> +				goto reload_action_stats_nest_cancel;
> +		}
> +	}
> +
> +	for (j = 0; j <= DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX; j++) {
> +		for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
> +			if (i == DEVLINK_RELOAD_ACTION_UNSPEC ||
> +			    devlink_reload_combination_is_invalid(i, j))
> +				continue;
> +
> +			stat_idx = j * __DEVLINK_RELOAD_ACTION_MAX + i;
> +			value = devlink->remote_reload_action_stats[stat_idx];
> +			if (devlink_reload_action_stat_put(msg, i, j, true, value))
> +				goto reload_action_stats_nest_cancel;
> +		}
> +	}

This calls for a helper.

Re: [PATCH net-next RFC v5 01/15] devlink: Add reload action option to devlink reload command

[Moshe Shemesh]

On 9/23/2020 9:25 PM, Jakub Kicinski wrote:
>
> On Fri, 18 Sep 2020 19:06:37 +0300 Moshe Shemesh wrote:
>> Add devlink reload action to allow the user to request a specific reload
>> action. The action parameter is optional, if not specified then devlink
>> driver re-init action is used (backward compatible).
>> Note that when required to do firmware activation some drivers may need
>> to reload the driver. On the other hand some drivers may need to reset
>> the firmware to reinitialize the driver entities. Therefore, the devlink
>> reload command returns the actions which were actually performed.
>> Reload actions supported are:
>> driver_reinit: driver entities re-initialization, applying devlink-param
>>                 and devlink-resource values.
>> fw_activate: firmware activate.
>>
>> command examples:
>> $devlink dev reload pci/0000:82:00.0 action driver_reinit
>> reload_actions_performed:
>>    driver_reinit
>>
>> $devlink dev reload pci/0000:82:00.0 action fw_activate
>> reload_actions_performed:
>>    driver_reinit fw_activate
>>
>> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
>> @@ -3971,15 +3972,19 @@ static int mlx4_devlink_reload_up(struct devlink *devlink,
>>        int err;
>>
>>        err = mlx4_restart_one_up(persist->pdev, true, devlink);
>> -     if (err)
>> +     if (err) {
>>                mlx4_err(persist->dev, "mlx4_restart_one_up failed, ret=%d\n",
>>                         err);
>> +             return err;
>> +     }
>> +     *actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);
> FWIW I think drivers should be able to assign this even if they return
> an error. On error there is no certainty what actions were actually
> performed (e.g. when timeout happened but the device did the reset a
> little later) so this argument should not be interpreted in presence of
> errors, anyway.


Not sure I got it. Do you mean driver can assign it anyway and devlink 
should ignore in case of failure ?

As I implemented here devlink already ignores actions_performed in case 
driver returns with error.

> Also consider providing a second enum for the BIT(xyz)s.
OK.
>> -static bool devlink_reload_supported(const struct devlink *devlink)
>> +static bool devlink_reload_supported(const struct devlink_ops *ops)
>>   {
>> -     return devlink->ops->reload_down && devlink->ops->reload_up;
>> +     return ops->reload_down && ops->reload_up;
>>   }
> Please make the change to devlink_reload_supported() a separate patch.


Ack.

>> -
>> +
> What is this white space funk? 🤔


Missed that.

>>   static void devlink_reload_failed_set(struct devlink *devlink,
>>                                      bool reload_failed)
>>   {
>> @@ -2969,32 +2975,79 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
>>   EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
>>
>>   static int devlink_reload(struct devlink *devlink, struct net *dest_net,
>> -                       struct netlink_ext_ack *extack)
>> +                       enum devlink_reload_action action, struct netlink_ext_ack *extack,
>> +                       unsigned long *actions_performed)
>>   {
>>        int err;
>>
>>        if (!devlink->reload_enabled)
>>                return -EOPNOTSUPP;
>>
>> -     err = devlink->ops->reload_down(devlink, !!dest_net, extack);
>> +     err = devlink->ops->reload_down(devlink, !!dest_net, action, extack);
>>        if (err)
>>                return err;
>>
>>        if (dest_net && !net_eq(dest_net, devlink_net(devlink)))
>>                devlink_reload_netns_change(devlink, dest_net);
>>
>> -     err = devlink->ops->reload_up(devlink, extack);
>> +     err = devlink->ops->reload_up(devlink, action, extack, actions_performed);
>>        devlink_reload_failed_set(devlink, !!err);
>> -     return err;
>> +     if (err)
>> +             return err;
>> +
>> +     WARN_ON(!test_bit(action, actions_performed));
>> +     return 0;
>> +}
>> +
>> +static int
>> +devlink_nl_reload_actions_performed_fill(struct sk_buff *msg,
>> +                                      struct devlink *devlink,
>> +                                      unsigned long actions_performed,
>> +                                      enum devlink_command cmd, u32 portid,
>> +                                      u32 seq, int flags)
>> +{
>> +     struct nlattr *actions_performed_attr;
>> +     void *hdr;
>> +     int i;
>> +
>> +     hdr = genlmsg_put(msg, portid, seq, &devlink_nl_family, flags, cmd);
>> +     if (!hdr)
>> +             return -EMSGSIZE;
>> +
>> +     if (devlink_nl_put_handle(msg, devlink))
>> +             goto genlmsg_cancel;
>> +
>> +     actions_performed_attr = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED);
>> +     if (!actions_performed_attr)
>> +             goto genlmsg_cancel;
>> +
>> +     for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
>> +             if (!test_bit(i, &actions_performed))
>> +                     continue;
>> +             if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION, i))
>> +                     goto actions_performed_nest_cancel;
> Why not just return a mask? You need a special attribute for the nest,
> anyway..
>
> User space would probably actually prefer to have a single attr than an
> iteration over a nest...
OK.
>> +     }
>> +     nla_nest_end(msg, actions_performed_attr);
>> +     genlmsg_end(msg, hdr);
>> +     return 0;
>> +
>> +actions_performed_nest_cancel:
>> +     nla_nest_cancel(msg, actions_performed_attr);
>> +genlmsg_cancel:
>> +     genlmsg_cancel(msg, hdr);
>> +     return -EMSGSIZE;
>>   }
>>
>>   static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>>   {
>>        struct devlink *devlink = info->user_ptr[0];
>> +     enum devlink_reload_action action;
>> +     unsigned long actions_performed;
>>        struct net *dest_net = NULL;
>> +     struct sk_buff *msg;
>>        int err;
>>
>> -     if (!devlink_reload_supported(devlink))
>> +     if (!devlink_reload_supported(devlink->ops))
>>                return -EOPNOTSUPP;
>>
>>        err = devlink_resources_validate(devlink, NULL, info);
>> @@ -3011,12 +3064,43 @@ static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>>                        return PTR_ERR(dest_net);
>>        }
>>
>> -     err = devlink_reload(devlink, dest_net, info->extack);
>> +     if (info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
>> +             action = nla_get_u8(info->attrs[DEVLINK_ATTR_RELOAD_ACTION]);
>> +     else
>> +             action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT;
>> +
>> +     if (action == DEVLINK_RELOAD_ACTION_UNSPEC) {
>> +             NL_SET_ERR_MSG_MOD(info->extack, "Invalid reload action");
>> +             return -EINVAL;
>> +     } else if (!devlink_reload_action_is_supported(devlink, action)) {
>> +             NL_SET_ERR_MSG_MOD(info->extack, "Requested reload action is not supported by the driver");
>> +             return -EOPNOTSUPP;
>> +     }
>> +
>> +     err = devlink_reload(devlink, dest_net, action, info->extack, &actions_performed);
> Perhaps we can pass the requested action to the driver via
> actions_performed already, and then all the drivers which
> only do what they're asked to don't have to touch it?


Not sure about it. Note that in the next patch I add here limit_level 
and that has only input param, so I think it would be confusing.

>>        if (dest_net)
>>                put_net(dest_net);
>>
>> -     return err;
>> +     if (err)
>> +             return err;
>> +     /* For backward compatibility generate reply only if attributes used by user */
>> +     if (!info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
>> +             return 0;
>> +
>> +     msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
>> +     if (!msg)
>> +             return -ENOMEM;
>> +
>> +     err = devlink_nl_reload_actions_performed_fill(msg, devlink, actions_performed,
>> +                                                    DEVLINK_CMD_RELOAD, info->snd_portid,
>> +                                                    info->snd_seq, 0);
>> +     if (err) {
>> +             nlmsg_free(msg);
>> +             return err;
>> +     }
>> +
>> +     return genlmsg_reply(msg, info);
> Are you using devlink_nl_reload_actions_performed_fill() somewhere else?
No
> I'd move the nlmsg_new() / genlmsg_reply() into the helper.


Can do it, but there are many _fill() functions in devlink.c code to 
fill the data, none of them include nlmsg_new() and genlmsg_reply() 
that's always in the calling function, even if the calling function adds 
only that. So I guess I will leave it for consistency.

>>   }
>>
>>   static int devlink_nl_flash_update_fill(struct sk_buff *msg,
>> @@ -7069,6 +7153,7 @@ static const struct nla_policy devlink_nl_policy[DEVLINK_ATTR_MAX + 1] = {
>>        [DEVLINK_ATTR_TRAP_POLICER_RATE] = { .type = NLA_U64 },
>>        [DEVLINK_ATTR_TRAP_POLICER_BURST] = { .type = NLA_U64 },
>>        [DEVLINK_ATTR_PORT_FUNCTION] = { .type = NLA_NESTED },
>> +     [DEVLINK_ATTR_RELOAD_ACTION] = { .type = NLA_U8 },
> Why not just range validation here?


All devlink attributes that pass here go through devlink_nl_poicy this 
way, including other enums.

I think changing that should be in a different patch for all, not in 
this patchset.

>>   };
>>
>>   static const struct genl_ops devlink_nl_ops[] = {
>> @@ -7402,6 +7487,20 @@ static struct genl_family devlink_nl_family __ro_after_init = {
>>        .n_mcgrps       = ARRAY_SIZE(devlink_nl_mcgrps),
>>   };
>>
>> +static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
>> +{
>> +     if (!devlink_reload_supported(ops)) {
>> +             if (WARN_ON(ops->supported_reload_actions))
>> +                     return false;
>> +             return true;
>> +     }
>> +
>> +     if (WARN_ON(ops->supported_reload_actions >= BIT(__DEVLINK_RELOAD_ACTION_MAX) ||
>> +                 ops->supported_reload_actions <= BIT(DEVLINK_RELOAD_ACTION_UNSPEC)))
> This won't protect you from ACTION_UNSPEC being set..
>
> WARN_ON(ops->supported_reload_actions & ~GENMASK(...))


Right, I will fix.

>> +             return false;
>> +     return true;
>> +}

Re: [PATCH net-next RFC v5 02/15] devlink: Add reload action limit level

[Moshe Shemesh]

On 9/23/2020 9:36 PM, Jakub Kicinski wrote:
> On Fri, 18 Sep 2020 19:06:38 +0300 Moshe Shemesh wrote:
>> Add reload action limit level to demand restrictions on actions.
>> Reload action limit levels supported:
>> none (default): No constrains on actions. Driver implementation may
>>                  include reset or downtime as needed to perform the
>>                  actions.
>> no_reset: No reset allowed, no down time allowed, no link flap and no
>>            configuration is lost.
>>
>> Some combinations of action and limit level are invalid. For example,
>> driver can not reinitialize its entities without any downtime.
>>
>> The no_reset limit level will have usecase in this patchset to
>> implement restricted fw_activate on mlx5.
>>
>> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
>> ---
>> v4 -> v5:
>> - Remove check DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX
>> - Added list of invalid action-limit_level combinations and add check to
>>    supported actions and levels and check user request
>> v3 -> v4:
>> - New patch
>> ---
>>   drivers/net/ethernet/mellanox/mlx4/main.c     |  3 +
>>   .../net/ethernet/mellanox/mlx5/core/devlink.c |  3 +
>>   drivers/net/ethernet/mellanox/mlxsw/core.c    |  3 +
>>   drivers/net/netdevsim/dev.c                   |  6 +-
>>   include/net/devlink.h                         |  6 +-
>>   include/uapi/linux/devlink.h                  | 17 +++++
>>   net/core/devlink.c                            | 76 +++++++++++++++++--
>>   7 files changed, 107 insertions(+), 7 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
>> index 1a482120cc0a..f0ef295af477 100644
>> --- a/drivers/net/ethernet/mellanox/mlx4/main.c
>> +++ b/drivers/net/ethernet/mellanox/mlx4/main.c
>> @@ -3947,6 +3947,7 @@ static int mlx4_restart_one_up(struct pci_dev *pdev, bool reload,
>>
>>   static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
>>                                    enum devlink_reload_action action,
>> +                                 enum devlink_reload_action_limit_level limit_level,
>>                                    struct netlink_ext_ack *extack)
>>   {
>>        struct mlx4_priv *priv = devlink_priv(devlink);
>> @@ -3964,6 +3965,7 @@ static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
>>   }
>>
>>   static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
>> +                               enum devlink_reload_action_limit_level limit_level,
>>                                  struct netlink_ext_ack *extack, unsigned long *actions_performed)
>>   {
>>        struct mlx4_priv *priv = devlink_priv(devlink);
>> @@ -3985,6 +3987,7 @@ static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_a
>>   static const struct devlink_ops mlx4_devlink_ops = {
>>        .port_type_set  = mlx4_devlink_port_type_set,
>>        .supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
>> +     .supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),
> Please cut down the name lenghts, this is just lazy.
>
> 'supported_reload_limits' or 'cap_reload_limits' is perfectly
> sufficient.
>
> 'reload_limits' would be even better. Cause what else would it be if
> not a capability.


Sounds good.

So instead of supported_reload_actions_limit_levels will have reload_limits.

Instead of supported_reload_actions will have reload_actions, OK ?

May also use reload_limit_level instead of reload_action_limit_level 
everywhere if its clear enough.

> Besides I don't think drivers should have to fill negative attributes
> (that they don't support something). Everyone is always going to
> support NONE, since it's "unspecified" / "pick your favorite", right?


Good point, will remove it.

>>        .reload_down    = mlx4_devlink_reload_down,
>>        .reload_up      = mlx4_devlink_reload_up,
>>   };
>> diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
>> index fdba7ab58a79..0c5d942dcbd5 100644
>> --- a/include/uapi/linux/devlink.h
>> +++ b/include/uapi/linux/devlink.h
>> @@ -289,6 +289,22 @@ enum devlink_reload_action {
>>        DEVLINK_RELOAD_ACTION_MAX = __DEVLINK_RELOAD_ACTION_MAX - 1
>>   };
>>
>> +/**
>> + * enum devlink_reload_action_limit_level - Reload action limit level.
>> + * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE: No constrains on action. Action may include
>> + *                                          reset or downtime as needed.
>> + * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET: No reset allowed, no down time allowed,
>> + *                                              no link flap and no configuration is lost.
>> + */
>> +enum devlink_reload_action_limit_level {
> You reserved UNSPEC for actions but not for limit level?


Yes, I used LIMIT_LEVEL_NONE = 0 as no limit needed, so I skipped UNSPEC.

Maybe should add UNSPEC and use UNSPEC as no limit needed. But UNSPEC is 
kind of invalid.


>> +     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
>> +     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
>> +
>> +     /* Add new reload actions limit level above */
>> +     __DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX,
>> +     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX = __DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX - 1
>> +};
>> +
>>   enum devlink_attr {
>>        /* don't change the order or add anything between, this is ABI! */
>>        DEVLINK_ATTR_UNSPEC,
>> @@ -480,6 +496,7 @@ enum devlink_attr {
>>
>>        DEVLINK_ATTR_RELOAD_ACTION,             /* u8 */
>>        DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,  /* nested */
>> +     DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL, /* u8 */
>>
>>        /* add new attributes above here, update the policy in devlink.c */
>>
>> diff --git a/net/core/devlink.c b/net/core/devlink.c
>> index 318ef29f81f2..fee6fcc7dead 100644
>> --- a/net/core/devlink.c
>> +++ b/net/core/devlink.c
>> @@ -462,12 +462,45 @@ static int devlink_nl_put_handle(struct sk_buff *msg, struct devlink *devlink)
>>        return 0;
>>   }
>>
>> +struct devlink_reload_combination {
>> +     enum devlink_reload_action action;
>> +     enum devlink_reload_action_limit_level limit_level;
>> +};
>> +
>> +static const struct devlink_reload_combination devlink_reload_invalid_combinations[] = {
>> +     {
>> +             /* can't reinitialize driver with no down time */
>> +             .action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
>> +             .limit_level = DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
>> +     },
>> +};
>> +
>> +static bool
>> +devlink_reload_combination_is_invalid(enum devlink_reload_action action,
>> +                                   enum devlink_reload_action_limit_level limit_level)
>> +{
>> +     int i;
>> +
>> +     for (i = 0 ; i <  ARRAY_SIZE(devlink_reload_invalid_combinations) ; i++)
> Whitespace. Did you checkpatch?


Yes, checked it again now, it still pass. I think checkpatch doesn't see 
double space.

But anyway, I missed it, I will fix.

>> +             if (devlink_reload_invalid_combinations[i].action == action &&
>> +                 devlink_reload_invalid_combinations[i].limit_level == limit_level)
>> +                     return true;
>> +     return false;
>> +}
>> +
>>   static bool
>>   devlink_reload_action_is_supported(struct devlink *devlink, enum devlink_reload_action action)
>>   {
>>        return test_bit(action, &devlink->ops->supported_reload_actions);
>>   }
>>
>> +static bool
>> +devlink_reload_action_limit_level_is_supported(struct devlink *devlink,
>> +                                            enum devlink_reload_action_limit_level limit_level)
>> +{
>> +     return test_bit(limit_level, &devlink->ops->supported_reload_action_limit_levels);
>> +}
> This single-use helper just grows LoC and muddies the code IMHO.


It is used again in patch 4.

>
>>   static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
>>                           enum devlink_command cmd, u32 portid,
>>                           u32 seq, int flags)
>> @@ -2975,22 +3008,23 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
>>   EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
>>
>>   static int devlink_reload(struct devlink *devlink, struct net *dest_net,
>> -                       enum devlink_reload_action action, struct netlink_ext_ack *extack,
>> -                       unsigned long *actions_performed)
>> +                       enum devlink_reload_action action,
>> +                       enum devlink_reload_action_limit_level limit_level,
>> +                       struct netlink_ext_ack *extack, unsigned long *actions_performed)
>>   {
>>        int err;
>>
>>        if (!devlink->reload_enabled)
>>                return -EOPNOTSUPP;
>>
>> -     err = devlink->ops->reload_down(devlink, !!dest_net, action, extack);
>> +     err = devlink->ops->reload_down(devlink, !!dest_net, action, limit_level, extack);
>>        if (err)
>>                return err;
>>
>>        if (dest_net && !net_eq(dest_net, devlink_net(devlink)))
>>                devlink_reload_netns_change(devlink, dest_net);
>>
>> -     err = devlink->ops->reload_up(devlink, action, extack, actions_performed);
>> +     err = devlink->ops->reload_up(devlink, action, limit_level, extack, actions_performed);
>>        devlink_reload_failed_set(devlink, !!err);
>>        if (err)
>>                return err;
>> @@ -3040,6 +3074,7 @@ devlink_nl_reload_actions_performed_fill(struct sk_buff *msg,
>>
>>   static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>>   {
>> +     enum devlink_reload_action_limit_level limit_level;
>>        struct devlink *devlink = info->user_ptr[0];
>>        enum devlink_reload_action action;
>>        unsigned long actions_performed;
>> @@ -3077,7 +3112,21 @@ static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
>>                return -EOPNOTSUPP;
>>        }
>>
>> -     err = devlink_reload(devlink, dest_net, action, info->extack, &actions_performed);
>> +     if (info->attrs[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL])
>> +             limit_level = nla_get_u8(info->attrs[DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL]);
>> +     else
>> +             limit_level = DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE;
>> +
>> +     if (!devlink_reload_action_limit_level_is_supported(devlink, limit_level)) {
>> +             NL_SET_ERR_MSG_MOD(info->extack, "Requested limit level is not supported by the driver");
>> +             return -EOPNOTSUPP;
>> +     }
>> +     if (devlink_reload_combination_is_invalid(action, limit_level)) {
>> +             NL_SET_ERR_MSG_MOD(info->extack, "Requested limit level is invalid for this action");
>> +             return -EINVAL;
>> +     }
>> +     err = devlink_reload(devlink, dest_net, action, limit_level, info->extack,
>> +                          &actions_performed);
>>
>>        if (dest_net)
>>                put_net(dest_net);
>> @@ -7154,6 +7203,7 @@ static const struct nla_policy devlink_nl_policy[DEVLINK_ATTR_MAX + 1] = {
>>        [DEVLINK_ATTR_TRAP_POLICER_BURST] = { .type = NLA_U64 },
>>        [DEVLINK_ATTR_PORT_FUNCTION] = { .type = NLA_NESTED },
>>        [DEVLINK_ATTR_RELOAD_ACTION] = { .type = NLA_U8 },
>> +     [DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL] = { .type = NLA_U8 },
> range
>
>>   };
>>
>>   static const struct genl_ops devlink_nl_ops[] = {
>> @@ -7489,6 +7539,9 @@ static struct genl_family devlink_nl_family __ro_after_init = {
>>
>>   static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
>>   {
>> +     const struct devlink_reload_combination *comb;
>> +     int i;
>> +
>>        if (!devlink_reload_supported(ops)) {
>>                if (WARN_ON(ops->supported_reload_actions))
>>                        return false;
>> @@ -7498,6 +7551,18 @@ static bool devlink_reload_actions_valid(const struct devlink_ops *ops)
>>        if (WARN_ON(ops->supported_reload_actions >= BIT(__DEVLINK_RELOAD_ACTION_MAX) ||
>>                    ops->supported_reload_actions <= BIT(DEVLINK_RELOAD_ACTION_UNSPEC)))
>>                return false;
>> +
>> +     if (WARN_ON(!ops->supported_reload_action_limit_levels ||
>> +                 ops->supported_reload_action_limit_levels >=
>> +                 BIT(__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX)))
>> +             return false;
>> +
>> +     for (i = 0; i <  ARRAY_SIZE(devlink_reload_invalid_combinations); i++)  {
> White space, again.
Same double space, I will fix.
>
>> +             comb = &devlink_reload_invalid_combinations[i];
>> +             if (ops->supported_reload_actions == BIT(comb->action) &&
>> +                 ops->supported_reload_action_limit_levels == BIT(comb->limit_level))
>> +                     return false;
>> +     }
>>        return true;
>>   }
>>
>> @@ -9793,6 +9858,7 @@ static void __net_exit devlink_pernet_pre_exit(struct net *net)
>>                                continue;
>>                        err = devlink_reload(devlink, &init_net,
>>                                             DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
>> +                                          DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
>>                                             NULL, &actions_performed);
>>                        if (err && err != -EOPNOTSUPP)
>>                                pr_warn("Failed to reload devlink instance into init_net\n");

Re: [PATCH net-next RFC v5 03/15] devlink: Add reload action stats

[Moshe Shemesh]

On 9/23/2020 9:42 PM, Jakub Kicinski wrote:
> External email: Use caution opening links or attachments
>
>
> On Fri, 18 Sep 2020 19:06:39 +0300 Moshe Shemesh wrote:
>> Add reload action stats to hold the history per reload action type and
>> limit level.
>>
>> For example, the number of times fw_activate has been performed on this
>> device since the driver module was added or if the firmware activation
>> was performed with or without reset.
>>
>> The function devlink_remote_reload_actions_performed() is exported to
>> enable also drivers update on reload actions performed, for example in
>> case firmware activation with reset finished successfully but was
>> initiated by remote host.
>>
>> Add devlink notification on stats update.
>>
>> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
> I'd split this and the next patch by stat type (remote vs local).
> Rather than split them by collect / report.


OK.

>> diff --git a/include/net/devlink.h b/include/net/devlink.h
>> index d8c62d605381..f09f55a47d09 100644
>> --- a/include/net/devlink.h
>> +++ b/include/net/devlink.h
>> @@ -20,6 +20,9 @@
>>   #include <uapi/linux/devlink.h>
>>   #include <linux/xarray.h>
>>
>> +#define DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE \
>> +     (__DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX * __DEVLINK_RELOAD_ACTION_MAX)
>> +
>>   struct devlink_ops;
>>
>>   struct devlink {
>> @@ -38,6 +41,8 @@ struct devlink {
>>        struct list_head trap_policer_list;
>>        const struct devlink_ops *ops;
>>        struct xarray snapshot_ids;
>> +     u32 reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
>> +     u32 remote_reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
>>        struct device *dev;
>>        possible_net_t _net;
>>        struct mutex lock; /* Serializes access to devlink instance specific objects such as
>> @@ -1400,6 +1405,9 @@ void
>>   devlink_health_reporter_recovery_done(struct devlink_health_reporter *reporter);
>>
>>   bool devlink_is_reload_failed(const struct devlink *devlink);
>> +void devlink_remote_reload_actions_performed(struct devlink *devlink,
>> +                                          enum devlink_reload_action_limit_level limit_level,
>> +                                          unsigned long actions_performed);
>>
>>   void devlink_flash_update_begin_notify(struct devlink *devlink);
>>   void devlink_flash_update_end_notify(struct devlink *devlink);
>> diff --git a/net/core/devlink.c b/net/core/devlink.c
>> index fee6fcc7dead..1509c2ffb98b 100644
>> --- a/net/core/devlink.c
>> +++ b/net/core/devlink.c
>> @@ -3007,16 +3007,74 @@ bool devlink_is_reload_failed(const struct devlink *devlink)
>>   }
>>   EXPORT_SYMBOL_GPL(devlink_is_reload_failed);
>>
>> +static void
>> +__devlink_reload_action_stats_update(struct devlink *devlink,
>> +                                  u32 *reload_action_stats,
>> +                                  enum devlink_reload_action_limit_level limit_level,
>> +                                  unsigned long actions_performed)
>> +{
>> +     int stat_idx;
>> +     int action;
>> +
>> +     if (!actions_performed)
>> +             return;
>> +
>> +     if (WARN_ON(limit_level > DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX))
>> +             return;
>> +     for (action = 0; action <= DEVLINK_RELOAD_ACTION_MAX; action++) {
>> +             if (!test_bit(action, &actions_performed))
>> +                     continue;
>> +             stat_idx = limit_level * __DEVLINK_RELOAD_ACTION_MAX + action;
>> +             reload_action_stats[stat_idx]++;
>> +     }
>> +     devlink_notify(devlink, DEVLINK_CMD_NEW);
>> +}
>> +
>> +static void
>> +devlink_reload_action_stats_update(struct devlink *devlink,
>> +                                enum devlink_reload_action_limit_level limit_level,
>> +                                unsigned long actions_performed)
>> +{
>> +     __devlink_reload_action_stats_update(devlink, devlink->reload_action_stats,
>> +                                          limit_level, actions_performed);
>> +}
>> +
>> +/**
>> + *   devlink_remote_reload_actions_performed - Update devlink on reload actions
>> + *     performed which are not a direct result of devlink reload call.
>> + *
>> + *   This should be called by a driver after performing reload actions in case it was not
>> + *   a result of devlink reload call. For example fw_activate was performed as a result
>> + *   of devlink reload triggered fw_activate on another host.
>> + *   The motivation for this function is to keep data on reload actions performed on this
>> + *   function whether it was done due to direct devlink reload call or not.
>> + *
>> + *   @devlink: devlink
>> + *   @limit_level: reload action limit level
>> + *   @actions_performed: bitmask of actions performed
>> + */
>> +void devlink_remote_reload_actions_performed(struct devlink *devlink,
>> +                                          enum devlink_reload_action_limit_level limit_level,
>> +                                          unsigned long actions_performed)
>> +{
>> +     __devlink_reload_action_stats_update(devlink, devlink->remote_reload_action_stats,
>> +                                          limit_level, actions_performed);
>> +}
>> +EXPORT_SYMBOL_GPL(devlink_remote_reload_actions_performed);
>> +
>>   static int devlink_reload(struct devlink *devlink, struct net *dest_net,
>>                          enum devlink_reload_action action,
>>                          enum devlink_reload_action_limit_level limit_level,
>>                          struct netlink_ext_ack *extack, unsigned long *actions_performed)
>>   {
>> +     u32 remote_reload_action_stats[DEVLINK_RELOAD_ACTION_STATS_ARRAY_SIZE];
>>        int err;
>>
>>        if (!devlink->reload_enabled)
>>                return -EOPNOTSUPP;
>>
>> +     memcpy(remote_reload_action_stats, devlink->remote_reload_action_stats,
>> +            sizeof(remote_reload_action_stats));
>>        err = devlink->ops->reload_down(devlink, !!dest_net, action, limit_level, extack);
>>        if (err)
>>                return err;
>> @@ -3030,6 +3088,10 @@ static int devlink_reload(struct devlink *devlink, struct net *dest_net,
>>                return err;
>>
>>        WARN_ON(!test_bit(action, actions_performed));
>> +     /* protect from driver updating the remote action within devlink reload */
> s/protect from/catch/


Ack.

>> +     WARN_ON(memcmp(remote_reload_action_stats, devlink->remote_reload_action_stats,
>> +                    sizeof(remote_reload_action_stats)));
>> +     devlink_reload_action_stats_update(devlink, limit_level, *actions_performed);
>>        return 0;
>>   }
>>

Re: [PATCH net-next RFC v5 04/15] devlink: Add reload actions stats to dev get

[Moshe Shemesh]

On 9/23/2020 9:50 PM, Jakub Kicinski wrote:
> On Fri, 18 Sep 2020 19:06:40 +0300 Moshe Shemesh wrote:
>> Expose devlink reload actions stats to the user through devlink dev
>> get command.
>>
>> Examples:
>> $ devlink dev show
>> pci/0000:82:00.0:
>>    stats:
>>        reload_action_stats:
>>          driver_reinit 2
>>          fw_activate 1
>>          fw_activate_no_reset 0
>>          remote_driver_reinit 0
>>          remote_fw_activate 0
>>          remote_fw_activate_no_reset 0
>> pci/0000:82:00.1:
>>    stats:
>>        reload_action_stats:
>>          driver_reinit 0
>>          fw_activate 0
>>          fw_activate_no_reset 0
>>          remote_driver_reinit 1
>>          remote_fw_activate 1
>>          remote_fw_activate_no_reset 0
>>
>> $ devlink dev show -jp
>> {
>>      "dev": {
>>          "pci/0000:82:00.0": {
>>              "stats": {
>>                  "reload_action_stats": [ {
>>                          "driver_reinit": 2
>>                      },{
>>                          "fw_activate": 1
>>                      },{
>>                          "fw_activate_no_reset": 0
>>                      },{
>>                          "remote_driver_reinit": 0
>>                      },{
>>                          "remote_fw_activate": 0
>>                      },{
>>                          "remote_fw_activate_no_reset": 0
>>                      } ]
>>              }
>>          },
>>          "pci/0000:82:00.1": {
>>              "stats": {
>>                  "reload_action_stats": [ {
>>                          "driver_reinit": 0
>>                      },{
>>                          "fw_activate": 0
>>                      },{
>>                          "fw_activate_no_reset": 0
>>                      },{
>>                          "remote_driver_reinit": 1
>>                      },{
>>                          "remote_fw_activate": 1
>>                      },{
>>                          "remote_fw_activate_no_reset": 0
>>                      } ]
>>              }
>>          }
>>      }
>> }
>>
>> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
>> ---
>> v4 -> v5:
>> - Add remote actions stats
>> - If devlink reload is not supported, show only remote_stats
>> v3 -> v4:
>> - Renamed DEVLINK_ATTR_RELOAD_ACTION_CNT to
>>    DEVLINK_ATTR_RELOAD_ACTION_STAT
>> - Add stats per action per limit level
>> v2 -> v3:
>> - Add reload actions counters instead of supported reload actions
>>    (reload actions counters are only for supported action so no need for
>>     both)
>> v1 -> v2:
>> - Removed DEVLINK_ATTR_RELOAD_DEFAULT_LEVEL
>> - Removed DEVLINK_ATTR_RELOAD_LEVELS_INFO
>> - Have actions instead of levels
>> ---
>>   include/uapi/linux/devlink.h |  5 +++
>>   net/core/devlink.c           | 70 ++++++++++++++++++++++++++++++++++++
>>   2 files changed, 75 insertions(+)
>>
>> diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
>> index 0c5d942dcbd5..648d53be691e 100644
>> --- a/include/uapi/linux/devlink.h
>> +++ b/include/uapi/linux/devlink.h
>> @@ -497,7 +497,12 @@ enum devlink_attr {
>>   	DEVLINK_ATTR_RELOAD_ACTION,		/* u8 */
>>   	DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,	/* nested */
>>   	DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL,	/* u8 */
>> +	DEVLINK_ATTR_RELOAD_ACTION_STATS,	/* nested */
>> +	DEVLINK_ATTR_RELOAD_ACTION_STAT,	/* nested */
>> +	DEVLINK_ATTR_RELOAD_ACTION_STAT_REMOTE,	/* flag */
>> +	DEVLINK_ATTR_RELOAD_ACTION_STAT_VALUE,	/* u32 */
>>   
>> +	DEVLINK_ATTR_DEV_STATS,			/* nested */
>>   	/* add new attributes above here, update the policy in devlink.c */
> I'd propose this nesting:
>
> 	[DEV_STATS]
> 		[RELOAD_STATS]
> 			[DEV_STATS_ENTRY]
> 				[ACTION]
> 				[LIMIT]
> 				[VALUE]
> 			[DEV_STATS_ENTRY]
> 				[...]
> 		[REMOTE_RELOAD_STATS]
> 			[DEV_STATS_ENTRY]
> 				[ACTION]
> 				[LIMIT]
> 				[VALUE]
> 			[DEV_STATS_ENTRY]
> 				[...]
>
> Then you can fill in the inside of the [REMOTE_]RELOAD_STATS nests with
> a helper, and similarly user space can separate the two in JSON more
> cleanly than string concat.


Right, will fix, thanks.

>>   	__DEVLINK_ATTR_MAX,
>> diff --git a/net/core/devlink.c b/net/core/devlink.c
>> index 1509c2ffb98b..71aeda259e6a 100644
>> --- a/net/core/devlink.c
>> +++ b/net/core/devlink.c
>> @@ -501,10 +501,39 @@ devlink_reload_action_limit_level_is_supported(struct devlink *devlink,
>>   	return test_bit(limit_level, &devlink->ops->supported_reload_action_limit_levels);
>>   }
>>   
>> +static int devlink_reload_action_stat_put(struct sk_buff *msg, enum devlink_reload_action action,
>> +					  enum devlink_reload_action_limit_level limit_level,
>> +					  bool is_remote, u32 value)
>> +{
>> +	struct nlattr *reload_action_stat;
>> +
>> +	reload_action_stat = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT);
>> +	if (!reload_action_stat)
>> +		return -EMSGSIZE;
>> +
>> +	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION, action))
>> +		goto nla_put_failure;
>> +	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL, limit_level))
>> +		goto nla_put_failure;
>> +	if (is_remote && nla_put_flag(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT_REMOTE))
>> +		goto nla_put_failure;
>> +	if (nla_put_u32(msg, DEVLINK_ATTR_RELOAD_ACTION_STAT_VALUE, value))
>> +		goto nla_put_failure;
>> +	nla_nest_end(msg, reload_action_stat);
>> +	return 0;
>> +
>> +nla_put_failure:
>> +	nla_nest_cancel(msg, reload_action_stat);
>> +	return -EMSGSIZE;
>> +}
>> +
>>   static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
>>   			   enum devlink_command cmd, u32 portid,
>>   			   u32 seq, int flags)
>>   {
>> +	struct nlattr *dev_stats, *reload_action_stats;
>> +	int i, j, stat_idx;
>> +	u32 value;
>>   	void *hdr;
>>   
>>   	hdr = genlmsg_put(msg, portid, seq, &devlink_nl_family, flags, cmd);
>> @@ -516,9 +545,50 @@ static int devlink_nl_fill(struct sk_buff *msg, struct devlink *devlink,
>>   	if (nla_put_u8(msg, DEVLINK_ATTR_RELOAD_FAILED, devlink->reload_failed))
>>   		goto nla_put_failure;
>>   
>> +	dev_stats = nla_nest_start(msg, DEVLINK_ATTR_DEV_STATS);
>> +	if (!dev_stats)
>> +		goto nla_put_failure;
>> +	reload_action_stats = nla_nest_start(msg, DEVLINK_ATTR_RELOAD_ACTION_STATS);
>> +	if (!reload_action_stats)
>> +		goto dev_stats_nest_cancel;
>> +
>> +	for (j = 0; j <= DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX; j++) {
>> +		if (!devlink_reload_action_limit_level_is_supported(devlink, j))
>> +			continue;
>> +		for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
>> +			if (!devlink_reload_action_is_supported(devlink, i) ||
>> +			    devlink_reload_combination_is_invalid(i, j))
>> +				continue;
>> +
>> +			stat_idx = j * __DEVLINK_RELOAD_ACTION_MAX + i;
>> +			value = devlink->reload_action_stats[stat_idx];
>> +			if (devlink_reload_action_stat_put(msg, i, j, false, value))
>> +				goto reload_action_stats_nest_cancel;
>> +		}
>> +	}
>> +
>> +	for (j = 0; j <= DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX; j++) {
>> +		for (i = 0; i <= DEVLINK_RELOAD_ACTION_MAX; i++) {
>> +			if (i == DEVLINK_RELOAD_ACTION_UNSPEC ||
>> +			    devlink_reload_combination_is_invalid(i, j))
>> +				continue;
>> +
>> +			stat_idx = j * __DEVLINK_RELOAD_ACTION_MAX + i;
>> +			value = devlink->remote_reload_action_stats[stat_idx];
>> +			if (devlink_reload_action_stat_put(msg, i, j, true, value))
>> +				goto reload_action_stats_nest_cancel;
>> +		}
>> +	}
> This calls for a helper.
Ack.

Re: [PATCH net-next RFC v5 01/15] devlink: Add reload action option to devlink reload command

[Jakub Kicinski]

On Thu, 24 Sep 2020 22:01:42 +0300 Moshe Shemesh wrote:
> On 9/23/2020 9:25 PM, Jakub Kicinski wrote:
> >> Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
> >> @@ -3971,15 +3972,19 @@ static int mlx4_devlink_reload_up(struct devlink *devlink,
> >>        int err;
> >>
> >>        err = mlx4_restart_one_up(persist->pdev, true, devlink);
> >> -     if (err)
> >> +     if (err) {
> >>                mlx4_err(persist->dev, "mlx4_restart_one_up failed, ret=%d\n",
> >>                         err);
> >> +             return err;
> >> +     }
> >> +     *actions_performed = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT);  
> > FWIW I think drivers should be able to assign this even if they return
> > an error. On error there is no certainty what actions were actually
> > performed (e.g. when timeout happened but the device did the reset a
> > little later) so this argument should not be interpreted in presence of
> > errors, anyway.  
> 
> Not sure I got it. Do you mean driver can assign it anyway and devlink 
> should ignore in case of failure ?

Yup.

> As I implemented here devlink already ignores actions_performed in case 
> driver returns with error.

Right, but you're changing all bunch of drivers like this:

 static void reload()
 {		
-	return do_it();
+	int err;
+
+	err = do_it();
+	if (err)
+		return err;
+
+	*actions_performed = SOMETHING;
+	return 0;
 }

When you can instead:

 static void reload()
 {		
+	*actions_performed = SOMETHING;
 	return do_it();
 }

> >> @@ -3011,12 +3064,43 @@ static int devlink_nl_cmd_reload(struct sk_buff *skb, struct genl_info *info)
> >>                        return PTR_ERR(dest_net);
> >>        }
> >>
> >> -     err = devlink_reload(devlink, dest_net, info->extack);
> >> +     if (info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
> >> +             action = nla_get_u8(info->attrs[DEVLINK_ATTR_RELOAD_ACTION]);
> >> +     else
> >> +             action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT;
> >> +
> >> +     if (action == DEVLINK_RELOAD_ACTION_UNSPEC) {
> >> +             NL_SET_ERR_MSG_MOD(info->extack, "Invalid reload action");
> >> +             return -EINVAL;
> >> +     } else if (!devlink_reload_action_is_supported(devlink, action)) {
> >> +             NL_SET_ERR_MSG_MOD(info->extack, "Requested reload action is not supported by the driver");
> >> +             return -EOPNOTSUPP;
> >> +     }
> >> +
> >> +     err = devlink_reload(devlink, dest_net, action, info->extack, &actions_performed);  
> > Perhaps we can pass the requested action to the driver via
> > actions_performed already, and then all the drivers which
> > only do what they're asked to don't have to touch it?  
> 
> Not sure about it. Note that in the next patch I add here limit_level 
> and that has only input param, so I think it would be confusing.

I don't think it'd be, but don't feel strongly either.

> >>        if (dest_net)
> >>                put_net(dest_net);
> >>
> >> -     return err;
> >> +     if (err)
> >> +             return err;
> >> +     /* For backward compatibility generate reply only if attributes used by user */
> >> +     if (!info->attrs[DEVLINK_ATTR_RELOAD_ACTION])
> >> +             return 0;
> >> +
> >> +     msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
> >> +     if (!msg)
> >> +             return -ENOMEM;
> >> +
> >> +     err = devlink_nl_reload_actions_performed_fill(msg, devlink, actions_performed,
> >> +                                                    DEVLINK_CMD_RELOAD, info->snd_portid,
> >> +                                                    info->snd_seq, 0);
> >> +     if (err) {
> >> +             nlmsg_free(msg);
> >> +             return err;
> >> +     }
> >> +
> >> +     return genlmsg_reply(msg, info);  
> > Are you using devlink_nl_reload_actions_performed_fill() somewhere else?  
> No
> > I'd move the nlmsg_new() / genlmsg_reply() into the helper.  
> 
> Can do it, but there are many _fill() functions in devlink.c code to 
> fill the data, none of them include nlmsg_new() and genlmsg_reply() 
> that's always in the calling function, even if the calling function adds 
> only that. So I guess I will leave it for consistency.

Don't call the helper _fill() and you'll be good.

> >>   }
> >>
> >>   static int devlink_nl_flash_update_fill(struct sk_buff *msg,
> >> @@ -7069,6 +7153,7 @@ static const struct nla_policy devlink_nl_policy[DEVLINK_ATTR_MAX + 1] = {
> >>        [DEVLINK_ATTR_TRAP_POLICER_RATE] = { .type = NLA_U64 },
> >>        [DEVLINK_ATTR_TRAP_POLICER_BURST] = { .type = NLA_U64 },
> >>        [DEVLINK_ATTR_PORT_FUNCTION] = { .type = NLA_NESTED },
> >> +     [DEVLINK_ATTR_RELOAD_ACTION] = { .type = NLA_U8 },  
> > Why not just range validation here?  
> 
> All devlink attributes that pass here go through devlink_nl_poicy this 
> way, including other enums.
> 
> I think changing that should be in a different patch for all, not in 
> this patchset.

I don't think this is on purpose. Please use range validation in new
code from the start. We support dumping policies to user space, it's
useful to know the range of parameters from the policy.

Re: [PATCH net-next RFC v5 02/15] devlink: Add reload action limit level

[Jakub Kicinski]

On Thu, 24 Sep 2020 22:29:55 +0300 Moshe Shemesh wrote:
> >> @@ -3964,6 +3965,7 @@ static int mlx4_devlink_reload_down(struct devlink *devlink, bool netns_change,
> >>   }
> >>
> >>   static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_action action,
> >> +                               enum devlink_reload_action_limit_level limit_level,
> >>                                  struct netlink_ext_ack *extack, unsigned long *actions_performed)
> >>   {
> >>        struct mlx4_priv *priv = devlink_priv(devlink);
> >> @@ -3985,6 +3987,7 @@ static int mlx4_devlink_reload_up(struct devlink *devlink, enum devlink_reload_a
> >>   static const struct devlink_ops mlx4_devlink_ops = {
> >>        .port_type_set  = mlx4_devlink_port_type_set,
> >>        .supported_reload_actions = BIT(DEVLINK_RELOAD_ACTION_DRIVER_REINIT),
> >> +     .supported_reload_action_limit_levels = BIT(DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE),  
> > Please cut down the name lenghts, this is just lazy.
> >
> > 'supported_reload_limits' or 'cap_reload_limits' is perfectly
> > sufficient.
> >
> > 'reload_limits' would be even better. Cause what else would it be if
> > not a capability.  
> 
> Sounds good.
> 
> So instead of supported_reload_actions_limit_levels will have reload_limits.
> 
> Instead of supported_reload_actions will have reload_actions, OK ?

Sounds good.

> May also use reload_limit_level instead of reload_action_limit_level 
> everywhere if its clear enough.

I think reload_limits is clear. I'd also cut down the length of the
defines / enum names.

> > Besides I don't think drivers should have to fill negative attributes
> > (that they don't support something). Everyone is always going to
> > support NONE, since it's "unspecified" / "pick your favorite", right?  
> 
> Good point, will remove it.
> 
> >>        .reload_down    = mlx4_devlink_reload_down,
> >>        .reload_up      = mlx4_devlink_reload_up,
> >>   };
> >> diff --git a/include/uapi/linux/devlink.h b/include/uapi/linux/devlink.h
> >> index fdba7ab58a79..0c5d942dcbd5 100644
> >> --- a/include/uapi/linux/devlink.h
> >> +++ b/include/uapi/linux/devlink.h
> >> @@ -289,6 +289,22 @@ enum devlink_reload_action {
> >>        DEVLINK_RELOAD_ACTION_MAX = __DEVLINK_RELOAD_ACTION_MAX - 1
> >>   };
> >>
> >> +/**
> >> + * enum devlink_reload_action_limit_level - Reload action limit level.
> >> + * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE: No constrains on action. Action may include
> >> + *                                          reset or downtime as needed.
> >> + * @DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET: No reset allowed, no down time allowed,
> >> + *                                              no link flap and no configuration is lost.
> >> + */
> >> +enum devlink_reload_action_limit_level {  
> > You reserved UNSPEC for actions but not for limit level?  
> 
> 
> Yes, I used LIMIT_LEVEL_NONE = 0 as no limit needed, so I skipped UNSPEC.
> 
> Maybe should add UNSPEC and use UNSPEC as no limit needed. But UNSPEC is 
> kind of invalid.

Yeah, if we have UNSPEC then it should be invalid.

I'm mostly asking for consistency, either have UNSPEC for both actions
and limits or for neither.

> >> +     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NONE,
> >> +     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
> >> +
> >> +     /* Add new reload actions limit level above */
> >> +     __DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX,
> >> +     DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX = __DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_MAX - 1
> >> +};
> >> +
> >>   enum devlink_attr {
> >>        /* don't change the order or add anything between, this is ABI! */
> >>        DEVLINK_ATTR_UNSPEC,
> >> @@ -480,6 +496,7 @@ enum devlink_attr {
> >>
> >>        DEVLINK_ATTR_RELOAD_ACTION,             /* u8 */
> >>        DEVLINK_ATTR_RELOAD_ACTIONS_PERFORMED,  /* nested */
> >> +     DEVLINK_ATTR_RELOAD_ACTION_LIMIT_LEVEL, /* u8 */
> >>
> >>        /* add new attributes above here, update the policy in devlink.c */
> >>
> >> diff --git a/net/core/devlink.c b/net/core/devlink.c
> >> index 318ef29f81f2..fee6fcc7dead 100644
> >> --- a/net/core/devlink.c
> >> +++ b/net/core/devlink.c
> >> @@ -462,12 +462,45 @@ static int devlink_nl_put_handle(struct sk_buff *msg, struct devlink *devlink)
> >>        return 0;
> >>   }
> >>
> >> +struct devlink_reload_combination {
> >> +     enum devlink_reload_action action;
> >> +     enum devlink_reload_action_limit_level limit_level;
> >> +};
> >> +
> >> +static const struct devlink_reload_combination devlink_reload_invalid_combinations[] = {
> >> +     {
> >> +             /* can't reinitialize driver with no down time */
> >> +             .action = DEVLINK_RELOAD_ACTION_DRIVER_REINIT,
> >> +             .limit_level = DEVLINK_RELOAD_ACTION_LIMIT_LEVEL_NO_RESET,
> >> +     },
> >> +};
> >> +
> >> +static bool
> >> +devlink_reload_combination_is_invalid(enum devlink_reload_action action,
> >> +                                   enum devlink_reload_action_limit_level limit_level)
> >> +{
> >> +     int i;
> >> +
> >> +     for (i = 0 ; i <  ARRAY_SIZE(devlink_reload_invalid_combinations) ; i++)  
> > Whitespace. Did you checkpatch?  
> 
> 
> Yes, checked it again now, it still pass. I think checkpatch doesn't see 
> double space.

And the spaces before semicolons? It's sad if checkpatch misses such
basic stuff :(

> But anyway, I missed it, I will fix.