diff options
28 files changed, 516 insertions, 254 deletions
diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_main.c b/drivers/net/ethernet/aquantia/atlantic/aq_main.c index 77609dc0a08d..0b2a52199914 100644 --- a/drivers/net/ethernet/aquantia/atlantic/aq_main.c +++ b/drivers/net/ethernet/aquantia/atlantic/aq_main.c @@ -21,6 +21,7 @@ #include <linux/ip.h> #include <linux/udp.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <linux/filter.h> MODULE_LICENSE("GPL v2"); diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h index be96f1dc0372..d4a862a9fd7d 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.h @@ -4,7 +4,7 @@ #ifndef __CXGB4_TC_MQPRIO_H__ #define __CXGB4_TC_MQPRIO_H__ -#include <net/pkt_cls.h> +#include <net/pkt_sched.h> #define CXGB4_EOSW_TXQ_DEFAULT_DESC_NUM 128 diff --git a/drivers/net/ethernet/engleder/tsnep_tc.c b/drivers/net/ethernet/engleder/tsnep_tc.c index c4c6e1357317..d083e6684f12 100644 --- a/drivers/net/ethernet/engleder/tsnep_tc.c +++ b/drivers/net/ethernet/engleder/tsnep_tc.c @@ -403,12 +403,33 @@ static int tsnep_taprio(struct tsnep_adapter *adapter, return 0; } +static int tsnep_tc_query_caps(struct tsnep_adapter *adapter, + struct tc_query_caps_base *base) +{ + switch (base->type) { + case TC_SETUP_QDISC_TAPRIO: { + struct tc_taprio_caps *caps = base->caps; + + if (!adapter->gate_control) + return -EOPNOTSUPP; + + caps->gate_mask_per_txq = true; + + return 0; + } + default: + return -EOPNOTSUPP; + } +} + int tsnep_tc_setup(struct net_device *netdev, enum tc_setup_type type, void *type_data) { struct tsnep_adapter *adapter = netdev_priv(netdev); switch (type) { + case TC_QUERY_CAPS: + return tsnep_tc_query_caps(adapter, type_data); case TC_SETUP_QDISC_TAPRIO: return tsnep_taprio(adapter, type_data); default: diff --git a/drivers/net/ethernet/freescale/enetc/enetc.c b/drivers/net/ethernet/freescale/enetc/enetc.c index 1c0aeaa13cde..2fc712b24d12 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc.c +++ b/drivers/net/ethernet/freescale/enetc/enetc.c @@ -2609,63 +2609,95 @@ out: return err; } -int enetc_setup_tc_mqprio(struct net_device *ndev, void *type_data) +static void enetc_debug_tx_ring_prios(struct enetc_ndev_priv *priv) +{ + int i; + + for (i = 0; i < priv->num_tx_rings; i++) + netdev_dbg(priv->ndev, "TX ring %d prio %d\n", i, + priv->tx_ring[i]->prio); +} + +static void enetc_reset_tc_mqprio(struct net_device *ndev) { struct enetc_ndev_priv *priv = netdev_priv(ndev); - struct tc_mqprio_qopt *mqprio = type_data; struct enetc_hw *hw = &priv->si->hw; struct enetc_bdr *tx_ring; int num_stack_tx_queues; - u8 num_tc; int i; num_stack_tx_queues = enetc_num_stack_tx_queues(priv); - mqprio->hw = TC_MQPRIO_HW_OFFLOAD_TCS; - num_tc = mqprio->num_tc; - if (!num_tc) { - netdev_reset_tc(ndev); - netif_set_real_num_tx_queues(ndev, num_stack_tx_queues); - priv->min_num_stack_tx_queues = num_possible_cpus(); - - /* Reset all ring priorities to 0 */ - for (i = 0; i < priv->num_tx_rings; i++) { - tx_ring = priv->tx_ring[i]; - tx_ring->prio = 0; - enetc_set_bdr_prio(hw, tx_ring->index, tx_ring->prio); - } + netdev_reset_tc(ndev); + netif_set_real_num_tx_queues(ndev, num_stack_tx_queues); + priv->min_num_stack_tx_queues = num_possible_cpus(); - return 0; + /* Reset all ring priorities to 0 */ + for (i = 0; i < priv->num_tx_rings; i++) { + tx_ring = priv->tx_ring[i]; + tx_ring->prio = 0; + enetc_set_bdr_prio(hw, tx_ring->index, tx_ring->prio); } - /* Check if we have enough BD rings available to accommodate all TCs */ - if (num_tc > num_stack_tx_queues) { - netdev_err(ndev, "Max %d traffic classes supported\n", - priv->num_tx_rings); - return -EINVAL; + enetc_debug_tx_ring_prios(priv); +} + +int enetc_setup_tc_mqprio(struct net_device *ndev, void *type_data) +{ + struct enetc_ndev_priv *priv = netdev_priv(ndev); + struct tc_mqprio_qopt *mqprio = type_data; + struct enetc_hw *hw = &priv->si->hw; + int num_stack_tx_queues = 0; + u8 num_tc = mqprio->num_tc; + struct enetc_bdr *tx_ring; + int offset, count; + int err, tc, q; + + if (!num_tc) { + enetc_reset_tc_mqprio(ndev); + return 0; } - /* For the moment, we use only one BD ring per TC. - * - * Configure num_tc BD rings with increasing priorities. - */ - for (i = 0; i < num_tc; i++) { - tx_ring = priv->tx_ring[i]; - tx_ring->prio = i; - enetc_set_bdr_prio(hw, tx_ring->index, tx_ring->prio); + err = netdev_set_num_tc(ndev, num_tc); + if (err) + return err; + + for (tc = 0; tc < num_tc; tc++) { + offset = mqprio->offset[tc]; + count = mqprio->count[tc]; + num_stack_tx_queues += count; + + err = netdev_set_tc_queue(ndev, tc, count, offset); + if (err) + goto err_reset_tc; + + for (q = offset; q < offset + count; q++) { + tx_ring = priv->tx_ring[q]; + /* The prio_tc_map is skb_tx_hash()'s way of selecting + * between TX queues based on skb->priority. As such, + * there's nothing to offload based on it. + * Make the mqprio "traffic class" be the priority of + * this ring group, and leave the Tx IPV to traffic + * class mapping as its default mapping value of 1:1. + */ + tx_ring->prio = tc; + enetc_set_bdr_prio(hw, tx_ring->index, tx_ring->prio); + } } - /* Reset the number of netdev queues based on the TC count */ - netif_set_real_num_tx_queues(ndev, num_tc); - priv->min_num_stack_tx_queues = num_tc; + err = netif_set_real_num_tx_queues(ndev, num_stack_tx_queues); + if (err) + goto err_reset_tc; - netdev_set_num_tc(ndev, num_tc); + priv->min_num_stack_tx_queues = num_stack_tx_queues; - /* Each TC is associated with one netdev queue */ - for (i = 0; i < num_tc; i++) - netdev_set_tc_queue(ndev, i, 1, i); + enetc_debug_tx_ring_prios(priv); return 0; + +err_reset_tc: + enetc_reset_tc_mqprio(ndev); + return err; } EXPORT_SYMBOL_GPL(enetc_setup_tc_mqprio); diff --git a/drivers/net/ethernet/freescale/enetc/enetc_qos.c b/drivers/net/ethernet/freescale/enetc/enetc_qos.c index fcebb54224c0..130ebf6853e6 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_qos.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_qos.c @@ -136,29 +136,21 @@ int enetc_setup_tc_taprio(struct net_device *ndev, void *type_data) { struct tc_taprio_qopt_offload *taprio = type_data; struct enetc_ndev_priv *priv = netdev_priv(ndev); - struct enetc_hw *hw = &priv->si->hw; - struct enetc_bdr *tx_ring; - int err; - int i; + int err, i; /* TSD and Qbv are mutually exclusive in hardware */ for (i = 0; i < priv->num_tx_rings; i++) if (priv->tx_ring[i]->tsd_enable) return -EBUSY; - for (i = 0; i < priv->num_tx_rings; i++) { - tx_ring = priv->tx_ring[i]; - tx_ring->prio = taprio->enable ? i : 0; - enetc_set_bdr_prio(hw, tx_ring->index, tx_ring->prio); - } + err = enetc_setup_tc_mqprio(ndev, &taprio->mqprio); + if (err) + return err; err = enetc_setup_taprio(ndev, taprio); if (err) { - for (i = 0; i < priv->num_tx_rings; i++) { - tx_ring = priv->tx_ring[i]; - tx_ring->prio = taprio->enable ? 0 : i; - enetc_set_bdr_prio(hw, tx_ring->index, tx_ring->prio); - } + taprio->mqprio.qopt.num_tc = 0; + enetc_setup_tc_mqprio(ndev, &taprio->mqprio); } return err; @@ -1611,6 +1603,13 @@ int enetc_qos_query_caps(struct net_device *ndev, void *type_data) struct enetc_si *si = priv->si; switch (base->type) { + case TC_SETUP_QDISC_MQPRIO: { + struct tc_mqprio_caps *caps = base->caps; + + caps->validate_queue_counts = true; + + return 0; + } case TC_SETUP_QDISC_TAPRIO: { struct tc_taprio_caps *caps = base->caps; diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index 17137de9338c..40f4306449eb 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -32,6 +32,7 @@ #include <linux/pkt_sched.h> #include <linux/types.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #define HNAE3_MOD_VERSION "1.0" diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index b4c4fb873568..25be7f8ac7cd 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -20,6 +20,7 @@ #include <net/gro.h> #include <net/ip6_checksum.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <net/tcp.h> #include <net/vxlan.h> #include <net/geneve.h> diff --git a/drivers/net/ethernet/intel/i40e/i40e.h b/drivers/net/ethernet/intel/i40e/i40e.h index 60e351665c70..38c341b9f368 100644 --- a/drivers/net/ethernet/intel/i40e/i40e.h +++ b/drivers/net/ethernet/intel/i40e/i40e.h @@ -33,6 +33,7 @@ #include <linux/net_tstamp.h> #include <linux/ptp_clock_kernel.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <net/tc_act/tc_gact.h> #include <net/tc_act/tc_mirred.h> #include <net/udp_tunnel.h> diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h index 23bc000e77b8..232bc61d9eee 100644 --- a/drivers/net/ethernet/intel/iavf/iavf.h +++ b/drivers/net/ethernet/intel/iavf/iavf.h @@ -30,6 +30,7 @@ #include <linux/jiffies.h> #include <net/ip6_checksum.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <net/udp.h> #include <net/tc_act/tc_gact.h> #include <net/tc_act/tc_mirred.h> diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h index 5f696ae5eaa0..7bd50e49312c 100644 --- a/drivers/net/ethernet/intel/ice/ice.h +++ b/drivers/net/ethernet/intel/ice/ice.h @@ -41,6 +41,7 @@ #include <linux/dim.h> #include <linux/gnss.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <net/tc_act/tc_mirred.h> #include <net/tc_act/tc_gact.h> #include <net/ip.h> diff --git a/drivers/net/ethernet/intel/igc/igc_main.c b/drivers/net/ethernet/intel/igc/igc_main.c index 6ddcbc8b7b6a..cf7f6a5eea3d 100644 --- a/drivers/net/ethernet/intel/igc/igc_main.c +++ b/drivers/net/ethernet/intel/igc/igc_main.c @@ -6205,12 +6205,35 @@ static int igc_tsn_enable_cbs(struct igc_adapter *adapter, return igc_tsn_offload_apply(adapter); } +static int igc_tc_query_caps(struct igc_adapter *adapter, + struct tc_query_caps_base *base) +{ + struct igc_hw *hw = &adapter->hw; + + switch (base->type) { + case TC_SETUP_QDISC_TAPRIO: { + struct tc_taprio_caps *caps = base->caps; + + if (hw->mac.type != igc_i225) + return -EOPNOTSUPP; + + caps->gate_mask_per_txq = true; + + return 0; + } + default: + return -EOPNOTSUPP; + } +} + static int igc_setup_tc(struct net_device *dev, enum tc_setup_type type, void *type_data) { struct igc_adapter *adapter = netdev_priv(dev); switch (type) { + case TC_QUERY_CAPS: + return igc_tc_query_caps(adapter, type_data); case TC_SETUP_QDISC_TAPRIO: return igc_tsn_enable_qbv_scheduling(adapter, type_data); diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index f8925cac61e4..a48588c80317 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -38,7 +38,7 @@ #include <net/ipv6.h> #include <net/tso.h> #include <net/page_pool.h> -#include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <linux/bpf_trace.h> /* Registers */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 27f90baac768..3973d86905e8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -39,6 +39,7 @@ #include <linux/if_bridge.h> #include <linux/filter.h> #include <net/page_pool.h> +#include <net/pkt_sched.h> #include <net/xdp_sock_drv.h> #include "eswitch.h" #include "en.h" diff --git a/drivers/net/ethernet/microchip/lan966x/lan966x_tc.c b/drivers/net/ethernet/microchip/lan966x/lan966x_tc.c index 80625ba0b354..cf0cc7562d04 100644 --- a/drivers/net/ethernet/microchip/lan966x/lan966x_tc.c +++ b/drivers/net/ethernet/microchip/lan966x/lan966x_tc.c @@ -1,6 +1,7 @@ // SPDX-License-Identifier: GPL-2.0+ #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include "lan966x_main.h" diff --git a/drivers/net/ethernet/microchip/sparx5/sparx5_tc.c b/drivers/net/ethernet/microchip/sparx5/sparx5_tc.c index 205246b5af82..e80f3166db7d 100644 --- a/drivers/net/ethernet/microchip/sparx5/sparx5_tc.c +++ b/drivers/net/ethernet/microchip/sparx5/sparx5_tc.c @@ -5,6 +5,7 @@ */ #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include "sparx5_tc.h" #include "sparx5_main.h" diff --git a/drivers/net/ethernet/stmicro/stmmac/hwif.h b/drivers/net/ethernet/stmicro/stmmac/hwif.h index 592b4067f9b8..16a7421715cb 100644 --- a/drivers/net/ethernet/stmicro/stmmac/hwif.h +++ b/drivers/net/ethernet/stmicro/stmmac/hwif.h @@ -567,6 +567,7 @@ struct tc_cbs_qopt_offload; struct flow_cls_offload; struct tc_taprio_qopt_offload; struct tc_etf_qopt_offload; +struct tc_query_caps_base; struct stmmac_tc_ops { int (*init)(struct stmmac_priv *priv); @@ -580,6 +581,8 @@ struct stmmac_tc_ops { struct tc_taprio_qopt_offload *qopt); int (*setup_etf)(struct stmmac_priv *priv, struct tc_etf_qopt_offload *qopt); + int (*query_caps)(struct stmmac_priv *priv, + struct tc_query_caps_base *base); }; #define stmmac_tc_init(__priv, __args...) \ @@ -594,6 +597,8 @@ struct stmmac_tc_ops { stmmac_do_callback(__priv, tc, setup_taprio, __args) #define stmmac_tc_setup_etf(__priv, __args...) \ stmmac_do_callback(__priv, tc, setup_etf, __args) +#define stmmac_tc_query_caps(__priv, __args...) \ + stmmac_do_callback(__priv, tc, query_caps, __args) struct stmmac_counters; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 1a5b8dab5e9b..f44e4e4b4f16 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -5992,6 +5992,8 @@ static int stmmac_setup_tc(struct net_device *ndev, enum tc_setup_type type, struct stmmac_priv *priv = netdev_priv(ndev); switch (type) { + case TC_QUERY_CAPS: + return stmmac_tc_query_caps(priv, priv, type_data); case TC_SETUP_BLOCK: return flow_block_cb_setup_simple(type_data, &stmmac_block_cb_list, diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c index 2cfb18cef1d4..9d55226479b4 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c @@ -1107,6 +1107,25 @@ static int tc_setup_etf(struct stmmac_priv *priv, return 0; } +static int tc_query_caps(struct stmmac_priv *priv, + struct tc_query_caps_base *base) +{ + switch (base->type) { + case TC_SETUP_QDISC_TAPRIO: { + struct tc_taprio_caps *caps = base->caps; + + if (!priv->dma_cap.estsel) + return -EOPNOTSUPP; + + caps->gate_mask_per_txq = true; + + return 0; + } + default: + return -EOPNOTSUPP; + } +} + const struct stmmac_tc_ops dwmac510_tc_ops = { .init = tc_init, .setup_cls_u32 = tc_setup_cls_u32, @@ -1114,4 +1133,5 @@ const struct stmmac_tc_ops dwmac510_tc_ops = { .setup_cls = tc_setup_cls, .setup_taprio = tc_setup_taprio, .setup_etf = tc_setup_etf, + .query_caps = tc_query_caps, }; diff --git a/drivers/net/ethernet/ti/am65-cpsw-qos.c b/drivers/net/ethernet/ti/am65-cpsw-qos.c index e162771893af..8dc2c3085dcf 100644 --- a/drivers/net/ethernet/ti/am65-cpsw-qos.c +++ b/drivers/net/ethernet/ti/am65-cpsw-qos.c @@ -585,6 +585,26 @@ static int am65_cpsw_setup_taprio(struct net_device *ndev, void *type_data) return am65_cpsw_set_taprio(ndev, type_data); } +static int am65_cpsw_tc_query_caps(struct net_device *ndev, void *type_data) +{ + struct tc_query_caps_base *base = type_data; + + switch (base->type) { + case TC_SETUP_QDISC_TAPRIO: { + struct tc_taprio_caps *caps = base->caps; + + if (!IS_ENABLED(CONFIG_TI_AM65_CPSW_TAS)) + return -EOPNOTSUPP; + + caps->gate_mask_per_txq = true; + + return 0; + } + default: + return -EOPNOTSUPP; + } +} + static int am65_cpsw_qos_clsflower_add_policer(struct am65_cpsw_port *port, struct netlink_ext_ack *extack, struct flow_cls_offload *cls, @@ -765,6 +785,8 @@ int am65_cpsw_qos_ndo_setup_tc(struct net_device *ndev, enum tc_setup_type type, void *type_data) { switch (type) { + case TC_QUERY_CAPS: + return am65_cpsw_tc_query_caps(ndev, type_data); case TC_SETUP_QDISC_TAPRIO: return am65_cpsw_setup_taprio(ndev, type_data); case TC_SETUP_BLOCK: diff --git a/drivers/net/ethernet/ti/cpsw_priv.c b/drivers/net/ethernet/ti/cpsw_priv.c index 758295c898ac..e966dd47e2db 100644 --- a/drivers/net/ethernet/ti/cpsw_priv.c +++ b/drivers/net/ethernet/ti/cpsw_priv.c @@ -20,6 +20,7 @@ #include <linux/skbuff.h> #include <net/page_pool.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include "cpsw.h" #include "cpts.h" diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h index 4cabb32a2ad9..cd410a87517b 100644 --- a/include/net/pkt_cls.h +++ b/include/net/pkt_cls.h @@ -788,16 +788,6 @@ struct tc_cls_bpf_offload { bool exts_integrated; }; -struct tc_mqprio_qopt_offload { - /* struct tc_mqprio_qopt must always be the first element */ - struct tc_mqprio_qopt qopt; - u16 mode; - u16 shaper; - u32 flags; - u64 min_rate[TC_QOPT_MAX_QUEUE]; - u64 max_rate[TC_QOPT_MAX_QUEUE]; -}; - /* This structure holds cookie structure that is passed from user * to the kernel for actions and classifiers */ diff --git a/include/net/pkt_sched.h b/include/net/pkt_sched.h index 38207873eda6..fd889fc4912b 100644 --- a/include/net/pkt_sched.h +++ b/include/net/pkt_sched.h @@ -160,8 +160,23 @@ struct tc_etf_qopt_offload { s32 queue; }; +struct tc_mqprio_caps { + bool validate_queue_counts:1; +}; + +struct tc_mqprio_qopt_offload { + /* struct tc_mqprio_qopt must always be the first element */ + struct tc_mqprio_qopt qopt; + u16 mode; + u16 shaper; + u32 flags; + u64 min_rate[TC_QOPT_MAX_QUEUE]; + u64 max_rate[TC_QOPT_MAX_QUEUE]; +}; + struct tc_taprio_caps { bool supports_queue_max_sdu:1; + bool gate_mask_per_txq:1; }; struct tc_taprio_sched_entry { @@ -173,6 +188,7 @@ struct tc_taprio_sched_entry { }; struct tc_taprio_qopt_offload { + struct tc_mqprio_qopt_offload mqprio; u8 enable; ktime_t base_time; u64 cycle_time; diff --git a/net/sched/Kconfig b/net/sched/Kconfig index de18a0dda6df..f5acb535413d 100644 --- a/net/sched/Kconfig +++ b/net/sched/Kconfig @@ -195,8 +195,14 @@ config NET_SCH_ETF To compile this code as a module, choose M here: the module will be called sch_etf. +config NET_SCH_MQPRIO_LIB + tristate + help + Common library for manipulating mqprio queue configurations. + config NET_SCH_TAPRIO tristate "Time Aware Priority (taprio) Scheduler" + select NET_SCH_MQPRIO_LIB help Say Y here if you want to use the Time Aware Priority (taprio) packet scheduling algorithm. @@ -253,6 +259,7 @@ config NET_SCH_DRR config NET_SCH_MQPRIO tristate "Multi-queue priority scheduler (MQPRIO)" + select NET_SCH_MQPRIO_LIB help Say Y here if you want to use the Multi-queue Priority scheduler. This scheduler allows QOS to be offloaded on NICs that have support diff --git a/net/sched/Makefile b/net/sched/Makefile index dd14ef413fda..7911eec09837 100644 --- a/net/sched/Makefile +++ b/net/sched/Makefile @@ -52,6 +52,7 @@ obj-$(CONFIG_NET_SCH_DRR) += sch_drr.o obj-$(CONFIG_NET_SCH_PLUG) += sch_plug.o obj-$(CONFIG_NET_SCH_ETS) += sch_ets.o obj-$(CONFIG_NET_SCH_MQPRIO) += sch_mqprio.o +obj-$(CONFIG_NET_SCH_MQPRIO_LIB) += sch_mqprio_lib.o obj-$(CONFIG_NET_SCH_SKBPRIO) += sch_skbprio.o obj-$(CONFIG_NET_SCH_CHOKE) += sch_choke.o obj-$(CONFIG_NET_SCH_QFQ) += sch_qfq.o diff --git a/net/sched/sch_mqprio.c b/net/sched/sch_mqprio.c index 4c68abaa289b..48ed87b91086 100644 --- a/net/sched/sch_mqprio.c +++ b/net/sched/sch_mqprio.c @@ -17,6 +17,8 @@ #include <net/sch_generic.h> #include <net/pkt_cls.h> +#include "sch_mqprio_lib.h" + struct mqprio_sched { struct Qdisc **qdiscs; u16 mode; @@ -27,6 +29,62 @@ struct mqprio_sched { u64 max_rate[TC_QOPT_MAX_QUEUE]; }; +static int mqprio_enable_offload(struct Qdisc *sch, + const struct tc_mqprio_qopt *qopt, + struct netlink_ext_ack *extack) +{ + struct tc_mqprio_qopt_offload mqprio = {.qopt = *qopt}; + struct mqprio_sched *priv = qdisc_priv(sch); + struct net_device *dev = qdisc_dev(sch); + int err, i; + + switch (priv->mode) { + case TC_MQPRIO_MODE_DCB: + if (priv->shaper != TC_MQPRIO_SHAPER_DCB) + return -EINVAL; + break; + case TC_MQPRIO_MODE_CHANNEL: + mqprio.flags = priv->flags; + if (priv->flags & TC_MQPRIO_F_MODE) + mqprio.mode = priv->mode; + if (priv->flags & TC_MQPRIO_F_SHAPER) + mqprio.shaper = priv->shaper; + if (priv->flags & TC_MQPRIO_F_MIN_RATE) + for (i = 0; i < mqprio.qopt.num_tc; i++) + mqprio.min_rate[i] = priv->min_rate[i]; + if (priv->flags & TC_MQPRIO_F_MAX_RATE) + for (i = 0; i < mqprio.qopt.num_tc; i++) + mqprio.max_rate[i] = priv->max_rate[i]; + break; + default: + return -EINVAL; + } + + err = dev->netdev_ops->ndo_setup_tc(dev, TC_SETUP_QDISC_MQPRIO, + &mqprio); + if (err) + return err; + + priv->hw_offload = mqprio.qopt.hw; + + return 0; +} + +static void mqprio_disable_offload(struct Qdisc *sch) +{ + struct tc_mqprio_qopt_offload mqprio = { { 0 } }; + struct mqprio_sched *priv = qdisc_priv(sch); + struct net_device *dev = qdisc_dev(sch); + + switch (priv->mode) { + case TC_MQPRIO_MODE_DCB: + case TC_MQPRIO_MODE_CHANNEL: + dev->netdev_ops->ndo_setup_tc(dev, TC_SETUP_QDISC_MQPRIO, + &mqprio); + break; + } +} + static void mqprio_destroy(struct Qdisc *sch) { struct net_device *dev = qdisc_dev(sch); @@ -41,37 +99,17 @@ static void mqprio_destroy(struct Qdisc *sch) kfree(priv->qdiscs); } - if (priv->hw_offload && dev->netdev_ops->ndo_setup_tc) { - struct tc_mqprio_qopt_offload mqprio = { { 0 } }; - - switch (priv->mode) { - case TC_MQPRIO_MODE_DCB: - case TC_MQPRIO_MODE_CHANNEL: - dev->netdev_ops->ndo_setup_tc(dev, - TC_SETUP_QDISC_MQPRIO, - &mqprio); - break; - default: - return; - } - } else { + if (priv->hw_offload && dev->netdev_ops->ndo_setup_tc) + mqprio_disable_offload(sch); + else netdev_set_num_tc(dev, 0); - } } -static int mqprio_parse_opt(struct net_device *dev, struct tc_mqprio_qopt *qopt) +static int mqprio_parse_opt(struct net_device *dev, struct tc_mqprio_qopt *qopt, + const struct tc_mqprio_caps *caps, + struct netlink_ext_ack *extack) { - int i, j; - - /* Verify num_tc is not out of max range */ - if (qopt->num_tc > TC_MAX_QUEUE) - return -EINVAL; - - /* Verify priority mapping uses valid tcs */ - for (i = 0; i < TC_BITMASK + 1; i++) { - if (qopt->prio_tc_map[i] >= qopt->num_tc) - return -EINVAL; - } + int err; /* Limit qopt->hw to maximum supported offload value. Drivers have * the option of overriding this later if they don't support the a @@ -80,31 +118,23 @@ static int mqprio_parse_opt(struct net_device *dev, struct tc_mqprio_qopt *qopt) if (qopt->hw > TC_MQPRIO_HW_OFFLOAD_MAX) qopt->hw = TC_MQPRIO_HW_OFFLOAD_MAX; - /* If hardware offload is requested we will leave it to the device - * to either populate the queue counts itself or to validate the - * provided queue counts. If ndo_setup_tc is not present then - * hardware doesn't support offload and we should return an error. + /* If hardware offload is requested, we will leave 3 options to the + * device driver: + * - populate the queue counts itself (and ignore what was requested) + * - validate the provided queue counts by itself (and apply them) + * - request queue count validation here (and apply them) */ - if (qopt->hw) - return dev->netdev_ops->ndo_setup_tc ? 0 : -EINVAL; - - for (i = 0; i < qopt->num_tc; i++) { - unsigned int last = qopt->offset[i] + qopt->count[i]; - - /* Verify the queue count is in tx range being equal to the - * real_num_tx_queues indicates the last queue is in use. - */ - if (qopt->offset[i] >= dev->real_num_tx_queues || - !qopt->count[i] || - last > dev->real_num_tx_queues) - return -EINVAL; - - /* Verify that the offset and counts do not overlap */ - for (j = i + 1; j < qopt->num_tc; j++) { - if (last > qopt->offset[j]) - return -EINVAL; - } - } + err = mqprio_validate_qopt(dev, qopt, + !qopt->hw || caps->validate_queue_counts, + false, extack); + if (err) + return err; + + /* If ndo_setup_tc is not present then hardware doesn't support offload + * and we should return an error. + */ + if (qopt->hw && !dev->netdev_ops->ndo_setup_tc) + return -EINVAL; return 0; } @@ -130,6 +160,67 @@ static int parse_attr(struct nlattr *tb[], int maxtype, struct nlattr *nla, return 0; } +static int mqprio_parse_nlattr(struct Qdisc *sch, struct tc_mqprio_qopt *qopt, + struct nlattr *opt) +{ + struct mqprio_sched *priv = qdisc_priv(sch); + struct nlattr *tb[TCA_MQPRIO_MAX + 1]; + struct nlattr *attr; + int i, rem, err; + + err = parse_attr(tb, TCA_MQPRIO_MAX, opt, mqprio_policy, + sizeof(*qopt)); + if (err < 0) + return err; + + if (!qopt->hw) + return -EINVAL; + + if (tb[TCA_MQPRIO_MODE]) { + priv->flags |= TC_MQPRIO_F_MODE; + priv->mode = *(u16 *)nla_data(tb[TCA_MQPRIO_MODE]); + } + + if (tb[TCA_MQPRIO_SHAPER]) { + priv->flags |= TC_MQPRIO_F_SHAPER; + priv->shaper = *(u16 *)nla_data(tb[TCA_MQPRIO_SHAPER]); + } + + if (tb[TCA_MQPRIO_MIN_RATE64]) { + if (priv->shaper != TC_MQPRIO_SHAPER_BW_RATE) + return -EINVAL; + i = 0; + nla_for_each_nested(attr, tb[TCA_MQPRIO_MIN_RATE64], + rem) { + if (nla_type(attr) != TCA_MQPRIO_MIN_RATE64) + return -EINVAL; + if (i >= qopt->num_tc) + break; + priv->min_rate[i] = *(u64 *)nla_data(attr); + i++; + } + priv->flags |= TC_MQPRIO_F_MIN_RATE; + } + + if (tb[TCA_MQPRIO_MAX_RATE64]) { + if (priv->shaper != TC_MQPRIO_SHAPER_BW_RATE) + return -EINVAL; + i = 0; + nla_for_each_nested(attr, tb[TCA_MQPRIO_MAX_RATE64], + rem) { + if (nla_type(attr) != TCA_MQPRIO_MAX_RATE64) + return -EINVAL; + if (i >= qopt->num_tc) + break; + priv->max_rate[i] = *(u64 *)nla_data(attr); + i++; + } + priv->flags |= TC_MQPRIO_F_MAX_RATE; + } + + return 0; +} + static int mqprio_init(struct Qdisc *sch, struct nlattr *opt, struct netlink_ext_ack *extack) { @@ -139,9 +230,7 @@ static int mqprio_init(struct Qdisc *sch, struct nlattr *opt, struct Qdisc *qdisc; int i, err = -EOPNOTSUPP; struct tc_mqprio_qopt *qopt = NULL; - struct nlattr *tb[TCA_MQPRIO_MAX + 1]; - struct nlattr *attr; - int rem; + struct tc_mqprio_caps caps; int len; BUILD_BUG_ON(TC_MAX_QUEUE != TC_QOPT_MAX_QUEUE); @@ -160,61 +249,18 @@ static int mqprio_init(struct Qdisc *sch, struct nlattr *opt, if (!opt || nla_len(opt) < sizeof(*qopt)) return -EINVAL; + qdisc_offload_query_caps(dev, TC_SETUP_QDISC_MQPRIO, + &caps, sizeof(caps)); + qopt = nla_data(opt); - if (mqprio_parse_opt(dev, qopt)) + if (mqprio_parse_opt(dev, qopt, &caps, extack)) return -EINVAL; len = nla_len(opt) - NLA_ALIGN(sizeof(*qopt)); if (len > 0) { - err = parse_attr(tb, TCA_MQPRIO_MAX, opt, mqprio_policy, - sizeof(*qopt)); - if (err < 0) + err = mqprio_parse_nlattr(sch, qopt, opt); + if (err) return err; - - if (!qopt->hw) - return -EINVAL; - - if (tb[TCA_MQPRIO_MODE]) { - priv->flags |= TC_MQPRIO_F_MODE; - priv->mode = *(u16 *)nla_data(tb[TCA_MQPRIO_MODE]); - } - - if (tb[TCA_MQPRIO_SHAPER]) { - priv->flags |= TC_MQPRIO_F_SHAPER; - priv->shaper = *(u16 *)nla_data(tb[TCA_MQPRIO_SHAPER]); - } - - if (tb[TCA_MQPRIO_MIN_RATE64]) { - if (priv->shaper != TC_MQPRIO_SHAPER_BW_RATE) - return -EINVAL; - i = 0; - nla_for_each_nested(attr, tb[TCA_MQPRIO_MIN_RATE64], - rem) { - if (nla_type(attr) != TCA_MQPRIO_MIN_RATE64) - return -EINVAL; - if (i >= qopt->num_tc) - break; - priv->min_rate[i] = *(u64 *)nla_data(attr); - i++; - } - priv->flags |= TC_MQPRIO_F_MIN_RATE; - } - - if (tb[TCA_MQPRIO_MAX_RATE64]) { - if (priv->shaper != TC_MQPRIO_SHAPER_BW_RATE) - return -EINVAL; - i = 0; - nla_for_each_nested(attr, tb[TCA_MQPRIO_MAX_RATE64], - rem) { - if (nla_type(attr) != TCA_MQPRIO_MAX_RATE64) - return -EINVAL; - if (i >= qopt->num_tc) - break; - priv->max_rate[i] = *(u64 *)nla_data(attr); - i++; - } - priv->flags |= TC_MQPRIO_F_MAX_RATE; - } } /* pre-allocate qdisc, attachment can't fail */ @@ -241,36 +287,9 @@ static int mqprio_init(struct Qdisc *sch, struct nlattr *opt, * supplied and verified mapping */ if (qopt->hw) { - struct tc_mqprio_qopt_offload mqprio = {.qopt = *qopt}; - - switch (priv->mode) { - case TC_MQPRIO_MODE_DCB: - if (priv->shaper != TC_MQPRIO_SHAPER_DCB) - return -EINVAL; - break; - case TC_MQPRIO_MODE_CHANNEL: - mqprio.flags = priv->flags; - if (priv->flags & TC_MQPRIO_F_MODE) - mqprio.mode = priv->mode; - if (priv->flags & TC_MQPRIO_F_SHAPER) - mqprio.shaper = priv->shaper; - if (priv->flags & TC_MQPRIO_F_MIN_RATE) - for (i = 0; i < mqprio.qopt.num_tc; i++) - mqprio.min_rate[i] = priv->min_rate[i]; - if (priv->flags & TC_MQPRIO_F_MAX_RATE) - for (i = 0; i < mqprio.qopt.num_tc; i++) - mqprio.max_rate[i] = priv->max_rate[i]; - break; - default: - return -EINVAL; - } - err = dev->netdev_ops->ndo_setup_tc(dev, - TC_SETUP_QDISC_MQPRIO, - &mqprio); + err = mqprio_enable_offload(sch, qopt, extack); if (err) return err; - - priv->hw_offload = mqprio.qopt.hw; } else { netdev_set_num_tc(dev, qopt->num_tc); for (i = 0; i < qopt->num_tc; i++) @@ -387,7 +406,7 @@ static int mqprio_dump(struct Qdisc *sch, struct sk_buff *skb) struct nlattr *nla = (struct nlattr *)skb_tail_pointer(skb); struct tc_mqprio_qopt opt = { 0 }; struct Qdisc *qdisc; - unsigned int ntx, tc; + unsigned int ntx; sch->q.qlen = 0; gnet_stats_basic_sync_init(&sch->bstats); @@ -411,15 +430,9 @@ static int mqprio_dump(struct Qdisc *sch, struct sk_buff *skb) spin_unlock_bh(qdisc_lock(qdisc)); } - opt.num_tc = netdev_get_num_tc(dev); - memcpy(opt.prio_tc_map, dev->prio_tc_map, sizeof(opt.prio_tc_map)); + mqprio_qopt_reconstruct(dev, &opt); opt.hw = priv->hw_offload; - for (tc = 0; tc < netdev_get_num_tc(dev); tc++) { - opt.count[tc] = dev->tc_to_txq[tc].count; - opt.offset[tc] = dev->tc_to_txq[tc].offset; - } - if (nla_put(skb, TCA_OPTIONS, sizeof(opt), &opt)) goto nla_put_failure; diff --git a/net/sched/sch_mqprio_lib.c b/net/sched/sch_mqprio_lib.c new file mode 100644 index 000000000000..c58a533b8ec5 --- /dev/null +++ b/net/sched/sch_mqprio_lib.c @@ -0,0 +1,117 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#include <linux/net.h> +#include <linux/netdevice.h> +#include <linux/netlink.h> +#include <linux/types.h> +#include <net/pkt_sched.h> + +#include "sch_mqprio_lib.h" + +/* Returns true if the intervals [a, b) and [c, d) overlap. */ +static bool intervals_overlap(int a, int b, int c, int d) +{ + int left = max(a, c), right = min(b, d); + + return left < right; +} + +static int mqprio_validate_queue_counts(struct net_device *dev, + const struct tc_mqprio_qopt *qopt, + bool allow_overlapping_txqs, + struct netlink_ext_ack *extack) +{ + int i, j; + + for (i = 0; i < qopt->num_tc; i++) { + unsigned int last = qopt->offset[i] + qopt->count[i]; + + if (!qopt->count[i]) { + NL_SET_ERR_MSG_FMT_MOD(extack, "No queues for TC %d", + i); + return -EINVAL; + } + + /* Verify the queue count is in tx range being equal to the + * real_num_tx_queues indicates the last queue is in use. + */ + if (qopt->offset[i] >= dev->real_num_tx_queues || + last > dev->real_num_tx_queues) { + NL_SET_ERR_MSG_FMT_MOD(extack, + "Queues %d:%d for TC %d exceed the %d TX queues available", + qopt->count[i], qopt->offset[i], + i, dev->real_num_tx_queues); + return -EINVAL; + } + + if (allow_overlapping_txqs) + continue; + + /* Verify that the offset and counts do not overlap */ + for (j = i + 1; j < qopt->num_tc; j++) { + if (intervals_overlap(qopt->offset[i], last, + qopt->offset[j], + qopt->offset[j] + + qopt->count[j])) { + NL_SET_ERR_MSG_FMT_MOD(extack, + "TC %d queues %d@%d overlap with TC %d queues %d@%d", + i, qopt->count[i], qopt->offset[i], + j, qopt->count[j], qopt->offset[j]); + return -EINVAL; + } + } + } + + return 0; +} + +int mqprio_validate_qopt(struct net_device *dev, struct tc_mqprio_qopt *qopt, + bool validate_queue_counts, + bool allow_overlapping_txqs, + struct netlink_ext_ack *extack) +{ + int i, err; + + /* Verify num_tc is not out of max range */ + if (qopt->num_tc > TC_MAX_QUEUE) { + NL_SET_ERR_MSG(extack, + "Number of traffic classes is outside valid range"); + return -EINVAL; + } + + /* Verify priority mapping uses valid tcs */ + for (i = 0; i <= TC_BITMASK; i++) { + if (qopt->prio_tc_map[i] >= qopt->num_tc) { + NL_SET_ERR_MSG(extack, + "Invalid traffic class in priority to traffic class mapping"); + return -EINVAL; + } + } + + if (validate_queue_counts) { + err = mqprio_validate_queue_counts(dev, qopt, + allow_overlapping_txqs, + extack); + if (err) + return err; + } + + return 0; +} +EXPORT_SYMBOL_GPL(mqprio_validate_qopt); + +void mqprio_qopt_reconstruct(struct net_device *dev, struct tc_mqprio_qopt *qopt) +{ + int tc, num_tc = netdev_get_num_tc(dev); + + qopt->num_tc = num_tc; + memcpy(qopt->prio_tc_map, dev->prio_tc_map, sizeof(qopt->prio_tc_map)); + + for (tc = 0; tc < num_tc; tc++) { + qopt->count[tc] = dev->tc_to_txq[tc].count; + qopt->offset[tc] = dev->tc_to_txq[tc].offset; + } +} +EXPORT_SYMBOL_GPL(mqprio_qopt_reconstruct); + +MODULE_LICENSE("GPL"); diff --git a/net/sched/sch_mqprio_lib.h b/net/sched/sch_mqprio_lib.h new file mode 100644 index 000000000000..63f725ab8761 --- /dev/null +++ b/net/sched/sch_mqprio_lib.h @@ -0,0 +1,18 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __SCH_MQPRIO_LIB_H +#define __SCH_MQPRIO_LIB_H + +#include <linux/types.h> + +struct net_device; +struct netlink_ext_ack; +struct tc_mqprio_qopt; + +int mqprio_validate_qopt(struct net_device *dev, struct tc_mqprio_qopt *qopt, + bool validate_queue_counts, + bool allow_overlapping_txqs, + struct netlink_ext_ack *extack); +void mqprio_qopt_reconstruct(struct net_device *dev, + struct tc_mqprio_qopt *qopt); + +#endif diff --git a/net/sched/sch_taprio.c b/net/sched/sch_taprio.c index c322a61eaeea..1c95785932b9 100644 --- a/net/sched/sch_taprio.c +++ b/net/sched/sch_taprio.c @@ -26,6 +26,8 @@ #include <net/sock.h> #include <net/tcp.h> +#include "sch_mqprio_lib.h" + static LIST_HEAD(taprio_list); #define TAPRIO_ALL_GATES_OPEN -1 @@ -924,7 +926,7 @@ static int taprio_parse_mqprio_opt(struct net_device *dev, struct netlink_ext_ack *extack, u32 taprio_flags) { - int i, j; + bool allow_overlapping_txqs = TXTIME_ASSIST_IS_ENABLED(taprio_flags); if (!qopt && !dev->num_tc) { NL_SET_ERR_MSG(extack, "'mqprio' configuration is necessary"); @@ -937,52 +939,17 @@ static int taprio_parse_mqprio_opt(struct net_device *dev, if (dev->num_tc) return 0; - /* Verify num_tc is not out of max range */ - if (qopt->num_tc > TC_MAX_QUEUE) { - NL_SET_ERR_MSG(extack, "Number of traffic classes is outside valid range"); - return -EINVAL; - } - /* taprio imposes that traffic classes map 1:n to tx queues */ if (qopt->num_tc > dev->num_tx_queues) { NL_SET_ERR_MSG(extack, "Number of traffic classes is greater than number of HW queues"); return -EINVAL; } - /* Verify priority mapping uses valid tcs */ - for (i = 0; i <= TC_BITMASK; i++) { - if (qopt->prio_tc_map[i] >= qopt->num_tc) { - NL_SET_ERR_MSG(extack, "Invalid traffic class in priority to traffic class mapping"); - return -EINVAL; - } - } - - for (i = 0; i < qopt->num_tc; i++) { - unsigned int last = qopt->offset[i] + qopt->count[i]; - - /* Verify the queue count is in tx range being equal to the - * real_num_tx_queues indicates the last queue is in use. - */ - if (qopt->offset[i] >= dev->num_tx_queues || - !qopt->count[i] || - last > dev->real_num_tx_queues) { - NL_SET_ERR_MSG(extack, "Invalid queue in traffic class to queue mapping"); - return -EINVAL; - } - - if (TXTIME_ASSIST_IS_ENABLED(taprio_flags)) - continue; - - /* Verify that the offset and counts do not overlap */ - for (j = i + 1; j < qopt->num_tc; j++) { - if (last > qopt->offset[j]) { - NL_SET_ERR_MSG(extack, "Detected overlap in the traffic class to queue mapping"); - return -EINVAL; - } - } - } - - return 0; + /* For some reason, in txtime-assist mode, we allow TXQ ranges for + * different TCs to overlap, and just validate the TXQ ranges. + */ + return mqprio_validate_qopt(dev, qopt, true, allow_overlapping_txqs, + extack); } static int taprio_get_start_time(struct Qdisc *sch, @@ -1203,7 +1170,8 @@ static u32 tc_map_to_queue_mask(struct net_device *dev, u32 tc_mask) static void taprio_sched_to_offload(struct net_device *dev, struct sched_gate_list *sched, - struct tc_taprio_qopt_offload *offload) + struct tc_taprio_qopt_offload *offload, + const struct tc_taprio_caps *caps) { struct sched_entry *entry; int i = 0; @@ -1217,7 +1185,11 @@ static void taprio_sched_to_offload(struct net_device *dev, e->command = entry->command; e->interval = entry->interval; - e->gate_mask = tc_map_to_queue_mask(dev, entry->gate_mask); + if (caps->gate_mask_per_txq) + e->gate_mask = tc_map_to_queue_mask(dev, + entry->gate_mask); + else + e->gate_mask = entry->gate_mask; i++; } @@ -1261,7 +1233,8 @@ static int taprio_enable_offload(struct net_device *dev, return -ENOMEM; } offload->enable = 1; - taprio_sched_to_offload(dev, sched, offload); + mqprio_qopt_reconstruct(dev, &offload->mqprio.qopt); + taprio_sched_to_offload(dev, sched, offload, &caps); for (tc = 0; tc < TC_MAX_QUEUE; tc++) offload->max_sdu[tc] = q->max_sdu[tc]; @@ -1981,18 +1954,11 @@ static int taprio_dump(struct Qdisc *sch, struct sk_buff *skb) struct sched_gate_list *oper, *admin; struct tc_mqprio_qopt opt = { 0 }; struct nlattr *nest, *sched_nest; - unsigned int i; oper = rtnl_dereference(q->oper_sched); admin = rtnl_dereference(q->admin_sched); - opt.num_tc = netdev_get_num_tc(dev); - memcpy(opt.prio_tc_map, dev->prio_tc_map, sizeof(opt.prio_tc_map)); - - for (i = 0; i < netdev_get_num_tc(dev); i++) { - opt.count[i] = dev->tc_to_txq[i].count; - opt.offset[i] = dev->tc_to_txq[i].offset; - } + mqprio_qopt_reconstruct(dev, &opt); nest = nla_nest_start_noflag(skb, TCA_OPTIONS); if (!nest) |