diff options
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_lib.c | 174 | ||||
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_lib.h | 4 | ||||
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_main.c | 1 | ||||
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_sched.c | 130 | ||||
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_sched.h | 6 | ||||
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c | 160 | ||||
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h | 15 |
7 files changed, 486 insertions, 4 deletions
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c index f981e77f72ad..acff8d3b1c97 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_lib.c @@ -3601,6 +3601,180 @@ int ice_clear_dflt_vsi(struct ice_sw *sw) } /** + * ice_get_link_speed_mbps - get link speed in Mbps + * @vsi: the VSI whose link speed is being queried + * + * Return current VSI link speed and 0 if the speed is unknown. + */ +int ice_get_link_speed_mbps(struct ice_vsi *vsi) +{ + switch (vsi->port_info->phy.link_info.link_speed) { + case ICE_AQ_LINK_SPEED_100GB: + return SPEED_100000; + case ICE_AQ_LINK_SPEED_50GB: + return SPEED_50000; + case ICE_AQ_LINK_SPEED_40GB: + return SPEED_40000; + case ICE_AQ_LINK_SPEED_25GB: + return SPEED_25000; + case ICE_AQ_LINK_SPEED_20GB: + return SPEED_20000; + case ICE_AQ_LINK_SPEED_10GB: + return SPEED_10000; + case ICE_AQ_LINK_SPEED_5GB: + return SPEED_5000; + case ICE_AQ_LINK_SPEED_2500MB: + return SPEED_2500; + case ICE_AQ_LINK_SPEED_1000MB: + return SPEED_1000; + case ICE_AQ_LINK_SPEED_100MB: + return SPEED_100; + case ICE_AQ_LINK_SPEED_10MB: + return SPEED_10; + case ICE_AQ_LINK_SPEED_UNKNOWN: + default: + return 0; + } +} + +/** + * ice_get_link_speed_kbps - get link speed in Kbps + * @vsi: the VSI whose link speed is being queried + * + * Return current VSI link speed and 0 if the speed is unknown. + */ +static int ice_get_link_speed_kbps(struct ice_vsi *vsi) +{ + int speed_mbps; + + speed_mbps = ice_get_link_speed_mbps(vsi); + + return speed_mbps * 1000; +} + +/** + * ice_set_min_bw_limit - setup minimum BW limit for Tx based on min_tx_rate + * @vsi: VSI to be configured + * @min_tx_rate: min Tx rate in Kbps to be configured as BW limit + * + * If the min_tx_rate is specified as 0 that means to clear the minimum BW limit + * profile, otherwise a non-zero value will force a minimum BW limit for the VSI + * on TC 0. + */ +int ice_set_min_bw_limit(struct ice_vsi *vsi, u64 min_tx_rate) +{ + struct ice_pf *pf = vsi->back; + enum ice_status status; + struct device *dev; + int speed; + + dev = ice_pf_to_dev(pf); + if (!vsi->port_info) { + dev_dbg(dev, "VSI %d, type %u specified doesn't have valid port_info\n", + vsi->idx, vsi->type); + return -EINVAL; + } + + speed = ice_get_link_speed_kbps(vsi); + if (min_tx_rate > (u64)speed) { + dev_err(dev, "invalid min Tx rate %llu Kbps specified for %s %d is greater than current link speed %u Kbps\n", + min_tx_rate, ice_vsi_type_str(vsi->type), vsi->idx, + speed); + return -EINVAL; + } + + /* Configure min BW for VSI limit */ + if (min_tx_rate) { + status = ice_cfg_vsi_bw_lmt_per_tc(vsi->port_info, vsi->idx, 0, + ICE_MIN_BW, min_tx_rate); + if (status) { + dev_err(dev, "failed to set min Tx rate(%llu Kbps) for %s %d\n", + min_tx_rate, ice_vsi_type_str(vsi->type), + vsi->idx); + return -EIO; + } + + dev_dbg(dev, "set min Tx rate(%llu Kbps) for %s\n", + min_tx_rate, ice_vsi_type_str(vsi->type)); + } else { + status = ice_cfg_vsi_bw_dflt_lmt_per_tc(vsi->port_info, + vsi->idx, 0, + ICE_MIN_BW); + if (status) { + dev_err(dev, "failed to clear min Tx rate configuration for %s %d\n", + ice_vsi_type_str(vsi->type), vsi->idx); + return -EIO; + } + + dev_dbg(dev, "cleared min Tx rate configuration for %s %d\n", + ice_vsi_type_str(vsi->type), vsi->idx); + } + + return 0; +} + +/** + * ice_set_max_bw_limit - setup maximum BW limit for Tx based on max_tx_rate + * @vsi: VSI to be configured + * @max_tx_rate: max Tx rate in Kbps to be configured as BW limit + * + * If the max_tx_rate is specified as 0 that means to clear the maximum BW limit + * profile, otherwise a non-zero value will force a maximum BW limit for the VSI + * on TC 0. + */ +int ice_set_max_bw_limit(struct ice_vsi *vsi, u64 max_tx_rate) +{ + struct ice_pf *pf = vsi->back; + enum ice_status status; + struct device *dev; + int speed; + + dev = ice_pf_to_dev(pf); + if (!vsi->port_info) { + dev_dbg(dev, "VSI %d, type %u specified doesn't have valid port_info\n", + vsi->idx, vsi->type); + return -EINVAL; + } + + speed = ice_get_link_speed_kbps(vsi); + if (max_tx_rate > (u64)speed) { + dev_err(dev, "invalid max Tx rate %llu Kbps specified for %s %d is greater than current link speed %u Kbps\n", + max_tx_rate, ice_vsi_type_str(vsi->type), vsi->idx, + speed); + return -EINVAL; + } + + /* Configure max BW for VSI limit */ + if (max_tx_rate) { + status = ice_cfg_vsi_bw_lmt_per_tc(vsi->port_info, vsi->idx, 0, + ICE_MAX_BW, max_tx_rate); + if (status) { + dev_err(dev, "failed setting max Tx rate(%llu Kbps) for %s %d\n", + max_tx_rate, ice_vsi_type_str(vsi->type), + vsi->idx); + return -EIO; + } + + dev_dbg(dev, "set max Tx rate(%llu Kbps) for %s %d\n", + max_tx_rate, ice_vsi_type_str(vsi->type), vsi->idx); + } else { + status = ice_cfg_vsi_bw_dflt_lmt_per_tc(vsi->port_info, + vsi->idx, 0, + ICE_MAX_BW); + if (status) { + dev_err(dev, "failed clearing max Tx rate configuration for %s %d\n", + ice_vsi_type_str(vsi->type), vsi->idx); + return -EIO; + } + + dev_dbg(dev, "cleared max Tx rate configuration for %s %d\n", + ice_vsi_type_str(vsi->type), vsi->idx); + } + + return 0; +} + +/** * ice_set_link - turn on/off physical link * @vsi: VSI to modify physical link on * @ena: turn on/off physical link diff --git a/drivers/net/ethernet/intel/ice/ice_lib.h b/drivers/net/ethernet/intel/ice/ice_lib.h index b6c429c5875d..d395bd590e84 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.h +++ b/drivers/net/ethernet/intel/ice/ice_lib.h @@ -116,7 +116,9 @@ bool ice_is_vsi_dflt_vsi(struct ice_sw *sw, struct ice_vsi *vsi); int ice_set_dflt_vsi(struct ice_sw *sw, struct ice_vsi *vsi); int ice_clear_dflt_vsi(struct ice_sw *sw); - +int ice_set_min_bw_limit(struct ice_vsi *vsi, u64 min_tx_rate); +int ice_set_max_bw_limit(struct ice_vsi *vsi, u64 max_tx_rate); +int ice_get_link_speed_mbps(struct ice_vsi *vsi); int ice_vsi_update_security(struct ice_vsi *vsi, void (*fill)(struct ice_vsi_ctx *)); diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index f531691a3e12..99647dceefc4 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -7390,6 +7390,7 @@ static const struct net_device_ops ice_netdev_ops = { .ndo_set_vf_vlan = ice_set_vf_port_vlan, .ndo_set_vf_link_state = ice_set_vf_link_state, .ndo_get_vf_stats = ice_get_vf_stats, + .ndo_set_vf_rate = ice_set_vf_bw, .ndo_vlan_rx_add_vid = ice_vlan_rx_add_vid, .ndo_vlan_rx_kill_vid = ice_vlan_rx_kill_vid, .ndo_setup_tc = ice_setup_tc, diff --git a/drivers/net/ethernet/intel/ice/ice_sched.c b/drivers/net/ethernet/intel/ice/ice_sched.c index 9f07b6641705..560e52b99f83 100644 --- a/drivers/net/ethernet/intel/ice/ice_sched.c +++ b/drivers/net/ethernet/intel/ice/ice_sched.c @@ -3771,6 +3771,136 @@ ice_cfg_q_bw_dflt_lmt(struct ice_port_info *pi, u16 vsi_handle, u8 tc, } /** + * ice_sched_get_node_by_id_type - get node from ID type + * @pi: port information structure + * @id: identifier + * @agg_type: type of aggregator + * @tc: traffic class + * + * This function returns node identified by ID of type aggregator, and + * based on traffic class (TC). This function needs to be called with + * the scheduler lock held. + */ +static struct ice_sched_node * +ice_sched_get_node_by_id_type(struct ice_port_info *pi, u32 id, + enum ice_agg_type agg_type, u8 tc) +{ + struct ice_sched_node *node = NULL; + + switch (agg_type) { + case ICE_AGG_TYPE_VSI: { + struct ice_vsi_ctx *vsi_ctx; + u16 vsi_handle = (u16)id; + + if (!ice_is_vsi_valid(pi->hw, vsi_handle)) + break; + /* Get sched_vsi_info */ + vsi_ctx = ice_get_vsi_ctx(pi->hw, vsi_handle); + if (!vsi_ctx) + break; + node = vsi_ctx->sched.vsi_node[tc]; + break; + } + + case ICE_AGG_TYPE_AGG: { + struct ice_sched_node *tc_node; + + tc_node = ice_sched_get_tc_node(pi, tc); + if (tc_node) + node = ice_sched_get_agg_node(pi, tc_node, id); + break; + } + + default: + break; + } + + return node; +} + +/** + * ice_sched_set_node_bw_lmt_per_tc - set node BW limit per TC + * @pi: port information structure + * @id: ID (software VSI handle or AGG ID) + * @agg_type: aggregator type (VSI or AGG type node) + * @tc: traffic class + * @rl_type: min or max + * @bw: bandwidth in Kbps + * + * This function sets BW limit of VSI or Aggregator scheduling node + * based on TC information from passed in argument BW. + */ +static enum ice_status +ice_sched_set_node_bw_lmt_per_tc(struct ice_port_info *pi, u32 id, + enum ice_agg_type agg_type, u8 tc, + enum ice_rl_type rl_type, u32 bw) +{ + enum ice_status status = ICE_ERR_PARAM; + struct ice_sched_node *node; + + if (!pi) + return status; + + if (rl_type == ICE_UNKNOWN_BW) + return status; + + mutex_lock(&pi->sched_lock); + node = ice_sched_get_node_by_id_type(pi, id, agg_type, tc); + if (!node) { + ice_debug(pi->hw, ICE_DBG_SCHED, "Wrong id, agg type, or tc\n"); + goto exit_set_node_bw_lmt_per_tc; + } + if (bw == ICE_SCHED_DFLT_BW) + status = ice_sched_set_node_bw_dflt_lmt(pi, node, rl_type); + else + status = ice_sched_set_node_bw_lmt(pi, node, rl_type, bw); + +exit_set_node_bw_lmt_per_tc: + mutex_unlock(&pi->sched_lock); + return status; +} + +/** + * ice_cfg_vsi_bw_lmt_per_tc - configure VSI BW limit per TC + * @pi: port information structure + * @vsi_handle: software VSI handle + * @tc: traffic class + * @rl_type: min or max + * @bw: bandwidth in Kbps + * + * This function configures BW limit of VSI scheduling node based on TC + * information. + */ +enum ice_status +ice_cfg_vsi_bw_lmt_per_tc(struct ice_port_info *pi, u16 vsi_handle, u8 tc, + enum ice_rl_type rl_type, u32 bw) +{ + return ice_sched_set_node_bw_lmt_per_tc(pi, vsi_handle, + ICE_AGG_TYPE_VSI, + tc, rl_type, bw); +} + +/** + * ice_cfg_vsi_bw_dflt_lmt_per_tc - configure default VSI BW limit per TC + * @pi: port information structure + * @vsi_handle: software VSI handle + * @tc: traffic class + * @rl_type: min or max + * + * This function configures default BW limit of VSI scheduling node based on TC + * information. + */ +enum ice_status +ice_cfg_vsi_bw_dflt_lmt_per_tc(struct ice_port_info *pi, u16 vsi_handle, u8 tc, + enum ice_rl_type rl_type) +{ + return ice_sched_set_node_bw_lmt_per_tc(pi, vsi_handle, + ICE_AGG_TYPE_VSI, + tc, rl_type, + ICE_SCHED_DFLT_BW); +} + +/** * ice_cfg_rl_burst_size - Set burst size value * @hw: pointer to the HW struct * @bytes: burst size in bytes diff --git a/drivers/net/ethernet/intel/ice/ice_sched.h b/drivers/net/ethernet/intel/ice/ice_sched.h index 9beef8f0ec76..f89b80ba3499 100644 --- a/drivers/net/ethernet/intel/ice/ice_sched.h +++ b/drivers/net/ethernet/intel/ice/ice_sched.h @@ -103,6 +103,12 @@ ice_cfg_q_bw_lmt(struct ice_port_info *pi, u16 vsi_handle, u8 tc, enum ice_status ice_cfg_q_bw_dflt_lmt(struct ice_port_info *pi, u16 vsi_handle, u8 tc, u16 q_handle, enum ice_rl_type rl_type); +enum ice_status +ice_cfg_vsi_bw_lmt_per_tc(struct ice_port_info *pi, u16 vsi_handle, u8 tc, + enum ice_rl_type rl_type, u32 bw); +enum ice_status +ice_cfg_vsi_bw_dflt_lmt_per_tc(struct ice_port_info *pi, u16 vsi_handle, u8 tc, + enum ice_rl_type rl_type); enum ice_status ice_cfg_rl_burst_size(struct ice_hw *hw, u32 bytes); void ice_sched_replay_agg_vsi_preinit(struct ice_hw *hw); void ice_sched_replay_agg(struct ice_hw *hw); diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c index d90a3b7be713..86f265268ac8 100644 --- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c +++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c @@ -5,6 +5,7 @@ #include "ice_base.h" #include "ice_lib.h" #include "ice_fltr.h" +#include "ice_dcb_lib.h" #include "ice_flow.h" #include "ice_eswitch.h" #include "ice_virtchnl_allowlist.h" @@ -885,6 +886,40 @@ static int ice_calc_vf_first_vector_idx(struct ice_pf *pf, struct ice_vf *vf) } /** + * ice_vf_rebuild_host_tx_rate_cfg - re-apply the Tx rate limiting configuration + * @vf: VF to re-apply the configuration for + * + * Called after a VF VSI has been re-added/rebuild during reset. The PF driver + * needs to re-apply the host configured Tx rate limiting configuration. + */ +static int ice_vf_rebuild_host_tx_rate_cfg(struct ice_vf *vf) +{ + struct device *dev = ice_pf_to_dev(vf->pf); + struct ice_vsi *vsi = ice_get_vf_vsi(vf); + int err; + + if (vf->min_tx_rate) { + err = ice_set_min_bw_limit(vsi, (u64)vf->min_tx_rate * 1000); + if (err) { + dev_err(dev, "failed to set min Tx rate to %d Mbps for VF %u, error %d\n", + vf->min_tx_rate, vf->vf_id, err); + return err; + } + } + + if (vf->max_tx_rate) { + err = ice_set_max_bw_limit(vsi, (u64)vf->max_tx_rate * 1000); + if (err) { + dev_err(dev, "failed to set max Tx rate to %d Mbps for VF %u, error %d\n", + vf->max_tx_rate, vf->vf_id, err); + return err; + } + } + + return 0; +} + +/** * ice_vf_rebuild_host_vlan_cfg - add VLAN 0 filter or rebuild the Port VLAN * @vf: VF to add MAC filters for * @@ -1420,6 +1455,11 @@ static void ice_vf_rebuild_host_cfg(struct ice_vf *vf) if (ice_vf_rebuild_host_vlan_cfg(vf)) dev_err(dev, "failed to rebuild VLAN configuration for VF %u\n", vf->vf_id); + + if (ice_vf_rebuild_host_tx_rate_cfg(vf)) + dev_err(dev, "failed to rebuild Tx rate limiting configuration for VF %u\n", + vf->vf_id); + /* rebuild aggregator node config for main VF VSI */ ice_vf_rebuild_aggregator_node_cfg(vsi); } @@ -4747,8 +4787,8 @@ ice_get_vf_cfg(struct net_device *netdev, int vf_id, struct ifla_vf_info *ivi) ivi->linkstate = IFLA_VF_LINK_STATE_ENABLE; else ivi->linkstate = IFLA_VF_LINK_STATE_DISABLE; - ivi->max_tx_rate = vf->tx_rate; - ivi->min_tx_rate = 0; + ivi->max_tx_rate = vf->max_tx_rate; + ivi->min_tx_rate = vf->min_tx_rate; return 0; } @@ -4927,6 +4967,122 @@ int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state) } /** + * ice_calc_all_vfs_min_tx_rate - calculate cumulative min Tx rate on all VFs + * @pf: PF associated with VFs + */ +static int ice_calc_all_vfs_min_tx_rate(struct ice_pf *pf) +{ + int rate = 0, i; + + ice_for_each_vf(pf, i) + rate += pf->vf[i].min_tx_rate; + + return rate; +} + +/** + * ice_min_tx_rate_oversubscribed - check if min Tx rate causes oversubscription + * @vf: VF trying to configure min_tx_rate + * @min_tx_rate: min Tx rate in Mbps + * + * Check if the min_tx_rate being passed in will cause oversubscription of total + * min_tx_rate based on the current link speed and all other VFs configured + * min_tx_rate + * + * Return true if the passed min_tx_rate would cause oversubscription, else + * return false + */ +static bool +ice_min_tx_rate_oversubscribed(struct ice_vf *vf, int min_tx_rate) +{ + int link_speed_mbps = ice_get_link_speed_mbps(ice_get_vf_vsi(vf)); + int all_vfs_min_tx_rate = ice_calc_all_vfs_min_tx_rate(vf->pf); + + /* this VF's previous rate is being overwritten */ + all_vfs_min_tx_rate -= vf->min_tx_rate; + + if (all_vfs_min_tx_rate + min_tx_rate > link_speed_mbps) { + dev_err(ice_pf_to_dev(vf->pf), "min_tx_rate of %d Mbps on VF %u would cause oversubscription of %d Mbps based on the current link speed %d Mbps\n", + min_tx_rate, vf->vf_id, + all_vfs_min_tx_rate + min_tx_rate - link_speed_mbps, + link_speed_mbps); + return true; + } + + return false; +} + +/** + * ice_set_vf_bw - set min/max VF bandwidth + * @netdev: network interface device structure + * @vf_id: VF identifier + * @min_tx_rate: Minimum Tx rate in Mbps + * @max_tx_rate: Maximum Tx rate in Mbps + */ +int +ice_set_vf_bw(struct net_device *netdev, int vf_id, int min_tx_rate, + int max_tx_rate) +{ + struct ice_pf *pf = ice_netdev_to_pf(netdev); + struct ice_vsi *vsi; + struct device *dev; + struct ice_vf *vf; + int ret; + + dev = ice_pf_to_dev(pf); + if (ice_validate_vf_id(pf, vf_id)) + return -EINVAL; + + vf = &pf->vf[vf_id]; + ret = ice_check_vf_ready_for_cfg(vf); + if (ret) + return ret; + + vsi = ice_get_vf_vsi(vf); + + /* when max_tx_rate is zero that means no max Tx rate limiting, so only + * check if max_tx_rate is non-zero + */ + if (max_tx_rate && min_tx_rate > max_tx_rate) { + dev_err(dev, "Cannot set min Tx rate %d Mbps greater than max Tx rate %d Mbps\n", + min_tx_rate, max_tx_rate); + return -EINVAL; + } + + if (min_tx_rate && ice_is_dcb_active(pf)) { + dev_err(dev, "DCB on PF is currently enabled. VF min Tx rate limiting not allowed on this PF.\n"); + return -EOPNOTSUPP; + } + + if (ice_min_tx_rate_oversubscribed(vf, min_tx_rate)) + return -EINVAL; + + if (vf->min_tx_rate != (unsigned int)min_tx_rate) { + ret = ice_set_min_bw_limit(vsi, (u64)min_tx_rate * 1000); + if (ret) { + dev_err(dev, "Unable to set min-tx-rate for VF %d\n", + vf->vf_id); + return ret; + } + + vf->min_tx_rate = min_tx_rate; + } + + if (vf->max_tx_rate != (unsigned int)max_tx_rate) { + ret = ice_set_max_bw_limit(vsi, (u64)max_tx_rate * 1000); + if (ret) { + dev_err(dev, "Unable to set max-tx-rate for VF %d\n", + vf->vf_id); + return ret; + } + + vf->max_tx_rate = max_tx_rate; + } + + return 0; +} + +/** * ice_get_vf_stats - populate some stats for the VF * @netdev: the netdev of the PF * @vf_id: the host OS identifier (0-255) diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h index 3115284e5411..5ff93a08f54c 100644 --- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h +++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h @@ -125,7 +125,8 @@ struct ice_vf { * the main LAN VSI for the PF. */ u16 lan_vsi_num; /* ID as used by firmware */ - unsigned int tx_rate; /* Tx bandwidth limit in Mbps */ + unsigned int min_tx_rate; /* Minimum Tx bandwidth limit in Mbps */ + unsigned int max_tx_rate; /* Maximum Tx bandwidth limit in Mbps */ DECLARE_BITMAP(vf_states, ICE_VF_STATES_NBITS); /* VF runtime states */ u64 num_inval_msgs; /* number of continuous invalid msgs */ @@ -172,6 +173,10 @@ int ice_set_vf_port_vlan(struct net_device *netdev, int vf_id, u16 vlan_id, u8 qos, __be16 vlan_proto); +int +ice_set_vf_bw(struct net_device *netdev, int vf_id, int min_tx_rate, + int max_tx_rate); + int ice_set_vf_trust(struct net_device *netdev, int vf_id, bool trusted); int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state); @@ -304,6 +309,14 @@ ice_set_vf_link_state(struct net_device __always_unused *netdev, } static inline int +ice_set_vf_bw(struct net_device __always_unused *netdev, + int __always_unused vf_id, int __always_unused min_tx_rate, + int __always_unused max_tx_rate) +{ + return -EOPNOTSUPP; +} + +static inline int ice_calc_vf_reg_idx(struct ice_vf __always_unused *vf, struct ice_q_vector __always_unused *q_vector) { |